artifact_id;status_id;status_name;priority;submitter_id;submitter_name;assigned_to_id;assigned_to_name;open_date;close_date;last_modified_date;summary;details;"Category";"Group";"Resolution" 839;1;"Open";3;1995;"Todd Koeckeritz";103;"Ton Roosendaal";"2004-01-03 09:28";"1970-01-01 01:00";"2007-05-24 12:03";"Vertex group management conflict w/ undo";"There are several problems with managing vertex groups and using the mesh undo feature. The first and easiest to reproduce causes newly created vertex groups to already have vertices assigned to them (the most obvious sympton anyway). The following steps reproduce this bug in 2.30 and the current CVS version as of 02 Jan 2004. 2.29 was tested also using ""Reload Original Data"" as it has no undo. 2.29 did not produce these results. Start a fresh blender and add a cube if necessary. TABKEY to edit mode, select a vertice, create a new vertex group and assign the vertice to that group. GKEY to grab the vertex and move to a different position. AKEY to deselect the vertice. Delete the vertex group. UKEY to undo the move of the vertice. Create a new vertex group and before assigning any vertices to it, press the Select button. You will select the vertice(s) that were in the deleted group. This is at the very least inconvenient, but I suspect there is a significant mismanagement of memory or data involved in this. I will post a follow up describing another vertex group/undo problem that I believe is another sympton of the under lying bug. I will do my best to either create a .blend file I can upload or hopefully a procedure to reproduce this other problem, but it may take a few days and I may only be able to describe the problem as it happens to me infrequently.";"Tools";"None";"New" 1368;1;"Open";3;2449;"Paolo Furani";103;"Ton Roosendaal";"2004-06-10 16:15";"1970-01-01 01:00";"2007-05-24 12:03";"Wrong rendering, of wire object.";"When the button, RayTrasnp are active, the wire object, behind, an transparent, is rendered with normal shading, only in the intersecting parts.";"Render (internal)";"None";"Investigate" 1403;1;"Open";3;617;"Yann Vernier";139;"Martin Poirier";"2004-06-26 15:40";"1970-01-01 01:00";"2007-05-24 12:03";"Grab won't move objects with size=0";"Resizing an object to 0 along any axis makes the Grab command stop working. It simply won't move the object. Rotate is also affected, while Scale works as expected (0*anything=0, so the 0-axis won't be scaled). Values that are very nearly 0, such as those represented by ""-0.0000"", do not trigger this problem. ";"None";"None";"New" 1419;1;"Open";3;1993;"Chris M ";103;"Ton Roosendaal";"2004-07-07 14:14";"1970-01-01 01:00";"2007-05-24 12:03";"CTRL - left|right not working";"A small bug, but a little annoying sometimes. When the cursor is over a windowborder (the cursor changes in a arrow pointing tow ways), CTRL LEFT and CTRL RIGHT do not work anymore. You have to move the cursor in a window before you can move to another screen. AMD 2000+ Linux Debian unstable NVIDIA geforce 4 ti (asus) Videodrivers nvidia binary version 1.0-53.36 Blender 2.33a Regards, Chris ";"None";"None";"Postponed" 1444;1;"Open";1;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2004-07-16 00:06";"1970-01-01 01:00";"2007-05-24 12:03";"Spiky points in weighted creases (subsurf)";"the new feature of ""Weighted Creases for Subsurfs"" has bad behaviors: - sometimes the viewport shows another result as the rendering (image 1+2) - instead of sharper edges things get little horns (image 3) - bad abort behavior discribed in bug 1439 with addition that pressing right mouse button should abort too - missing special options by hold Ctrl for Creasing in 0.100 Steps and Shift for fine adjustment";"None";"None";"Postponed" 1470;1;"Open";3;2778;"andrea barolo";103;"Ton Roosendaal";"2004-07-24 10:43";"1970-01-01 01:00";"2007-05-24 12:03";"autosmooth not visible in 3dview";"when i press setsmooth an object it's or completely smooth or completely solid in 3d view, autosmooth feature doesn't work as it would in 3d view, in rendering it's working right, but I dont think it's a good thing render everytime i want to see if the autosmooth is set at the right angle. A notice: if you give different smoothing groups, on different faces the smoothing works right with seleceted faces, it's autosmooth that is not rendered/updated in 3d view even if i switch from edit to object mode. my specs AMD Athlon XP 2200+ 512 Megs Geforce FX5900 128 Megs Windows Xp pro Latest Driver (61 etc.) blender 2.33a";"Render (internal)";"None";"New" 1501;1;"Open";3;2119;"Gabriel Beloin";3002;"Joseph Eagar";"2004-08-06 23:03";"1970-01-01 01:00";"2007-05-24 12:03";"Various UI beaviour for sound";"Windowz Xp.. For the first time today i tryed to sound feature and bumpped across some stuff(all releated to interface): -In Sound block buttons, when loading a wav file, the file browser and data browser(when there is too much datablock) open in the same windows as the buttons windows, which is rarely big. some loading button, like load texture, do it nicely by openning the file browser in the first windows on the top. -When the wav file is loaded, there is a ""number of use"" side to the file name(the 2rd dropdown): it's always n+1 of the reality. and some time even bigger.(like I just loaded it and it show 3...) also the number of use of the datablock also show itself 1/10 time.(this one is for the first dropdown) -How can we remove these datablock? There is no X beside the addblock dropdown menu.(on user side, this might confuse some: what i'm doing now is loading a file or using it: this interface look like using it. -This lead to an other note: where is the sound datablock in the oops windows? -Next: when saving the file once: the drop down is now all red.(maybe a feature i'm not aware of). So you have to save twice to remove datablock. -the loop and ping pong button is not following the use you intend on the datablock. for exemple you use it in an actuator. you can set volume and pitch, these tow values follow in the sound block button but if you set a mode for playback, ex: loop or ping pong, the button don't get pressed in the sound block buttons and vise versa. - and on a side note i'm asking myself what is the use of the audio timeline: such windows could be good for audio edition or even audio sequence edition, instead of having it in the video sequencer. But since we can't... We can't even ear the sound under the cursor. I remember clearly it worked once. -last: mixdown should open a file browser to save the file correctly since it dont save the file at the same place as the .blend AND don't have the same name. that's it for my analyse of the sound feature so far.";"None";"None";"New" 1601;1;"Open";3;2990;"David Becker";139;"Martin Poirier";"2004-09-15 13:59";"1970-01-01 01:00";"2007-05-24 12:03";"Bug with multiple ""Copy Rotation"" constraints in single hierarchy";"If you have multiple bones that are parented to each other and give each of them a ""copy Rotation"" constraint with different strength (for example, to bend multiple bones in a smooth bow and controlling that with a single bone) the bones get all weird. The child bones don't use the parent's rotation to determine their position (and instead assume the parent to be rotated much stronger) and only the lowest bone in the hierarchy gets updated in real time when you rotate the controller bone. I've attached an example of such a hierarchy. The separate bone to the left (Bone.003) is the controller, if you rotate it the hierarchy on the right freaks out.";"Tools";"None";"Investigate" 1629;1;"Open";3;665;"styken ";103;"Ton Roosendaal";"2004-10-08 23:36";"1970-01-01 01:00";"2007-05-24 12:03";"Hooks on curve has unwanted behaviour";"1) When adding a hook to a curve and then subdividing the curve the hook gets assigned to the newly added subdivision instead of the end point which I assigned from the beginning. 2) Also when adding a bevel object to a curve with hooks and the convert to mesh the result is abnormal, looks like rubbish on screen! ";"Tools";"None";"New" 1643;1;"Open";3;3056;"Kiernan Holland";103;"Ton Roosendaal";"2004-10-15 02:27";"1970-01-01 01:00";"2007-05-24 12:03";"Alt-C on NURBS with UV texturing doesn't carry UV texture coodinates over to mesh. ";"It used to be the case that when you did an Alt-C on a surface type (aka NURBS), the UV coordinates that are automatically given to NURBS are no longer passed to the mesh in conversion.. It used to be that I could convert a NURBS surface to a mesh, and the UV texturing coordinates would carry over to the mesh.. This is not the case anymore.. This used to be the case, evidently someone removed it.. Note, I'm take a NURBS Surface Curve (not a NURBS Curve) and extruding it into a cylinder, because the basic cylinder doesn't have the correct parameterization for a uniformly weighted cylinder.. You will notice that quad placement is pin-cushioned in a ""+"" pattern, that's because of the use of Square-roots and such for the weighting.. Anyhow.. Hint: Use (x * cos(theta),z * sin(theta)) for placement of the control points at rim of cylinder.. That should place the points in a circle. If you weighting is 1.0, you should get a perfectly uniform cylinder. The last isoparm reuses the first set of control points (this is when the surface's U is made cyclic, the cyclic parameterization tends to create this wonky pattern on the last isoparm pair. Its done this in blender forever.. I suppose why other people don't complain about it is NURBS are a bit advanced for some modellers. I use NURBS more just for preceision and control of tesselation, then I take the NURBS convert them to meshes, use subsurfaces, etc.. Its probably about time someone implemented a CV control system in blender, as its NURBS surfaces currently use Hull points, not true CV's surface points like Alias, Maya and Rhino.. A combination of CV on a surface NURBS and subdisvision surfaces could be very powerful, that should be a future blender primitive. ";"None";"None";"New" 1654;1;"Open";3;3043;"rednelb ";103;"Ton Roosendaal";"2004-10-15 18:24";"1970-01-01 01:00";"2007-05-24 12:03";"when leaving camera view, viewport is always in perspective";"regardless it was before orthographic or not.";"None";"None";"New" 1768;1;"Open";3;375;"Paul Lunneberg";103;"Ton Roosendaal";"2004-11-09 21:41";"1970-01-01 01:00";"2007-05-24 12:03";"Slow Redraw of different items";"Windows 2000 SP 4 ATI Radeon 7000 Series 1280x1024 16Bit Color 1Gb Ram 2.35 Prerelease build from testing forum. When rendering the top text bar that shows current scene specs and also what sample is being worked on for rendering redraws very slowly when pressing F12 or the Render button. This slowdown also happens in the file dialogue window while scrolling or moving the mouse over the items in the list. It is very similar to what was happening in the previous version with the menu drawing.";"None";"None";"New" 1770;1;"Open";2;665;"styken ";139;"Martin Poirier";"2004-11-10 14:04";"1970-01-01 01:00";"2007-05-24 12:03";"Rotating in texture space do not work";"Version: 2.35 RC 1 OS: WinXP When I try to rotate in texture space nothing happens. To reproduce 1 add a cube with a texture, ex wood with default setting 2 press T to enter texture space. Choose grab/move 3 press R to rotate now you see the rotation angle in lower left but nothing happens to texture This might be a todo thing, or it might be a bug? //styken";"None";"None";"Investigate" 1975;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2004-12-06 01:11";"1970-01-01 01:00";"2007-05-24 12:03";"""Random"" particles move into cube shape";"This file has a round mesh emitting particles on one frame with a ""random"" starting velocity. I expect them to move away from the mesh into a randomly fuzzy-looking shape roughly like a sphere, but instead they move into almost a perfect cube. The longer they live, the more the shape becomes a cube. Blender 2.35a Windows";"None";"None";"New" 1999;1;"Open";3;2423;"Chris Burt";103;"Ton Roosendaal";"2004-12-10 02:50";"1970-01-01 01:00";"2007-05-24 12:03";"Overlapping panels render incorrectly";"Essentially the problem is that any overlapping panel in the interface has no ""on top"" state. When the mouse cursor is places over an area on the screen where a button from the ""bottom"" panel exists, that button, or group of buttons is drawn as if it is on top. I hope the attached screen shot explains the issue well enough. It occurs even if alpha for the panel background is set to 1. My setup: Blender Foundation Limited Edition ;) ATI Radeon 9600 Windows XP Home";"None";"None";"New" 2051;1;"Open";3;980;"Alexander Ewering ";3957;"Erwin Coumans";"2004-12-23 01:25";"1970-01-01 01:00";"2007-05-24 12:03";"Game engine mixrate fixed";"The game engine seems to always mix at 22.05 kHz. This is pretty bad quality in times of 96 kHz 5.1 sound ;-) Also, not all soundcards even support 22.05 kHz. For example, my RME HDSP only supports 44.1 kHz and up. Should be changable... I wanted to test if sound in the engine works on windows (it doesn't on Linux), but heh, I've been stopped by my soundcard not supporting 22.05 kHz. ";"None";"None";"New" 2076;1;"Open";3;3366;"Thomas Buschhardt";103;"Ton Roosendaal";"2004-12-30 16:36";"1970-01-01 01:00";"2007-05-24 12:03";"No taper change by change the transformation properties";"Hallo, I've found a minimal bug in the taper-function in ""Curve and Surface"". The taper dosn't change when I change the transformations properties of the taper objects by value. Ok here come the repeat: 1. Open a new Blender Project 2. Add a ""Bezier Curve"" (thats our final object) 3. Exit Editmode (Key ""Tab"") 4. Add a ""Bezier Circle"" 5. Open ""Transformations Properties"" (Key ""N"") 6. Rename the Circle to ""bevel"" 7. Exit Editmode (Key ""Tab"") 8. Add a ""Bezier Curve"" 9. Rename the Curve to ""taper"" 10. Exit Editmode (Key ""Tab"") 11. Select the first Bezier Curve (Nr.2) 12. Change in the ""Buttons Window"" (Key ""Shift-F7"" then Key ""F9"") under the panel ""Curve and Surface"" the property ""BevOb"" to ""bevel"" and the property ""TaperOb"" to ""taper"" OK now you can change the ""taper""-Curve by selecting a vertex and moving them (Key ""G""). But if you change the vertex coordinates by the ""Transformations Properties""-Window the final object dosn't change immediate. 13. select a vertex in the object ""taper"" and change the Y-Property in the ""Transformations Properties"" Window -> No change to the final object. Bye Thomas";"Tools";"None";"New" 2109;1;"Open";1;555;"Omar Modesto ";103;"Ton Roosendaal";"2005-01-06 16:05";"1970-01-01 01:00";"2007-05-24 12:03";"Ray-trace with image-textures not filtered.";"Again I return with the transparent shadow issue. Previously (Blender 2.30-2.35a), transparent shadows using textures resulted in very bad filtering (corrupt scanlines, white dots...), but now it seems that they are not at all textured since the 2.36 release (in which the above mentioned problems were said to be fixed). The non-filtering results in a very pixelated shadow (which can be solved, for the moment, by using a very high-resolution image). Again I supply the coders with a .blend file for your testing needs. P.S.: Please bear with me, Ton :)";"None";"None";"Postponed" 2189;1;"Open";3;2930;"douglas ollivier";103;"Ton Roosendaal";"2005-01-28 00:06";"1970-01-01 01:00";"2007-05-24 12:03";"panorama mode and edge settings not functioning correctly";"using pano mode and edge settings (blender 2.36 Mach OSX version 10.3 (i assume all OS's)) results in one of two things. Unified renderer, black edges are crossing the render at every boarder between slices. normal renderer, gaps are at every boarder of the slices. also on the same note, DispView does not work with panorama rendering (not a big problem) thanks again guys. Douglas Ollivier";"Render (internal)";"None";"New" 2217;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2005-02-07 01:53";"1970-01-01 01:00";"2007-05-24 12:03";"""Set Solid"" for curve with BevOb appears solid on U but not V";"Blender 2.36 nVidia GeForce6800 Ultra Mac OS X 10.3.7 This scene is a bezier curve with another bezier curve as a BevOb. The bevel object looks like the profile of a stair. Clicking ""Set Solid"" only makes it appear solid along the U direction, but not the V direction. This only affects drawing, not rendering. The render looks fine. p.s.: Why does the curve object have ""Set Solid"" and ""Set Smooth"", but no ""Auto Smooth""?";"None";"None";"New" 2244;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2005-02-15 23:59";"1970-01-01 01:00";"2007-05-24 12:03";"Changing ""All Edges"" for one mesh changes it for others too";"1. Add two mesh objects to a scene. 2. With one selected, enter edit mode. 3. Select ""All Edges"" (Mesh Tools 1 panel). 4. Leave edit mode. 5. Select other mesh object. 6. Enter edit mode. 7. Behold! It also now has ""All Edges"" selected. 8. Leave edit mode. 9. Add another mesh object to the scene. 10. Behold! It too has ""All Edges"" selected. If you turn ""All Edges"" off for one of them, it will turn off for the rest too. You won't see the change until you enter and leave edit mode for each. Blender post-2.36 CVS Mac OS X 10.3.8 nVidia GeForce 6800 Ultra";"None";"None";"New" 2277;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2005-02-28 23:11";"1970-01-01 01:00";"2007-05-24 12:03";"If mesh has two particle effects, only the first one works";"According to this page: http://download.blender.org/documentation/htmlII/x6358.html ""An Object might have multiple effects. A row of radiobuttons without labels below the New Effect button toggles between them."" But if I add two particle effects to a single mesh, only the first one works. The second never appears. Blender 2.36 CVS Mac OS X 10.3.8 nVidia geForce 6800 Ultra";"None";"None";"New" 2282;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2005-03-02 00:36";"1970-01-01 01:00";"2007-05-24 12:03";"cmir broken";"Load this attached file in Blender 2.25. Hit F12. Spherical reflection map using cmir! Now load it in Blender 2.36. Hit F12. No reflection map. When did this functionality disappear? Sometime since 2.25. I don't have all old versions to test. It's a mystery...";"Render (internal)";"None";"Investigate" 2343;1;"Open";2;980;"Alexander Ewering ";103;"Ton Roosendaal";"2005-03-19 14:53";"1970-01-01 01:00";"2007-05-24 12:03";"Fastshade is broken for scaled/transformed objects";"I'm starting to think that none of what I'm reporting here was ever meant to be implemented in fastshade, so maybe i should stop reporting. While I do think that the missing image pixel RGB-value-clipping in my previous report IS a bug, THIS one is probably just a missing feature: Fastshade (shaded view) ALWAYS uses ""Orco"" projection, no matter what you setup (Glob, UV, etc.) ";"None";"None";"New" 2444;1;"Open";3;2419;"Dominic Agoro-Ombaka";139;"Martin Poirier";"2005-04-13 02:29";"1970-01-01 01:00";"2007-05-24 12:03";"Animated constraints aren't baked in actions + Copy Rotation baking bug";"Hi. For actions with animated constraints, only the constraint's influence at the first frame is used. Please see the attached file. My apologies if this has been posted before, but a search in the bug tracker, elysiun, and the official docs turned up nothing about this issue. I'm using the baker extensively in a large blender game engine project for college, due in about 2 weeks, so I would be eternally grateful for this bugfix =) Tested with: bf-blender April 4, 2.36, 2.34 Win2000 pro sp4 and Geforce3 ti200 Thanks";"None";"None";"Investigate" 2505;1;"Open";3;3212;"Randall Rickert";100;"Nobody";"2005-05-01 19:11";"1970-01-01 01:00";"2012-10-29 12:27";"if user scripts folder exists, bundled scripts don't appear in menus";"If I make a user scripts directory as ~/.blender/scripts, the scripts that come bundled in the Blender .app package don't appear in the menus, including the import/export options. If I make the directory while Blender is running and then refresh the scripts from the script window, everything works the way it should (bundled scripts appear alongside user scripts). However, after restarting Blender only the user scripts will appear. Mac OS X 10.3.9 Blender 2.36";"None";"None";"Postponed" 2527;1;"Open";3;2807;"Nick Winters";103;"Ton Roosendaal";"2005-05-08 02:21";"1970-01-01 01:00";"2007-05-24 12:03";"Zero area faces and reversed faces in decimator result";"The decimator has an annoying habit of generating flipped faces in the result [particularly on flat surfaces], and zero area faces [also often on flat surfaces] somewhat related [mostly so you can see my instructions give different results for you], it seems to give different results in different builds/versions to see what I mean, subdivide a plane 3 times, then decimate to 115 triangles I've attached a pic of the problems, zero area faces circled in yellow, flipped face circled in blue";"None";"None";"New" 2529;1;"Open";3;3739;"Michal Sudolsky";103;"Ton Roosendaal";"2005-05-09 14:18";"1970-01-01 01:00";"2007-05-24 12:03";"Curve deform with dupliframes animation";"Windows XP professional If you render this file blender generate error (this was tested on many computers with os windows). Feature ""curve deform"" has bad behavior. Object ""Cube"" has 10 copies created with feature ""dupliframes"". But these copies are distributed along vertical line no curve. ";"None";"None";"New" 2609;1;"Open";3;980;"Alexander Ewering ";103;"Ton Roosendaal";"2005-05-23 11:54";"1970-01-01 01:00";"2007-05-24 12:03";"+/- keys and expose events not handled anymore in Renderwin";"I think the Summary says it all :) +/- keys and expose events are not handled anymore in Renderwin during render, only after finishing. In contrast to 2.36, where they are always handled pretty well. ";"None";"None";"New" 2679;1;"Open";3;3663;"Patrick O'Toole";103;"Ton Roosendaal";"2005-06-03 01:10";"1970-01-01 01:00";"2007-05-24 12:03";"NURBS sphere mapping error";"Specs of Computers Tested: Operating system: Microsoft Windows XP Professional Version 2002 Service Pack 2 3-D graphics card: Mobility Radeon 7500 Versions of Blender: 2.36 and 2.37 ~ Operating system: Microsoft Windows XP Home Edition Version 2002 Service Pack 2 3-D graphics card: NVIDIA GeForce 4 MX 440 with APG8X Version of Blender: 2.37 Summary of problem: When mapping on a NURBS sphere an image texture, a stripe of color can appear. The band of color does not react to ""nor"" when it is set. Instructions to reproduce: 1.) Open your image editor or browser that and find or make an image that has a central spot of color and a background that is easily discernible from the dot in the middle. (I have verified that the following file types are affected by this bug: PNG, JPEG, GIF. There may be more.) 2.) Open Blender version 2.37 or 2.36. 3.) Delete the default cube and insert a NURBS sphere. 4.) Assign a material to the sphere. 5.) Create an image type texture and locate the image with the dot that you found/made. 6.) Make sure that the material and texture are linked. 7.) In the map settings, make sure that ""orco,"" ""flat,"" and ""col"" are selected. 8.) Move your lamp and camera as necessary to see either of the poles of the sphere. 9.) Render using either engine (Blender internal or Yafray WITHOUT xml) and observe the mapping error on the surface of the sphere.";"None";"None";"Investigate" 2743;1;"Open";3;3275;"adrian llewellyn";103;"Ton Roosendaal";"2005-06-13 14:43";"1970-01-01 01:00";"2007-05-24 12:03";"hsv colour sliders not always moving in 0.01 increments when clicked";"Blender 2.37 version, windows xp pro - amd barton, nvidia mx4 440 card. I noticed that when using the colour sliders, only in HSV mode, the S & V sliders don't always move in 0.01 increments when clicked - H value seems ok.";"None";"None";"New" 2775;1;"Open";3;173;"Enrico Valenza";103;"Ton Roosendaal";"2005-06-22 17:42";"1970-01-01 01:00";"2007-05-24 12:03";"Joining two or more armatures loose all the constrains";"Joining two or more armatures (Cntrl+J) loose all the applied constrains, including the Ik solvers. EnV";"None";"None";"Postponed" 2783;1;"Open";3;139;"Martin Poirier";103;"Ton Roosendaal";"2005-06-24 03:58";"1970-01-01 01:00";"2007-05-24 12:03";"Link / Append Conflicts";"Two things, but related, so I'm making a single report. 1) Linking from a file and then appending from a file, makes all the previously linked data local (because append calls all_local(lib)) 2) Linking and then appending the same object (not restricted to objects I guess) from the same file creates two objects linked to the same data. The data is local, but both object links to it. Obviously, moving one of them moves the other, the only way to realise that there's two objects there is to use boundbox select, lasso, ... Assigning to Jesty since I believe he did some work on that lately.";"None";"None";"Investigate" 2902;1;"Open";1;389;"Daniel Dunbar";103;"Ton Roosendaal";"2005-08-07 21:06";"1970-01-01 01:00";"2007-05-24 12:03";"Material texface option doesn't work in fastshade mode";"Create mesh with tfaces, enable material texface option, enter fastshade mode (shift-z), compare difference to render. Possibly annoying to fix...";"None";"None";"New" 2914;1;"Open";3;3487;"Campbell Barton";103;"Ton Roosendaal";"2005-08-10 10:35";"1970-01-01 01:00";"2007-05-24 12:03";"Changing material colours + shaded view,";"I noticed that changing the material color on a mesh while in shaded view dosent update the colour of the shaded mesh until you enter/exit edit mode. ";"None";"None";"New" 2936;1;"Open";3;3164;"Guillaume Lecocq";103;"Ton Roosendaal";"2005-08-15 21:52";"1970-01-01 01:00";"2007-05-24 12:03";"Object disappear in wireframe after a subsurf 6";"Hello, the bug is: - On a new scene with the cube - Add a Subsurf modifier on the cube - Do level 6 - Push Apply buttons - Pass in wireframe - Press Tab twice to enter and exit Edit Mode - Normally the mesh disappear - But if you go in Solid Mode the mesh is here I test on windows XP, 1 Go, Geforce FX 5800 with 2.37a (without modifier only subsurf and mesh convert) and actual CVS ";"None";"None";"New" 2953;1;"Open";3;2572;"Krzysiek ";4246;"Peter Schlaile";"2005-08-22 02:18";"1970-01-01 01:00";"2007-05-24 12:03";"wrong frame numbers when rendering in sequence editor";"check the attached file. it`s all explained there. I`ve checked it on linux with blender 2.37 and the testing build from 2005-08-16. both the same thing";"None";"None";"New" 3058;1;"Open";3;665;"styken ";100;"Nobody";"2005-09-14 23:33";"1970-01-01 01:00";"2010-12-07 15:29";"Underlining don't follow curve";"When underlining chars and using TextOnCurve, the underline doesn't deform as the text does but remain horizontal! //styken";"None";"None";"Investigate" 3110;1;"Open";3;3995;"Eric Forsythe";103;"Ton Roosendaal";"2005-09-27 21:17";"1970-01-01 01:00";"2007-05-24 12:03";"Wire Shade in Edit Mode";"The patch adds shading to wire frames while in Edit Mode. This makes viewing complex meshes much easier and aids in the selection of vertices, edges, and faces. More information can be found at: http://www.ericmforsythe.com/Blender/Development/WireShade/WireShade.html Someone in the forum recommended that I submit the patch soon because a release is upcoming. So, here it is for you to use if you wish to do so.";"None";"None";"Investigate" 3132;1;"Open";3;3164;"Guillaume Lecocq";103;"Ton Roosendaal";"2005-10-02 22:54";"1970-01-01 01:00";"2007-05-24 12:03";"With quicktime codec, if choose h264, blender crash";"Hello, I try in officiel release and current CVS. If I choose quicktime codec and choose h.264 compression, at end of the second frame rendered, Blender crash. In 2.37a compiled with qt6 sdk and current cvs with qt7 sdk, same think blender crash.";"None";"None";"Investigate" 3204;1;"Open";3;2990;"David Becker";103;"Ton Roosendaal";"2005-10-17 11:14";"1970-01-01 01:00";"2007-05-24 12:03";"Wireframe rendering with multiple material IDs causes errors";"A mesh object with multiple materials cannot have some materials rendered in wireframe and others as solid. Materials other than the first are no longer drawn when their wire option is checked. If the first material has its wire option checked the entire mesh will get a wireframe overlayed. This can be useful when you want to show a model for getting critique, just add a wireframe material in the first material slot and you get a wireframe overlayed over the entire model but it doesn't seem to be intentional. Optimally the bug would be fixed but an option (modifier?) for overlaying the wireframe over the object would be offered instead. Reproduction: -Create a mesh object. -Add a material to it. -Assign all faces to the second material. -Set the first material to render as wireframe (giving it another color and using ZOffset for visibility might be useful). -Render.";"Render (internal)";"None";"New" 3230;1;"Open";3;4244;"Jeremy Matthews";4290;"Joshua Leung";"2005-10-20 13:21";"1970-01-01 01:00";"2011-03-19 04:23";"Action Baking does not work with constraints";"When baking a action from an armature that has IK Constraints (and copy location/rotation), the constraints will not bake to the joints. The IPO's of the IK controller object will though, just not the joints being controlled with constraints.";"None";"None";"New" 3243;1;"Open";2;4274;"Franz Rogar";100;"Nobody";"2005-10-22 16:14";"1970-01-01 01:00";"2010-12-07 15:27";"l10n bug in "User preferences" window - Textured Fonts";"When using l10n (localization) and the scrings related to Objects (cube, isosphere...) aren't shown in that menu. It shows the info of the MO object. OS tested: Windows XP SP2 Graphic Card: ATI RADEON 9200 128DDR Blender version: From 2.37 till 2.40-alpha1 Ocurrency: Always when using l10n (translated GUI) This bug should be corrected asap because it's one of the main characteristics of Blender and is the support for i18n. Please, correct it. ";"None";"None";"New" 3290;1;"Open";3;3295;"Jonathan Hudson";100;"Nobody";"2005-10-29 17:17";"1970-01-01 01:00";"2012-10-29 12:18";"dupliverts and hooks";"I'm using Linux, cvs version (Oct 29, 2005 10:20 am EST), Geforce 6600GT Dupliverts won't work correctly for vertices that have hooks. 1. create a cube 2. extrude a face twice 3. add a hook connecting the vertices at the end of the extrusion and vertices at the front face 4. duplivert the shape to another cube The vertices of the dupliverted copies which are not connected to hooks will act normally. But the vertices of the dupliverted copies which have a hook will remain at the same location as the parent vertices. I've uploaded a blend file that shows the problem.";"None";"None";"Investigate" 3299;1;"Open";3;980;"Alexander Ewering ";139;"Martin Poirier";"2005-10-31 00:54";"1970-01-01 01:00";"2007-05-24 12:03";"Transform Behavior/UI Review";"When I set the 3d view header to ""Local"" transform, and I select a bone and press R-Z to rotate it around its LOCAL Z axis, the global Z axis is used instead. Likewise, when the header is set to ""global"", pressing R-Z-Z should switch to local Z transform, but does not. ";"Interface";"None";"Investigate" 3302;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2005-10-31 09:24";"1970-01-01 01:00";"2010-12-07 15:33";"Boolean modifier does not retain material indices";"I think the summary pretty much says it all. Material indices get lost with boolean modifiers, instead of being retained like with the regular booleans. ";"None";"None";"New" 3309;1;"Open";3;2119;"Gabriel Beloin";103;"Ton Roosendaal";"2005-10-31 15:09";"1970-01-01 01:00";"2007-05-24 12:03";"No easy way to add the first action strip in NLA";"That is kinda weird. Previously in 2.37 when you had an armature, it had an entry in the NLA so you could add a strip by puting your mouse over and doing shift-a (this should be changed for a ""add button"" btw) but since object can be included too, there is nothing in the NLA. so you can't add a strip even if you select the object, even if you select an armature you can't add the first strip! You have to open the action window, load an action and then go in nla add a strip and unload the action. Also when you do shift-a on nothing it give you a kinda mix of 2 messages ""Object has not an action"" load demo and try to add an action to the object without openning the action or IPO window ;)";"None";"None";"Investigate" 3311;1;"Open";3;4300;"Juan Carlos Rodrigo";103;"Ton Roosendaal";"2005-10-31 17:30";"1970-01-01 01:00";"2007-05-24 12:03";"Screen Name: ""Model "" bytes in B.blend.c should be ""Model""";"PabloMarmol reported this to me and I've tried to fix it. The Model Screen Name in /source/blender/src/B.blend.c: 77,111,100,101,108, 32 ""Model "" Must be: 77,111,100,101,108, 0 ""Model"" People should resolve their ""Model "" ~/.B.blend files by hand. Changing the name in the Screen Names box and saving the file. When loading the ~/.B.blend (Or any other .blend) the problem could be 'fixed' using a strip( id->name ) function in: /source/blender/blenloader/intern/readfile.c static BHead *read_libblock( ... ) Removing the the spaces automagically could make the Screen Names collide. ";"None";"None";"New" 3312;1;"Open";3;4301;"Ralf Hölzemer";103;"Ton Roosendaal";"2005-10-31 18:11";"1970-01-01 01:00";"2007-05-24 12:03";"IK chain update problem";"IK chains do not update their positions if the IK target is constrained to one of the bones in the chain AND the translation of the IK target is aborted via RMB or ESCKEY. Tested on: Ubuntu Breezy, NVIDIA Quadro XGL 980, Current CVS ( 051031 ) Windows XP SP2, "", "" See attached .blend file for instructions.";"None";"None";"New" 3318;1;"Open";3;4252;"rob dude";103;"Ton Roosendaal";"2005-11-01 05:22";"1970-01-01 01:00";"2007-05-24 12:03";"lighting update issue";"when you change the colour of a light in shaded mode, using ipo drivers, it doesnt update in the 3d view till you change frame. from levon_, posted by wavez";"None";"None";"New" 3384;1;"Open";3;2923;"Tom Musgrove";100;"Nobody";"2005-11-09 22:58";"1970-01-01 01:00";"2008-09-09 20:19";"subdived smooth gives incorrect results";"The result from a subdivide smooth at 100% gives different results compared to 2.37a and compared to doing subsurf level 1. The cube on the left is a subdivide smooth once and on the right is subsurf level 1, ";"Tools";"None";"Investigate" 3392;1;"Open";3;2119;"Gabriel Beloin";103;"Ton Roosendaal";"2005-11-12 16:20";"1970-01-01 01:00";"2007-05-24 12:03";"Snap selection to cursor doesn't work in posemode.";"Snap selection to cursor doesn't work in posemode. Simply put. Then trying to snap a bone to the cursor it doesn't work. I'm trying to snap a bone that can be moved, like a root of chain, before you ask. ;) Check demo! open and do shift-s-2.";"None";"None";"New" 3393;1;"Open";3;2119;"Gabriel Beloin";103;"Ton Roosendaal";"2005-11-12 16:27";"1970-01-01 01:00";"2007-05-24 12:03";"Renaming constraint doesn't update action channel.";"Renaming constraint doesn't update action channel. So for example I animated something and keyed some constraint's influcence. If I rename the constraint. The entry in the Action datablock don't get updated. So the influence doesn't get animated animore. Renaming it to the name in Action work, so it's really just about looking for an innexistant channel name. check the demo! change the name of one constraint for fun.";"None";"None";"New" 3408;1;"Open";3;187;"Mika Saari";100;"Nobody";"2005-11-15 17:43";"1970-01-01 01:00";"2012-10-29 12:40";"IPOs not evaluated before executing a framechanged scriptlinked script";"I read the really great BMG tutorial from the blenderart magazine and noticed that the sample code with setEuler is not working correctly in blender 2.40 alpha X but works nicely in 2.37. After debugging a bit noticed that the animation is lacking a one frame from the very start. So when the first wheel in the attached example blend is rotated along the IPO from frame 1 to 2, the python still claims that the rotation of the object (ob.rot or ob.getEuler()) is [0,0,0] even the NKEY dialog says that the object rotation is [0,0,2.34232] (or something). -Mika ";"None";"None";"Approved" 3413;1;"Open";3;2923;"Tom Musgrove";103;"Ton Roosendaal";"2005-11-17 00:25";"1970-01-01 01:00";"2007-05-24 12:03";"moving the rootbone after a mirrored extrude shows the mirrored bone unattached to the root";"This only happens on the first mirrored extrude on the root bone the left bone is attached the right bone is dashed lined. If you look at a later area where i branched both bones are attached. ";"None";"None";"New" 3433;1;"Open";3;4345;"Jerry Farm";103;"Ton Roosendaal";"2005-11-21 03:29";"1970-01-01 01:00";"2007-05-24 12:03";"ztransp renders with wrong alpha";"Also present in 2.4a2 When ZTransp is enabled for a material (on the Mirror Transp button panel), the object using the material becomes partially transparent, even if Alpha is set to 1.0 (The 'A' button on the Material button panel). The object is in effect rendered with a lower Alpha than desired. Decreasing Alpha will make the object more transparent, but an Alpha of 1.0 should be opaque, but is not. This means that a user can't set their material to be as opaque as they might desire. In my case, I wanted to use Zoffs (Z offset) to help the renderer resolve z-buffer ambiguities of details in a deep scene that requires wide camera clipping planes. But Zoffs is ignored unless ZTransp is enabled. Since my objects were not supposed to be transparent, I couldn't use Zoffs this way. :-( (It would be nice if Zoffs could be used independently of ZTransp. Though I can understand why they are linked from what I've read of their history, I'm not sure there is any reason they need to be linked, and it is a better user interface to decouple them to eliminate the current hidden link that prevents Zoffs from doing anything unless the user knows to activate ZTransp. But I digress.) A .blend file is attached (I hope!) showing a rectangle floating over three staggered cubes (2 white, one black). The rectangle has ZTransp turned on. Doing a Render will show the cubes as visible through the rectangle, even though the rectangle has Alpha=1. I've only looked at the code a little so far, and haven't found anything obvious. There is a chance this is a UI issue that prevents settings for complete opaqueness from reaching the renderer. ";"None";"None";"New" 3437;1;"Open";3;3089;"fred van der essen";103;"Ton Roosendaal";"2005-11-22 00:03";"1970-01-01 01:00";"2007-05-24 12:03";"Mirror Modifier doesn't work with particles";"The Mirrored part of a modified mesh doesn't emit particles. ";"None";"None";"New" 3441;1;"Open";3;4356;"ram matityahu";100;"Nobody";"2005-11-22 10:41";"1970-01-01 01:00";"2008-09-09 20:18";"cant add hebrew text in blender";"when add a text i select hebrew font (from culmus.sf.net), then i change my keyboard layout to hebrew and when i try to write i don't see nothing been written. btw, if you want to support laungages like hebrew and arabic you should add a option for Right To Left text...";"None";"None";"New" 3476;1;"Open";3;3993;"Fabien Devaux";146;"Brecht Van Lommel ";"2005-11-27 01:25";"1970-01-01 01:00";"2007-05-24 12:03";"Random incorrect rendering of textures";"Hard to replicate so I joined two successives renders. To reproduce it, here I must change the ""map to"" mode of the texture (I switch between emit, reverse-emit and col) and then render. Sometimes the bug takes long to happend, very weired.";"Tools";"None";"New" 3499;1;"Open";3;4380;"Levi Schooley";100;"Nobody";"2005-11-30 05:33";"1970-01-01 01:00";"2010-12-07 15:33";"Boolean modifiers only consider raw mesh data";"I have noticed that boolean modifiers only consider raw mesh data. This means that multiple boolean modifiers cannot be combined in the modifier stack, and any boolean operation between two objects that may have modifiers (such as subsurf) produces incorrect results. A sample file is attached that demonstrates that multiple boolean modifiers do not work as expected. If this is a chosen feature, then I appologize. It merely struck me as being incorrect behaviour. Thanks! Levi Schooley";"None";"None";"New" 3501;1;"Open";3;1819;"Carlos Lopez ";103;"Ton Roosendaal";"2005-11-30 13:14";"1970-01-01 01:00";"2007-05-24 12:03";"Vertexparent in a softbody Curve";"Make a curve softbody Parent an object to a curve's node. The object doesnt follow the node.";"Tools";"None";"New" 3523;1;"Open";3;3138;"Olivier Debroise";103;"Ton Roosendaal";"2005-12-02 16:42";"1970-01-01 01:00";"2007-05-24 12:03";"Ray transparency and World exposure/range";"Sorry for my poor English (a blend is joined :) : The exposure and the range channel disturb the Alpha Channel in Ray transparency mode. In the blend file you'll see that if you modify the Range Cursor or the Exposure cursor and you take a render. (The two cube are the same material with the exception of the transparency mode, of course). PS : Congratulation has all the team for your great works.";"Render (internal)";"None";"New" 3553;1;"Open";3;3059;"Olaf Arnold";100;"Nobody";"2005-12-07 15:22";"1970-01-01 01:00";"2012-10-29 12:18";"Loopcuting on deformed Mesh by Curve in Editmode";"Hello, i don´t know if it is a bug or just not yet implemented: -Add a Bezier Circle or Curve - Add a Cube - Add a MeshModifier to the Cube - Now press the button ""Enable Modifier during Editmode"" and also this little ""Enable Modifier to editing cage during ""Editmode""-button. OK everything looks OK to me, except if i try to loopcut the mesh in Editmode (+), the preview lines for loopcuting are shown on the original Mesh not on the deformed Mesh, this is a little disturbung in my humple opinion, can it be changed so that we it works on the deformed Mesh? Note: If the Original Mesh is out of 3D View, you even haven´t a chance to loopcut the deformed Mesh... I hope you´ve understand what i mean here, my english isn´t that good... I´ve attached a blendfile, to demonstrate that i mean, download it and try to loopcut in that situation... Bye, Olaf. ";"None";"None";"Postponed" 3595;1;"Open";3;3206;"Rui Campos";103;"Ton Roosendaal";"2005-12-14 02:56";"1970-01-01 01:00";"2007-05-24 12:03";"""Frames per second"" setting doesn't save in AVI or movie files.";"In Anim/Playback buttons panel, setting up the frames per second will only affect the sequence editor visually. When rendering these settings do not seem to take any effect. I tested this by importing a video capture (in jpeg sequence files) done with 10fps to the sequencer. Changed the frames per second from 25 to 10 in the Anim/Playback panel and then rendered the sequence. The end result was an avi file with 25fps, meaning it had less time then it should.";"None";"None";"New" 3629;1;"Open";3;4279;"Stephan Kassemeyer";103;"Ton Roosendaal";"2005-12-19 13:08";"1970-01-01 01:00";"2007-05-24 12:03";"static particle deflection";"static particles don't get deflected.";"None";"None";"New" 3641;1;"Open";3;3993;"Fabien Devaux";103;"Ton Roosendaal";"2005-12-20 16:45";"1970-01-01 01:00";"2007-05-24 12:03";"Deflection not working for self-generated particles";"A mesh emitting particles can't deflect them. It's probably harder to do since at begining the particles can be part of the mesh, but once they are outsite, they shouldn't be able to re-enter it. Hari' on #blendercoders suggested the use of a bias button... 16:44 < hari`> If my theory (about particle birth problems) is correct, then it should have a ""bias"" number you can adjust for how old the particle should be before it deflects off its own emitter. 16:45 < hari`> That's the same reason there are raytracing biases. 16:45 < hari`> You don't want a ray that's leaving a polygon to hit the same polygon again somehow. Don't know if it helps or if it's near the real situation.";"None";"None";"New" 3650;1;"Open";3;3993;"Fabien Devaux";103;"Ton Roosendaal";"2005-12-22 17:55";"1970-01-01 01:00";"2007-05-24 12:03";"CurveGuide properties aren't scaled when the curve is scaled";"Everything is in the title. So we can't select a whole scene to rescale it [I remember some fixes were applied to allow it for area lights]. ";"Tools";"None";"New" 3676;1;"Open";3;980;"Alexander Ewering ";103;"Ton Roosendaal";"2005-12-28 13:28";"1970-01-01 01:00";"2007-05-24 12:03";"""Delete edgeloop"" and undo glitch";"Hi, probably just a question of placing the correct undo calls: When undoing 'delete edgeloop', you need to undo twice. ";"None";"None";"Investigate" 3691;1;"Open";3;1819;"Carlos Lopez ";103;"Ton Roosendaal";"2005-12-30 12:12";"1970-01-01 01:00";"2007-05-24 12:03";"Objects selected at Info bar: select linked doesn't refresh";"1-Select some objects 2-Look at info bar: ""Ob:#total:#selected"" shows correct numbers 3-Use Shift+L>any option 4-Info bar doesn't get counting ";"Interface";"None";"New" 3696;1;"Open";3;4265;"hua wong";100;"Nobody";"2005-12-31 11:57";"1970-01-01 01:00";"2012-10-29 12:27";"Minimization and user interface problems.";"Using OS X 10.4.3 I found that windowed mode could lead blender to have a strange behavior regarding the cursor position and interaction. The cursor is inaccurate and can even refuse to interact with the interface, or interact uncorrectly. This happen often, when blender is minimized and brought back. What I did : On startup, I put blender in full screen. Blender works well in full screen, no problem. Until... Then dock it. When I maximize it back, user interface doesn't respond to cursor action, or interact weirdly. And I found no way to return it to it normal state. Full screen doesn't resolve the issue. Thanks";"None";"None";"New" 3717;1;"Open";3;4230;"Stefan Jurisch";103;"Ton Roosendaal";"2006-01-04 19:28";"1970-01-01 01:00";"2007-05-24 12:03";"animation/modelling: error in updating objects when interrupting transform operations";"I think, it's widely the animation system; and I put it here because I did not find a category named ""modelling"". To repdroduce the fault I have attached a scene. but I think it is reproducable with lot of others. When going into grab-mode for the activated Empty (in the scene), moving a little bit around, and then aborting the action with RMB or ESC-key, the objects will be placed on their old position, but the child-lattice (used to deform the wheel) now deforms the whole wheel in a ugly manner, because it (and other childed objects) do not go back correctly into their old locations. when i grab again and directly abort again without moving around, the wheel gets his old (and correct) shape again. shouldn't there be an additional object-update/redraw after aborting a translation? this problem comes up in other cases, too, eg. with constraints - I think, ever when there are objects with various dependencies. I'm not so good explaining the problem, so please test it with the attached blend-file (ZIP-arc). Just press Gkey, move a few centimeters up and then press ESC. FastEddy";"None";"None";"Approved" 3726;1;"Open";3;2600;"Carl Olsson";103;"Ton Roosendaal";"2006-01-07 13:40";"1970-01-01 01:00";"2007-05-24 12:03";"Children of looped TrackTo constrained parents do not update";"Windows, 2.4 and recent Orange build. When you have and object with a TrackTo contraint on another object which has a TrackTo constraint on the first object (looped TrackTo constraints), then neither object's children get update (neither position nor rotation) when the parent objects move. This happen both in the 3D view and during render. Entering then leaving edit mode, or grabing and cancelling, on a child, will update its position, but it will then stay in that position. This works correctly in 2.37a.";"None";"None";"Approved" 3744;1;"Open";3;2630;"Eckhard M. Jäger";2923;"Tom Musgrove";"2006-01-13 01:31";"1970-01-01 01:00";"2007-05-24 12:03";"Outliner did not use ""Select with"" option of ""User Preferences"" ""View & Controls""";"Selecting object in the outliner works always with LMB wether or not you setup ""Select with"" option of ""User Preferences"" ""View & Controls"" RMB or LMB. This is a little bit confusing because the Oops View care about this and if you setup ""Select With"" as RMB then Oops works with RMB and the Outliner works with LMB. BTW it would be cool if the Outliner/Oops could support short ""X"" for deleting objects, which is very usefull to delete objects from a complex scene without switching from Outliner/Oops to 3D View and back to delete objects. Or just deleting object that has just an pivot and nothing more.";"None";"None";"New" 3753;1;"Open";3;4522;"Kevin Foley";100;"Nobody";"2006-01-16 00:30";"1970-01-01 01:00";"2012-10-29 12:27";"Blender locks up when switching to dual monitors";"I have Windows XP SP2, an ATI Radeon 7000 64mb, and the ATI control panel installed. I also have a dual monitor setup, with the second monitor being a TV from the Radeon's S-Video out. If I have the second monitor disabled when I start Blender, and then go into the Display Properties window (right click desktop, goto Properties) and the settings tab, and choose ""extend my Windows desktop onto this monitor"" in the second monitor settings screen, the second monitor starts working fine but Blender hangs (the window goes black and it says ""Not Responding"", and never starts responding again). I don't know whether this problem is isolated to my particular card or setup or it occurs in general when activating a second monitor. It's not a big issue because it's easy to avoid, but it's annoying.";"None";"None";"New" 3755;1;"Open";3;4523;"Jonathan Cottrill";2923;"Tom Musgrove";"2006-01-16 03:45";"1970-01-01 01:00";"2007-05-24 12:03";"Only single bone created when adding armature";"My computer: AMD Athlon XP 2400+ 768MB RAM VisionTek ATI Radeon 9250 Windows XP Professional, SP 2, all updates Blender 2.40 The problem: When I add an armature (SPACE, Add, Armature), a single bone appears, pointing straight up. Note that this is straight up relative to the screen, and regardless of the current viewpoint. This happens with 100% consistency. The bone is ""locked,"" in that it does not move in any way when I move the mouse. Pressing GKEY and moving the mouse allows the end of the bone to be moved. Left-clicking does not produce a new bone in the chain, regardless of whether I have previously moved the end of the bone using GKEY. Using versions 2.37a and 2.34, I have not been able to reproduce the problem. It seems specific to 2.40. On 2.37a and 2.34, Blender operates exactly as described in the tutorial I have been using, in that the initial bone can be swung around until the first left click, then a new bone is added and joined to the first, etc. The problem occurs in all files, even the default new file when Blender starts. Therefore, I have not attached a file. I can try to offer any more information you need. Thanks for your time.";"None";"None";"New" 3762;1;"Open";3;3138;"Olivier Debroise";103;"Ton Roosendaal";"2006-01-17 11:36";"1970-01-01 01:00";"2007-05-24 12:03";"Problem with the size of lamps.";"Windows XP SP2 GeForce 4 Blender 2.40 I have resized a lamp and the lamp do not point to the good direction when you take a render. I have posted a blend file for more comprehension. In this file, the light cone is into the plan but in the render of the first camera it is not the 'case'. It is more visible with the second camera. (I note a refresh problem when you change the track settings in the 'anim settings' panel).";"None";"None";"New" 3765;1;"Open";3;3138;"Olivier Debroise";103;"Ton Roosendaal";"2006-01-18 14:05";"1970-01-01 01:00";"2007-05-24 12:03";"DupliVerts and Build";"Blender 2.40 Windows XP SP2 GeForce 4 Not really a bug :-) Take two cube, one is parented to the other. The parent have a build modifier and the dupliverts options activated. The cube become visible, but the dupliverts object are present all the time. Sorry for my bad English, this is comprehensible ? ";"None";"None";"New" 3770;1;"Open";3;3138;"Olivier Debroise";103;"Ton Roosendaal";"2006-01-19 13:38";"1970-01-01 01:00";"2007-05-24 12:03";"Texture Space rotation";"With the menu 'T key' you can not use the rotation... If you type 'T key' then 'R key' you go in rotate mode, but... I know it is not a bug :-)";"None";"None";"Approved" 3780;1;"Open";1;3191;"Charles Wardlaw";100;"Nobody";"2006-01-20 17:20";"1970-01-01 01:00";"2012-10-29 12:27";"OSX doesn't overwrite in-scene memory with Finder-opened files";"I'm putting this down as a bug, not a feature request, because of the way all other programs work in OSX. This is expected behaviour. If you double-click a blend file in the Finder and Blender is not running, it launches and opens the file. However, if Blender is already running and you double-click a blend file, Blender becomes the active program but does not open the file you clicked. From the standpoint of most Mac users, I think what should happen is that the in-memory file gets overwritten with what's been opened in the Finder, or another window with the new scene data opens. Since having multiple windows breaks Blender's own single-window convention, the former solution would seem best. Thanks.";"None";"None";"Investigate" 3820;1;"Open";3;3212;"Randall Rickert";100;"Nobody";"2006-01-27 08:18";"1970-01-01 01:00";"2012-10-29 12:18";"Boolean modifiers only use first material and throw away the rest";"See the attached file. The selected cube has two materials (red and green) and a boolean union with a blue cube. Tab into edit mode or disable the boolean modifier and you will see the blue and green materials. These two materials are discarded by the boolean modifier. The docs say this is supposed to work. Blender 2.41 Mac G5, OS X 10.4.4";"None";"None";"New" 3821;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-01-27 08:34";"1970-01-01 01:00";"2007-05-24 12:03";"subsurf modifier can't be chained before boolean modifier, only after";"See this file. One red cube has a boolan modifier after a subsurf modifier. The other has the boolean before the subsurf modifier. On the one that has a boolean after the subsurf, the boolean is being performed with the original unmodified mesh, not with the result of the subsurf. Blender 2.41 Mac G5, OS X 10.4.4";"None";"None";"New" 3844;1;"Open";3;4565;"tom hannen";103;"Ton Roosendaal";"2006-01-30 22:08";"1970-01-01 01:00";"2007-05-24 12:03";"International fonts display incorrectly inside text objects";"Cannot get Persian text to display correctly. The truetype font used displayed correctly in TextEdit, but not in Blender. I used Blender 2.41 for Python 2.3 on Mac OS X 10.4 Using TextEdit: I used a font called ""nesf2.ttf"". I saved the text into a text (txt) file using Unicode UTF-8. Then in Blender - I added a new text object, deleted the default text that appears, and went to the editing button (F9), on the Font tab, I clicked ""Load"", and loaded the ""nesf2.ttf"" font. Then clicked on the ""Insert Text"" button, and the text is displayed, but the characters are in reverse order (because persian text reads right-to-left). Persian text is all ""joined up"", which means that simply reversing the character order does not solve the problem because the letters connect together differently. (To all involved: Blender is fantastic). ";"None";"None";"Investigate" 3874;1;"Open";3;4584;"Bartłomiej Syguła";103;"Ton Roosendaal";"2006-02-05 00:16";"1970-01-01 01:00";"2007-05-24 12:03";"Oops...";"When using static particles on object that has a texture the color of the particles (hair/fur/etc) matches the texture on the object. But only if the mapping is one of: Glob, Orco or Win. If the mapping is UV - all particles on the object have the 'average' color. To reproduce this: o create an object, o create a matherial with image texture, UV-map it on the object, o add statis particles to the object, o check the 'Mesch' so the object will be rendered too o render. Particles do not match texture. Example is attached: four planes differ ony in the way the mapping is done. The one in bottom-right corner has UV-mapped cross. The particles ignore the texture on that plane. I have tried this in 2.40 and 2.41 My system: Linux 2.6.15-gentoo #3 PREEMPT i686 AMD Athlon(tm) XP 2400+ Graphic card: ATI Technologies Inc RV280 [Radeon 9200 SE] (rev 01) X.org version: 6.8.2 ";"None";"None";"New" 3878;1;"Open";3;3187;"David Schnur";103;"Ton Roosendaal";"2006-02-06 19:38";"1970-01-01 01:00";"2007-05-24 12:03";"Render window resize/restore artefacts";"I have noticed this for several version of Blender. It is a very minor cosmetic thing, but still a bug. Description: When the render window is minimized or resized, it will often be filled with random graphical artefacts when the next line of rendering completes. Usually, the already-rendered section of the image is no longer shown; only newly-completed lines will be displayed. Once the render is completed, the window refreshes and displays the full image with no artefacts. On Windows in 2.41 I notice this only when resizing the window (screenshot attached showing the result of maximizing the render window). My Mac laptop had the same thing happen on both resize and minimize/restore, although that still has version 2.39 on it. This has also happened on another windows computer with a different graphics card. Laptop #1 (screenshot) - Pentium M 1.6 gHz - Intel Extreme Graphics (Intel 82852/82855 GM/GME) - Windows XP Professional SP 2 Laptop #2 - Power PC G3 700mHz - ATI Rage Mobility 16MB - Mac OSX 10.4 Desktop - Pentium 4 2gHz - Sapphire Radeon 9600 256MB - Windows XP Professional SP 2 I have seen this problem on a number of other computers too whose specs I don't have with me. I believe it is something more related to how Blender refreshes/redraws the render window than any specific hardware profile. ";"None";"None";"New" 3893;1;"Open";3;3002;"Joseph Eagar";103;"Ton Roosendaal";"2006-02-08 18:24";"1970-01-01 01:00";"2007-05-24 12:03";"unexpeted lines in vector motion blur";"In the new vector motion blur node, for some reason strange lines are being drawn. It works most of the time, but something about the attached .blend is messing it up. I can't track down the problem to a specific thing, especially since I'm limited by the slowness and buggyness of this idiotic ATI card. open the .blend file and render the scene. joeedh";"None";"None";"New" 3895;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-02-09 04:25";"1970-01-01 01:00";"2007-05-24 12:03";"Crash in file browse window";"Since 2.41, Blender has crashed my whole system a few times now when I'm in the file browser window. The mouse stops responding or becomes very jerky, then sometimes I can get out of the file browser by hitting Esc and everything is normal again, but sometimes the system crashes. It doesn't seem to happen in any other type of Blender window. Tested in Blender 2.41 and current CVS OS X 10.4.4 dual G5 Unresolved kernel trap(cpu 1): 0x300 - Data access DAR=0x000000005C27A34E PC=0x00000000008E5E14 Latest crash info for cpu 1: Exception state (sv=0x472E6780) PC=0x008E5E14; MSR=0x00009030; DAR=0x5C27A34E; DSISR=0x40000000; LR=0x008E682C; R1=0x2C7E3740; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x008E76C4 0x008E682C 0x008F1528 0x008F0DC0 0x008EC744 0x008EC9D0 0x008CECDC 0x002E7F00 0x002E9DCC 0x0008C4B0 0x000291C0 0x000233AC 0x000AC02C 0x006B6B6B backtrace terminated - frame not mapped or invalid: 0xBFFFE900 Kernel loadable modules in backtrace (with dependencies): com.apple.driver.AppleTopazAudio(2.5.6b5)@0x6ac000 dependency: com.apple.driver.AppleOnboardAudio(2.5.6b5)@0x66a000 dependency: com.apple.iokit.IOAudioFamily(1.5.5b2)@0x59f000 com.apple.GeForce(4.1.8)@0x8c3000 dependency: com.apple.iokit.IOPCIFamily(1.7)@0x458000 dependency: com.apple.iokit.IOGraphicsFamily(1.4.1)@0x6c0000 dependency: com.apple.iokit.IONDRVSupport(1.4.1)@0x6e4000 dependency: com.apple.NVDAResman(4.1.8)@0x6fa000 Proceeding back via exception chain: Exception state (sv=0x472E6780) previously dumped as ""Latest"" state. skipping... Exception state (sv=0x46C56C80) PC=0x9000B208; MSR=0x0000F030; DAR=0x06DB7000; DSISR=0x42000000; LR=0x9000B15C; R1=0xBFFFE900; XCP=0x00000030 (0xC00 - System call) Kernel version: Darwin Kernel Version 8.4.0: Tue Jan 3 18:22:10 PST 2006; root:xnu-792.6.56.obj~1/RELEASE_PPC panic(cpu 1 caller 0xFFFF0003): 0x300 - Data access Latest stack backtrace for cpu 1: Backtrace: 0x00095718 0x00095C30 0x0002683C 0x000A8384 0x000ABD00 Proceeding back via exception chain: Exception state (sv=0x472E6780) PC=0x008E5E14; MSR=0x00009030; DAR=0x5C27A34E; DSISR=0x40000000; LR=0x008E682C; R1=0x2C7E3740; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x008E76C4 0x008E682C 0x008F1528 0x008F0DC0 0x008EC744 0x008EC9D0 0x008CECDC 0x002E7F00 0x002E9DCC 0x0008C4B0 0x000291C0 0x000233AC 0x000AC02C 0x006B6B6B backtrace terminated - frame not mapped or invalid: 0xBFFFE900 Kernel loadable modules in backtrace (with dependencies): com.apple.driver.AppleTopazAudio(2.5.6b5)@0x6ac000 dependencpModel: PowerMac7,3, BootROM 5.1.8f7, 2 processors, PowerPC G5 (3.0), 2.5 GHz, 2 GB Graphics: NVIDIA GeForce 6800 GT DDL, GeForce 6800 GT, AGP, 256 MB Memory Module: DIMM0/J11, 512 MB, DDR SDRAM, PC3200U-30330 Memory Module: DIMM1/J12, 512 MB, DDR SDRAM, PC3200U-30330 Memory Module: DIMM2/J13, 512 MB, DDR SDRAM, PC3200U-25330 Memory Module: DIMM3/J14, 512 MB, DDR SDRAM, PC3200U-25330 Network Service: Built-in Ethernet, Ethernet, en0 PCI Card: usb, usb, SLOT-4 PCI Card: usb, usb, SLOT-4 PCI Card: pci1735,e0, , SLOT-4 Serial ATA Device: ST3160023AS, 149.05 GB Serial ATA Device: ST3120026AS, 111.79 GB Parallel ATA Device: SONY DVD RW DW-U21A, USB Device: Hub in Apple Pro Keyboard, Mitsumi Electric, Up to 12 Mb/sec, 500 mA USB Device: Microsoft Wireless Optical Mouse® 1.0A, Microsoft, Up to 1.5 Mb/sec, 100 mA USB Device: Apple Pro Keyboard, Mitsumi Electric, Up to 12 Mb/sec, 250 mA USB Device: Hub, , Up to 480 Mb/sec, 500 mA USB Device: USB2223, SMSC, Up to 480 Mb/sec, 500 mA USB Device: Hub, , Up to 480 Mb/sec, 500 mA USB Device: Camera, , Up to 12 Mb/sec, 500 mA USB Device: USB Device, Tablet, Up to 12 Mb/sec, 500 mA ";"None";"None";"New" 3898;1;"Open";3;709;"Ken Hughes";3487;"Campbell Barton";"2006-02-09 18:42";"2006-03-02 17:29";"2008-04-07 08:50";"Scene.unlink() not updating datablock user count correctly";"Don't have time to look at this myself right now, so filing here: When Blender.Scene.Unlink() is called with an object, the object is deleted but the datablock linked to it is ignored. It should be decremented instead.";"None";"None";"Investigate" 3905;1;"Open";3;322;"Doug Bischoff ";103;"Ton Roosendaal";"2006-02-11 03:14";"1970-01-01 01:00";"2007-05-24 12:03";"Ztransp materials show through Env materials";"Setting an Env material in front of an object should allow the background to show through. If you set a Ztransp material behind it, though, that Ztransp material shows up through the Env material. Blend attached.";"None";"None";"Investigate" 3923;1;"Open";3;1819;"Carlos Lopez ";103;"Ton Roosendaal";"2006-02-15 12:57";"1970-01-01 01:00";"2007-05-24 12:03";"Env materials ignore alpha channels";"Make a alpha-ed material for an object (a tree, f.i.). Put a plane behind of the tree, and give a material to the plane. Set ENV on. Render, then the tree is visible. Even its alpha channel. Is it the same problem that the old alpha behind halos?";"Render (internal)";"None";"New" 3929;1;"Open";3;322;"Doug Bischoff ";103;"Ton Roosendaal";"2006-02-17 04:06";"1970-01-01 01:00";"2007-05-24 12:03";"Stucci Preview Not Rendering";"In the Texture Buttons' ""Preview"" pane, the Stucci texture no longer generates a preview. All other Texture Types work fine. Stucci continues to render correctly and show correctly in the Materials preview pane, but not in the Texture preview pane. Blend attached works in 2.41, not in CVS as of 15 Feb 06. -Doug ";"None";"None";"New" 3937;1;"Open";3;4280;"martin vestergaard";103;"Ton Roosendaal";"2006-02-18 22:17";"1970-01-01 01:00";"2007-05-24 12:03";"Render viewport";"Problem when pressing the ""Render viewport"" button, far right in the header of a viewport (not the F12!). The render then mixes with the prevous viewport-renders. Example: Make a box, pan so the box will be out left in the viewport. Press the render-viewport button of the specific viewport. Now pan so that the box is in the right side of the viewport, and press the render-viewport button again. The output will be the box in the left from before, and the box in the right. Also, when choosing ""DispView"" in the scene-menu (to display render output in the viewport), and then again pressing the render-viewport button severel times after changing the view, the window displays the menus in Blender, AND the viewport-rendered object. These errors sometimes seem to happen when splitting up the viewports in some different ways. It's tested on following systems: Ubuntu Breezy i686 Geforce 420 GO 32mb NVIDIA Driver version: 1.0-7667 Blender 2.41, python 2.4 -Bug as written above Pclinuxos radeon 9800 Blender 2.41, python 2.3 -Bug: Black screen when pressing render-viewport button MS Windows Home SP2 32bit ATi X800 Blender 2.31 -Bug: No bug, render-viewport button works perfect. Sincerely, Martin V";"None";"None";"New" 3971;1;"Open";3;3002;"Joseph Eagar";103;"Ton Roosendaal";"2006-02-28 09:33";"1970-01-01 01:00";"2007-05-24 12:03";"glBitmap replaced with FTGL (was: Significant Speed Increase on ATIs with File Selector)";"I have found a source of major slowdown on some ATI cards. Apparently some ATI's do *not* do bitmapped fonts very well. Certainly my laptop ATI radeon mobility xpress goes really slow. To test this, I opened the Node Editor which uses all BIF_DrawString calls. I edited nodes with the ""International Fonts"" user option both on and off, and turning international fonts on provided a major speedup. Short intro to those uninformed, Blender has 2 font rendering libraries, one that uses the standard OpenGL bitmapped-font methed, and bFTGL, a modified version of FTGL that uses textured non-bitmapped fonts. The first is used by default (and is used always by the file selector), while the second is only on is the ""International Fonts"" option is set. Anyway, this fix isn't exactly obvious, especially as the file selector doesn't use it, causing major headaches. I've attached a patch that modifies the file selector code to use all BIF_ string calls, along with modifying BIF_RasterPos to use glRasterPos2f if bFTGL is turned off. With this code, the file selector actually works if ""International fonts"" are on! Usually the selector lags about a half-second or more, but after changing all the text draws to BIF_DrawString (and turning ""International Fonts"" on, of course) it has no lag at all. ";"None";"None";"New" 3998;1;"Open";3;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2006-03-11 19:31";"1970-01-01 01:00";"2007-05-24 12:03";"menu in node editor does not comply wuth user setting ""Plain menu"" toggle";"The menu of the node editor (when pressing space) does not act with the user setting ""Plain menu"" like the menu in the ""3D View"" does. The menu in the node editor is always a plain menu wether the toggle is on or off. I personal did not like the plain menus.";"None";"None";"New" 4013;1;"Open";3;4681;"Sangwoo Hong";146;"Brecht Van Lommel ";"2006-03-17 19:34";"1970-01-01 01:00";"2007-05-24 12:03";"Stretch IK should only scale the stretching bone in one axis instead of all three axes.";"When Stretch IK is turned on, the stretching bone seems to scale in all three local axes instead of just the one local axis pointing at the IK target. This causes bound geometry to grow in size instead of just the one end streching to reach the IK target. This is easy enough to verify. Setup an IK chain with two bones and an empty. Bind a tube to it. Set the Stretch value for both bones to 0.5. and move the IK target out of reach. The further away from the original position the IK target gets, the bigger the bones and the bound tube grows. Probably not what's intended. I understand this might be a ""feature"". I suspect this because this is how bones seem to behave in edit mode. If this is so then the ability to turn off scaling based on axes should be provided with the default set to scale only in the direction the bone is pointing. It should also be possible to set min/max limits so it's not possible to make the bones shrink to nothing or grow to infinity. Also, it would be nice if the Stretch values of bones in a chain could be automatically filled out when Stretch IK is turned on. Since most applications of this technique would have all the bones in the chain stretch uniformly, it would be nice if turning Stretch IK on a chain with many bones, a tail comes to mind, would set the stretch value evenly to all the bones in the chain. Then, one can go in and tweak the values if the need be.";"None";"None";"New" 4027;1;"Open";3;2657;"Germán Alonso Tamayo";103;"Ton Roosendaal";"2006-03-22 13:32";"1970-01-01 01:00";"2007-05-24 12:03";"Library.Load losts the hieracity when importing ojects and children at the same time";"When using the API to load objects and it's childs from other blend file the hieracity is lost and then there's no parents or children. This happens even loading without update and after load all do the update. Library.Load(object,""Object"",0) Library.Update() This happens with 2.41 and also with a testing build from the builds forum, and under windows, other configs haven't been tested yet. In the file you'll find a library with a cone, rectangle and Suzanne (the parent), the test blend file with the script loaded and ready to use, and the scritp itself. ";"Python";"None";"New" 4041;1;"Open";3;4704;"John Nyquist";100;"Nobody";"2006-03-28 09:03";"1970-01-01 01:00";"2007-05-24 12:03";"Photoshop Transparency support";"In Photoshop, create a file with a transparent layer. Set the color to Red 255. Draw with a brush with 0% hardness. Load the file as a texture image map in Blender. Make sure the Use Alpha is clicked. Make the material of the object Red 255. Render with Blender's internal renderer. Note how you can see where the red image map does not blend well with the underlying red. It looks like instead of blending to transparency it is blending to white. (Side note: Yafray renders this fine) Go back to Photoshop and save the previous file as a 24-Bit PNG with Transparency. In Blender, replace the PSD with the PNG. Render Note how the image map blends perfectly. I remember Macromedia Director having a similiar problem many years ago. I'm sorry I don't have more details on how they fixed, but apparently it is common enough for Blender to have the same issue. I am using: Blender 2.41 Photoshop CS2 Windows XP I can reproduce this every time. Attached is a file with a simple .BLEND, .PSD, and .PNG which illustrates the above two scenarios. ";"None";"None";"New" 4045;1;"Open";3;4681;"Sangwoo Hong";103;"Ton Roosendaal";"2006-03-29 09:50";"1970-01-01 01:00";"2007-05-24 12:03";"3D View and Camera "Clip End" should not have programmed limits.";"I often have to work on extremely large environments to scale. Not being able to see past 1000 units can become very frustrating especially when the limit seems like an artificial barrier that's been programmed in. ""Clip End"" should default to a resonable value but the artificial limit should be eliminated.";"None";"None";"New" 4131;1;"Open";3;4754;"Allan Liddle";103;"Ton Roosendaal";"2006-04-20 14:24";"1970-01-01 01:00";"2007-05-24 12:03";"Static particles (hair) do not portray the colour of a UV map";"Blender does not allow the colours of a UV-image to portray as the colour on static particles (hair), whether loaded as a UV-map or a texture image. This is a definite requirement if one wants to define the colours of a furry animal, e.g. a tiger's stripes, where one cannot depend on ""random"" textures. File and textures attached. If necessary, see discussion thread: http://blenderartists.org/forum/showthread.php?t=64853";"None";"None";"New" 4142;1;"Open";3;4736;"Joun tongue";103;"Ton Roosendaal";"2006-04-23 18:53";"1970-01-01 01:00";"2007-05-24 12:03";"Clear Track to constraint with ALT + T does not works.";"Clear Track to constraint with ALT + T does not works. But you need to go into ""F7"" menu (anim) to see all constraints you have. I believe you have to select the objects having a track to constraint, thought... check by yourself! The menu name should be ""constraints"" thats it... ";"None";"None";"New" 4152;1;"Open";3;3093;"Anders Gudmundson";4246;"Peter Schlaile";"2006-04-28 21:38";"1970-01-01 01:00";"2008-04-07 08:50";"Set sequence start/end frame to current frame.";"Handy shortcuts when editing in the sequence editor. Keys: s set sequence start frame e set sequence end frame Keys (fast edit): shift + s set sequence start frame and remove gaps between clips shift + e set sequence end frame and remove gaps between clips";"None";"None";"New" 4168;1;"Open";3;3818;"Alessandro Pirovano";103;"Ton Roosendaal";"2006-05-02 18:57";"1970-01-01 01:00";"2007-05-24 12:03";"Potato mode: issue with alpha in png images";"Hi I've an issue with RGBA png real time textures in potato mode. Simply a plane with an UV-mapped png image: I'd like to see through the image parts with full transparency. It works correctly if the plane is selected: I can see thru transparent portions of the image the object behind. When the plane is not selected, the UVmapped image is shown but thru transparent portions of image I see directly the backdrop, not the objects behind. Win XP Blender 2.41 AND CVS (jms daily build) ";"None";"None";"New" 4185;1;"Open";3;3191;"Charles Wardlaw";146;"Brecht Van Lommel ";"2006-05-11 07:21";"1970-01-01 01:00";"2007-05-24 12:03";"Copy Rotation constraint does not work stacked";"I've noticed that if you apply copy rotation (local) to one bone (Say, Cog1 is copying the rotation of Shoulder) it works fine, but if you add another bone (say, Cog2) and have it copy the rotation of the bone previously constrained (Cog1) in local mode, nothing happens. Attached is a blend to better explain this. The problem is that the local rotation copy is, I think, working on unconstrained bone rotations, and it needs to take the constrained bone rotation into account. I'm guessing this patch would take a redo of the way constraints are applied, as it would need to know about dependencies and would also need to know to fail when cyclical dependencies come up. Anyway, thanks in advance.";"None";"None";"New" 4186;1;"Open";3;4457;"Tobias Regenbrecht";103;"Ton Roosendaal";"2006-05-11 11:09";"1970-01-01 01:00";"2007-05-24 12:03";"UV mapped textures do not affect the colour of particle strands (hair)";"If you use UV mapped textures you can't color hair with it. Version: all OS: Windows and Linux 3D Card: several Nvidia cards Steps to reproduce the problem: 1) UV map a cube. 2) Use the UV coordinates as Map Input for an image texture. 3) Switch on static particles and use the settings for hair. The strands are strangely coloured, it looks as if the texture is mapped to every single strand. If you use Orco coordinates (or any other Map Input setting), the hair is colored as expected. It works neither with the traditional material system, nor with material nodes. I've already posted this problems on blenderartists and in the german blender forum, but the problem could not be solved by anyone, so it seems to be a real bug.";"None";"None";"New" 4206;1;"Open";3;4321;"Ed Halley";4238;"Geoffrey Bantle";"2006-05-16 00:35";"1970-01-01 01:00";"2007-05-24 12:03";"Creating new faces (extrude, knife) doesn't retreive Vertex Color or UV information from edges.";" The Knife tool and the Extrude tool create new vertices related to existing ones. These new vertices need to have similar attributes, and this is not always the case. * vertex color paints * vertex weight paints (for each vertex group) * vertex uv coordinates Some Knife cases are giving reasonable vertex color paints, but not Extrude, and so on. Cloned UV coords and paints are not 100% ideal but far better than nothing. As discussed on IRC. Assigned by request. ";"None";"None";"New" 4248;1;"Open";3;4279;"Stephan Kassemeyer";146;"Brecht Van Lommel ";"2006-05-29 20:29";"1970-01-01 01:00";"2007-05-24 12:03";"UV live unwrap transform update";"It would be useful if the live unwrap transform updated after changing the coordinates of a pinned vertex via the n-key properties.";"None";"None";"New" 4252;1;"Open";3;3487;"Campbell Barton";4238;"Geoffrey Bantle";"2006-05-30 00:25";"1970-01-01 01:00";"2007-05-24 12:03";"Knife Exact wrong in perspective mode";"Its simple to test this, take the default cube in a perspective view and cut it with ""Knife Exact"" - the cut line does not lie exactly on the line themouse made.";"None";"None";"New" 4290;1;"Open";3;4301;"Ralf Hölzemer";146;"Brecht Van Lommel ";"2006-06-07 21:22";"1970-01-01 01:00";"2007-05-24 12:03";"wrong UV's with cyclic extruded curves";"UV coordinates of extruded curves get messed up if the curve is made cyclic. Steps to reproduce: -Open attached blendfile -render -make the curve cyclic (CKEY) -render There is a repeated pattern in the area of one of the connected endpoints. Unrelated to this, but still an issue to note here is that if i convert this extruded curve to a mesh, the generated UV's are dismissed and the resulting mesh has standard 1/1 uv coordinates. I would have expected the original coordinates from the curve extrude on the new generated mesh. Steps to reproduce: -change to Object Mode -convert the extruded curve to mesh(ALT+c) -render";"None";"None";"New" 4300;1;"Open";3;4836;"Ondrej Novak";103;"Ton Roosendaal";"2006-06-09 09:01";"1970-01-01 01:00";"2007-05-24 12:03";"Direct Entry of Loc/Rot/Size doesn't get autokeyed";"When using the Autokey function, numerically entered changes don't get autokeyed. To reproduce: 1. Open Timeline pane 2. Select Autokey 3. Insert LocRotSize keyframe for the default cube with I key 4. Go to a different frame (ex. 10) 5. Enter a different Loc/Rot/Size in the ""Transform properties"" window 6. Go to the first frame. 7. Alt+A In the submitted .blend file, there are the first four steps done. ";"None";"None";"New" 4319;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-06-13 02:22";"1970-01-01 01:00";"2007-05-24 12:03";"Library link with relative path isn't found during render from command line";"This zip has two files. Unzip them to the same directory. Open link_destination.blend. It has a lib link with a relative path pointing to //link_source.blend. The subsurf cube you see is linked from link_source.blend. It looks and renders fine. Now render link_destination.blend from the command line. The cube does not appear in the rendered image, and I get this error message: read_libraries: lib //link_source.blend ERROR: can't find lib /link_source.blend LIB ERROR: can't find OBCube LIB ERROR: base removed Blender CVS (pre-2.42) Mac OS X 10.4.6 nVidia GeForce 6800 GT";"None";"None";"New" 4336;1;"Open";3;1280;"malefico ";103;"Ton Roosendaal";"2006-06-15 00:40";"1970-01-01 01:00";"2007-05-24 12:03";"Bones get values for loc keyframes after being stopped by floor constraint";"When using floor constraint, let say for constraining Z axis traslation of a foot, the bone keeps getting different Z values under the floor, once the floor bone has been passed. This lead to unwanted residual animation of character feet for instance, lag in rising feet, displacement in other coords, etc. The expected (from me) behaviour was to the bone completely ""freeze"" at the Z loc of the floor bone. Best regards malefico. ";"None";"None";"New" 4375;1;"Open";3;3212;"Randall Rickert";139;"Martin Poirier";"2006-06-20 06:35";"1970-01-01 01:00";"2007-05-24 12:03";"Object with path constraint transforms in wrong space";"Open this file. The empty constrained to the path is already selected. Try grabbing and rotating the object. This should happen in view space, but instead it happens in the local space of the path object. Now try grabbing and rotating while constrained to global x/y/z. Instead of global space, this also happens in the path object's local space. Now try scaling constrained to global x/y/z. This is the only one that works as expected. Now turn down the influence of the path constraint. Notice that the constrained object returns to default orientation. Its orientation shouldn't change, since the constraint has ""CurveFollow"" turned off. Blender 2.41 and current CVS Mac OS X 10.4.6g ";"None";"None";"New" 4390;1;"Open";3;4917;"david gellhardt";103;"Ton Roosendaal";"2006-06-20 21:52";"1970-01-01 01:00";"2007-05-24 12:03";"no ""menu"" popup in node-editor when you hold down L/R-mousebotton ";"no ""menu"" pops up when u hold down l/R-mouse button in the node-editor. You have to push spacebar to reach the ""onscreen-menu"" (unlike the 3d-view where you can hold down L/R-mousebutton to get acess to the ""onscreen-menu"") Maybe its not a real bug but due to the interface-logic in the 3d-view it is.. windows xp wildcat vp880 blender 2.42rc1";"None";"None";"New" 4399;1;"Open";3;3440;"Barbie LeVile";103;"Ton Roosendaal";"2006-06-21 20:15";"1970-01-01 01:00";"2008-05-09 21:08";"Usability: Save Image dialog";"The save image dialog should enhanced for better usability: a) When saving over an existing file, a popup opens telling the file exists. If this popup is cancel the whole dialog closes and the save image dialog has to be called up again from the menu. Better would be to keep the dialog open since in many cases the user just wants to change the name to not conflict with an existing one. --> speed up of workflow, no loss of functionality, and less user frustration. b) The dialog should have the filename textentry field in focus when opend so the can start type the desired filename right away and don't has to selecte the textentry field first. --> speed up of workflow with no loss of functionality. ";"None";"None";"New" 4423;1;"Open";3;4544;"Kai Kostack";103;"Ton Roosendaal";"2006-06-23 12:21";"1970-01-01 01:00";"2007-05-24 12:03";"Halo texture channels: only 1 supported, button to enable/disable it doesn't work";"well, i'm not sure whether it's a bug or intended... displacement maps doesn't work right on particle emitter objects. although the emitter will be deformed correctly, the particles were emitted from the undisplaced surface. i'm aware of the fact, that it can't be seen in the interactive display, just when rendering. another issue is, that the displacements doesn't change the vertices of an object (just a mesh, no particle emitter), if the material has been set to halo. btw, an existing but disabled texture for halos in the texture tab of the shading menu, isn't in fact disabled. ";"None";"None";"New" 4432;1;"Open";3;4917;"david gellhardt";100;"Nobody";"2006-06-24 00:03";"1970-01-01 01:00";"2012-10-29 12:18";"dupliverts with modifiers does not work correct";"create plane, multisubdivide it, select build modifier, parent with any object, switch on dupliverts: No matter the frame position of the build effect, all dupliverted objects are drawn... winXp 3dlabs vp800 ...-2.42rc1";"None";"None";"New" 4437;1;"Open";3;2600;"Carl Olsson";103;"Ton Roosendaal";"2006-06-24 02:13";"1970-01-01 01:00";"2007-05-24 12:03";"Bump mapping with procedurals renders incorrectly with global and orco mappings";"For: 2.42 RC1 Bump mapping with procedurals (clouds, marble, stucci, musgrave, voronoi, distorted noise) renders incorrectly with global and orco mappings. See bump_yz_bug.blend They render fine with object and window mappings. Worst when viewing along Z axis and lighting near Y axis. With camera facing out of Z with Y up, when rotated 90 degrees about x axis and it look fine. See bump_zy_bug.blend System: Win XP, Athlon 2000+, ATI Radeon 9600 Pro ";"None";"None";"New" 4438;1;"Open";3;2600;"Carl Olsson";100;"Nobody";"2006-06-24 02:23";"1970-01-01 01:00";"2012-10-29 12:18";"Decimate modifier removes "set smooth" normal interpolation.";"For: 2.42 RC1, 2.41 When a mesh has a decimate modifier on it, its ""set smooth"" normal interpolation setting is ignored an it always displays/renders as with ""set solid"". System: Win XP, Athlon 2000+, ATI Radeon 9600 Pro";"None";"None";"New" 4456;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-06-24 21:05";"1970-01-01 01:00";"2007-05-24 12:03";"strange bright pixels appear at edge of mesh";"Open this file and render. The image should be completely black, but on my machine the image has some unexpected bright pixels. Blender CVS Mac OS X 10.4.6 Dual G5";"None";"None";"New" 4506;1;"Open";3;4951;"Toby Youngberg";103;"Ton Roosendaal";"2006-06-27 17:16";"1970-01-01 01:00";"2007-05-24 12:03";"Interface problems & Rendering. . .";"I am grouping these two together because they are both related to the Intel Mac Mini and Macbook. The interface problems are remnants of already existing interface problems with the Intel Mac Mini. A hack was submited and the menus now function, however the selection box and brush tool don't appear in the interface when being used. When the dashboard button is pressed they appear and stay at the same spot on the screen. Also, rendering times are absolutely awful at nearly 30 seconds to render a lit, untextured cube. This could most likely be corrected by releasing an intel build of blender.";"None";"None";"New" 4510;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-06-27 19:48";"1970-01-01 01:00";"2007-05-24 12:03";"Material setting ""Add"" does nothing";"Render this blend file. It has four spheres: 1. Ztransp and Add=0 2. Ray Transp and Add=0 3. Ztransp and Add=1 4. Ray Transp and Add=1 No matter if the transparency is ray-traced or Z-buffered, Add makes no difference. If ""Add"" only works for halos, maybe the button should only be available for halo materials.";"None";"None";"Investigate" 4512;1;"Open";3;2570;"Nicola Jelmorini ";3002;"Joseph Eagar";"2006-06-27 20:37";"1970-01-01 01:00";"2007-05-24 12:03";"Linux: Game sound disables Sequencer Editor sound";"I have imported in Blender a .WAV 16 bit audio strip into the Video Sequencer Editor, but I can't ear the sound. I have opened the same sound strip in the Audio Window and here too I can't ear the sound. In both cases in the console window this error message is displayed: Couldn't open audio: No available audio device Au contraire, in the ""Buttons window"" I activate the ""Sound block buttons"" and when I press the Play button I can ear the sound correctly. It's a bug, or I make an error using the sound functionalities in Blender? ";"None";"None";"Investigate" 4520;1;"Open";3;4949;"Christian Hirche";100;"Nobody";"2006-06-28 12:13";"1970-01-01 01:00";"2012-10-29 12:18";"Bool after subsurf modifier doesn't work";"In the attached file you can clearly sea what is meant: If your subsurf modifier is the first in the modifier list and boolean the second, subsurf isn't applied at all. the other way round it works perfectly (bool is applied and after it subsurf) System is OSX 10.4.6, RC2 build";"None";"None";"New" 4540;1;"Open";3;1862;"Ben Batt";100;"Nobody";"2006-06-29 18:55";"1970-01-01 01:00";"2012-10-29 12:18";"Boolean maintainer needed! This modifier segfaults with vertex colours, UV coordinates";"The Boolean modifier segfaults when both meshes have vertex colours or UV coordinates. The attached .blend demonstrates this: 1. Load booleantest01.blend 2. Enable the Boolean modifier 3. Segfault Running on Linux, nVidia, current CVS.";"None";"None";"Investigate" 4549;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-06-30 15:54";"1970-01-01 01:00";"2007-05-24 12:03";"Relative File paths break when file saved to new location";"Tested on CVS - Windows XP The file paths in the Image input node seem to be stored in some kind of relative fashion. - Add an Image node, connect it to a composite or viewer node - Load an image - Save the file - New - Load the file, save it to a new location - New - Load the previous file - Image node shows blank, rendering ""Do composite"" does not include the (missing) image Mike";"None";"None";"New" 4556;1;"Open";3;4911;"Jan Kalab";103;"Ton Roosendaal";"2006-07-01 18:13";"1970-01-01 01:00";"2007-05-24 12:03";"Animated curve guide only static particles, not dynamic";"When you want do guide static particles, there is no problem, just make guide curve and you can animate it without any problem. But it doesn't guide dynamic particles, when animated. If the guide doesn't move, it guide dynamic particles too. Blender 2.42 RC2, Python 2.4 dynamic Fedora Core 5, Windows 98";"None";"None";"New" 4566;1;"Open";3;3212;"Randall Rickert";100;"Nobody";"2006-07-02 02:41";"1970-01-01 01:00";"2008-09-09 20:18";"Simple subdivision acts like Catmull-Clark on edges with no faces";"Open this file. The subsurf modifier on the mesh is set to ""Simple Subdiv."", but it behaves like ""Catmull-Clark"" on any edges that are not part of a face. Blender CVS Mac OS X 10.4.7 Dual G5 2.5GHz ";"None";"None";"Investigate" 4568;1;"Open";3;3212;"Randall Rickert";103;"Ton Roosendaal";"2006-07-02 03:16";"1970-01-01 01:00";"2007-05-24 12:03";"Object Ipos not affected by TimeOffset";"Open this file. See the FStrength channel changing every 50 frames. The object has a TimeOffset of 25. The FStrength channel is part of the object ipo, and the second green line appears in the ipo window, indicating that the FStrength should respect the TimeOffset. However, the TimeOffset and Time ipo have no effect on the behavior of the force. Blender CVS Mac OS X 10.4.7 Dual G5 2.5GHz";"None";"None";"New" 4619;1;"Open";3;4994;"Massimo Spazian";4246;"Peter Schlaile";"2006-07-05 17:36";"1970-01-01 01:00";"2008-04-07 08:48";"FFMPEG output problems (2.42 RC3a linux)";"I have problems rendering an animation with FFMPEG. 1. With format AVI and codec DV there is the error: Error initializing video The animation is not rendered. 2. With format AVI and codec HuffYUV the animation is ruined. The attached file is an example of this problem.";"None";"None";"Postponed" 4632;1;"Open";3;4911;"Jan Kalab";103;"Ton Roosendaal";"2006-07-06 14:04";"1970-01-01 01:00";"2007-05-24 12:03";"Array modifier uses controller object parented to own object; cyclic issue!";"Everything is explained in attached .blend";"None";"None";"New" 4636;1;"Open";1;3487;"Campbell Barton";1013;"Jiri Hnidek";"2006-07-07 00:27";"1970-01-01 01:00";"2008-05-09 21:01";"metacube strange normals";"with this test blend spin the metacube in editmode, it has very strange normals. Only effects half-res updating with wiresize above 0.7";"None";"None";"None" 4646;1;"Open";3;3191;"Charles Wardlaw";103;"Ton Roosendaal";"2006-07-08 18:52";"1970-01-01 01:00";"2007-05-24 12:03";"bad preview render button, part 2 - animation";"Thanks for fixing the last one so quickly. This is another preview render button bug: 1) In any scene, click the preview button to do a preview still or control+click to do an animated preview. 2) Hit escape to go back to Blender. 3) Control+Click the preview render to generate a preview animation. Only a still will be generated. To fix the problem, close the render window by actually clicking the X in the corner and then Control+Click the preview button -- then the animation is generated. Also, the same problem happens if a render was done with the internal renderer -- you still have to close the window to get the animated preview to work. Blender CVS (2.42 RC3 abouts), Mac OS X 10.4.7, NVidia GeForce2Go 5200 (64 meg of ram), 1.25 gig of ram. Any other info, you know where to find me. ^_^; Thanks again, the latest release is really shaping up magnificently. ~ Charles";"Interface";"None";"New" 4648;1;"Open";3;2276;"pat buergi";139;"Martin Poirier";"2006-07-08 23:28";"1970-01-01 01:00";"2007-05-24 12:03";"Transform Input Consistency";"1. In size mode you can move your mouse up and down (vertical) and horizontal to size the object, but in shrink/fatten mode blender only accepts vertical mouse moves. 2. If you start sizing with the mouse on a certain point in the 3d space and move it up, the mesh would e.g. be scaled up. But if you start shrinking on the same point and move the mouse up the objects gets smaller. No real bugs but pretty annoying while modelling, at least for me. buergi";"Interface";"None";"Approved" 4654;1;"Open";3;4909;"andy goralczyk";103;"Ton Roosendaal";"2006-07-09 19:15";"1970-01-01 01:00";"2007-05-24 12:03";"image window render: 'save buffers' option doesnt store RenderResult on ESC";"happens when rendering to image window or fullscreen: when the 'save buffers' option is enabled and you abort a render, it neither shows the (incomplete) render result nor any render stats in image editor. this works when 'save buffers' is disabled and also when rendering to a seperate window.";"None";"None";"New" 4671;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2006-07-11 20:33";"1970-01-01 01:00";"2012-10-29 12:18";"Curve modifiers: still working on data level, doesn't work for multiple linked Curves.";"In the attached Blend file there are 3 objects 1 curve thats used as a curve blend curve, and 2 others that share linked data- both of these share curve data. they are both offset in different directions so they should sit on either side on the curve they follow. Now go into editmode on he different curves. As you exit, the others curves are effected. Theres no circular dependancy that should casue this.";"None";"None";"New" 4672;1;"Open";3;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2006-07-11 22:37";"2006-07-16 15:22";"2007-05-24 12:03";"Updating the 3D View";"WinXP, Ati9700pro, Blender 2.42 RC3 On a lot of actions (Like changing the blendmode of a texture) Blender is updating the 3D View for an object but two things are missing: - changing the texture type in ""Texture Buttons (F6)"" in panel ""Texture"" - changing a texture in ""Texture Buttons (F6)"" in panel ""Image""";"None";"None";"New" 4684;1;"Open";3;2682;"Tron Thomas";103;"Ton Roosendaal";"2006-07-13 05:50";"1970-01-01 01:00";"2007-05-24 12:03";"Animation playback is slow";"I have a scene I created in Blender 2.41. When I was using that version play back of the animation in the program was real time. Using RC3 the animation plays back in slow motion. Steps: 1. Open the attached file 2. Click on the play button in the time line window Results: The animation plays back in slow motion Expected: The animation should play back in real time. ";"None";"None";"New" 4716;1;"Open";3;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2006-07-17 22:14";"1970-01-01 01:00";"2007-05-24 12:03";"Raytraced alpha could become an option";"WinXP SP2, Ati9700pro, Blender 2.42 The scene has a complete object based background but i'm getting strange alpha effects, so i can't use png format for saving the result. Take a look at the attached JPG.";"Render (internal)";"None";"Investigate" 4717;1;"Open";3;3132;"Daniel Salazar";103;"Ton Roosendaal";"2006-07-17 23:23";"1970-01-01 01:00";"2007-05-24 12:03";"Transparency don't show on OGL window if object is on a group";"Hi, I got this model with transparent windows and they show ok in the real object, but I used a group for duplicating and then transparency is not showing on duplicated objects inside the group http://www.zanqdo.com/tmp/BugTransp.blend gracias :)";"None";"None";"New" 4737;1;"Open";3;2594;"Stephane SOPPERA";103;"Ton Roosendaal";"2006-07-19 19:58";"1970-01-01 01:00";"2007-05-24 12:03";"RGB curves do not work properly with alpha premultiplied pictures";"The ""RGB curves"" node seem to be applied without dividing RGB values with alpha. So they don't work on alpha-pre-multiplied images. Scenario: * open the linked .blender file * render -> in the ""Render Result"" you can notice that the egdes of the cube are badly composed. -> if you have a look at the output of the ""Viewer Node"", you can notice that the RGB curves are applied on the alpha-pre-multiplied values that's why the result is bad. ";"None";"None";"Approved" 4738;1;"Open";3;2594;"Stephane SOPPERA";103;"Ton Roosendaal";"2006-07-19 21:52";"1970-01-01 01:00";"2007-05-24 12:03";"Mix node does not output correct alpha";"When using a mix node in ""Add"" mode, the alpha generated is only the one of one of ths input images, not the combination of the alpha of the two images as it should. Actually all other modes (mix,...) have this problem except nodes for which it is the correct behavior of using the alpha of the first image like Hue,Saturation,... Scenario: * open the linked .blend * render -> check the render output in ""Draw image with alpha"", you can see that the alpha channel is the one of one of the input images, not the cobination of both. ";"None";"None";"New" 4758;1;"Open";3;4716;"andreas schumann";103;"Ton Roosendaal";"2006-07-21 16:12";"1970-01-01 01:00";"2007-05-24 12:03";"composite nodes don't update when rendering with yafray";"hey guys! since yafray 0.0.9 came out, i played around with the compatibility with newer blender features, to my suprise i noticed yafray works really good with compositing, even correct z values were given to blender so i could do dof and everything, i was also able to have vector blur by using a copy of the scene rendered with blender as input for speed normals. but! the problem is blender doesnt update the render output window. in order to make it apply compositing again you have to change a value for example the strength of the blur node... in the vector blur example it kinda worked when dis and reconnecting some nodes, so i could get a stilll picture out of it, rendering an animation just deleted the render buffer as the yafray scene was not updated every frame... also i noticed that after rendering with yafray the cursor over the render window stays a ""1"" like the one when viewing an animation, so it might be stuck waiting for yafray to finish rendering although it already is... i hope you understand my problem :) good day btw: great release, ton and everyone! <3 ";"None";"None";"New" 4759;1;"Open";3;3874;"Ervin Weber";103;"Ton Roosendaal";"2006-07-21 17:11";"1970-01-01 01:00";"2007-05-24 12:03";"Material OnlyShadow should render (optinally) backfacing shadow too";"Onlyshadow material doesn't handle well shadows coming with a high incident angle. Instructions: - add a UVsphere and set to ""smooth"" - add a material with: - ONLYSHADOW on - TRACEABLE off - SHADBUF off - ZTRANSP on (this is optional but shows the bug better) - add a lamp casting shadow - add a mesh to cast shadow on the spere - render Result: The shadow along the sphere's shadow boundary (AKA the zone where terminator problem occours) disappears. Tested on: 2.34: works OK only with a Spotlight with BufShadow 2.37: works with a Spotlight with BufShadow or RayShadow too 2.41: same as 2.37 2.42: doesn't work at all I attach a blend file and an image showing the problem Computer: AMD Athlox XP 1600+ nVidia GeForce3 Ti 200 WinXP Pro SP1 ";"None";"None";"New" 4761;1;"Open";3;4944;"mr Carnivore";103;"Ton Roosendaal";"2006-07-22 01:40";"1970-01-01 01:00";"2007-05-24 12:03";"When collapsing a grouped composition node the order of the inputs and outputs, respectively, are changed";"When collapsing (with the tab key or via ""edit node"" in the menu) a grouped node the order of the inputs as well as the order of the outputs is scrambled. Meaning the order is different when the node is in edit mode than when the node is not in edit mode. Also the order can not be influenced by changing the layout of the nodes inside of the grouped node.";"None";"None";"New" 4763;1;"Open";3;4944;"mr Carnivore";103;"Ton Roosendaal";"2006-07-22 01:50";"1970-01-01 01:00";"2007-05-24 12:03";"Undo (ctrl-z) in node view clears all preview images of the nodes";"When undo (ctrl-z) is used in the node window (node editor window), the preview images inside every node is cleared. The images are restored when a view node is clicked.";"None";"None";"New" 4776;1;"Open";3;555;"Omar Modesto ";139;"Martin Poirier";"2006-07-23 21:37";"1970-01-01 01:00";"2008-05-09 21:01";"Push-Pull in Object mode uses SHIFT+P, should this work?";"In Object Mode, the hotkeys for Push/Pull and Preview Window are the same (Shift+P), making the former accessible only through menu access (Object>Transform) since Shift+P only works for the latter. Note that Push/Pull is also Shift+P in Edit Mode, so making their hotkeys different in different modes would be a bit confusing. Maybe Shift+P can be kept for preview and something like Ctrl+Alt+P or Shift+Alt+P can be used for Push/Pull in Edit/Object modes.";"None";"None";"Postponed" 4785;1;"Open";3;2600;"Carl Olsson";103;"Ton Roosendaal";"2006-07-24 02:10";"1970-01-01 01:00";"2007-05-24 12:03";"View grid spacing / unit implementation vs min/max limits for settings like clipping";"For: 2.41, 2.42 In the 3D view View Properties float: - Clip End seems to be clamped to 1000 times the Grid Spacing setting. This has the effect of reducing the Clip End setting to an unusably small value when Grid Spacing is set to small values. - Grid spacing can be set to 0. Then changing view camera Clip End value by dragging sets it to -1.#IO The Grid Spacing value does not seem (and I don't think should) have any effect on the scene except grid drawing and grid snapping, so it seems strange that it should be tightly bound to view clipping. The only possible link I can think of is if Grid Spacing is actually a unit scaling factor, but if that is the case it should be renamed to reflect its actual purpose, and grid drawing/snapping should be seperate functionality. To test: - Display View Properties float - Drag Clip End to its maximum value. - Increase Grid Spacing value. - Drag Clip End to its maximum value. - Observe increase in Clip End maximum value. - Set Grid Spacing to 0. - Click and drag Clip End field to change value. - Observe Clip End value to be -1.#IO. Test Systems: Win XP, Athlon 2000+, ATI Radeon 9600 Pro Win XP, P4 3Ghz, Intel 82945G Express";"Tools";"None";"New" 4786;1;"Open";3;2593;"Peter Rehley";3709;"Andrea Weikert";"2006-07-24 02:15";"1970-01-01 01:00";"2007-05-24 12:03";"if space in the Application path name, system() doesn't work on some platforms";"I downloaded the Mac OS X PPC Python 2.3 10.3+ version of blender. I tried the array3.blend I found on blendernation. After the animation was done, I clicked the play button and nothing happened. I tried the same file with 2.42RC3 and the play button played the animation back. When I compiled 2.42 on my powerbook G4 the play button works. My system is Mac OS 10.4.7, with python 2.3 and ATI Mobility Radeon 9700 graphics card.";"None";"None";"Postponed" 4798;1;"Open";3;2930;"douglas ollivier";103;"Ton Roosendaal";"2006-07-25 06:43";"1970-01-01 01:00";"2007-05-24 12:03";"Env Preview render broken in 2.42";"Hi Guys, this is repeatable bug in 2.42 relating to images not loading at all/ incorrectly when trying to load an Environment map. 1)open Blender. 2)Add a texture to the default cube 3)note it doesn't load, or load properly 4)render.... note the blackness of the material this was tested and confirmed by briggs (Zarf/ Xarf) Doug Ollivier";"Render (internal)";"None";"Fixed" 4813;1;"Open";3;4255;"Carlos Folch";103;"Ton Roosendaal";"2006-07-28 18:56";"1970-01-01 01:00";"2007-05-24 12:03";"The BAKE function don´t works properly";"Sorry if it´s not a bug. But before (blender 2.37a)it was working fine. Now, since 2.40, if we make a BAKE in an Armature with IK solvers (bones or emptys) and after delete the IK solvers, we lost the transformations of the bones affected by IK. This was an important feature to export models and animations for games (torque, qradiant, etc..) that not support the ik solvers. I have found more people that need this function working again: http://www.blender3d.org/forum/viewtopic.php?t=7682&sid=e96ddd24f3e9921c12a20940113e0bb9 Sorry for my english. Thanks ;-)";"None";"None";"New" 4815;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-07-28 22:28";"1970-01-01 01:00";"2007-05-24 12:03";".b.blend - View properties dialog initally opens ""under"" window border - View Name";"2.42a Win Xp A minor but annoying bug (in the default .b.blend file ??) The first time the view properties dialog box is opened, the dialog box with the bottom 1/10 of the box or so ""under"" the window border, hiding the ""Relationship Lines"" button. I've fixed this in my default .b.blend file, but every time I open a scene from someone else and choose View properties I have to LMB/drag the box up to see the bottom. Also it would be nice if the default for the View name was on. New users would appreciate it, and experienced users could always turn it off. Mike ";"None";"None";"New" 4823;1;"Open";3;3138;"Olivier Debroise";103;"Ton Roosendaal";"2006-07-30 19:36";"1970-01-01 01:00";"2007-05-24 12:03";"HDR Aliasing problem, example with radiosity";"Aliasing problem with the radiosity option. The intersection of two object is not anti-aliased, apparently. A .blend file is joined.";"None";"None";"New" 4837;1;"Open";3;5088;"Mitch Hughes";103;"Ton Roosendaal";"2006-08-03 06:52";"1970-01-01 01:00";"2008-05-09 21:00";"Python: Lens Value delayed 1 Frame after setCurrentCamera during Animation";"This is a Bug with the python scene.setCurrentCamera() function This bug was introduced in 2.42 as this blend renders correctly in 2.41 If a script is used to change the current camera in a FrameChanged scriptlink the animation in the blender ui looks correct, however if you render the animation in blender using the Render button or render the animation on a command line the camera location is updated however the camera lens remains the same as the last camera for one frame, resulting in a zoom effect when the camera changes. When distributing a rendering on a renderfarm having a small step like 25 frames on a 1000 frame animation would result in 40 or so incorrect frames, depending on which camera you had selected when the blend file was saved and if you modify the camera lens value between cameras. This actually happened to me and I had to render the 40 frames again, by hand. If you render the animation you should see frames 1,5 and 8 rendered with an incorrect lens value.";"None";"None";"Investigate" 4877;1;"Open";3;4793;"Mike Stramba";100;"Nobody";"2006-08-12 22:06";"1970-01-01 01:00";"2012-10-29 12:18";"particle emitter parented or using Armature modifier prevents manipulating bones";"Tested with Blender 2.42a - Windows Xp If a particle emitter is parented to or uses an armature modifier : When an armature bone is keyframed in Pose Mode, (Loc, Loc/Rot) at any frame, then when the bone is manipulated (moved / rotated, when LMB is clicked to confirm the transform the bone ""snaps"" back the previous keyframe. The bones can be transformed if : 1) If the emitter plane is ""unparented"" (or the Armature modifier name is blanked out) Or 2) If Verts / Faces in the Paricles/Emit From: panel are both turned off, the bones can be transformed - Keyframes *can* be created with autokey switched on, or by ctr-LMB clicking in the IPO window ";"None";"None";"New" 4881;1;"Open";3;2570;"Nicola Jelmorini ";4246;"Peter Schlaile";"2006-08-13 09:22";"1970-01-01 01:00";"2008-04-07 08:47";"OGG preview in Video Sequence Editor cause Segmentation fault";"Hi, I have tried to import an OGG video file in the Video Sequence Editor and this operation function well (you can see the video strip details in the picture I have attached to this bug report). The problem is that if I activate the image preview in another Blender Video Sequence Editor window to watch the video strip, Blender exit with a Segmentation fault. The messages in the console window are these: nicola@cavergno:~$ /home/nicola/blender/blender Compiled with Python version 2.4. Checking for installed Python... got it! [theora @ 0x8edb800]Theora bitstream version 3.2.0 Input #0, ogg, from '/home/nicola/Files-ricevuti/page3-movie-35.ogg': Duration: 00:03:28.9, start: 0.000000, bitrate: 225 kb/s Stream #0.0, 15.00 fps(r): Video: theora, yuv420p, 800x608 Stream #0.1: Audio: vorbis, 22050 Hz, mono, 30 kb/s Segmentation fault I have tried other kind of video strips (mpeg, avi, divx, mov) and all function very well. It's only the OGG format that have this problem with the image preview. ";"None";"None";"Postponed" 4908;1;"Open";3;5121;"Adolf Mathias";103;"Ton Roosendaal";"2006-08-20 16:54";"1970-01-01 01:00";"2007-05-24 12:03";"Nurbs/Curve: option ""UV Orco"" is confusing... better is to create real UV coords for UV map-to channel";"I have a closed NURBS curve with a NURBS circle as bevel object, and when I try to texture-map this babe using UV coords, each triangle (not square!) in the tesselation has UV coords from 0 to 1, and the triangles forming a square additionally don't match. Consequently, the number of iterations changes with the UV resolutions of both curves, and the texture scaling buttons are essentially useless. Please render the attached file to see what I mean. I'd consider the right coords to be from 0 to 1 each around the circle and around the cyclic NURBS, no matter what the splines' UV resolutions are.";"None";"None";"New" 4926;1;"Open";3;4990;"Henry Schreiner";103;"Ton Roosendaal";"2006-08-24 22:11";"1970-01-01 01:00";"2007-05-24 12:03";"X-Ray option makes Transparant drawing wrong";"I found a problem that occurs when an object in the 3D view port is set to x-ray (like an armature, for instance) and either a transparent object or a set-chained object is in the view. When one or more objects are set to x-ray, transparent and active scene objects become x-ray too. I attached a .blend that is self explanatory. :) Sorry, I couldn't tell which category to put this in - interface or OpenGL, so I gave up and put ""None"". It's a 3D view only problem, with no effect on rendering, so not very pressing. I hope this helps. Oh, my system is a WinXP home Athlon64 3200+, nVidia 5700, 1 Gig ram, using Blender 2.42a and CVS 8-21. Same problem on a Pentium2 400Mz Ubuntu Linux also with Blender 2.42a. ";"None";"None";"Investigate" 4948;1;"Open";3;2423;"Chris Burt";103;"Ton Roosendaal";"2006-08-29 23:15";"1970-01-01 01:00";"2007-05-24 12:03";"Particle duplicators not displayed in instanced groups";"Particle duplicators do not instanciate correctly in groups. As a refresher to anyone reading this, static particle duplicators are created by parenting a mesh to another mesh with a static particle system assigned to it and activating ""DupliVerts"" on the parent mesh. Grouplicators, as I will call them, would be very useful for making quick copies of trees wherein the leaves are particle duplicators. Alas.. I attempted to use this functionality and it does not work. Only the standard static particles display on the instanced copy. I have attached a very simple demo file in which this problem is illustrated.";"None";"None";"New" 4962;1;"Open";3;4911;"Jan Kalab";100;"Nobody";"2006-09-04 13:14";"1970-01-01 01:00";"2012-10-29 12:18";"Array modifier doesn't support particles";"Objects which are made by array modifier doesn't emit particles. When you apply the modifier, they does. I think they should emit them as normal objects.";"None";"None";"New" 4964;1;"Open";3;5175;"grey tomorrow";100;"Nobody";"2006-09-06 11:53";"1970-01-01 01:00";"2012-10-29 12:18";"Dupli-vert doesn't duplicate hierarchies with deform (like armature + mesh)";"1) parent a mesh to an armature 2) animate the armature 3) create an duplivert emitter and parent the armature to it 4) animate the duplivert emitter 5) check animation in viewport (alt-A) 6) render the resulting animation is not what what seen in the viewport. dupliverted armature are disjointed and their meshes distorted. see http://blenderartists.org/forum/showthread.php?p=701811#post701811 for further explantion. thank you, gt";"None";"None";"New" 4987;1;"Open";3;4028;"Jorge Rocha";103;"Ton Roosendaal";"2006-09-13 20:52";"1970-01-01 01:00";"2007-05-24 12:03";"Render window is not refreshed when brought to the front while rendering.";"I know Blender never refreshes the main GUI window while rendering, but it used to refresh the render window. In versions older than 2.36, if I made a desktop arrangement which covered Blender's render window while rendering, such as changing desktop (in Linux) or bringing another window to the front, the render window was quickly refreshed when it got uncovered. But that doesn't happen anymore, I get totally lost on how my rendering is going if I use my computer to do a simple task such as checking my email. If I go back to Blender, only the most recent rendered lines show on top of screen garbage. This is particularly bad for a rendering which takes too long to complete, because we are forced to wait to the end just to find out that the rendering should have been interrupted because of a misplaced light or some other such detail. ";"None";"None";"New" 4988;1;"Open";3;3191;"Charles Wardlaw";146;"Brecht Van Lommel ";"2006-09-14 00:35";"1970-01-01 01:00";"2007-05-24 12:03";"texture paint images save with weird alpha";"Hi, Playing with the new paint tools (current Blender CVS) I came across an interesting issue. The alpha, when saved, seems to be inverted. This is best shown through images, so here we go: 1) Image of painted texture and 3D model (looks fine): http://ministryofdoom.org/cloud/blender/20060913/painted-1.jpg 2) Image loaded into Photoshop Elements: http://ministryofdoom.org/cloud/blender/20060913/painted-2.jpg As you can see the alpha channel is backwards. The image itself seems to have been saved properly otherwise. What I did: - Created a new 2048x2048 blank texture, and assigned it to the UVs on my object. - Painted. - Saved. Pretty simple. If you need the blend I can send it to you with the image packed.";"None";"None";"New" 4993;1;"Open";3;4793;"Mike Stramba";2923;"Tom Musgrove";"2006-09-15 11:06";"1970-01-01 01:00";"2007-05-24 12:03";"NLA ActionStrip / ActionStrips example code";"Since there is no example code for the NLA- ActionStrip/ActionStrips classes, perhaps this might be added to the docs. Took me about 10 hours to figure it out :) Mike (Attached file 'actionstrips.py' if formatting gets messed up) Mike import Blender from Blender import * from Blender.Armature import NLA ctr = 1 numStrips = 1 vflags ={32:'LOCK_ACTION',1:'SELECT',2:'STRIDE_PATH',8:'HOLD',16:'ACTIVE'} def tranflag(flag): global vflags if flag: print for v in vflags.keys(): t = flag & v if t: print '\t\t',v,vflags[t] else: return def showStrip(strip): global ctr,numStrips print ctr,'/',numStrips print strip.action.name print '\tstripStart',strip.stripStart print '\tstripEnd',strip.stripEnd print '\tactionStart',strip.actionStart print '\tactionEnd',strip.actionEnd print '\tblendin',strip.blendIn print '\tblendout',strip.blendOut print '\tflag',strip.flag, tranflag(strip.flag) print '\tmode',strip.mode print '\tbrepeat',strip.repeat print '\tstrideAxis',strip.strideAxis print '\tstrideBone',strip.strideBone print '\tstrideLength',strip.strideLength armOb=Object.Get('Char1Armature') actions=Armature.NLA.GetActions() # # Actions named 'rot', 'move', 'Run' assumed to exist, or substitute # your own action names # rotAct = actions['rot'] movAct = actions['move'] runAct = actions['Run'] # # get all NLA strips for this object # Char1NLAstrips = armOb.actionStrips # # set the current frame to where you want NLA strips to initially appear # in the NLA editor frame = 1 Blender.Set('curframe',frame) # # remove all NLA strips # for strip in Char1NLAstrips: Char1NLAstrips.remove(strip) # # some different ways of adding action strips to the NLA editor # Blender.Object.Get('Char1Armature').actionStrips.append(Blender.Armature.NLA.GetActions()['tester']) Char1NLAstrips.append(Blender.Armature.NLA.GetActions()['UpDown']) armOb.actionStrips.append(rotAct) Char1NLAstrips.append(movAct) Char1NLAstrips.append(actions['Run']) # # get a strip # strip0 = Char1NLAstrips[0] print '\nstrip0.action.name =""'+strip0.action.name+'""' # # show it's properties # showStrip(strip0) # # change it's stripStart, stripEND (add 50 frames) # (effectively moving the strip strip0.stripEnd += 50 strip0.stripStart += 50 # # show the changes # showStrip(strip0) # # select the strip in the NLA editor # strip0.flag += NLA.Flags['SELECT'] Blender.Window.RedrawAll() showStrip(strip0) # # move all strips by FrameOffset # def moveallStrips(FrameOffset): for strip in Char1NLAstrips: strip.stripEnd += FrameOffset strip.stripStart += FrameOffset moveallStrips(30) # # show all strips Properties # print print '============ ALL STRIPS ================' numStrips = len(Char1NLAstrips) print numStrips,' NLA strips for ',armOb ctr=1 for strip in Char1NLAstrips: showStrip(strip) ctr +=1";"None";"None";"New" 5007;1;"Open";3;3191;"Charles Wardlaw";103;"Ton Roosendaal";"2006-09-18 22:37";"1970-01-01 01:00";"2007-05-24 12:03";"Two Preview Render woes";"Hey again. Two things: 1) Preview Render renders at 1/4 scale (roughly) if you set the image size too high. Start a new scene and change the output resolution to 1920 x 1920. Now hit the preview render button on a 3D area. The render will come out weird. 2) Preview Renders do not take into account user preferences for where renders should turn up (window, image editor, full screen) -- they *always* pop up in the window. These issues are present in 2.42a.";"None";"None";"New" 5032;1;"Open";3;2923;"Tom Musgrove";146;"Brecht Van Lommel ";"2006-09-25 04:27";"1970-01-01 01:00";"2007-05-24 12:03";"3d paint stroke gets confused";"unpack the image from the blend, draw a stroke quickly from A to B - the stroke path on the 3d view is way different from the stroke path actually taken. (It gives a linear line in the UV image from A to B). The 3d painting is blazingly fast now - thanks :) LetterRip ";"None";"None";"New" 5039;1;"Open";3;4300;"Juan Carlos Rodrigo";103;"Ton Roosendaal";"2006-09-27 04:24";"1970-01-01 01:00";"2007-05-24 12:03";"JKey does not show previous render when Border (Render) is selected";"Open Blender Go into Render F10 Set the Border switch on Go into Camera mode 0 Select a border SHIFT-B Render F12 (A window pops up) Press J Render F12 Press J [Previous Lost] ";"None";"None";"New" 5066;1;"Open";3;5239;"giovanni lanza";103;"Ton Roosendaal";"2006-10-04 16:06";"1970-01-01 01:00";"2007-05-24 12:03";"UVMapping rendering";"The transparence of the UVMapping is not supported in the rendering. ";"None";"None";"New" 5078;1;"Open";3;2611;"nemyax ";103;"Ton Roosendaal";"2006-10-08 16:22";"1970-01-01 01:00";"2007-05-24 12:03";"Mixed-up transform axes for deformed meshes (""crazy space"")";"This is a known bug (been there since 2.40), but it doesn't seem like anyone has actually reported it in the tracker. If a deforming modifier (armature, curve) is used on a mesh, and the modifier is enabled in edit mode, then transformation of vertices is unpredictable. Attached is a file where this behaviour is reproduced. Select either cylinder, go to edit mode and try moving the selected vertex.";"None";"None";"New" 5108;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-10-18 22:37";"1970-01-01 01:00";"2007-05-24 12:03";"Dupliverted Object with Armature modifier differs on rendering...";"Tested with 2.42a / CVS (2006-10-15) Compare the 3d viewport with the rendered result when the second bone in the armature is not in it's rest position. Cube has an armature modifier with multi (2) bones. Cube mesh then parented to plane, dupliverts turned on for plane. When rendered, if the second bone is not in rest position (i.e. frame 20,40 in the attached file) (deforming the mesh), then the mesh gets severly distorted. Mike";"None";"None";"New" 5114;1;"Open";3;5256;"Nitro Furano";103;"Ton Roosendaal";"2006-10-20 15:18";"1970-01-01 01:00";"2007-05-24 12:03";"pure vectorial rendering instead of raster, as option";"While the most common way to get rendering were getting raster results (like png, etc.), very useful would be just vectorial rendering, without raster images (svg, swf, ai, dxf, pdf, eps, etc.) (in a first stage, reprojecting the 3d coordinates into the 2d camera is more than enough) - animations rendered into sequences of .svg or .ai, or into a .swf or animated .svg The code would be not that hard, since would be no much more than translating the 3d coordinates into the 2d display projection If useful, i have some code at http://nitrofurano.linuxkafe.com/sdlbasic, which would not be very hard to recode on Python or C - of course maybe some improvements must be needed... ";"None";"None";"New" 5131;1;"Open";3;2923;"Tom Musgrove";1013;"Jiri Hnidek";"2006-10-26 10:03";"2006-11-04 09:53";"2007-05-24 12:03";"straggler metaelements after converting meta to mesh";"I created a object from metas (I used mirror mode when editing). Then converted to mesh. The resulting mesh was unable to enter editmode. (Also in the file it had left the radius stuff of the metaelements even though I chose delete original.) I've attached two blends, the first is the straggler metaelement stuff, the second is the mesh that can't be edited. ";"None";"None";"New" 5145;1;"Open";3;4544;"Kai Kostack";100;"Nobody";"2006-10-29 08:05";"1970-01-01 01:00";"2012-10-29 12:18";"dupliverts are applied before modifiers are executed";"i have to convert the modifiers of an object (for example an array) into a mesh, before i can use it for dupliverts. that makes not really sense imho. is this intended?";"None";"None";"New" 5161;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-11-02 06:14";"1970-01-01 01:00";"2007-05-24 12:03";"RGBA (Alpha channel) button in Scene/Format panel available for unsupported formats";"2.42a / CVS The RGBA button should be disabled when formats that do not support an alpha channel are chosen (jpg, avi-jpg ... others ??). Mike";"None";"None";"New" 5173;1;"Open";2;2923;"Tom Musgrove";3709;"Andrea Weikert";"2006-11-06 04:02";"1970-01-01 01:00";"2007-05-24 12:03";"file browser unable to browse folder 'shortcuts'";"The Blender browser is unable to navigate via folder symbolic links on windows as reported in this thread, http://blenderartists.org/forum/showthread.php?t=81454";"None";"None";"Investigate" 5222;1;"Open";3;5320;"thomas g";146;"Brecht Van Lommel ";"2006-11-14 00:16";"1970-01-01 01:00";"2007-05-24 12:03";"Ik Solver Problem -very important!-";"the problem is included in the blend file that is attached. basically, the ik solver does not track the ik solver correctly when it is directly above/below the armature. also the ik solver switches the armature when moving horizontally (making disjointed knees and painfully looking walk-cycles) better details are writen in the blend file, just open it and it will become apparent. it includes a short animation and the text will make mention to it to see the problem more precisly ";"None";"None";"New" 5228;1;"Open";3;3914;"Erwan Jacq";103;"Ton Roosendaal";"2006-11-14 18:01";"1970-01-01 01:00";"2007-05-24 12:03";"Transparency in textured mode not working properly";"In the linked file, Trees are UVmapped with a PNG transparent picture. Transparency works in textured mode but: - sometimes we se the third tree behind the first in transparency (but not the 2nc which is just behind?) - The ground plane is not drawn behind the trees. I link a picture of what i see on my screen too. Linux, NVIDIA fx5200 with 9626 beta drivers.";"None";"None";"New" 5300;1;"Open";3;3577;"Roger Prefontaine";103;"Ton Roosendaal";"2006-11-23 21:17";"1970-01-01 01:00";"2007-05-24 12:03";"Parented Hooked verts: cyclid dependency";"1) Select an Object (like default cube). 2) Enter the EditMode and Select a vertex. 3) Add Hooks([Ctrl]+[H]) and Select ""Add Hook, To New Empty"". 4) Leave the EditMode. 5) Select new Empty first (unless already selected), then select base object, and Ctrl-P parent them. (so that empty moves when you move cube). 6) Grab cube and drag mouse to left side of screen and press ESC. 7) The base object moves back to it's original position, but the hooked verts stay on screen wherever you pressed ESC. - Rendered version of this will look correct, though... press F12 to see. - Preview animations of this base object will look wild as the hook fails to update properly (try moving base object very fast in animation and alt-a). - Grab & Esc without moving restores the display, so does ""Reset"" in Hook Modifier panel. - noticed that hooks are now a modifier, but there is no hook option in modifier panel dropdown box. If I've missed something regarding the developments involving hooks, I apologize. Couldn't really find anything super-relevant in the tracker. I'm using Win2k & nVidia w/ CVS 22/11, and 2.42a. Roger";"None";"None";"New" 5302;1;"Open";5;5346;"M. Bashir Al-Noimi";103;"Ton Roosendaal";"2006-11-23 23:08";"1970-01-01 01:00";"2007-05-24 12:03";"Arabic interface bugs";"1- The interface was wrong (1.jpg and its correction) although i used UTF-8 encoding for *.po file. what's wrong !? http://www.fileden.com/files/2006/7/22/140272/01.jpg http://www.fileden.com/files/2006/7/22/140272/correction.jpg 2- Arabic language need Bidi mode (Right to left), the default layout of blender is Left to right, how we can solve this problem ? remark: ""Blender localize arabic.zip"" include Arabic strings into *.po & *.mo file format. http://www.fileden.com/files/2006/7/22/140272/Blender%20localize%20arabic.zip";"None";"None";"New" 5306;1;"Open";3;5348;"Karl Erlandsen";103;"Ton Roosendaal";"2006-11-24 19:36";"1970-01-01 01:00";"2007-05-24 12:03";"Compositor interface doesn't update image sequences on frame change";"Hi everyone, BUG DESCRIPTION: If you load an image sequence in the compositor, the interface doesn't update when you change frames - the only way to force the interface to update is to render with F12. STEPS TO REPRODUCE: - Open Blender afresh, and change views so you have a Node Editor and a UV/Image Editor - In the Node Editor, click on the ""Composite Nodes"" button and click ""Use Nodes"" - Create an Image node, a Viewer node and a Composite node - Connect the Image node's ""Image"" output to both the Viewer node and the Composite node's ""Image"" inputs. Also set the UV/Image Editor to preview the viewer node - click on the image node's ""Load New"" button, browse to a directory with a numbered image sequence in it, tap ""a"" to select all images, and middle mouse click to load them in (I was loading in 400 targa frames, named as ""Scene3_Intro.0001.tga"", ""Scene3_Intro.0002.tga"" etc) - click on the movie button on the image node to open up the image sequence options, and set the number of frames to 400 - change frame, and neither the image viewer node preview nor the image view itself update unless you tap F12 to render the compositor. AFFECTED VERSIONS: Tested in Blender 2.42a, CVS (23rd November) and the orange build available from German Server 1 listed on the orange download page. The bug was present in all 3. OS: Linux (Debian Etch) Graphics Card: NVidia GeForce 6800, driver 8776 Hope this helps guys, I'll be happy to provide further details if you need them. Karl";"None";"None";"New" 5318;1;"Open";3;2923;"Tom Musgrove";4827;"Nicholas Bishop";"2006-11-28 03:05";"1970-01-01 01:00";"2008-04-07 08:46";"sculpt sculpts on hidden mesh";"If you hide geometry by doing Alt-B selection, if a sculpt is started near the edge of the visible mesh it will sculpt onto hidden mesh. Or if a stroke ends into a hidden part of mesh. Also the part of the hidden mesh is temporarily revealed if a stroke continues through an area of hidden mesh (it disappears again after the stroke completes).";"None";"None";"New" 5350;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-12-03 20:36";"1970-01-01 01:00";"2007-05-24 12:03";"mouse record in IPO doesn't play other animations";"Ok, I'll admit this a more a feature request than a bug ... I think. When using the ""record movement"" feature in the IPO window, and choosing ""play animation"" it would be very cool if any other animation in the scene (i.e. alt-a) would play at the same time. If that's not possible (what ? .. isn't everything ? :) ), then at least having the video-sequencer /i.e sound track play would be great. If I wanted to try programming a patch for this, where would I find the ""alt-a"" / play animation code? Is this doable ? Mike";"None";"None";"New" 5352;1;"Open";3;3475;"Renato Perini";100;"Nobody";"2006-12-04 02:54";"1970-01-01 01:00";"2007-05-24 12:03";"Strange ""Crease Subsurf"" behaviour.";"1) Open the uploaded blend file. 2) Enter Edit Mode (TAB). Note an edge loop selection. 3) Now crease subsurf pressing Shift + E and move the mouse around a little bit. At a crease level of about +0.504, I receive a strange behaviour, like the image I submitted. Alternatively, you can subsurf the model with values higher than 3 to observe the phenomena, without creasing. ";"None";"None";"Investigate" 5356;1;"Open";3;3163;"Oliver Göpferich";103;"Ton Roosendaal";"2006-12-04 20:29";"1970-01-01 01:00";"2007-05-24 12:03";"Outliner - buttons disappear, using short coordinates still";"I'm working on a scene with more than 4,300 objects. If I scroll down the outliner, after the first third of the list the visible/selectable/renderable icons disappear (see screenshot and scene file). ";"None";"None";"Fixed" 5405;1;"Open";3;2923;"Tom Musgrove";4827;"Nicholas Bishop";"2006-12-11 04:18";"2007-01-16 08:43";"2008-04-07 08:52";"3D texture strange artifacts";"If you set your brush to 3D there is significant stretching of the texture as you approach the poles. Also concentric rings appear on the brush for all 3D textures. I've attached a screen shot and the image i used as a brush. Scale is set to 1000. ";"None";"None";"Rejected" 5409;1;"Open";3;3547;"Sean Farrell";100;"Nobody";"2006-12-11 15:03";"1970-01-01 01:00";"2012-10-29 12:18";"usability: mesh modifiers lock interface ";"This is a usability bug. With large meshes the mesh modifiers take to long to compute for interactive display. Although this is logical and may be switched of explicitly, once the modifier is computing the interface is dead for that time. Waiting for the operation to terminate is ok, if all is in a desired state but during an modification of the mesh, like moving it. Here is my proposal, inspired from tecplot, while performing a modification, like moving, scaling or rotating switch the modifiers of, once done compute them once. Note: As said this is a usability bug, so low priority...";"None";"None";"New" 5412;1;"Open";3;5298;"bassam kurdali";4246;"Peter Schlaile";"2006-12-11 21:06";"1970-01-01 01:00";"2008-04-07 08:51";"can't output multiplexed audio with mpeg4 from sequencer / ffmpeg";"I can't do audio multiplexing if I pick ""mpeg4"" as the video codec- working at ntsc, 1/2 ntsc and 1/4 HD resolutions, choosing ""multiplex audio"" with any of the audio codec options causes the error ""couldn't find a valid audio codec"" followed by the error ""error initializing audio stream"". My framerates are 24fps in one project and 30fps in another. speculation: for aac encoding I glean from the web that faac has two output formats, one for mpeg2 and one compatible with mpeg4; could it be that the wrong one is being used? platform: linux/ubuntu edgy, current cvs release, build date: 2006-12-10 ";"None";"None";"Postponed" 5415;1;"Open";3;177;"Matt Ebb ";103;"Ton Roosendaal";"2006-12-12 04:43";"1970-01-01 01:00";"2007-05-24 12:03";"Shadow buffer auto-clipend is inaccurate on larger faces";"On even a simple example of a sphere on a plane, the auto-clipend fails quite obviously. See the attached image and blend file for a better description. I found I could make the problem go away by subdividing the plane a few times, but this is not an acceptable solution because: * Having the shadow working correctly or not should not be dependent on mesh topology. This is a bad thing from a philosophical perspective (my mesh topology should be the way I like it for my own considerations, such as ease of working with it, animation deformation, etc) and also from a practical perspective (trying to manage things like this if a scene involved animated lights moving over objects with changing topology/polygon density would be very painful) * The whole point of the auto clipping is for the software to make things easier and faster so artists can concentrate on art, not on massaging the software to make up for technical problems. Having to tweak meshes to make the auto clipping work goes back to the original situation again, and it's worse than tweaking the clipping manually anyway since it's destructive to the mesh.";"None";"None";"New" 5416;1;"Open";3;177;"Matt Ebb ";146;"Brecht Van Lommel ";"2006-12-12 04:57";"1970-01-01 01:00";"2008-05-09 20:20";"Solid view shows the wrong material colour, when overridden using nodes";"In solid view, the colour that is shown is a node tree's active material base colour, even when it is being overridden by connecting an RGB input into the Material's Col input. I am trying to make a range of materials that are tinted different colours, but sharing the same basic shading settings, so it's easier to tweak them all in one go. I can do this easily by just modifying the RGB input into a material node, but I can't see the results in solid view, making it difficult to find the objects that I'm looking for (each coloured differently). When a material node has an external colour input, conceptually the 'colour' of the material is no longer the data stored in the material block, but it is the input colour. The solid view should visualise this. test file attached";"None";"None";"Investigate" 5426;1;"Open";3;3914;"Erwan Jacq";146;"Brecht Van Lommel ";"2006-12-12 18:40";"1970-01-01 01:00";"2008-04-07 08:43";"Transp Drawextra option does'nt work in Textured mode";"All others drawextra options work both in solid and textured modes. But Transp doesn't. Transp is useful to show glass objects and textured ones in a same view. Tested in 2.42a and CVS (12 dec 2006) ";"None";"None";"New" 5428;1;"Open";3;3191;"Charles Wardlaw";103;"Ton Roosendaal";"2006-12-12 22:48";"1970-01-01 01:00";"2008-04-07 08:43";"button light intensity updates don't refresh Textured view";"I was trying interactive light placement, and I found that while you can move lights and have their position be updated in the realtime 3D display, changing the intensity (and possibly other values) has no effect unless you rechoose ""Textured Solid"" from the DKey menu. To repeat this issue: 1) Start a new scene, switch to Textured Solid view mode. 2) Note the brightness of the default cube, and change the light's intensity. Nothing changes. To work around this issue you can hit the DKey and reselect ""Textured Solid"" mode after changing light values.";"None";"None";"New" 5440;1;"Open";3;4793;"Mike Stramba";103;"Ton Roosendaal";"2006-12-14 20:20";"1970-01-01 01:00";"2007-05-24 12:03";"add toggle all bone layers hotkey";"The SH-M bone-layers hotkey (in CVS 2006-12-xx_ is great ! It would be very handy if the ""~"" or other key worked to toggle all bone layers on/off. Mike";"Interface";"None";"New" 5478;1;"Open";3;2923;"Tom Musgrove";146;"Brecht Van Lommel ";"2006-12-19 23:08";"1970-01-01 01:00";"2008-04-07 07:47";"stencils doesn't work with normalmaps or displacement maps";"It appears that stencils is bypassed (unintentionally?) for normalmaps and displacement maps - I've see this reported in forums a few times but I don't use any of the three features myself so can't create a nice test blend for ya ;)";"None";"None";"New" 5502;1;"Open";3;4301;"Ralf Hölzemer";4042;"Janne Karhu";"2006-12-23 15:54";"1970-01-01 01:00";"2008-04-07 07:47";"Particles - incorrect placement of particles on a subsurfed mesh";"When using a subsurface modifier on a mesh that is also a particle emitter, the new particle positions don't follow the smoothed surface. They are taking the new, slightly changed positions of the base mesh into account, but ignore all extra data produced by the modifier. See attached blendfile to reproduce.";"None";"None";"New" 5509;1;"Open";3;4301;"Ralf Hölzemer";103;"Ton Roosendaal";"2006-12-24 03:20";"1970-01-01 01:00";"2008-05-09 20:19";"Image Editor - wrong representation of alpha values";"I am not sure if this is really a bug. The main issue, however, is confusing. See attached blendfile. The machine i tested this runs on Ubuntu Linux 7.04 with an NVIDIA Quadro FX 3000 running on driver version 1.0-9631.";"None";"None";"New" 5536;1;"Open";3;3462;"Jędrzej Slewczuk";4827;"Nicholas Bishop";"2006-12-28 04:35";"1970-01-01 01:00";"2008-04-07 07:46";"Sculpt; Strange Shape keys";"I noticed strange behavior for Shape keys conjunction with Sculpt Mode and Object Mode (for sculpted mesh with Shape keys). Please change Shape keys in my BLEND file in Sculpt Mode and in Object Mode. On PNG file I show more detailes. My PC: WinXP SP1, 1GB RAM, AMD Athlon 1700+, ATI Radeon 9600 pro.";"None";"None";"Investigate" 5564;1;"Open";3;5443;"Daniel Seres";146;"Brecht Van Lommel ";"2006-12-31 11:03";"1970-01-01 01:00";"2008-04-07 07:46";"There is TraShadow. Why isn't there MirShadow too?";"Rays bounce off objects - thus we see objects. Rays - that bounce into the camera - bounce off mirrors (after they bounce off objects) - thus we see reflections of objects. Why don't rays - that don't bounce into the camera - bounce off mirrors? In other words: Why don't rays bounce off objects (after they bounce off mirrors)? In the picture: Why doesn't light from the lamp in the near end of the corridor bounce off the mirror half-way and light the far end of the corridor? In my opinion: This is because mirrors don't care about rays that come directly from the lamp. They only care about rays that bounce off objects.";"Render (internal)";"None";"New" 5594;1;"Open";3;5443;"Daniel Seres";4290;"Joshua Leung";"2007-01-04 09:23";"1970-01-01 01:00";"2008-05-09 20:18";""Clear Menu" in "Open Recent" and remove invalid entries when user clicks on them?";"It would be great to have a ""Clear Menu"" at the bottom of ""Open Recent"" in File Or at least to remove entries that no longer exist when the user clicks on them";"Interface";"None";"New" 5622;1;"Open";3;4754;"Allan Liddle";146;"Brecht Van Lommel ";"2007-01-06 21:19";"1970-01-01 01:00";"2008-04-07 07:46";"Transparent shadow does not support node shaders yet";"SYMPTOMS: When one generates a node-based material, which is UV-unwrapped, and one then adds a blend-texture, the resultant transparent shadow onto another object is incorrect: A) Using ONLY node-based materials, the rendered shadow onto another object is not transparent (red JPG image) B) If one adds another ""normal"" (NON-node-based) material, the transparency now shows up in the NODE-based material's render , but... B.1) Is misaligned due to an incorrect coordinate system (misaligned shadow in green jpg image), or B.2) Can be fixed if UV coordinates are selected for the NON-node-based material (aligned shadow in green jpg image) FILES: A) The file ""AlphaShadeTestUsingOnlyNodeMaterials.blend"" is the file that produces A above. No transparent shadow is produced on the white background plane, regardless of coordinate system. The rendered output of this file (RED material with NO transparent shadow) is shown in the file ""AlphaShadeTestUsingOnlyNodeMaterials.jpg"". B) The file ""AlphaShadeTestUsing_NodeMaterials+OldMaterials.blend"" is the same as as the previous blend file, except that a normal/old material has been added to the node-based material. To avoid confusion, I only changed the RED node-based material's colour to GREEN. I added the normal/old material (in BLUE) as well as its associated texture. B.1) In order to test the dependency of the node-based material's coordinate system on the NON-node-based material's, I left the NON-node-based material's coordinate system as ORCO (default as you open the file). Using the NODE-based material, it produced the transparent image ""AlphaShadeTestUsing_NodeMaterials+NormalMaterial_ORCO_mapped.jpg"". B.2) If one then changes the *NON*-node-mapped material's coordinate system to UV, the *NODE*-based material generates the correct transparent image ""AlphaShadeTestUsing_NodeMaterials+NormalMaterial_UV_mapped.jpg"". I tested in both Blender version 2.42a and 2.43RC1. This problem exists in both versions, but they do not produce 100% the same results. CONCLUSION: Unless I totally misunderstand the dependencies between NODE-based materials and NON-node-based materials, I believe the NODE-based material is referencing a NON-node-based material's coordinate system. SYSTEM: I am working on Windows XP, using latest release + a GForce 7600 GT graphics card, but I don't think is would matter.";"None";"None";"New" 5644;1;"Open";3;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2007-01-08 12:38";"1970-01-01 01:00";"2008-05-09 20:18";"Inconsistant: Texture channels count starting at zero and at one...";"WinXP, 2.43RC1, AtiX800 256mb In the particle option texture channels will be counted from 1 to 10 but in the IPO curves at material will be counted from 0 to 9. ";"None";"None";"New" 5645;1;"Open";3;980;"Alexander Ewering ";4246;"Peter Schlaile";"2007-01-08 13:43";"1970-01-01 01:00";"2008-04-07 07:45";"Inconsistencies in output file naming with FFMPEG";"Hi! There's appearently different output file naming going on depending on the Codec/format selected. When I export to an AVI, the ""Output"" filename is taken literally. But, if I change to MPEG2 output format, the output filename is taken as a *prefix* (just like with non-FFMPEG animation formats), and thus, having ""/tmp/foo.avi"" as the output filename (I know, makes no sense if I want an mpeg, I just left it untouched from the previous AVI I wrote), the resulting filename is ""/tmp/foo.avi0001_1780.dvd"". ";"None";"None";"Postponed" 5649;1;"Open";3;4635;"Thomas Knight";4246;"Peter Schlaile";"2007-01-08 22:12";"1970-01-01 01:00";"2008-04-07 07:45";"Thorough ffmpeg render formats test results...";"Here is a list of problems that I've found with blender243rc1 and recent CVS when using ffmpeg to render video files in various formats - sorry if some are not bugs (I don't really know about the format limitations of some formats etc...). Anyway, here goes: Problem: Formats affected: --------------------------------------------------------------------------------- Won't render at all: XVid, DV, H264, AVI[Xvid], AVI[DV], AVI[H264], Quicktime[XVid], Quicktime[DV], Quicktime[H264]. (these may just be format limitations) Won't play: Quicktime[HuffYUV]. First frame screwed up: Quicktime[MPEG-4], Quicktime[MPEG-2], Quicktime[MPEG-1], AVI[MPEG-4], AVI[MPEG-2], AVI[MPEG-1], MPEG-4, MPEG-2, MPEG-1. Blackout on last frame: Quicktime[MPEG-2], Quicktime[MPEG-1], AVI[MPEG-2], AVI[MPEG-1]. No problem: AVI[HuffYUV]. All tested with my original video testfile (included) using the blender video player - render size the same throughout. Problems can be clearly seen in ping-pong play mode... -epat.";"None";"None";"Postponed" 5691;1;"Open";3;2923;"Tom Musgrove";146;"Brecht Van Lommel ";"2007-01-12 10:56";"2007-01-12 17:21";"2008-04-07 07:45";"texture paint undo - undos one level then undo again 'redos'";"If you texture paint, do two strokes then undo twice - the second stroke in undone, then redone. ";"None";"None";"New" 5713;1;"Open";3;2160;"Klaus Brinkmann";103;"Ton Roosendaal";"2007-01-14 21:06";"1970-01-01 01:00";"2008-05-09 20:17";"Object-positioning issues when using text-ObFamilies";"Bug: When using the technique of applying objects to fontcharacters of an text-object by using the duplivert-feature, then blender does not correctly position the applied objects for characters having an unicode-ID of 144 and beyond. In order to test it yourself you need to do the following: - create a text-object - go to f9-menu and enter an ""Ob Family""-name. I will use the name ""Obfont"" in this example. - increase the textbox-dimension by raising ""Width""- and ""Heigth""-values to for example 30. - go to f7-menu while the text-object is selected and enable ""DupliVerts"". - create an mesh-plane and go to f9-menu - enter 'Obfonta' in the objectname-field. Explanation: This will apply the object to the character 'a' of the text-object, which uses 'Obfont' as 'Ob family'-name. - select the text-object, enter edit-mode (tab) and press 'a' several times. You should see a copy of the mesh-plane appear each. To this point everything is just as it should be. I had no problems while doing this for characters having an unicode-id below 144 (see the 'Char'-table in f9-menu while the text-object is selected). Problems start when i try to do the same thing for characters above ID 144. After assigning the mesh-object and typing the character several times. The first time the mesh-object will appear way off the position it should have - most of the times near the upper-left corner of the text-box, the second time it does position correct (or at least nearly correct, not sure), the third time it is way off again ... and so on. ";"None";"None";"Investigate" 5723;1;"Open";3;2923;"Tom Musgrove";4827;"Nicholas Bishop";"2007-01-15 21:14";"1970-01-01 01:00";"2008-04-07 07:44";"retopo - with multiple 3d views - paint strokes only shown and creatable in original view";"with multiple 3d views I'd expect the paint stroke to be shown after each stroke in the other 3d views as happens with verts. also could a selected paint stroke be deleted by xkey - i could have sworn this used to work...?";"None";"None";"New" 5724;1;"Open";3;4544;"Kai Kostack";103;"Ton Roosendaal";"2007-01-16 00:26";"1970-01-01 01:00";"2008-04-07 07:44";"multilayer images only save speed vector pass when ""do composite"" is enabled";"it seems that it's only the case with the speed vector pass and it results in an empty speed pass in the multilayer .exr file when composite is off. so i think it's not intended.";"None";"None";"New" 5747;1;"Open";3;5434;"Luca Bonavita";103;"Ton Roosendaal";"2007-01-18 18:39";"1970-01-01 01:00";"2008-04-07 07:42";"Composite nodes: flawed output from texture mapped with repeated mirrored tiles";"When a texture has repeated mirrored tiles, when applied on a 3D object it gives normal render. If the same texture is used in nodes, the Composite output of that texture shows a flaw when Xrepeat or Yrepeat are even numbers. In this blend file a texture has been applied on a cube and doesn't show problems In Composite nodes editor, if that same texture is connected to a Composite output: - with Xrepeat=2 the central column of pixels is not mirrored - adding Yrepeat=2 the central row of pixels won't be mirrored either - Xrepeat=4 makes 2 flawed columns, 6 makes 3 and so on (same for Yrepeat) - with odd numbers it seems fine This has been tested on two 64 bit Mandriva and a 32 bit Debian. Sorry atm I can't provide a 32 bit file (no problems have been reported here reading the file on a 32 Debian)";"None";"None";"Investigate" 5797;1;"Open";3;3487;"Campbell Barton";103;"Ton Roosendaal";"2007-01-23 10:27";"1970-01-01 01:00";"2008-04-07 07:41";"make local image does nothing";"pressing the make local button for a linked image does nothing.";"None";"None";"New" 5831;1;"Open";2;4948;"Thomas Brandt";3487;"Campbell Barton";"2007-01-25 19:32";"1970-01-01 01:00";"2008-04-07 07:40";"Update error in the UV/Image Editor";"I am using the blender 2.43-RC2-linux-glibc232-py24-i386 Build. My System: Ubuntu Linux Dapper Drake Nvidia Card Wacom Tablet ============================================================== There is an update Error in the UV/Image Editor. To get this fault you just 1. have to enter the UV/Image Editor 2. create a new image 3. enable ""texture paint mode"" and make some art ;o) 4. open View -> Properties 5. go to the ""Movie"" tab and back to the ""Generated"" tab 6. now the image is lost until you go to another image or delete the link to the datablock and assign it again. Just press the Movie tab in my ErrorFile and go back to still.. then you will see this tiny fault. ============================================================== The preview beside the Imagename does not update when i paint something and then save the Image, this would be a good function.. but it's secondary! ============================================================== Keep up the great work!";"None";"None";"New" 5833;1;"Open";3;4709;"Manuel Esguerra";4290;"Joshua Leung";"2007-01-25 21:27";"2007-02-05 02:30";"2008-04-07 07:39";"Transform Rotation: pose mode mess";"[Note: I am submitting this report on behalf of Philippe Roubal, who is having problems accessing the bugtracker - CD38] I have encountered a problem in an armature, using V2.43 RC2 and I cant acces to the bugtracker for an unknown reason (for several weeks), I have posted it on the forum, but could you check this problem and if no one has yet done it,submit it to the Bugtracker please? Here is my report : I get a wrong Bone Rotation when selecting the 3D cursor as rotation center. The selected bone doesn't rotates around the cursor. I have tried all modes (Local,Global, etc...), opening the file with 2.43.RC2 and 2.42a, the problem is the same with the two versions if the file has been created in V2.43 RC2, but if the file has been created with 2.42a it works normally in 2.42a and 2.43 RC2 as well. Here is a simple blend file made from scratch in V 2.43RC2, with only two bones : http://3d-synthesis.com/blender_miscellaneous/Bones_Bug.blend Is it a bug or have I missed something new in V2.43? Tested on Pentium 4 - 2GHz - 1.5 GB Ram. Nvidia Quadro FX1100 on AGP4x. Running Windows XP home Edition SP1. Is it a bug or something new in 2.43? Thanks in advance. Philippe.";"None";"None";"Approved" 5989;1;"Open";3;5659;"Michael Schierl";100;"Nobody";"2007-02-11 13:56";"2008-04-15 07:10";"2008-05-18 11:08";"2.43rc3: Problems with multiuser installation on Windows";"Since 2.42 is really unusable in a multi-user environment (see #4713), I tried if 2.43rc3 is better. That particular bug is fixed, but there are still some bugs left: 1. When I tell the installer to install user settings to appdata folder, it will install them to ""%CSIDL_PROFILE%\Application Data"" instead of ""%CSIDL_APPDATA%"". On a non-English version of Windows, the folder is named differently, which causes the Blender application data to appear in another folder which (1) is visible to the user even if application data is hidden and (2) is not backed up in a backup that includes application data only. See http://msdn.microsoft.com/library/default.asp?url=/library/en-us/shellcc/platform/Shell/reference/enums/csidl.asp for more information about system folders. 2. The start menu and desktop shortcuts always end up in the start menu folder of the user who installed the program, not in the all users start menu folder. This can be fixed manually, but is a nuisance if you have to do that on lots of PCs. 3. When another user (different from the one who installed the program) wants to save a file, it defaults to %CSIDL_WINDOWS% folder. %CSIDL_PERSONAL% would be a better default I guess. Tested on Windows XP Professional SP2 and Windows 2000 SP4.";"None";"None";"None" 5997;1;"Open";3;3487;"Campbell Barton";103;"Ton Roosendaal";"2007-02-12 05:06";"1970-01-01 01:00";"2008-05-09 20:15";"View Selected (dot key) doesn't work with deformed data";"Using the NumPad Del key (.) dosnt work when an armature has deformed verts. In face select and editmode, testfile has an example.";"None";"None";"New" 6001;1;"Open";3;5377;"Jan Devera";146;"Brecht Van Lommel ";"2007-02-12 18:01";"1970-01-01 01:00";"2008-04-07 07:38";"Mirror and subsurf ";"Hi, this is smaller display bug. It visiblle when with subsurf and mirror modifer (each face have 4 dots). I uploaded simple picture and blend";"None";"None";"None" 6041;1;"Open";3;5672;"mathieu auvray";146;"Brecht Van Lommel ";"2007-02-14 00:24";"1970-01-01 01:00";"2008-04-07 07:37";"DISPLACE modifier with animated texture";"When I put an animated texture (black and white noise on a tif or jpeg sequence) that I've made with another software, in the displace modifier, it works properly when I render it but in the viewport, I can't see any animation. The object does have some displace on it, but a still one, like my texture is a still frame... I've tried this on a linux and windows, with Blender 2.43 RC1 and RC3 on different computers. ";"None";"None";"New" 6048;1;"Open";3;4486;"Petar Velkovski";4246;"Peter Schlaile";"2007-02-15 03:08";"1970-01-01 01:00";"2008-04-07 07:37";"Video Sequence Editor bug in Image preview option";"Version: Blender RC3 OS: Windows XP SP2 GPU: Nvidia GeForce2 MX400 If Image preview is activated clicking on the displayed image changes the displayed frame (as if Sequence is still displayed). Same happens for Luma Waveform and Chroma Vectorspace display modes. A bit anoying and disorienting.";"None";"None";"New" 6050;1;"Open";3;5676;"Richard Vrijland";3487;"Campbell Barton";"2007-02-15 15:51";"1970-01-01 01:00";"2008-04-07 22:34";"Edge length doesn't update when scaling in object mode";"When scaling an object in object mode, the displayed edge length (activated within mesh tools 1) doesn't update. You can check this by entering edit mode after scaling in object mode, the length from before the scaling is still displayed.";"None";"None";"New" 6063;1;"Open";3;4457;"Tobias Regenbrecht";3709;"Andrea Weikert";"2007-02-17 17:59";"1970-01-01 01:00";"2008-04-07 07:36";"Filenames with special characters are not displayed correctly.";"If you have a filename with äöüß or another special character, it is not displayed correctly, whether you activate ""international font"" in user preferences or not. This error occurs at least since Blender Version 2.42, both in windows as in linux versions. Regards";"None";"None";"New" 6086;1;"Open";3;177;"Matt Ebb ";146;"Brecht Van Lommel ";"2007-02-22 00:16";"1970-01-01 01:00";"2008-05-09 20:15";"Deformed (modifed) curves don't deform other objects";"(reported on the blenderartists forums, I made the test file) It seems that the curve modifier only uses the original unmodified data for its deformations. If a deformed curve (eg. by an armature or wave modifier) is used to deform another object, only the location of the origial curve's CVs has any effect. This isn't the case for (eg) a lattice. You can deform a lattice with an armature and the deformed lattice will then be used to deform another object. See the attached test file for a better description.";"None";"None";"None" 6108;1;"Open";3;401;"Nathan Vegdahl";3487;"Campbell Barton";"2007-02-25 11:48";"1970-01-01 01:00";"2008-05-09 20:15";"Redraw appears to occur before scripts execute.";"This is a bug that makes python scripting pretty much useless for any sort of non-baked animation or rigging use. When you set a python script to execute on redraw, it appears that the redraw happens *before*, not after, the python script execution. This leaves the screen in a *not-up-to-date* state, since the python script modified the scene *after* the redraw. I don't know if that's actually what's going on at the code level, but it's what appears to be happening without having looked at the code. An example .blend file is attached. Load the blend file and enable script links. Grab the ball and drag it around. If the scene were kept properly up to date, the following behavior is what would (and should) happen: 1) The left side of the ball should always be *precisely* on the center line (the ball stretching in a volume-preserving way to accommodate). 2) The left side of the cube should always line up *precisely* with the right side of the ball. But instead, both of those effects lag behind by one or two pixels (or more, if you move the ball rapidly), and if you right click to abandon the grab operation the cube does *not* follow the sphere back to the original position. The .blend file also has a bit of animation on it. If you frame through the animation the effect of the bug is even more exaggerated. And what's worse, these errors are propagated into the render if you render the animation. When the script execution is set to happen on frame change, this bug occurs as well. I have verified this bug on both the Linux and Windows platforms.";"None";"None";"None" 6154;1;"Open";3;5779;"Steven Buscher";103;"Ton Roosendaal";"2007-03-03 15:25";"1970-01-01 01:00";"2008-04-07 07:36";"2.43: DupliGroup+Timeoffset doens't work for rendering";"Dear Blender Team, I found that DupliGroup+Timeoffset doens't work for rendering. In the 3D preview groups duplicated with a mesh or empty do show the effect of the TimeOffset, however the effect of the TimeOffset is not visible in the render. See also the attached .blend file (using Blender 2.43, Windows XP, ATI Radeon X600) Greetings, Uncia Unica";"None";"None";"Investigate" 6155;1;"Open";3;5779;"Steven Buscher";4290;"Joshua Leung";"2007-03-03 15:42";"1970-01-01 01:00";"2008-04-07 07:24";"Duplivert+Armature doesn't result in rendering of deformations as displayed in 3D preview.";"Dear Blender Team, I found that using Duplivert with a child mesh deformed by armature displays the expected deformations of the duplicated objects in the 3D view, but doesn't show the deformations in the render. See also the attached .blend file (using Blender 2.43, Windows XP, ATI Radeon X600) Greetings";"None";"None";"New" 6160;1;"Open";3;5781;"Mike Pan";146;"Brecht Van Lommel ";"2007-03-04 06:48";"1970-01-01 01:00";"2008-04-07 07:23";"automatic octree resolution setting";"since many user are not familiar with the octree resolution setting, is it possible to add a 5th option to the octree resolution which basically 'intelligently guess' what's the best octree resolution to use? i believe a simple case switch statements based on the total number of polygons in the scene (after rendertime mesh modifiers) can easily guess the best octree to use, no? also, if this is implemented, the default setting for the .blend file should of course be set to 'auto'. Thanks for your time.";"Render (internal)";"None";"New" 6162;1;"Open";3;5785;"Ziggy Spooner";103;"Ton Roosendaal";"2007-03-05 02:18";"1970-01-01 01:00";"2008-04-07 07:23";"DupliGroup Loop";"Looped DupliGroup references result in objects losing their location. ""Transform Properties"" still displays the correct location, but the objects won't render or display normally. Moving an object in the loop may allow viewing of one nested loop, but after the move command is ended the objects quickly 'fly' off the screen again. Example file included, it has three objects, but the error occurs with two as well. I didn't try with more than three. Found in Windows XP";"None";"None";"Investigate" 6207;1;"Open";3;980;"Alexander Ewering ";146;"Brecht Van Lommel ";"2007-03-11 20:32";"1970-01-01 01:00";"2008-04-07 07:22";"Fastshade, curve objects and texture space";" Curve objects do not take into account their texture space in shaded view. To reproduce: Add->Curve->Bezier Circle, Add material, Add texture (refresh a few times to avoid the update bugs ;-), select texture type ""Blend"". The preview is roughly accurate. Now, move texture space. Notice how nothing changes in the preview. Convert to mesh. Move texture space. Preview changes as expected. ";"None";"None";"New" 6237;1;"Open";3;4914;"Samuel Black";103;"Ton Roosendaal";"2007-03-13 23:34";"1970-01-01 01:00";"2008-04-07 07:22";"Objects Dissapear";"I can't reproduce this bug, but I attached a blend file that opens up to the bug. It causes a mesh with an armature modifier to disappear when zooming in. This has been quite annoying and I managed to trap the bug in this file so it could be examined by others. Thank you very much for all your hard work. ";"None";"None";"New" 6263;1;"Open";3;5855;"Michael Robinson";4246;"Peter Schlaile";"2007-03-14 22:28";"1970-01-01 01:00";"2008-04-07 07:22";"Button to disable video sequencer previews when moving a sequence.";"I have added a toggle button to the VSE header. When a user grabs a sequence, the button can turn off updates to the chroma vectorscope, and the luma waveform. After the user releases the sequence the windows will be updated. This can make the program much more responsive, especially if the project has a scene embedded in sequencer.";"None";"None";"None" 6295;1;"Open";3;2630;"Eckhard M. Jäger";103;"Ton Roosendaal";"2007-03-16 11:06";"2007-03-29 13:36";"2008-04-07 07:21";"Setting keys for procedural texture animations didn't work corect";"WinXP SP2, Ati X600, 2GB Ram, 3 GHZ P4, Blender 2.43 When you try inserting animation keys for the procedual textures shipping with Blender it only works correct with Voronoi. In other procedural textures keys for NoiseSize (NSize) and IScale will be ignored. It seems that there is a rendering problem too when animating NoiseSize and IScale no change of the procedural texture will be rendered :(";"None";"None";"New" 6366;1;"Open";3;4254;"Ari Häyrinen";7473;"Benoit Bolsee";"2007-03-19 17:03";"1970-01-01 01:00";"2008-05-11 21:09";"using getReactionForce() crashes engine ";" Blend file is attached. OS: Linux version: latest CVS + 2.43 Backtrace: blender: source/gameengine/Ketsji/KX_BulletPhysicsController.cpp:138: virtual MT_Vector3 KX_BulletPhysicsController::getReactionForce(): Assertion `0' failed. Program received signal SIGABRT, Aborted. [Switching to Thread -1220077904 (LWP 19503)] 0xffffe410 in __kernel_vsyscall () (gdb) bt #0 0xffffe410 in __kernel_vsyscall () #1 0xb78ed770 in raise () from /lib/tls/i686/cmov/libc.so.6 #2 0xb78eeef3 in abort () from /lib/tls/i686/cmov/libc.so.6 #3 0xb78e6dbb in __assert_fail () from /lib/tls/i686/cmov/libc.so.6 #4 0x0894aa9e in KX_BulletPhysicsController::getReactionForce (this=0x99d63c0) at source/gameengine/Ketsji/KX_BulletPhysicsController.cpp:138 #5 0x089071e6 in KX_GameObject::PyGetReactionForce (this=0x998b5c8, self=0x998b5cc, args=0xb743302c, kwds=0xbf9ad4a8) at source/gameengine/Ketsji/KX_GameObject.cpp:931 #6 0x0890a5fc in KX_GameObject::sPyGetReactionForce (self=0x998b5cc, args=0xb743302c, kwds=0xbf9ad4a8) at source/gameengine/Ketsji/KX_GameObject.h:598 #7 0xb7e35a0d in PyCFunction_Call () from /usr/lib/libpython2.4.so.1.0 #8 0xb7e7167d in PyEval_EvalFrame () from /usr/lib/libpython2.4.so.1.0 #9 0xb7e72a0a in PyEval_EvalCodeEx () from /usr/lib/libpython2.4.so.1.0 #10 0xb7e72a83 in PyEval_EvalCode () from /usr/lib/libpython2.4.so.1.0 #11 0x08963506 in SCA_PythonController::Trigger (this=0x99d7818, logicmgr=0x984a1f8) at source/gameengine/GameLogic/SCA_PythonController.cpp:300 #12 0x08957d66 in SCA_LogicManager::BeginFrame (this=0x984a1f8, curtime=1.1841409399174161, fixedtime=0.016666666666666666) at source/gameengine/GameLogic/SCA_LogicManager.cpp:237 #13 0x0892b293 in KX_Scene::LogicBeginFrame (this=0x9849d38, ---Type to continue, or q to quit--- curtime=1.1841409399174161) at source/gameengine/Ketsji/KX_Scene.cpp:1082 #14 0x089119f1 in KX_KetsjiEngine::NextFrame (this=0x9847b20) at source/gameengine/Ketsji/KX_KetsjiEngine.cpp:431 #15 0x088b0c1d in StartKetsjiShell (area=0x9875b80, scenename=0x9938732 ""Scene"", maggie1=0x986c818, sipo=0x0, always_use_expand_framing=1) at source/gameengine/BlenderRoutines/BL_KetsjiEmbedStart.cpp:369 #16 0x083369b5 in start_game () at source/blender/src/space.c:524 #17 0x0833b808 in winqreadview3dspace (sa=0x9875b80, spacedata=0x9875e98, evt=0xbf9adec2) at source/blender/src/space.c:2108 #18 0x0834c1a0 in scrarea_do_winhandle (area=0x9875b80, evt=0xbf9adec2) at source/blender/src/spacetypes.c:146 #19 0x082516df in scrarea_dispatch_events (sa=0x9875b80) at source/blender/src/editscreen.c:593 #20 0x08252f91 in screen_dispatch_events () at source/blender/src/editscreen.c:1198 #21 0x082539b5 in screenmain () at source/blender/src/editscreen.c:1477 #22 0x0820c57c in main (argc=1, argv=0xbf9ae1c4) at source/creator/creator.c:715 ";"None";"None";"None" 6399;1;"Open";3;5248;"Trevor Harris";146;"Brecht Van Lommel ";"2007-03-22 02:25";"1970-01-01 01:00";"2008-04-07 07:21";"Sky Color appears along bottom edge when rendering";"Whenever I render an image, the sky color shows through. I set the color to red so I could see it easier. It only happens when I use oversampling, though it does not happen with box or CatRom mode of oversampling. It is just a red 1 pixel wide line running across the bottom of the render, semitransparent. When I increase the resolution, the line stays 1 pixel wide, so I think it is a glitch with the renderer.";"None";"None";"New" 6410;1;"Open";3;4317;"gsr b3d";103;"Ton Roosendaal";"2007-03-22 17:17";"1970-01-01 01:00";"2008-04-07 07:20";"RGB Curves node in Compositor draws Extrapolate/Horizontal inconsistently";"In Compositor, add one RGB Curves (maybe other have similar issues if the ""curve widget"" is reused) and zoom out (+- controls in the node), check that Horizontal draws the start and end parts fully to the edge, while Extrapolate does not, just a small line. Both should do the same, and preferably, both should behave like Horizontal and all IPO curves, that is draw until it reaches the edge of the view.";"None";"None";"New" 6428;1;"Open";3;5894;"Susanne H.";146;"Brecht Van Lommel ";"2007-03-23 12:16";"1970-01-01 01:00";"2008-04-07 07:20";"No smooth halo gradient in pano render";"When I use a halo with a large halo size in a pano render (e.g. to simulate a sun), there's a visible step between the normal background and the halo area. Expected: smooth gradient. This happens with the official Blender binary from blender.org as well as with my self compiled 64Bit version (I know it's not safe yet). The attached blend is saved with the 32Bit Blender version. CPU: AMD Athlon64 X2 3800+ RAM: 2048 MB GFX: NVidia GeForce 7600 GT, 256MB OS: Kubuntu Linux 6.06.1 LTS 64Bit Drivers: Nvidia 1.0-8762 Blender: 2.43 32Bit from Blender.org / 64Bit self compiled Thanks, Sanne";"None";"None";"New" 6459;1;"Open";3;5162;"Markus Kothe";100;"Nobody";"2007-03-27 13:47";"1970-01-01 01:00";"2007-05-24 12:03";"Boolean in boolean bug";"I don't actualy know if this is supposed to work, but I thought it should. When you assign a boolean modifier to a object that is used in another boolean modifier, the first modifier doesn't takes in to count. (I have a realy hard time trying express myself clearly right now) I attach a very clear file with what I meen. //Markus Kothe";"None";"None";"New" 6500;1;"Open";3;4709;"Manuel Esguerra";146;"Brecht Van Lommel ";"2007-04-06 20:28";"1970-01-01 01:00";"2008-04-07 06:51";"Suzanne preview makes Texface materials render black";"UV-map an image to an unwrapped mesh, assign a material and turn on Texface. Activate the Suzanne preview in the Materials panel. Material renders black. No other preview mode seems to do this. Also, only quitting Blender and restarting seems to correct the problem; ie if you Ctrl-X to create a new file, and repeat the process on a ""new"" object, it will render black *even if you don't activate the monkey preview.* Discussed briefly in this blenderartists thread http://blenderartists.org/forum/showthread.php?t=93179 Version: Bf-Blender 2.43 OS: Tested on OSX 10.3.3 and WinXP Card: ATI Rage 128 (Mac), GeForce FX 5500 (PC) ";"None";"None";"New" 6540;1;"Open";3;4679;"Juho Vepsalainen";146;"Brecht Van Lommel ";"2007-04-10 08:36";"1970-01-01 01:00";"2008-04-07 06:51";"Build modifier before curve modifier causes the object to jump";"See the provided .blend. It has a suzanne mesh object that has build and curve modifiers on it. Notice what it does when you animate (alt-a) it.";"None";"None";"New" 6544;1;"Open";3;3818;"Alessandro Pirovano";103;"Ton Roosendaal";"2007-04-10 15:30";"1970-01-01 01:00";"2008-04-07 06:50";"Omly files with .blend extension can be used as library... make it using magic number";"Hi from some investigations it seems that the new lib interface (through bpy modules) works only with filenames ending with "".blend"" extension. I know this could not be considered strictly a bug (begging pardon in advance for this) but I'm in trouble switching to the new API cause I've a lot of library files whose name do not end with "".blend"" extension. They are distributed as a part of a VR application where a script-customized blender is one of the modules. Is it possible to relax this constraint on file names and test validity of library files examining the file format (the usual ""magic"" chars) ? Thank you in advance regards Alessandro Pirovano ";"None";"None";"Investigate" 6563;1;"Open";3;3487;"Campbell Barton";103;"Ton Roosendaal";"2007-04-14 16:15";"1970-01-01 01:00";"2008-04-07 06:50";"DupliGroup depends on last scene";"With Groups, the dupligroup layer is undefined when using more objects in more then 1 scene. Since the object's layer is scene dependant With the blend file, go to. Scene, then Scene.002 == 4 planes in the dupli. Scene.001, then Scene.002 == 3 planes in the dupli. The display of a scene should not depend on the last used scene. ";"None";"None";"New" 6619;1;"Open";3;5365;"Roger Wickes";4246;"Peter Schlaile";"2007-04-28 17:05";"1970-01-01 01:00";"2008-04-07 06:50";"VSE 4-digit filename bugs emerge with five-digit strips";"Windows 2.43 graphicall build sse2 from a few weeks ago. I am compositing video tutorials with > 9999 frames. I render out the captured avi to a frame sequence set, and load that set into VSE as an image sequence set. When I started working on a set that has 16900 images, in the VSE, Last: is still 9999.jpg (even though the strip is 16900) and Cur: (also displayed below channel 0) as you arrow right: 0999.jpg, 1000.jpg, 10000.jpg, 10001.jpg, 10002...10009, 1001, 10010, 10011, etc. I seem to remember Ton saying somewhere that some fixes were put in, so this report may be redundant with something already open (but I could not find it in the bug list). WORKAROUND: split up image sequnces into separate subdirectories that each have <9999 images (8 minuted video strips). ";"None";"None";"New" 6622;1;"Open";3;3439;"Ramon Carlos Ruiz";100;"Nobody";"2007-04-28 23:20";"1970-01-01 01:00";"2008-09-18 19:00";"Fast Caustics";"Hi, well I did this small path that do an approximation for caustics. The use is very simple. -Activate the FakeCaustic in the lamp. -In the material I put 3 extra parameters for control. CaustFact, take care of the concentration style of the caustic. CaustMult, it multiply in a global way the caustic effect. CaustInt, control the amount of refraction rays that will be use to calculate the approximation, if 0 will try to do as much interactions as the distance of the shi to the lamp allow. In A practical way I will say that you use a lamp with very small energy then increes the values of the caustmult to big values like using the causfact between 0 and 1. Best Regards, Carlos ";"None";"None";"None" 6638;1;"Open";3;3914;"Erwan Jacq";3709;"Andrea Weikert";"2007-05-02 05:45";"1970-01-01 01:00";"2008-04-07 06:49";"Desktop environment: aliased paths not suported";"I'm using linux. Let say there a blend file at /home/me/3d/test.blend Let say I put a shortcut called ""3d"" on my desktop to go directly to the /home/me/3d/ folder. When I open the folder from my desktop and then open test.blend; the path for Blender is /home/me/Desktop/3d/test.blend The problem is: Blender cannot open files (images, textures, etc...) which are in /home/me/textures/ for example. I guess it's because use related paths so ../textures goes to /home/me/Desktop/textures and not /home/me/textures/ It's really annoying as you must make sure you start blender from the physical path. ";"None";"None";"New" 6678;1;"Open";3;5199;"Michael Crawford";4042;"Janne Karhu";"2007-05-10 11:33";"1970-01-01 01:00";"2008-04-07 06:48";"particle collision and dupliverts";"Blend file: http://homepage.mac.com/psyborgue/.Public/colision.blend.zip problem duplication: open file, play animation (no collision) turn on dupliverts (red dots, no colission) unparent object ""blood"" from object ""plane"" (the emitter) colission works! Blend file: http://homepage.mac.com/psyborgue/.Public/colision.blend.zip Irrelevant note: If you're wondering what it's about... Among other things: vengeance causing violence without end, looking at things from the ""enemy""'s perspective... etc... It's not quite done yet animation wise. PS: it's meant to be played on loop.. first frame same as end. ";"None";"None";"New" 6679;1;"Open";3;5199;"Michael Crawford";103;"Ton Roosendaal";"2007-05-10 11:37";"1970-01-01 01:00";"2008-04-07 06:48";"action of 0.07 frame (multiplied in NLA) does not work for motion blur";"http://homepage.mac.com/psyborgue/.Public/colision.blend.zip 3d mblur seems to be broken. Rendering at frame 72 produces a result for the gun hand nowhere near that of the viewport (far different than either previous or next frame). If mblur is off, it works fine and the hand is in the correct position. This might be an issue with inter-frame interpolation of nla actions, but all other actions seem to work fine. System details current cvs (as of a day or so ago), OSX Tiger, PPC, Python 2.3";"None";"None";"New" 6681;1;"Open";3;4679;"Juho Vepsalainen";113;"Kent Mein";"2007-05-10 16:16";"1970-01-01 01:00";"2007-07-05 22:35";"Corrupt jpeg causes system to hang up on compositor";"Loading provided image to Image input node in compositor caused Blender to hang up on my system (Ubuntu Feisty Fawn). A backtrace about this has been provided. An user reported that it segfaults on his system in case he sets the image as a background of camera view but I could repeat this behaviour here. The image showed up just fine. It worked on the UV editor just fine on his Linux based system. It does so here also. It was hinted also that this sort of behaviour might be present also with Targa (tga) file format so that might be worth checking out as well. Backtrace: [Thread -1318589552 (LWP 19604) exited] [New Thread -1318589552 (LWP 19605)] Corrupt JPEG data: premature end of data segment Program received signal SIGINT, Interrupt. [Switching to Thread -1218820400 (LWP 19586)] 0xffffe410 in __kernel_vsyscall () (gdb) bt #0 0xffffe410 in __kernel_vsyscall () #1 0xb78b7e66 in nanosleep () from /lib/tls/i686/cmov/libc.so.6 #2 0xb78f0c3c in usleep () from /lib/tls/i686/cmov/libc.so.6 #3 0x086387f8 in PIL_sleep_ms (ms=50) at source/blender/blenlib/intern/time.c:102 #4 0x085b01f8 in ntreeCompositExecTree (ntree=0x8d77590, rd=0x8d76c2c, do_preview=1) at source/blender/blenkernel/intern/node.c:2100 #5 0x0830174c in snode_handle_recalc (snode=0x8ceb448) at source/blender/src/editnode.c:160 #6 0x08301914 in load_node_image ( str=0xbfd13eb8 ""/home/juho/Desktop/200408_greatwall_juyong.jpg"") at source/blender/src/editnode.c:209 #7 0x083b77e9 in filesel_execute (sfile=0x8cebdb0) at source/blender/src/filesel.c:1689 #8 0x083b7c9b in do_filesel_buttons (event=5, sfile=0x8cebdb0) at source/blender/src/filesel.c:1758 #9 0x083b8952 in winqreadfilespace (sa=0x8ceb130, spacedata=0x8cebdb0, evt=0xbfd142a2) at source/blender/src/filesel.c:2009 #10 0x0827ac98 in scrarea_do_winhandle (area=0x8ceb130, evt=0xbfd142a2) at source/blender/src/spacetypes.c:146 #11 0x0812831b in scrarea_dispatch_events (sa=0x8ceb130) at source/blender/src/editscreen.c:593 #12 0x08129e5c in screen_dispatch_events () ---Type to continue, or q to quit--- at source/blender/src/editscreen.c:1223 #13 0x0812a880 in screenmain () at source/blender/src/editscreen.c:1502 #14 0x08126fdc in main (argc=1, argv=0xbfd145a4) at source/creator/creator.c:694 ";"None";"None";"New" 6689;1;"Open";3;3496;"Pascal Tempier";100;"Nobody";"2007-05-13 16:31";"1970-01-01 01:00";"2008-09-21 00:19";"Memory usage increase when roatating the viewport";"Linux pascalou 2.6.20-gentoo-r3 #5 SMP Wed Mar 21 21:40:35 CET 2007 i686 AMD Athlon(tm) 64 X2 Dual Core Processor 4600+ AuthenticAMD GNU/Linux Issue happen in version 2.43 , 2.44rc, do not happen in 2.42, 2.42a i think the issue is only noticable with huges scenes with a lot of texture orogor: there were some changes with texture cache cleaning, it doesn't cleanup memory was well as it did (IIRC). That is, it doesn't remove texture from mem once they stop being used i can not trace it specifically to texture usage, but ""fell"" like it s related like, if the texture has been used onbly one, even switching the wireframe, somhow keep the texture in use and provoke the issue theeth, would it be that rotating the viewport, make a change in the mipmap in use and that then the mip map would not be freed as well ? and that then because in what i said earlier, rottating the viewport even in wireframe view would triger the issue i noticed than changing th emip map option whas the only yhing which could reduce memory usage but yeah, the mipmap wouldn't be freed for the same reasons that the image aren't that is, freeing all of them and then reallocating them as needed also an odd thing, memory do not increase slowly, it ioncrease by steps, hene at the begin i thougth it was autosave option, the undo options or whatever but then i noticed it was only increasing when i was rotating viewport or zooming ";"None";"None";"None" 6710;1;"Open";3;4353;"Amir Taaki";4290;"Joshua Leung";"2007-05-16 19:47";"1970-01-01 01:00";"2008-04-07 06:47";"dodgy rig that doesn't obey copy constraint for a certain pose";"move the foot ik along the y axis a little left and right and you'll see that the foot should have a copy constraint on the leg but it comes unstuck.";"None";"None";"None" 6732;1;"Open";3;5069;"Gavin Stewart";146;"Brecht Van Lommel ";"2007-05-22 00:17";"1970-01-01 01:00";"2008-05-09 20:11";"Image based texture maps not seen correctly in ray-traced reflections.";"If an image is used as a bump map, it is not seen in ray-traced reflections in other objects when OSA is enabled. Disable OSA and the bump map is seen in the reflection OK. Use a procedural texture as a bump map with OSA, and the reflection is OK. This problem appears in both 2.44 and 2.43 (win32 and linux), but is not a problem in 2.42a See attached .blend for a simple example. (Try toggling OSA on and off). Cheers, gavin. ";"None";"None";"None" 6742;1;"Open";3;4545;"David Millán Escrivá";100;"Nobody";"2007-05-24 09:14";"1970-01-01 01:00";"2008-09-18 19:06";"Save and load view3d space properties";"It's a basic feature, Spanish people request me it. The video can explain the feature better: http://www.artresnet.com/david/save_view.ogg You can save some properties of view3d to load it in same or other 3d view. ";"None";"None";"None" 6745;1;"Open";3;324;"Matthew Plough";3709;"Andrea Weikert";"2007-05-25 08:30";"1970-01-01 01:00";"2008-04-07 06:47";"Incorrect relative path handling with -b command";"Category: System related (command line interface) When relative paths to the desired .blend file to act upon are used on the command line, Blender does not handle internal relative paths to dynamically linked assets correctly. One must specify the *absolute* path to the desired file. The following commands do not work when x.blend uses assets dynamically linked from other files using relative paths. $ blender x.blend $ blender -b x.blend These two commands are unable to find filenames with spaces in them. They result in error output of the form: read library: lib //file with spaces.blend ERROR: can't find lib /file with spaces.blend LIB ERROR: can't find MEAsset Notice that Blender looks in the root directory, /. ------------ $ blender ./x.blend $ blender -b ./x.blend $ blender ../x.blend $ blender -b ../x.blend These four commands fail on all files, and produce output of the form: read library: lib //y.blend ERROR: can't find lib /y.blend read library: lib //file with spaces.blend ERROR: can't find lib /file with spaces.blend LIB ERROR: can't find MEAsset1 LIB ERROR: can't find MEAsset2 The following commands do work: $ blender ~/x.blend $ blender -b ~/x.blend $ blender /Absolute/Path/To/x.blend $ blender -b /Absolute/Path/To/x.blend This is out of line with every Unix application; Blender should be smart enough to look relative to the current working directory if an absolute path is not specified. References: ================ - [1] http://projects.blender.org/tracker/?group_id=9&atid=125&func=detail&aid=6680 - [2] http://projects.blender.org/tracker/?group_id=9&atid=125&func=detail&aid=4021 ";"None";"None";"None" 6746;1;"Open";3;324;"Matthew Plough";100;"Nobody";"2007-05-25 08:56";"1970-01-01 01:00";"2012-10-29 12:27";"Titlebar and state when opening a .blend file on Mac by double click or drag and drop";"Steps to reproduce: 1. Start Blender, and open a .blend file using the Open dialog (F1). 2. Notice that the .blend file's name is set in the window title. 3. Close Blender. Double click on the .blend file in the interface. 4. Notice that Blender's title bar contains only ""Blender"". 5. Close Blender. Place an icon for Blender in the dock. 6. Drag the .blend file to the dock icon, thereby opening Blender. 7. Notice that Blender's title bar contains only ""Blender"". In states (4) and (7), the current file name is known, but one cannot link a file using relative paths until the .blend file is saved. ";"None";"None";"None" 6752;1;"Open";3;6205;"Tom Bech";100;"Nobody";"2007-05-26 14:18";"1970-01-01 01:00";"2008-09-18 19:10";"Full Cook-Torrance model";"Hi! I'm new to Blender, so bear with me :) I noticed that Blenders Cook-Torrance (CT) specular shader isn't really a ""true"" CT BRDF model as it lacks it's microfacet base. I'm not sure why it's designed this way, but sometimes it's useful to be able to specify the ""standard"" set of parameters for a CT model, for example when trying to reproduce the results from tutorials for other programs or from a research paper. I've written a ""full"" CT model (see diff's attached) for those who might be interested in this kind of thing. It's a pretty trivial piece of code, and I'm not even sure if it's of interest to anyone else but me :) I don't know Blenders code well at all - for example, where does the export to yafray occur? I've only modified the shader panel and the file DNA and the shader code, but I'd like to be able to export the ""full"" settings to yafray as well. Tom";"None";"None";"None" 6754;1;"Open";3;3817;"Dolf Veenvliet";100;"Nobody";"2007-05-26 16:40";"1970-01-01 01:00";"2008-09-17 19:56";"zooming by holding ctrl + draggin with mmb is inverted";"Right simple enough. If you roll your mousewheel up in the 3d view you zoom in, the same happens if you hold crtl and roll up in other windows. When zooming by holding the ctrl key and dragging your mouse up with the middle mouse button held in 3d view you zoom out. But doing exactly the same in a buttons/nodes window it zooms in. This seems very unnatural. Here's a list in which I test what happens when I drag up to zoom: Scripts window: no zoom File browser: no zoom Image browser: no zoom Node editor: zooms in Buttons window: zooms in Outliner (regular): no zoom Outliner (Oops): zooms in User preferences: no zoom Text editor: no zoom Audio window: doesn't seem to like drag-zoom at all (unreliable results)? Timeline: same as Audio window Video sequence editor: zooms in UV/Image editor: zooms out NLA editor: same as Audio window Action editor: same as Audio window Ipo curve editor: zooms in 3D view: zooms out So in the end it appears that the 3d view and UV/Image editor are the inverted ones, since scrolling up and dragging up give the opposing results in those, whilst most everywhere else up is always the same, whether your drag or roll. I don't think many people notice this, but since I have no scroll wheel on my tablet is was feeling a bit strange.";"None";"None";"None" 6757;1;"Open";3;6207;"Christian Ordig";100;"Nobody";"2007-05-26 20:12";"1970-01-01 01:00";"2008-07-20 11:35";"linked objects made ""local"" still not editable";"Blender version and plattforms: Blender 2.44 on MacOS (PPC) and Windows. Problem: when linking an object from an external blend file it is not editable - so far it's correct. When making the object ""local"" so the mesh and the materials etc. are still linked, but the object is local, you still cannot change anything belonging to the object. (it's impossible to move it, add game logic bricks etc.). Blender says I cannot modify externally linked material. As said above ... I am not trying to modify anything belonging to the object's mesh or material which intentionally is still linked, only things belonging to the object itself - which is made local. In Blender 2.43 this still works as expected. Cheers ";"None";"None";"None" 6784;1;"Open";3;5365;"Roger Wickes";3487;"Campbell Barton";"2007-06-01 00:30";"1970-01-01 01:00";"2008-04-07 06:35";"Textured Viewport Mode shows Vertex over UV";"An opaque UV Texture should overlay any vertex paint. If the 3D View is in Shaded Draw Type, it shows correctly. However, if it is in Textured viewport shading Draw Type, the Vertex Color appears on top of the UV Texture. In the attached file, which opens in Textured mode, shows the top of the toilet tank as black. Switch to Shaded view, and it is cream colored. Delete the Vertex Color layer, and Textured mode is correct. It also crashed my display driver while messing around with it, and response is noticeably slow in Textured mode, for example changing frames.";"Interface";"None";"None" 6821;1;"Open";3;6260;"Daniel Lima";100;"Nobody";"2007-06-06 19:18";"1970-01-01 01:00";"2007-09-04 22:04";"Audio Wav only plays on Windows Version";"Hi, I'm using Blender 2.44 on a Linux Ubuntu 7.04 and the sound does not play in the sequencer. The funny thing is, I install the ""windows"" blender 2.43 and run via WINE HQ (a kind of emulator...) And the same sound file plays normally. * CPU: pentium III * RAM: 784 MB * GFX: NVidia geforce 4 64MB * OS: Ubuntu 7.04 * Blender: 2.44";"None";"None";"None" 6827;1;"Open";3;1993;"Chris M ";4246;"Peter Schlaile";"2007-06-08 14:33";"1970-01-01 01:00";"2008-05-09 20:11";"Framenumbers hardly visible in sequence editor on grabbing";"Hello, Nothing major, but a little irritating when grabbing video strips in the sequence window a lot. The frame numbers of the begin and end of the sequence strips are shown under the sequence strips itself (at least if the visible part of the strip is smaller the the video source. I attached a screenshot for clarification. I do not know if it is my videocard or hardware, or just a Blender thing. Athlon 64 X2 4400+ 1024MB nVidia Geforce 9700GTX Ubuntu 7.04 Feisty Fawn (32 bits) nVidia 1.0.9755 driver Blender 2.44 Regards, Chris ";"None";"None";"None" 6835;1;"Open";3;6279;"dan williams";4042;"Janne Karhu";"2007-06-11 02:17";"1970-01-01 01:00";"2008-04-07 06:33";"Static particles life one don't render";"When a static particle has life of one and vert enabled it dosn't render. I am using the latest Blender version under Win XP and a Nividia GeForce 6 (intergrated)";"None";"None";"None" 6845;1;"Open";3;3129;"ILdar AKHmetgaleev";146;"Brecht Van Lommel ";"2007-06-14 15:07";"1970-01-01 01:00";"2008-04-07 06:33";"Memory overflow with viewport's background image";"Blender's cache grows up when I am playing long and large tga image sequence (720x576 1188 frames) in viewport background. After 400 frames cache takes about 1Gb and my system kills blender. Please add more efficient cache management for viewport background image.";"None";"None";"None" 6858;1;"Open";3;6315;"Armin Zingler";146;"Brecht Van Lommel ";"2007-06-18 00:46";"1970-01-01 01:00";"2008-05-09 20:10";"Lattice deform: Deformation affects unrelated objects in 3D view";"WinXP/SP2, nVidia 6800, Blender 2.44 Steps to reproduce: 1. Open the file attached (test6.blend) 2. Press Tab to enter edit mode 3. Press g key, then move the mouse to move the selected points. While doing this, in 3D view, you see *both* spheres being deformed. Press LMB to apply the deformation. Problem: In step 3, only the right sphere must be shown deformed. Note: Both sphere objects use the same NURBS sphere, though the left sphere must not be deformed in 3D view because the left sphere is linked to a different lattice that is not deformed. 4. Press F12 to render it. You will see that only the right sphere has been deformed - as expected. This proves that it is only a display problem in 3D view. Additional information: The wrong display in 3D view would not appear if the NURBS sphere was a Mesh. So, select the sphere and press Alt+C to transform it to a Mesh. Repeat step 3 and you see the correct behavior now.";"None";"None";"None" 6865;1;"Open";3;4679;"Juho Vepsalainen";4290;"Joshua Leung";"2007-06-20 13:56";"1970-01-01 01:00";"2008-04-07 06:32";"Cyclic dependency; Curve hook missing update";"See provided .blend. Move (g key) the selected empty (bend.down) and cancel moving. Notice curve status. Try same procedure on upper empty (one with parent). It works as expected in that case. SVN Blender (06-20-07) Kubuntu Feisty Fawn ATI Mobility Radeon 9700";"None";"None";"None" 6872;1;"Open";3;6337;"rey nae";100;"Nobody";"2007-06-22 13:18";"1970-01-01 01:00";"2012-10-29 12:27";"OSX + mouse keys + blender = loss of mouse";"MacOSX 10.4.9 Intel (thus dualcore) No separate graphics card (it's a macmini) 1. First of all, go into System Preferences > System > Universal Access > Mouse and turn on Mouse Keys. 2. Fire up blender. If you now use the numpad, your mouse moves. Good 3. Now hit 0 on the numpad. The mouse (at least mine..) stops working (It MOVES, but click has no effect). Asking me why would I hit 0? Simple, when using size. Now, to get you out of trouble without needing to reboot: Quit blender (apple+q), go into system preferences, and disable mouses keys (using the numpad...). Now your mouse should work again, else try switching on/off again..";"None";"None";"None" 6930;1;"Open";3;4679;"Juho Vepsalainen";146;"Brecht Van Lommel ";"2007-07-10 18:53";"1970-01-01 01:00";"2008-05-09 20:09";"Lattice + Copy Location -> Missing update on cancel";"Open the provided .blend, move the sphere and cancel moving. Note what it looks like after cancelling. SVN Blender (07-10-07) Kubuntu (Feisty) ATI MR 9700";"None";"None";"None" 6938;1;"Open";3;6433;"Panos Christopoulos-Charitos";3318;"Robert Holcomb";"2007-07-12 22:06";"1970-01-01 01:00";"2008-05-09 20:09";"MD2 exporter doesn't export animation";"The current MD2 exporter (0.18.1) doesn't export animations. The steps... 1) I imported a standard working md2 model 2) and then I tried to (re)export it The problem is that there is no animation exported, only the current shape key. I've tried to export it from blender 2.42 and it worked so probably it is a bug. OS: WinXP SP2 VGA: ATI 9200 mobile Blender ver: 2.44 PS: I've attached a .blend with the md2 imported model";"None";"None";"None" 6942;1;"Open";3;6438;"Peter Kaufman";3487;"Campbell Barton";"2007-07-13 21:40";"1970-01-01 01:00";"2008-04-07 06:30";"import of multypart VRML 1.0 files reference points lost";"As I tried to import a vrml 1.0 file into Blender something went wrong. The parts are all over the place. They should be joined. The orange hook, the pink part the helicoil, the yellow part and the cyan part all should be aligned the big hole in the purple part. The small red and green screws should be in the smaller holes of the purple part. The parts them selves are displayed correct, its just their place that is wrong. As I made the 3D model I know some parts reference points are related to holes, surfaces and points of other parts. I think this referencing is taken into the vrml file. As the viewers I have seen never displayed the displacement of parts this is a Blender problem. I include the ""gebit.wrl"" file here so you can see for yourself. The model was build in and exported to VRML with ""radan"" http://www.planit.com/uk/radan/ or http://www.somatech.nl/Producten/Plaatwerk/Radan/tabid/233/Default.aspx ";"None";"None";"None" 6957;1;"Open";3;6456;"Gary Ritchie";100;"Nobody";"2007-07-19 05:00";"1970-01-01 01:00";"2007-09-05 03:52";"UV Map node seams";"""Single edge pixel that does not map the corresponding image XY pixel; it chooses the first (last) pixel for like that row only. This results in the smeared edge (again only 1 pixel wide) where the UV starts/ends."" You can follow some discussion here: http://blenderartists.org/forum/showthread.php?t=89670 BF 2.44 SUSE 10.2 GeForce 6600, driver: 1.0-9631 Reproduce by mapping an image with the UV Map node -- with the goal of a seamless transition. There's two scenes in the sample file that illustrate the UV Map node seams whether OSA is active or not.";"None";"None";"None" 6968;1;"Open";3;3129;"ILdar AKHmetgaleev";4290;"Joshua Leung";"2007-07-21 20:07";"1970-01-01 01:00";"2008-04-07 06:19";"script link and pose mode";"script link attached to armature object to redraw event doesn't executes when I transform bones in pose mode. But I think it should.";"None";"None";"None" 6973;1;"Open";3;3002;"Joseph Eagar";103;"Ton Roosendaal";"2007-07-24 18:01";"1970-01-01 01:00";"2008-04-07 06:18";"Cannot use multilayer EXR sequences in either the sequencer or with anim playback";"For some reason, multilayer EXR sequences can't be used either with anim playback or in the sequencer, yet they can be used in the compositor or loaded as individual images in the UV image editor. The follow error message is printed in the console: Error: can't process EXR multilayer file Unknown fileformat couldn't open C:\blender\Furry\RenderImages\RenderImages0001.exr Joe";"None";"None";"None" 6979;1;"Open";3;2276;"pat buergi";100;"Nobody";"2007-07-26 11:04";"1970-01-01 01:00";"2008-06-06 11:51";"Active layer becomes invalid";"while playing a little with the array modifier in blender i discovered a strange behavior in local view. i have 2 3dviews open one in local view one in global view. if one inserts an object in global view its of course not visible in local view. but if an object is added in local view i can't see it in neither of the 2 windows although you should see it in both of them. the only problem is that i couldn't reproduce it from scratch, but i saved the blender file directly after i discovered the bug. with this blendfile i can reproduce the behavior in any blender version (tried 2.25-2.44) on WinXP and openSuse 10.2. tried to reproduce from scratch but i wasn't successful, perhaps someone else is. sry again for missing instructions for reproducing, but i guess its anyway better posting this than just keeping still. buergi btw you could remove all objects from the blenderfile and the bug is still reproducible.";"None";"None";"None" 7008;1;"Open";3;6530;"Jean-Luc Stevens";146;"Brecht Van Lommel ";"2007-08-05 00:33";"1970-01-01 01:00";"2008-04-07 06:17";"Large, black, triangular artefact rendered on plane.";" Upon rendering, a large triangular black area appears across the diagonal of the plane in both the render window and the rendered file. Originally noticed on a texture plane - however, no textures needed for the problem to occur. The problem manifests itself in the .blend provided. Easy to replicate using 2.43, 2.44 and 2.45 (rc1) on the two Intel Pentinum 4 machines I have to hand. Bug confirmed as reproducible by *one* other user on IRC #blendercoders but several other users reported no problems. This bug may therefore be hard to pin down. The fix is to subdivide the plane suggesting to me that the issue may be to do with the viewed size of the polygon. Perhaps this is a know limitation of the render engine? Hope that helps. Koba ";"None";"None";"None" 7058;1;"Open";3;6614;"Heiko Harders";3709;"Andrea Weikert";"2007-08-16 13:07";"1970-01-01 01:00";"2008-04-07 06:17";"Custom Application Data location causes crash when opening/saving files";"I'm using a small domain to log on (Samba on Linux and Windows XP/Linux clients). Therefore I have my Windows 'Application Data' directory redirected to the samba server and I have a custom location for the Windows 'Application Data'. Whenever I try to do anything which has to do with files (open, save, export) Blender closes imediatly. During the blender installation I have chosen to use the 'WinXP/Win2K Application Data directory' for saving user related files. The crash is always reproduceable, I've tried it on two machines (both Windows XP service pack 2, Blender 2.44). When I'm using the administrator account everything runs fine. I guess the problem is my custom Application Data location (the Administrator account uses the default location), in my registry I've set the following location: HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\User Shell Folders\AppData = %LOGONSERVER%\userdata\%USERNAME%\AppData In a windows command prompt, when I give the command 'echo %AppData%' this gives: \\Thor\userdata\tdummy\AppData (where Thor is the logonserver and tdummy is the username). This directory does exist and the userrights are also ok. There are no other programs which cause problems with this. Blender does create a folder in the default location: C:\Documents and Settings\tdummy\Application Data\Blender Foundation But this doesn't seem to help. The folder is empty (except for the 'Blender' folder which is also empty). In my domain setup I've also excluded the Application Data folder from copying into the userprofile (because its already redirected), I the following registry key for that: HKEY_CURRENT_USER\Software\Microsoft\Windows NT\CurrentVersion\Winlogon\ExcludeProfileDirs (with 'Application Data' among one of the values). And besides that, all my roaming cache is deleted after logout (so even if Blender would succeed to save anything in the default AppData folder, it would be deleted after logout and it would not be copied into the userprofile, so the data is lost). I've also read this (which I think is the cause for the problem): http://projects.blender.org/tracker/index.php?func=detail&aid=1664&group_id=9&atid=264 And this (bottom of the page): http://www.blender.org/development/coding-guides/installation-policy/ But I think this is a problem in any environment which uses domain logons with folder redirection, so I think this issue should be taken seriously.";"None";"None";"None" 7112;1;"Open";3;6675;"Yuri Shchedov";100;"Nobody";"2007-08-26 13:34";"1970-01-01 01:00";"2007-09-02 15:20";"Boolean tool is too memory hungry";"hi, I have a mesh with 200.000 more faces. When I'm trying to use boolean operations on this mesh, Blender allocates 1.5Gm of memory or more and goes to swap. Blender 2.44 (official release) Linux Debian Sarge 3.1 Pentium D 2.8, 1Gb RAM, Nvidia 7600gt Yuri ";"None";"None";"None" 7136;1;"Open";3;4462;"Ron Taulbee";146;"Brecht Van Lommel ";"2007-08-29 01:59";"1970-01-01 01:00";"2008-04-07 06:16";"Halo Material with HaloTex only uses the first texture slot";"Reference #1481 This has been around a while. I can see a need for multiple textures on halo materials, as I can see a need for multiple halo materials (ref #7117) From this thread: http://blenderartists.org/forum/showthread.php?t=104075 I created a basic cube, and assigned a halo material with HaloTex. I then created a default cloud texture, in the second texture slot, mapped to color only (default shocking pink). Render, and I don't get any color - see pic 1. I then copy to the buffer, clear the texture, then paste from the buffer to the first slot & I get color. It seems as though halo materials will not recognize any color or even texture in anything but the first slot. I have verified this behavior in 2.43 & 2.42A System Specs: Windows XP Pro 2ghz 2gb ram nvidia quadro FX Go1000";"None";"None";"None" 7316;1;"Open";3;6675;"Yuri Shchedov";146;"Brecht Van Lommel ";"2007-09-08 11:04";"1970-01-01 01:00";"2008-04-07 06:15";"3d view window doesn't update during texture changes";"hi, I'm experiencing a problem duiring mesh texturing. What I'm doing 1. Adding Image texture to a mesh 2. Switching from Solid to Shaded mode 3. Making changes of Brightness/Colors/Contrast of textures or some Map image properties No changes in 3D view window. The model still looks like it was after loading a texture image. Looks like no way to update the screen. Blender 2.44 Debian Sarge linux 2.6.22 Pentium D 2.8 1Gb RAM Thanks, Yuri ";"Render (internal)";"None";"New" 7381;1;"Open";3;4836;"Ondrej Novak";4290;"Joshua Leung";"2007-09-20 23:18";"1970-01-01 01:00";"2008-04-07 06:14";"Can't animate when using render preview";"Hello, using render preview and armature animation at the same time doesn't work. Render preview is probably using the keyframed data and not the actual data, what's not very much of a problem. However when the redraw of the preview is finished, the actual data gets dumped and the scene returns to the keyframed state. Therefore animating and render preview at the same time is impossible. Steps to reproduce 1. Create an armature with two bones 2. Go to pose mode, create LocPosRot keyframe 3. Move to another frame 4. Split window in two. 5. Shift+P to open render preview 6. Move one bone 7. Render preview rerenders. 8. Click the other bone and voila the first bone returns to the keyframed position. Attached is very simple scene with first 5 steps. ";"None";"None";"None" 7421;1;"Open";3;4679;"Juho Vepsalainen";146;"Brecht Van Lommel ";"2007-09-27 18:30";"1970-01-01 01:00";"2008-04-07 06:14";"Mat and Light options found in Render Layers panel do not work for preview";"Open provided .blend. Note how preview render (shift-p in 3D viewport) differs from real render. This is due to Mat and Light settings found on Buttons Window -> Render Layers panel.";"None";"None";"None" 7422;1;"Open";3;4679;"Juho Vepsalainen";146;"Brecht Van Lommel ";"2007-09-27 18:34";"1970-01-01 01:00";"2008-04-07 06:13";"Border render does not work for animation";"See attached .blend. Note that border render works as it should only for the first frame in case you render animation.";"None";"None";"None" 7426;1;"Open";3;4882;"Keir Mierle";100;"Nobody";"2007-09-28 08:14";"1970-01-01 01:00";"2008-09-19 14:22";"Add a World Coordinate vector render layer, AND a camera matrix output node";"This patch adds a vector pass (like the normal pass) which produces at each (x,y) pixel location the (X,Y,Z) world coordinates of the corresponding scene point. This is extremely useful for evaluating the performance of my reconstruction work; however, I imagine it is useful to others in the rendering pipeline. When output to a file via OpenEXR, it produces XYZW = RGBA, where W/A indicating a point at infinity, i.e. no scene visible at that point. The patch also adds a node which outputs the view and window matrix. I access that data by also outputting it to an OpenEXR image (4x4 pixels!). I haven't found the 'industry standard' way of doing this so if you know please tell and I'll do that instead of this. See http://wiki.blender.org/index.php/Image:Ztoworld_example.png for an example of what the world coordinate layer looks like. It's obviously not intended to be visualized directly but nevertheless. ";"None";"None";"None" 7446;1;"Open";3;4908;"Alberto Torres Ruiz";139;"Martin Poirier";"2007-10-02 11:00";"1970-01-01 01:00";"2008-05-09 20:03";"Cursor -> Selection is useless in most curve points.";"Select a point of a curve with handles with different lengths or asymetric. Shift+S, Cursor -> Selection. It's impossible to snap the cursor to the point because it always takes in account the handles.";"None";"None";"None" 7487;1;"Open";3;6844;"Scott Wilson";103;"Ton Roosendaal";"2007-10-08 08:06";"1970-01-01 01:00";"2008-04-07 06:13";"Diagonal Line in orthographic rendering";"Computer system: OS: Ubuntu 7.04 64 bit CPU: AMD64 Dual Core 2.4 ghz RAM: 2 gigs Blender 2.45 Steps: 1) Create a new scene 2) Delete the cube, and add a plane. 3) Change the camera into an orthographic camera and line it up so the plane is completely facing the camera. 4) Turn off OSA and render. You should see a diagonal line on the plane.";"None";"None";"None" 7517;1;"Open";3;6859;"Roger Frye";100;"Nobody";"2007-10-13 00:51";"1970-01-01 01:00";"2008-05-24 15:42";"'blenderplayer -p' doesn't pass focus to window";"I run blenderplayer V 2.45 on OSX(Intel) with command line ./blenderplayer -p 0 0 1200 750 KF8x6.blend This opens the game in a window and the game runs except that the player window does not respond to mouse clicks or keyboard. By contrast ./blenderplayer -f fb 1 KF8x6.blend opens the game in full screen mode, and the player window does respond. Note that V 2.44 for OSX Intel didn't work at all, but V 2.44 for OSX PowerPC running on an Intel Mac had no problems. I have not checked the 2.45 PowerPC version. ";"None";"None";"None" 7532;1;"Open";3;3155;"Doc Holiday";7473;"Benoit Bolsee";"2007-10-15 01:59";"1970-01-01 01:00";"2008-05-15 10:51";"TrackTo Actuator does not work correctly if it has a Vertex Parent.";"Hello I've noticed, the the TrackTo Actuator does not work correctly, if the tracking Objekt has a Vertex Parent. I hope the added File explain the Problem. Also try this: Make the Base Object regular Parent to the Target, and see what happens. The tracking Object now tracks the Startposition of the Target. Thanks for helping. Doc Blender 2.45 Windows XP Sp2 GeForce 8600 GT";"None";"None";"None" 7618;1;"Open";3;6890;"Nicolas Wu ";100;"Nobody";"2007-10-21 03:27";"1970-01-01 01:00";"2008-09-19 14:44";"Threshold node";"Hi. This is a patch for a composite threshold node with smoothness control. Screenshots: (without smoothness) http://blender.astrometry.net/attachment/ticket/30/threshold.2.JPG (with smoothness) http://blender.astrometry.net/attachment/ticket/30/threshold1.JPG ";"None";"None";"None" 7653;1;"Open";3;3487;"Campbell Barton";4246;"Peter Schlaile";"2007-10-25 11:07";"1970-01-01 01:00";"2009-06-14 13:13";"Sequencer prefetch cache - not working with effects";"When using prefetch cache, the effects IPO's are not evaluated based on their ipo/time Here is a test file... Open the test file, set the prefetch frame setting to 500 and the memory limit to 1024. Press Alt+A... The final render will not animate. ReOpen the scene and set the prefetch to 0. Press Alt+A and the transform effect works properly";"None";"None";"None" 7669;1;"Open";3;6877;"Niels Krook";4290;"Joshua Leung";"2007-10-27 08:40";"2007-10-27 09:52";"2008-04-07 06:11";"AR:Bones(/including constraints) update-lag ANIM and 3Dview";"I uploaded a file and included a info-Text. Points: - Bones don't update onscreen in 3Dview - Render with ANIM, will render the wrong bone-positions Niels ps. The RENDER-Button, does render with the correct bone-positions and also updates the 3Dview correctly afterwards... For the record: Although I use 2 armatures in this file, I noticed with more 'extensive rigs', the same problem occurs with 'just one rig'... File is also at download here: http://niels.degooier.net/blender_scripts/bugreport_motorJob.zip";"Tools";"None";"None" 7699;1;"Open";3;4679;"Juho Vepsalainen";103;"Ton Roosendaal";"2007-10-31 12:09";"1970-01-01 01:00";"2008-05-09 20:02";"Dumping screen and cancelling the operation gives memory error";"1. Dump Screen (ctrl-shift-f3). 2. Cancel the operation. This should print memory error ""Error Totblock: 1 dumprect len: 3795648 0xb1343028"" on the console. It seems that on the cancel case the memory doesn't get freed as it should. Tested this on current SVN on Kubuntu (Feisty Fawn).";"None";"None";"None" 7708;1;"Open";3;6877;"Niels Krook";4246;"Peter Schlaile";"2007-11-01 18:02";"1970-01-01 01:00";"2008-05-09 20:02";"Blender VideoPlayer without Audio?";"The Blender VideoPlayer playback is without audio... Reasons why there should be audio? - Lipsync - Dancing - 'Normal' walking/events is related to 'background' music Thanks, Niels ps. I doubt if audio was ever intergrated in the player, so it's really a feature than, but from the animators perspective a ... :)";"None";"None";"None" 7763;1;"Open";3;4679;"Juho Vepsalainen";3709;"Andrea Weikert";"2007-11-08 14:56";"1970-01-01 01:00";"2008-05-09 20:01";"Accessing appendable option after writing to directory field works weird";"1. Go to ""Append or Link"" 2. Find a .blend and access it. 3. Write something to the directory field (example: .../myblend.blend/foobar/). 4. Access some of the appendable properties seen below directory field. Note that this just appends the property to previous entry (example: .../myblend.blend/foobar/Material). I just wanted to check if this is the expected behaviour. Perhaps it would be better to erase bad entry (3.) or to change the way part four works.";"None";"None";"None" 7771;1;"Open";3;3132;"Daniel Salazar";103;"Ton Roosendaal";"2007-11-11 17:02";"1970-01-01 01:00";"2008-04-07 05:59";"Glitch in vector blur with Z Transparency";"Details in Blend file, thanks";"None";"None";"None" 7772;1;"Open";3;3132;"Daniel Salazar";103;"Ton Roosendaal";"2007-11-11 17:17";"1970-01-01 01:00";"2008-05-09 19:58";"Group dupli time-offset fails with original group visible";"There are two objects, an IPO animated cube that belongs to a group and an empty duplicating that group. The duplygroup object (empty) has a time offset in animation buttons. Try and select the Cube and next the Empty and then back to the Cube using the outliner and see how the Cube jumps from one location to another. Select the Cube using the outliner, the object is freezed, you can not transform using G for grab, rotate and scale do work fine";"None";"None";"None" 7819;1;"Open";3;6785;"Marco Frisan";100;"Nobody";"2007-11-21 15:45";"1970-01-01 01:00";"2012-10-29 12:27";"Mac OS X Tiger, 3D View, v-sinc not enabled (tearing problem)";"I just bought a Mac Pro, 2x Intel Quad Core 3Gh, 4GB RAM, ATI Radeon X1900, PCIe 16x, 512MB video RAM, monitor LCD, Apple Cinema Display, 1680 x 1050, 32 bit color depth, accelerated Hardware, Quartz Extreme enabled. I noticed that when I ""pan"" (Shift+MMB) objects, dispayed in 3D View, flicker as like as v-sync was not enabled. The problem is visible in perspective projection (NUMPAD 5) and all other projections, ortogonal (=NUMPAD 1, 3 and 7) and isometric (=rotating view with MMB). Already tested other 3D OpenGL applications like ""World of Warcraft"" and everything is stable. Ironically, under Linux (Nvidia GeForce 7300 GS with Nvidia Drivers), desktop and 2D applications have this problem, and I noticed it with several Linux+HW combinations; while Blender and OpenGL applications are stable (v-sync activated). Under Mac OS X, everything else is stable except for Blender. Anyway I think it is specific of my configuration. In a school lab where I worked last year, equiped with 3 eMac, 1 PPC dual G5 + Nvidia GeForce 5500 and an iMac Intel of the 2006 release (that has an ATI Radeon inside) I never had this problem. Anyway I am not sure of this and I don't work there anymore so I cannot test it.";"None";"None";"None" 7821;1;"Open";3;1890;"Roland Hess ";146;"Brecht Van Lommel ";"2007-11-22 18:12";"1970-01-01 01:00";"2008-04-07 05:46";"Mesh Deform gives bad results on full cage rotation.";"When rotating an entire mesh deform cage, the deformed mesh changes shape even though the relationship between meshdef faces does not. In the example .blend file is a beauty mesh of a dog on layer 1, a bound cage mesh on layer 11 and an armature that drivers the cage mesh only. Upon rotating the armature 180 degrees in pose mode, the cage rotates properly, maintaining its shape as you would expect, but the beauty mesh distorts. A mesh consisting of two quads has been added to the scene and made the child of the head bone, to show the original bounds of the cage mesh and positioning of the eyes. The effect can be seen by rotating around the z axis of the selected bone in the sample file. As for the attached images, they show the face stretching horizontally when the entire cage has been rotated. Image ""non-def.png"" shows the dog's face in the rest position. Image ""180baddef.png"" shows the same thing with a 180 degree rotation. Note in the second image how the beauty mesh has stretched, even though the cage mesh, the reference quads and the eyes (which are bone-parented and don't use meshdef) remain in the correct position. The effect is the same when the cage mesh is a simple cube around the head, or tightly follows its contours as in the sample file.";"Tools";"None";"None" 7843;1;"Open";3;5323;"Yorik van Havre";3709;"Andrea Weikert";"2007-11-28 01:59";"1970-01-01 01:00";"2008-03-04 22:48";"writing \\server\share paths don't work in file selector";"Well I'm really not sure if this can be considered a bug or not, you guys will know that better... Writing a pathname like \\server\share in windows in the file selector doesn't work, blender asks if I want to create a new directory. But if you double-click on a .blend file located on \\server\share in the windows explorer, blender opens, and in the file selector it appears \\server\share correctly. So it seems to me I should be able to specify \\server\share as a path for saving or retrieving files? Well, sorry to bother if this is not a bug... Cheers to all Yorik";"None";"None";"None" 7870;1;"Open";3;7112;"Дмитрий Васильев";100;"Nobody";"2007-12-04 20:38";"1970-01-01 01:00";"2008-04-18 03:13";"russian filenames and texts are broken";"System: Ubuntu 7.10 [AMD64] My russian filenames looks strange in Blender's Flile Browser. Texts loaded in Text Editor looks the same way. Also I can't type anything in russian, including 3d text :(. When I trying to type 'й' (on the same key with 'q') in Text Editor ""Quit Blender OK?"" appears. All russian menus and buttons looks correct. ";"None";"None";"None" 7887;1;"Open";3;3087;"Kenneth Strom";4042;"Janne Karhu";"2007-12-08 01:55";"1970-01-01 01:00";"2008-04-16 03:58";"Particles don't project from objects";" Updated svn earlier today. Using particles for hair seems ok, however, using particles for particles doesn't work. Most of the various attempts I made ended up with barely a few particles on the object to project them, even when the variables were turned up to 200. This being a problem, I thought to try an old version of a blend with particles in it. http://img178.imageshack.us/img178/5382/alienarraytest003is4.jpg Using the old system the above gave a nice effect, but I tried the blend with the new system and see that the new system doesn't project particles. http://img152.imageshack.us/img152/4979/guidenewparticlesyp8.png Normally, I can figure out how to get some effect from, but there is no recalc button for particles so it is hard to tell if there is any effect or not. Current test was set at frame 2500 (due to lack of negative starting frame), with start of 1 and end of 5000. There are maybe 2 dozen particles on the surface of the sphere. If I switch from emittor to hair, I get this huge thing of hair, but when I switch back I still don't get and particle projection. ";"None";"None";"Postponed" 7889;1;"Open";3;7130;"Christian Monfort";100;"Nobody";"2007-12-08 20:37";"1970-01-01 01:00";"2008-09-19 15:41";"Basic support for #define in makesdna";"Here is a patch that brings basic support for #define's in makesdna - I say basic, because there are some limitations: * Only ""full numeric"" #define are processed: #define OK_VALUE 32 /* is ok: only digits */ #define KO_VALUE 1024*4 /* is wrong: character other than digit */ #define KO_VALUE2 OK_VALUE /* is wrong: ref. to other define not allowed */ * define's are only replaced in array definitions #define ARRAY_SIZE 32 struct mystruct { ... char myarray[ARRAY_SIZE]; /* will be replaced with ""char myarray[32]"" ... } * multiple #define of the same tag with different values drops the tag: if you have in one file: #define MY_TAG 32 then later (in same file or in other one) #define MY_TAG 64 then the value of MY_TAG is marked invalid and MY_TAG will not be replaced * #ifdef and #undef are not taken into account, so something like the following will result in MY_TAG being marked invalid (this should not harm a lot, as array sizes in DNA should not depend of any #ifdef...) #ifdef SOMETHING #define MY_TAG 32 #else #define MY_TAG 64 /* MY_TAG seen twice with different values => invalid */ #endif - or - #define MY_TAG 32 ... #ifdef MY_TAG #undef MY_TAG #define MY_TAG 64 /* MY_TAG seen twice with different values => invalid */ #endif I also added a test in makesdna to check that array size are >= 0, and make the program fail if this is not the case... This to prevent something like the following not being detected by makesdna and producing zero length array... char array1[UNKNOWN_TAG]; /* UNKNOWN_TAG not being defined / invalid */ char array2[16*2]; etc... For now I only modified VS2003 project (I dont think Scons needs any change), and tested under Windows (MSVC only, as I failed to build with Scons under mingw - I'm checking my configuration as I never had any problem with 2.43). I only added a few #define, for ""proof of concept"": NAME_SZ in struct ID (field name) NAME_SZ in struct Sequence (field name) FILE_MAX in struct Library (fields name & filename) FILE_MAXFILE in struct StripElem (field name) FILE_MAXDIR in struct Strip (field dir) ... and could removed the double definitions of FILE_MAXDIR, FILE_MAXFILE and FILE_MAX from DNA_space_types.h and BKE_utildefines.h (moved them to new file DNA_nameSize.h) My ultimate goal is to be able to increase the size of ID's in Blender (21 chars is a bit short). This could also make the code easier to read, if one see something like NAME_SZ instead of integer values like 24 or 21... ---- I know this is not perfect; I wanted to use the preprocessor to handle all that stuff in a clean way: * create a new file DNA_prepro.c, which just includes all DNA_*.h files * preprocess DNA_prepro.c (with all project flags) to obtain a DNA_prepro.i file * change ""char *includefiles[]"" array in makesdna.c so that it contains only DNA_prepro.i (and only #include ""DNA_prepro.i"" at end of makesna.c file) * ... and let all the code in makesdna.c just the way it is now (except for testing array sizes). I know how to do this whith make (and I thing I can deal with Scons too, with some reading/googleing) but my main concern is MSCV projects; I was not able to make a ""preprocess only"" project in Visual Studio 2003 (I find a way in VS2005, by making an "".exe"" project, and preventing the linker to launch, but I don't know how to do the same in VS2003). So I rolled back to my ""second solution"" that I presented here... ";"None";"None";"None" 7895;1;"Open";3;639;"bastian salmela";3487;"Campbell Barton";"2007-12-10 14:24";"1970-01-01 01:00";"2008-04-07 05:45";"solid tex -view option turned on, you can't see wireframe";" if you are in solid mode, and view has ""solid tex"" turned on, blender doesnt draw wireframe in editmode. in vertex mode, you see dots, but not lines, in edge/face mode, you dont see anything.. except the one that that currently selected. running on linux, nvidia drivers (geforce 7600) .b ";"None";"None";"None" 7910;1;"Open";3;177;"Matt Ebb ";177;"Matt Ebb ";"2007-12-13 06:07";"1970-01-01 01:00";"2008-05-09 19:55";"Incorrect refraction behaviour";"I'm going to take a look at the code for this myself, but thought I'd make a report in case anyone else has any ideas too. The refraction raytracing in Blender seems to have a few problems. I'm trying to do the seemingly basic task of rendering liquid in a glass, but it's not working out. I made a test file and image that describes the problem pretty well, it's attached. Most renderers work correctly with the setup as in the rightmost glass - the liquid mesh should be overlapping the glass mesh. I had a look at the refraction code and it seems like there are a few potential problems, please correct me if I'm wrong: * Only the IOR of the material the ray is entering is considered, not the IOR of the material it's exiting. This would presumably give incorrect results for the interface between two materials with IOR > 1 * The total internal reflection seems to be a binary value, either reflecting or refracting. AFAIK, what should really happen is that the ray should split into two, with varying amounts of reflection and refraction, depending on the fresnel. * The normal gets flipped for odd depths, regardless of the direction that the normal is facing - so if there are two solids overlapping, the normal of the second 'internal' solid gets flipped, even if it's pointing at the camera. This seems to be assuming that all refraction rays will enter a solid and then exit that solid (to air) on the next bounce, which is incorrect. * Something else seems wrong with the filter (absorption) values too. if you put a heavily absorbing object behind a non-absorbing object (not necessarily overlapping), rays through the front, clear object will not pick up the changes in colour as they should...";"None";"None";"None" 7996;1;"Open";3;7015;"Gunnar Stahl";100;"Nobody";"2007-12-22 22:41";"1970-01-01 01:00";"2012-10-29 12:27";"Trouble with blender on leopard (macbook)";"Hi, I've installed blender on my macbook (intel core2 duo, 1,83 GHz, 3GB Ram. Graphics is an Intel GMA 950). The same blender worked fine on OS X 10.4 (tiger) on the same macbook. On Leopard blender is barely usable. After clicking on a button in the button nothing happes for about 3 seconds. When I move the cursor in this time the click activates the button which is under the cursor after the move. Adding a new sphere to a new scene takes again about 3 seconds to finish. Pressing ""s"" for scale - the same. Splitting the viewport - the same. Are there some workarounds for this? Yt, Gunnar ";"None";"None";"None" 8022;1;"Open";3;6623;"Aaron Digulla";139;"Martin Poirier";"2007-12-28 23:03";"1970-01-01 01:00";"2008-05-09 19:53";"Some operations on control points can introduce weird twists in paths";"1. Open the accompanied file 2. Select the strange loop 3. Edit mode 4. Move the point next to the text ""Move this point"" left and right You'll see some really strange twisting happening near the end point of the curve.";"None";"None";"None" 8054;1;"Open";3;5498;"Stefan Wilke";100;"Nobody";"2008-01-05 20:03";"1970-01-01 01:00";"2008-09-17 20:11";"New Point Light visualisation for raytraced soft shadows";"For me it looks like the ""soft size"" virtually extends the size of the point lamp. So for example if i set the ""soft size"" at 5.0 the point light becomes 5 blender units big. The problem you can get is obvious: if you create a wall 2 blender units away from the point light, a decent percentage of the point light emitters ending up on the other side of the wall, resulting in raytracing artefacts and lightning errors. I attached three pictures where the point light is 1 blender unit away from the wall. the first picture has a soft size 2.0 with 5 samples the second picture has a soft size 5.0 with 5 samples the third picture has a soft size 10.0 with 5 samples I think because thats a more or less expected behavior of the renderer, a outlined ball in the interface arround the point light could indicate the true point light size, similar to the sphere lightning outline visualisation.";"None";"None";"None" 8057;1;"Open";3;3837;"Mats Holmberg";146;"Brecht Van Lommel ";"2008-01-06 17:22";"1970-01-01 01:00";"2008-04-07 05:58";"Armature DOF doesn't work when using pole targets";"Rotation isn't limited for bones in an IK chain when using the new pole target option. File attached to demonstrate. -mats";"None";"None";"None" 8075;1;"Open";3;3132;"Daniel Salazar";100;"Nobody";"2008-01-10 18:56";"1970-01-01 01:00";"2008-09-09 19:50";"PET + armature deformed mesh";"Hi, the area of influence of PET is working on the original, non transformed position of vertices, even if the transform of modifiers such as armature is applied to editing cage";"None";"None";"None" 8077;1;"Open";3;7230;"Neil Kolban";100;"Nobody";"2008-01-11 03:14";"1970-01-01 01:00";"2008-09-09 19:39";"Composition Time Node not affecting the brightness";"I have created a composite node solution that should change the brightness level of the resulting image over time. Unfortunately, when I change frames, the resulting image does not change its brightness level. I asked some folks on the #blender IRC to also test and everyone else had the same problem and no-one could think of a reason and hence the bug submission. I am attaching a screen shot showing the composite nodes as well as the corresponding .blend. To see the issue ... Set the frame number to be 1 and render. Then set the farme number to be 250 and render. Notice that there is no change yet frame #1 should have a brightness fac of 100 while frame #250 should have a brightness face of 0.";"None";"None";"None" 8097;1;"Open";3;7316;"Patrick PIERRE";4246;"Peter Schlaile";"2008-01-16 22:25";"1970-01-01 01:00";"2008-03-02 16:39";"VSE: bad quality in a transform effect with scale greater than 1";"I have a big image (2592x1944) and I want to zoom in and move. A very simple manipulation is to use VSE with a transform effect. But when I zoom in (scale=3.4) I obtain an ugly result (blender_vse_quality02.png). I think Blender use a resized image to process the zoom and not the original one. However, I have a very good quality with the node editor. But, it needs a complicated manipulation for a so simple effect. Configuration: Mandriva Linux 2008.0 ; blender SVN 13261 ; AMD 4800 ; NVidia 6800";"None";"None";"None" 8129;1;"Open";3;6147;"David B.";146;"Brecht Van Lommel ";"2008-01-21 00:49";"1970-01-01 01:00";"2008-04-07 06:03";"SVN : Particles and Halo aren't correctly rendered";"There is a problem when ""halo"" material is assigned to particules : - add a plane and a particules system on it - increase ""normal"" value - add a new material and hit f12, render is ok (see picture 1.jpg) - activate halo button and change HaloSize to 0.100 - Hit f12, render is ok with halos (see picture 2.jpg) - now, disable halo button and hit f12 : the halos are always there (picture 3.jpg) while the render should look like picture 1 with halo disabled. Cache problem ? Note : When I made these tests, I had a lot of crash of blender. Either by making the render, or during the setup of the halo. I didn't report this separtly because that happen randomly... That happen especially when the test scene was not saved. (And please, sorry for my english... :/)";"None";"None";"None" 8142;1;"Open";3;177;"Matt Ebb ";146;"Brecht Van Lommel ";"2008-01-23 06:33";"2008-01-30 14:39";"2008-04-07 06:02";"Strand rendered hairs don't get speed vectors when in front of geometry";"Although it goes fine when in front of nothing (alpha 0), speed vectors don't seem to be getting generated properly for strand-rendered strands, when in front of geometry. Turning off the strand renderer goes fine. Check the attached demo file and image for info. Sorry if this is a known issue/being worked on, I couldn't find anything on it in the docs. cheers ";"None";"None";"None" 8144;1;"Open";3;1819;"Carlos Lopez ";146;"Brecht Van Lommel ";"2008-01-23 16:23";"1970-01-01 01:00";"2008-04-07 06:01";"Baking ignores textures when 'Clear' is active";"1-Setup a scene, assign a material with image texture to an object. 2-In Bake panel, active 'Clear'. 3-Make either a 'Texture' or 'Full Render' baking. I get just a lighting solution of the baking, i.e., the material color and lights. No textures have in count. 4-Disable 'Clear', then reload the image and press 'Bake' again. It works well.";"None";"None";"None" 8161;1;"Open";3;7450;"Brendan Bolles";103;"Ton Roosendaal";"2008-01-26 21:57";"1970-01-01 01:00";"2008-04-07 05:44";"MultiLayer EXR files being saved negative Y stride without lineOrder tag";"Images rendered in Blender 2.45 using the MultiLayer (OpenEXR) format are saved in opposite scan-line order. This would be OK if the file was properly tagged DECREASING_Y, but it is not. See commented-out line 446 of openexr_api.cpp: /* header.lineOrder() = DECREASING_Y; this crashes in windows for file read! */ When Blender reads a MultiLayer EXR that it rendered (it looks for the ""Blender MultiChannel"" attribute), it just assumes this line order. But when these files are opened in other programs, they appear upside down because the OpenEXR library tags files as INCREASING_Y unless told otherwise. This can be verified by viewing the files in ILM's exrdisplay utility or with the ProEXR plug-ins in Photoshop and After Effects. Files written with the regular OpenEXR output module are written correctly because the Blender buffer is read in reverse order. So I suggest either putting back the DECREASING_Y tag or scanning the buffer in reverse for MultiLayer files.";"None";"None";"None" 8181;1;"Open";3;4457;"Tobias Regenbrecht";146;"Brecht Van Lommel ";"2008-01-29 20:43";"1970-01-01 01:00";"2008-05-09 19:50";"Blender Crash with fewer particles";"SVN 13457 Linux 32 Bit and Windows XP 2 resp. 3 Gig RAM In the most recent builds I can't render as much static particles as I could with older builds. Simply add a particlesystem to the default cube, change it to hair, add 1000 children per particle and try to render. Instant crash. It may be related to this: http://lists.blender.org/pipermail/bf-blender-cvs/2008-January/012674.html";"None";"None";"None" 8238;1;"Open";3;4589;"Davide Vercelli";100;"Nobody";"2008-02-05 23:48";"1970-01-01 01:00";"2008-09-19 16:22";"mesh extrusion";"Patch to expose extrusion in Python. Includes updated epydoc documentation. mesh.extrude() takes two parameter, a translation and an extrusion type. translation can be a vector or a float extrusion type is one of the ExtrudeModes constants (ONLYVERTS, ONLYEDGES, REGION, INDIVFACES) selected verts/edges/faces get extruded and: - if translation is None, that's it - if translation is a float, extruded geometry gets translated of that amount along the ""extrusion normal"" (like it happens in 3d view) - if translation is a vector, extruded geometry gets translated along that vector INDIVFACES is different. in this case, if translation is a float, each faces gets translated of that amount along their own normal (like in ""Individual Faces"") Unfortunately, could not cleanly call extrusion functions since they are intertwined with input handling... so I guess in the end this is going to overlap with Blender 2.50... sigh.";"None";"None";"None" 8260;1;"Open";3;139;"Martin Poirier";103;"Ton Roosendaal";"2008-02-10 16:12";"1970-01-01 01:00";"2008-05-09 19:49";"Preview Window not updating";"The render preview window doesn't update correctly in a couple of situation: - When materials are assigned - When undo is used Other things (like transform) refreshes the preview correctly. How to reproduce: Use a mesh with multiple material, assign materials (preview doesn't update), get out of edit mode (preview doesn't update). Martin";"None";"None";"None" 8307;1;"Open";3;6890;"Nicolas Wu ";100;"Nobody";"2008-02-18 23:30";"1970-01-01 01:00";"2008-09-19 14:10";"Film grain composite node";"Hi all, I made a film grain composite node. It introduces simulated grain on an image. Check the attached pictures. Cheers, Nick";"None";"None";"None" 8337;1;"Open";3;2960;"Petri Rantanen";103;"Ton Roosendaal";"2008-02-23 20:23";"1970-01-01 01:00";"2008-05-09 19:48";"Camera clipping distance clips the camera too and so does passepartout";"Hi, This isn't so much of a bug as it is a fault. When inside a camera = in camera view it is impossible to select the camera and change its settings if the camera's clipping distance is above .50 = it becomes invisible/clipped (this may vary from 0.5 to 1.0 depedending on the lens size). I suggest that the camera clipping will NOT affect the camera object itself. Also the passepartout shouldn't affect the camera object itself, for this also makes it hard to select the camere behind the dark veil. It shouldn't bother anyone that the camera shows even through the passepartout because the camera's wireframe (which you use to select the camera) is within a good distance from the borders of the passepartout so it won't ""ruin"" the black veil if one zooms a bit. There's still more space between the visible camera wireframe and the actual picture border (where the passepartout starts), than there's between the picture border and the safeframe. So the black won't be spoiled if you just zoom a bit (this distance is actually zooming out so you can see a lot of the black already). Regards, Petri Rantanen aka Falgor";"None";"None";"None" 8360;1;"Open";3;6730;"Remigiusz Fiedler";100;"Nobody";"2008-02-25 20:12";"1970-01-01 01:00";"2012-10-29 12:40";"menu goes off screen: concerns registered script listing in Script-Window";"Menu goes off screen in Script-Window. It seems the standard menu functionality is incomplete implemented. Only a part of listed scripts can be displayed and is accessible. The problem: no access to the rest of scripts. Solution: by massive amount of listed items the menu should switch in multi-columns mode. Any chance to fix it for 2.46?";"None";"None";"None" 8372;1;"Open";3;5781;"Mike Pan";146;"Brecht Van Lommel ";"2008-02-27 01:10";"1970-01-01 01:00";"2008-05-09 19:47";"OGL Render viewport stops functioning when Blender is minimized";"The title pretty much says it all. When you start the viewport rendering (Ctrl+ render viewport button), and then minimize the render window, the final render comes out frozen at the instant one minimize the window. A minor inconvenience when one is trying to rendering a long pre-vis while trying to use the screenspace for something else. I can only confirm this problem on Windows XP with Nvidia driver at the moment. But I wonder if there is an easy fix for it.";"None";"None";"None" 8398;1;"Open";3;5434;"Luca Bonavita";103;"Ton Roosendaal";"2008-02-29 11:09";"1970-01-01 01:00";"2008-05-09 19:46";"Nodes editors: moving | zooming an editor could be stored in nodetree...";"A minor (but annoying) issue. Suppose you have nodes on the LEFT of both the editors. Drag the material editor on the RIGHT (drag with MMB), then switch to the compositor: now also the compo nodes are on the right. The same applies to zoom. Zoom in in an editor and switch to the other: also this editor has been zoomed. Of course this is also true for ""View all"" (HOME shortcut), because it pans and zooms the editor. ";"None";"None";"None" 8484;1;"Open";3;6689;"Ed Britton";146;"Brecht Van Lommel ";"2008-03-06 01:47";"1970-01-01 01:00";"2008-04-07 05:16";"UV Curves not updating UV texture on loaded blend files";"In current version/2.46RC1 When you save a blend file with a texture in the UV editor and a curve set up on the Curves window, then reload the blend file, your texture will not be affected until you touch the curve on the curves window again.";"None";"None";"None" 8525;1;"Open";3;5573;"Christoph Lücking";100;"Nobody";"2008-03-10 22:01";"1970-01-01 01:00";"2008-10-09 13:38";"Sequencer: Audio and Video are not alligned if movie-framerate does not match scene framerate";"Adding a video with sound in the sequencer results only in alligned channels, if the framerate for the output it set to the framerate of the video. Audio seem to always play with original speed, while the videoframes are mapped 1:1 to the set framerate, ignoring the framerate of the source file. So basically a 5s 30fps videofile added with the default output rate of 25fps results in 5s audio, but 6s video. If the framerate of the scene is adjusted to mach the framerate of the video-file before adding the file, the file gets added correctly. Changing the framerate back to the desired output frame rate afterwards does not change anything. So this work a workaround. This is probably only an issue if Blender is abused as a video cut-programm.";"None";"None";"None" 8534;1;"Open";3;7772;"D Charlesworth";100;"Nobody";"2008-03-12 02:58";"1970-01-01 01:00";"2012-10-29 12:27";"Crash on launch in OS X if path to app is too long";"Blender v2.45 (built 2007.09.15) silently fails on launch. The reason appears to be that the path to Blender.app is too long. Reducing the length of names of containing folders averted the crashes. ___ Version: 2.45 (2.45, 2007-Sep-15, Blender Foundation) PID: 1352 Thread: 0 Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_INVALID_ADDRESS (0x0001) at 0xc0000000 Thread 0 Crashed: 0 libSystem.B.dylib 0x9001c8a9 strcat + 25 1 org.blenderfoundation.blender 0x0045e459 BLI_make_file_string + 89 2 org.blenderfoundation.blender 0x002fe43d bpy_gethome + 269 Thread 0 crashed with X86 Thread State (32-bit): eax: 0x0000002f ebx: 0x002fe344 ecx: 0xbfffff60 edx: 0xc0000000 edi: 0xbffff600 esi: 0xbffff6a0 ebp: 0xbffff598 esp: 0xbffff594 ss: 0x0000001f efl: 0x00010206 eip: 0x9001c8a9 cs: 0x00000017 ds: 0x0000001f es: 0x0000001f fs: 0x00000000 gs: 0x00000037 ";"None";"None";"None" 8538;1;"Open";3;7779;"mino Jankus";146;"Brecht Van Lommel ";"2008-03-12 16:57";"1970-01-01 01:00";"2008-04-07 05:15";"crashes when opening a sound";"when I press ""select wav file"" it crash what to do?please help ";"None";"None";"None" 8549;1;"Open";3;7786;"Garth Zeglin";4332;"Daniel Genrich";"2008-03-13 05:04";"1970-01-01 01:00";"2008-04-07 05:14";"cloth simulation doesn't respect wind field radial fall-off or maxdist (2.46RC1)";" Blender version: blender-2.46RC1-OSX-10.3-py2.5-powerpc ""Rev-13942M"" system version: Mac OS 10.4.11, Dual 2GHz PowerPC G5, nVIDIA GeForce FX 5200 64M VRAM The cloth simulation appears not to respect the radial fall-off settings when responding to a tube-shaped wind field, but instead seems to assume a uniform field. This is in contrast to the particle system, which will respond only locally to the wind field within the specified radial bounds. I have prepared a simple example with a 9 unit wide vertical fabric mesh pinned at a set of top edge points, a 9 unit wide horizontal particle-emitting mesh, and a horizontal tubular wind field attached to a single empty. The wind field has a Radial Fall-off of 2.0 and a Radial MaxDist of 1.0. The emitted particles stream away only in front of the wind emitter, but the fabric moves symmetrically as if the wind were uniformly distributed. I would expect the fabric to move asymmetrically and look as if a fan were blowing on one area. I did try varying the mesh density and material properties, wind source placement, various geometries, clearing caches, and all exhibited similar behavior. In any case, thanks for including the cloth simulator in the upcoming release, I think this will be very useful. ";"Tools";"None";"None" 8566;1;"Open";3;4405;"Killian De Volder";146;"Brecht Van Lommel ";"2008-03-15 07:56";"1970-01-01 01:00";"2008-04-07 05:14";"Envmap does not render ambient occlusion";"When using envmaps the envmap is generated without abmient occlusion.";"Render (internal)";"None";"None" 8569;1;"Open";3;1819;"Carlos Lopez ";4332;"Daniel Genrich";"2008-03-15 12:24";"1970-01-01 01:00";"2008-04-07 05:13";"Cloth system: Stress texture projection doesn't work.";"Stress texture projection looks not to be implemented, as it is for softbodies.";"Tools";"None";"None" 8570;1;"Open";3;1819;"Carlos Lopez ";4332;"Daniel Genrich";"2008-03-15 12:28";"1970-01-01 01:00";"2008-04-07 05:12";"Cloth system: multiple parenting doesn't work.";"DupliFace and DupliVert look not to be implemented, as it is for softbodies.";"Tools";"None";"None" 8579;1;"Open";3;7812;"James Tuvell";146;"Brecht Van Lommel ";"2008-03-16 05:32";"1970-01-01 01:00";"2008-04-07 05:12";"Normal maped texture not shown in reflextion while OSA is on";"CPU: Intel Pentium 4 – HT 2.8GHz RAM: 512 MB GFX: NVidia Gforce FX 5200 256mb OS: Windows XP SR 2 Drivers: NVIDA 6.14.10.9371 Blender: 2.45.15 rinky 2008-03-02 20:17:29 I open Blender. Then delete the default cube. Change into top view. Hit space bar and add mesh UV sphere 32 segments 32 rings 1 radias. Hit space bar add mesh plain. Switch to front view. Select the Sphere. Using the blue transform widget I raise the sphere 2 Blender units. Select the Plain. I scale it up to 5.00 blender units. I add a reflective material to the plain and bump map (image of the moon in jpg format.) to the sphere mapped to the normal. Setting the normal slider to 8.31 then render using the blender renderer and OSA on. The bump map effects the sphere but the reflection is smooth. When I turn off the OSA then the reflection is the same as the sphere. ";"Render (internal)";"None";"None" 8593;1;"Open";3;177;"Matt Ebb ";4290;"Joshua Leung";"2008-03-18 07:05";"1970-01-01 01:00";"2008-04-07 05:11";"Action constraints in proxy armatures lose their targets";"Instructions to reproduce: * Start a new blend file * Link in the group from the attached blend file * Add -> Group -> Group * Ctrl alt P, make a proxy for the armature * Go into pose mode, move the single individual bone on the x axis, this should trigger the action constraint on the other 3 bones * Save this new file as test.blend * New file * Re-open test.blend * Move the single action-driver bone --> No action constraint The actions still seem to be in the file, though they're lost from the constraint itself.";"Tools";"None";"None" 8711;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2008-03-22 18:33";"1970-01-01 01:00";"2008-09-19 12:33";"DisplayList for UV's Window";"This patch is an experiment, to see how much faster display lists are. My tests show it to be about 20% faster.";"None";"None";"None" 8722;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-03-23 02:12";"1970-01-01 01:00";"2008-05-11 13:50";"bevel tool kills the UV";"make or load a model that have UVs enter edit mode and press W, then select bevel - so bevel something on your model your UV will disappears";"None";"None";"None" 8745;1;"Open";3;7569;"Igor Orekhoff";146;"Brecht Van Lommel ";"2008-03-25 12:55";"1970-01-01 01:00";"2008-04-07 05:11";"Crash when rendering a big image using compositing";"Blender crashes when rendering a big (3096x2322px) image. The attached test scene uses three render layers to compose the final image. Crash happens when memory consumption reaches the RAM capacity. My system: AMD AthlonXP 2600+ RAM 1.5MB NVIDIA GeForce FX 5700 Windows XP SP2 Blender SVN trunk rev. 14232 ";"Render (internal)";"None";"None" 8756;1;"Open";3;7890;"Dariusz Dawidowski";100;"Nobody";"2008-03-26 12:43";"1970-01-01 01:00";"2008-05-18 22:14";"3DConexion enhancements and bugfixing";"Background: We testing 3dconnexion device for Apricot Open Game - seems it's perfect tool for levelmaking. I have Ubuntu 8.04, and quite recent blender svn version. Bugs: 1. focus: when I'm using this device, then other things loosing focus, for example I can't use mouse wheel (I need to pan or clock on something and mouse wheel works again), it's also hard to make mouse gestures, you need to click twice. 2. independent speed: now speed of the moving depends on polycount (if it's high polycount then I'm moving very slow) it should drop frames like in games 3. axes are confused Y with Z in transform (I know that we can change this in setup but then, other axes are opposite, so we need one layout). 4. Transforming object is in world space, should be in view space 5. Turntable mode is much much more sensitive (moving, rotating) than other modes 6. I think rotating in Z should be opposite in turntable mode 7. maybe it's Linux driver bug, but left rotation is much faster than rotation to right side Enhancements: 1. Would be great to grab object and move it together with moving view using 3dconnexion. At this moment it's impossible because GKEY changes mode from fly to transform, and grabbing transform arrow stops 3dconnexion input. Greetings";"None";"None";"None" 8776;1;"Open";3;5676;"Richard Vrijland";100;"Nobody";"2008-03-27 18:37";"1970-01-01 01:00";"2009-04-20 15:17";"Bad AA on image textures 2.46 RC1 and SVN";"If the Ray Mirror button (Material Buttons, Mirror and Transparancy tab) isn't enabled, image textures render without any AA. Pressing the Ray Mirror without adjusting any settings solves the AA, same goes for Full OSA in the render pipeline. BUT either of them makes the rendering a lot slower than without those settings or compared to 2.45. To redo it: open attached Blender file, render it. Then render again with the Ray Mirror button enabled. Please also note the render times. OS: XP prof GPU: Nvidia 8600 GT Blender 2.46 rc1 and 2.45 15 Rinky ";"None";"None";"None" 8785;1;"Open";3;7908;"Scott Meador";103;"Ton Roosendaal";"2008-03-28 18:51";"1970-01-01 01:00";"2008-04-07 05:10";"Mapping node and Front/Back causes ""Back"" not to Anti-Alias ";"When using the Front/Back option in the Geometry node for a material, the ""back"" material must have its texture map flipped horizontally so it will display correctly. To do that in Blender, rather than an image editor, I added a Vector>Mapping node to the input Vector of the image texture. In the Mapping node I set the Y-Rotation to 180. Geometry node noodled from its UV output to the Mapping Node > Mapping output to the Texture Vector input > Color output of Texture to the Color input of the Material This works as expected, however, at render time the back is not anti-aliased, but the front is. The workaround is to not use the Mapping node and just flip the image in an editor. Attached is a blend file with front and back textures packed. Thanks! Apple Power Mac Quad G5 NVidia GeForce 6600 OS X.4.11 ""Tiger"" Blender 2.46-RC1 Python 2.5 PPC Rev-13904M Version darwin dynamic ";"Render (internal)";"None";"None" 8790;1;"Open";3;3284;"ys vry";103;"Ton Roosendaal";"2008-03-29 09:38";"1970-01-01 01:00";"2008-04-07 05:06";"Tangent shading doesn't have a terminator bias fix";"In blender 2.46 rc, I get a render error with tanget on. see picture. It seems to render holes where there are non? You can duplicate the error by adding a sphere to a new scene cut half of it away and add a mirror modifier then delete some random face to make holes and turn on tangent in the render options.";"Render (internal)";"None";"None" 8878;1;"Open";3;3132;"Daniel Salazar";103;"Ton Roosendaal";"2008-04-04 22:12";"1970-01-01 01:00";"2008-04-07 05:05";"Arealight shading is wrong";"Shading produced by area lights is plain unrealistic, leaves black areas that should not be with such a large light (see included blend file) Rendered in Blender Internal: http://www.zanqdo.com/tmp/AreaLight_BI.jpg Heres an example of how it should look (rendered in maya-mental ray): http://www.zanqdo.com/tmp/AreaLight_MR.jpg";"None";"None";"None" 8890;1;"Open";3;7246;"Markus Ilmola";146;"Brecht Van Lommel ";"2008-04-06 18:29";"1970-01-01 01:00";"2008-04-07 05:04";"Spot halo does not reflect from a ray mirror material";"The spot halo does not reflect from a ray mirror material (and does not show through ray transparent material (but works with z-transparent material)).";"None";"None";"None" 8911;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2008-04-09 13:22";"1970-01-01 01:00";"2008-10-09 12:59";"Texface Image Usercount is Broken (real bug - not a todo item)";"In this simple case blender will distroy a reference when it should not. * Add an image texture to a material. * Load any image for that texture. * In the UV/Image window and mesh editmode, apply the same image to a mesh (texface). * Now unlink the image texture from the material. * Notice the image name now displays red in the image browser - it has zero users. Save and reload the blend file, the image is now lost from the mesh. The problem is assigning an image dosn't add a user if there is alredy 1 or more users. If you assign the texface first an THEN assign the image to a texture, this wont break because the texface image will take a user. ";"None";"None";"None" 9020;1;"Open";3;8075;"Joseph Veillon";146;"Brecht Van Lommel ";"2008-04-16 06:50";"1970-01-01 01:00";"2008-05-09 19:45";"3D View Screen bugs out when in 1. edit mode 2. clipping is turned on 3. a button is highlighted 4. Video Card settings antialias the 3D view screen";"System Description: CPU- Intel Core Quad Q6600 2.66 Ghz RAM- 4 GB (Not all of it is availible because of 32 bit OS) GFX- GeForce 8800 GT OS- Windows Vista 32 Bit Blender- V. 2.45 Whenever a button is highlighted while in edit mode (doesn't matter mesh object; even happens with the default cube), with clipping on, and video card settings are set to antialias the edges, the 3D view will turn completly black except for the edges of the mesh, they turn red, but only breifly when a vertex/edge/face is selected or when a the mouse is over a button in the buttons window (it will stay messed up if the mouse is over a button). As a side note, I found out that this doesn't occur if there is no antialiasing so it might just be a compatibility issue. A screen shot of a space ship I was making when this occured is provided, sorry about quality but when i saved it using paint it compressed it. (The screenshot doesn't display the mouse but it was over Subdivide) If any more info is needed just contact me, I'm glad to help out Blender :).";"None";"None";"None" 9178;1;"Open";3;3132;"Daniel Salazar";3487;"Campbell Barton";"2008-04-17 08:02";"1970-01-01 01:00";"2008-05-09 19:42";"""Recover Last Session"" won't save edit mode changes";" - open Blender, add a mesh, optionally move some verts, go to object mode and close - open Blender again and use ""Recover Last Session""; everything fine - enter edit mode, move some verts and without leaving edit mode close Blender - open Blender and ""Recover Last Session"", the changes you did in edit mode are lost";"None";"None";"None" 9406;1;"Open";3;7460;"Jiri H";4290;"Joshua Leung";"2008-04-17 20:49";"1970-01-01 01:00";"2008-05-09 19:40";"Constraint Switches unfunctional with Bones(posemode)";"While playing with Bassam solution for Constraint switches (it uses keyed Constrained influence and IPO Driven Channel, see e.g. this Mancandy tutorial http://blendergroupie.blip.tv/file/723405/)it seems only standard objects can be used as the drivers for IPO Driven Channel Inf. See example Armature A in attached file. Cube is the Constraint switch. If you move Cube up the Location Limit Constraint shall start to influence bone of Armature A. Everything works OK. When I use Bones of Armatures in Pose mode as a Constraint Influence IPO Driven Channel Drivers it does not work as the driver/switch. See example of Armature B in attached file. Same solution as in Armature A but instead of Cube the switch/driver of Armature B is another Bone. It is also tricky which axes of bone are evaluated as bones axes are not in compliance with standard axes orientation. Not sure which axes are used for IPO Driven Channel. ";"None";"None";"None" 9494;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-21 23:30";"1970-01-01 01:00";"2008-04-22 09:02";"More precise Huffyuv options";"Please use more precise Huffyuv exporting options. Right now, when you select AVI/Huffyuv to export, you still have some options about bitrate, which are not really targeted towards Huffyuv as it is a lossless codec. Instead, colorspace options should be there. I understand that Huffyuv is a thrown-in codec for Blender, but it should not be underestimated. It is much more important to professionals than supporting let's say, XViD, as it's the only codec with intermediate purposes.";"None";"None";"None" 9495;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-21 23:40";"1970-01-01 01:00";"2008-04-22 09:18";"Usability problem with new strips";"When you load a new video+audio in the sequencer, the mouse cursor is one the one side of the screen, while the clips imported are on the other. This makes it impossible to direct the strips towards the end of the channel's timeline, but instead it gets sticky towards the right end of the monitor and then you have to release the mouse, and then manually press ""G"" to move the clips towards the beginning of the channel timeline! I hope I am making some sense. XP here btw.";"Interface";"None";"None" 9497;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-21 23:57";"2008-04-22 00:33";"2008-04-22 08:38";"Mouse context menus";"Please add *by default* mouse context menus (the user should not have to go to preferences to enable that) when someone right-clicks on a strip or the iPo curve editor (e.g. add/remove marker/key). This simple addition can skyrocket the usability of Blender as a video editor.";"Interface";"None";"None" 9498;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-22 00:00";"1970-01-01 01:00";"2008-04-22 09:01";"Lock sequencer timeline to positive numbers only";"One very annoying thing with Blender's sequencer is that when you use your wheel mouse to change the size of the timeline, the change now includes negative seconds/frames too. I wish there was a way to lock the timeline to only show positive numbers only, just like any other editor. While this negative number thing might be useful for other Blender usages, for a video editor is very annoying and completely useless. It would be very nice to introduce by default, an icon in that toolbar that says ""lock to positive values"" or something like that. ";"Interface";"None";"None" 9499;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-22 00:06";"2008-04-22 00:23";"2008-04-22 08:36";"Video preview on second monitor";"I have heard of Blender being able to use multiple monitors, but I don't see how this is done. I have two monitors, one 1680x1050 and the other one 1920x1200. I would like my secondary HD monitor to preview the video in full HD, fullscreen, on the second monitor (1:1 size). I found no obvious way to do that. It's the only way to color correct correctly.";"Interface";"None";"None" 9616;1;"Open";3;8130;"Eugenia Loli-Queru";4246;"Peter Schlaile";"2008-04-22 09:47";"2008-04-23 05:39";"2008-04-23 10:08";"Feature request: AviSynth support for Windows";"Please implement this for the Windows build: http://www.blender.org/forum/viewtopic.php?p=66675 ";"None";"None";"None" 9696;1;"Open";3;5409;"George Milnes II";146;"Brecht Van Lommel ";"2008-04-29 16:53";"1970-01-01 01:00";"2008-05-09 19:37";"Octree Fill uses only 1 thread (OSX Inetl/PPC)";"Mac Intel & PCC Not sure if this considered a bug. When rendering large scenes, the octree fill only uses 1 thread on our dual-core machines. This happens on both our core2 Intel iMac and our dual-proc G5 PPC. The octreee fill ends up taking longer than the actual render, which is supporting multiple threads. I have noticed that the octree goes up to over 1000000 and then resets back to nothing and starts over a few times. Any thoughts?";"None";"None";"None" 10062;1;"Open";3;5271;"Alex Fraser";3487;"Campbell Barton";"2008-05-01 07:37";"1970-01-01 01:00";"2008-05-09 19:36";"Mesh Corrupt (was Crashing bug in Edge Split modifier)";"When an object is in a certain state, the Edge Split modifier causes Blender to crash. I'm not sure what the state is or how to get to it - this happened in a complex scene. The affected object was on an invisible layer. When trying to switch back to that layer, the edge splitting was applied and a crash occurred. Attached are two files: the original (parliament18_gz.blend) in which I first noticed the bug, and a cut-down version (EdgeSplitBug_gz.blend) with only the affected object and two others it depends on. To trigger the bug: - In parliament18_gz.blend, switch to layer 1 to trigger a redraw of the object TunnelCladdingFlat. - In EdgeSplitBug_gz.blend, select TunnelCladdingFlat and enable the Edge Split modifier. In both cases blender crashes immediately in versions 2.45 and 2.45.17. Following is a backtrace of 2.45.17 (Rev-14610M) for parliament18_gz.blend: #0 0x08500492 in edgesplitModifier_applyModifier () #1 0x084af371 in mesh_calc_modifiers () #2 0x084b032a in mesh_build_data () #3 0x084b05fa in mesh_get_derived_final () #4 0x0812fbd2 in count_object () #5 0x08130205 in countall () #6 0x08209d66 in do_view3d_buttons () #7 0x081ed9d9 in do_headerbuttons () #8 0x0812ebd6 in screenmain () #9 0x080ede56 in main () And this is from EdgeSplitBug_gz.blend: #0 0x08500492 in edgesplitModifier_applyModifier () #1 0x084af371 in mesh_calc_modifiers () #2 0x084b032a in mesh_build_data () #3 0x084ba565 in object_handle_update () #4 0x082ceeff in do_modifier_panels () #5 0x08162100 in do_butspace () #6 0x0813f9f8 in winqreadbutspace () #7 0x081243e2 in scrarea_do_winhandle () #8 0x0812e3a2 in screenmain () #9 0x080ede56 in main () The bug can be worked around simply by entering edit mode. Perhaps the mesh became corrupt at some point. However, I think the error should be handed gracefully. CPU: Intel Core2Duo T7200 RAM: 1GB GFX: ATI Radeon Mobility x1400 OS: Ubuntu 8.04 Blender: 2.45 and 2.45.17 (Rev-14610M)";"None";"None";"None" 10101;1;"Open";3;4679;"Juho Vepsalainen";3487;"Campbell Barton";"2008-05-01 10:31";"1970-01-01 01:00";"2008-05-09 19:35";"PyDriver Rotation is not evaluated correctly";"See provided .blend. The cube on the left rotates as one would expect. The one on the right evaluated by Python does not. Tested on Ubuntu (Hardy), SVN (rev. 14634).";"None";"None";"None" 10429;1;"Open";3;8239;"Emilio Le Roux";100;"Nobody";"2008-05-02 21:00";"1970-01-01 01:00";"2008-05-05 19:38";"Left Button Select preference not working with Hair editing";" When editing hair, the hair strands can only be selected by right button, independently of the user's preference. Further info: This applies only for directly selecting the hair strand's 'vertices' by clicking on them. The BKEY (box) or BB (paint select) obbey the user's left/right preference correctly. Also, for all the other operations, like grabbing or combing the hair, the left mouse preference works fine. (except you can't of course select and ""drag immediately"" with the left mouse).";"None";"None";"None" 10477;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2008-05-03 03:23";"1970-01-01 01:00";"2009-04-20 15:26";"Windows Installer functions limited ";"After saving a file to Z:\foo\bar\blah.blend and then quitting Blender (no matter how often this is repeated), the reopen-last menu always says: Open Recent C:\Program Files (x86)\Blender Foundation\Blender\.blender\untitled.blend";"None";"None";"None" 10629;1;"Open";3;8269;"Alec Burgess";100;"Nobody";"2008-05-04 22:37";"1970-01-01 01:00";"2008-09-09 19:11";"Win32 taskbar vertical covers blender window when not maximized";"On WinXP with a taskbar on the right side (instead of normal position at bottom): When Blender is opened it always opens maximized (maybe this is a preference option I haven't found yet? - sorry new user) If Blender is ""restored"" (ie. changed from mazimized) the window is set to full screen dimensions with the right side obscured by the windows taskbar if it is positioned there. I believe Windows has a concept of maximum useable screen area - I don't think this is being honored in this case. (I tried to search for previous reporting of this bug w/o success)";"None";"None";"None" 10702;1;"Open";3;8238;"Claus Ilginnis";100;"Nobody";"2008-05-05 14:53";"1970-01-01 01:00";"2008-05-06 10:48";"SVN-Revision 14682: Font does not React on 3d Curve";" Ok, I am sure you will hate me by now :-) - so that's last test for this week .... Steps to reproduce: Open the attached file. Look at the top view and you will see there is a kurve that goes in all directions (xyz) but the corresponding font goes just in two directions (xy). ";"None";"None";"None" 10838;1;"Open";3;3169;"Vladimi­r Marek";4332;"Daniel Genrich";"2008-05-06 12:32";"1970-01-01 01:00";"2008-05-09 19:33";"Cloth and Time IPO curve do not play along";"In the attached .blend file, if you hit the ""Bake"" button (top right on the screen), blender computes 9 frames and then the cloth disapears. If I first erase the Time IPO curve for both monkey head and cloth, the simulation finishes as I suppose it to. I checked on #blendercoders, and I was told that it is inteded to work. I am running latest SVN compiled on development version of Solaris using SunCC, so theoretically there might be issue with (yet) unsupported compiler. I'm using Nvidia graphic card. Thank you for making Blender happen -- Vlad";"None";"None";"None" 10989;1;"Open";3;4446;"samran abed alsalam";4246;"Peter Schlaile";"2008-05-07 11:43";"1970-01-01 01:00";"2008-05-09 19:33";"When you set project to 25 frames, the play back run between 25 and 26, not stable";"When you set project to 25 frames, the play back run between 25 and 26, not stable This also happen in 50 FBS and work good in 30 I test it in default cube P4 HT vista SVN 8600GTS Thanks ";"None";"None";"None" 11111;1;"Open";3;5434;"Luca Bonavita";4246;"Peter Schlaile";"2008-05-10 03:33";"1970-01-01 01:00";"2008-05-12 22:21";"Sequencer: inconsistent audio playback with FLV clips";"Hi, I'm trying to edit some flash video files, but the audio is odd (attaching a file). Everywhere you click and play, the audio file start from the same audio sample, sometimes changing the starting sample but randomly. Maybe this is related to this one? http://projects.blender.org/tracker/index.php?func=detail&aid=8525&group_id=9&atid=125 Thanks, Luca - blender's shell output: Input #0, flv, from './ASD.flv': Duration: 00:01:21.0, start: 0.000000, bitrate: 56 kb/s Stream #0.0: Video: flv, yuv420p, 320x240, 25.00 fps(r) Stream #0.1: Audio: mp3, 22050 Hz, mono, 56 kb/s mdb:44, lastbuf:0 skipping granule 0 - mplayer's output: Playing ASD.flv. libavformat file format detected. VIDEO: [FLV1] 320x240 0bpp 25.000 fps 0.0 kbps ( 0.0 kbyte/s) ========================================================================== Trying to force audio codec driver family libmad... Opening audio decoder: [libmad] libmad mpeg audio decoder AUDIO: 22050 Hz, 1 ch, s16le, 8.0 kbit/2.27% (ratio: 1000->44100) Selected audio codec: [mad] afm: libmad (libMAD MPEG layer 1-2-3) ========================================================================== ========================================================================== Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family Selected video codec: [ffflv] vfm: ffmpeg (FFmpeg Flash video) ========================================================================== [AO OSS] audio_setup: Can't open audio device /dev/dsp: Device or resource busy alsa-init: using device default alsa: 48000 Hz/1 channels/2 bpf/32768 bytes buffer/Signed 16 bit Little Endian AO: [alsa] 48000Hz 1ch s16le (2 bytes per sample) Starting playback... VDec: vo config request - 320 x 240 (preferred colorspace: Planar YV12) VDec: using Planar YV12 as output csp (no 0) Movie-Aspect is undefined - no prescaling applied. VO: [xv] 320x240 => 320x240 Planar YV12 [zoom] Cannot sync MAD frame: 0.188 ct: -3.262 1910/1910 0% 0% 0.6% 0 0 Cannot sync MAD frame A: 80.5 V: 81.0 A-V: -0.532 ct: -2.890 2026/2026 0% 0% 0.6% 0 0 alsa-uninit: pcm closed Exiting... (End of file) ";"None";"None";"None" 11132;1;"Open";3;6147;"David B.";100;"Nobody";"2008-05-11 00:34";"1970-01-01 01:00";"2008-05-22 18:22";"ipo Time do not works anymore with particles and explode modifier";"Ipo ""Time"" do not work anymore with particles. Works well with test build 2 ""gamera"" http://download.blender.org/release/Blender2.46rc/old/ Time1.blend -> works well with test build2 but do not works with latest build Time1.blend -> video of the result Explode.blend -> example with explode modifier SVN for windows, revision 14768";"None";"None";"None" 11244;1;"Open";3;6183;"Troy Sobotka";4246;"Peter Schlaile";"2008-05-14 06:29";"2008-05-14 10:32";"2008-05-14 12:23";"BASIC EDITING: Fundamental RIPPLE EDIT Procedure Missing";"Summary: Blender is evolving into a tremendously powerful workhorse for non linear editing environments. Unfortunately, a few minor and yet key editing techniques are missing. SLIP EDITS, SLIDE EDITS, and ROLLING EDITS are all currently supported. However, the vital RIPPLE EDIT is missing or not clearly obvious. RIPPLE EDIT: Adjusting the time line such that adjacent clips are unaffected. Effectively this amounts to inserting and cutting slices of the time line. Assuming clip A, B, and C, shortening B should automatically collapse C toward A. Expanding or inserting frames into clip B would resultantly lengthen the overall time line and shift clip C further along. This would also effectively allow for a scene B2 to be inserted between B and C and adjust the overall time line accordingly. Assuming a time line of 200 frames consisting of clips A and C each of 100 frames. Choosing a point between the two clips and performing a RIPPLE EDIT insert of clip B (also 100 frames) will insert clip B and shift clip C further along the time line. The new time line would be 300 frames. ";"None";"None";"None" 11398;1;"Open";3;139;"Martin Poirier";100;"Nobody";"2008-05-14 18:43";"1970-01-01 01:00";"2008-05-18 09:50";"Recursive Dupliface not working correctly";"This is most likely a todo, but recursive (with more than one level of child) dupliface doesn't work right. See example file. Smaller cube parented to bigger cube from left to right. Dupliface on the bigger cube works ok. The smallest cube is replicated on the top parent as if it was directly duplied on it. Turning Dupliface on for the middle cube results in the mess that is show. Parenting offsets seem to affect the result (moving both smaller and middle cube change the position of the little cubes in a weird fashion) What I was expecting: The little cube to be duplicated on the middle cube, the result being duplicated on the big cube.";"None";"None";"None" 11708;1;"Open";3;8220;"Fintan Boyle";100;"Nobody";"2008-05-16 15:37";"1970-01-01 01:00";"2008-09-12 15:00";"Bevel Modifier impacts UV mapped textures";" Version: Blender 2.46.17 ""Bird"" OS: Windows XP AMD Turion 64 Behaviour: When a bevel modifier is applied to a mesh and a material using a uv projection for texturing is applied to the mesh, the texture is not rendered in either the 3d window (texture mode) or in the final render. If i switch to orco projection the problem goes away and the texture renders as expected.";"None";"None";"None" 11719;1;"Open";3;5434;"Luca Bonavita";100;"Nobody";"2008-05-17 02:50";"1970-01-01 01:00";"2008-05-17 19:48";"Sequencer: channels numbers move wrong while zooming vertically";"Hi, very often I had the sensation I wasn't able to say exactly which channel was the number N: now I figured out that this is because the channel number position depends on the vertical zoom of the sequence view.. Click with LMB on the channel numbers area and drag up-down: there's a point in which the numbers are ""decimated"", and they also change position inconsistently (see the capture). That's a lot confusing if you don't know it's happening! http://www.mindrones.com/downloads/Bbugs/blender_seq_channels_num.avi Cheers, Luca ---- Blender: Blender 2.44 + svn-20080515 OS: Mandriva Linux 2007.0 x86_64 Python: 2.4.3 OpenGL: GeForce 6600 GT/PCI/SSE2 NVIDIA Corporation Version: 2.1.0 NVIDIA 97.55";"None";"None";"None" 11724;1;"Open";3;4457;"Tobias Regenbrecht";100;"Nobody";"2008-05-17 08:45";"1970-01-01 01:00";"2008-05-18 09:47";"Can't unlink a linked scene";"Blender current SVN (#11717) Linux If you link a to a scene in another file, you can't unlink it via the Li-Button next to the scene selection field (though the tooltip states so ""Direct linked datablock. Click to make local""). I think that it hasn't worked before, so the tooltip is wrong. But of course it would be nice to unlink a scene. ";"None";"None";"None" 11730;1;"Open";3;6877;"Niels Krook";3487;"Campbell Barton";"2008-05-17 18:43";"1970-01-01 01:00";"2008-05-17 19:51";"Outliner Vertical Scrollbar, can't mouse-click position";"The Outline-Area, when having enough objects, the vertical-scrollbar can't be clicked and moved to a different position, it will now always select an item in the outline-list.(When having alot of objects, the mouse-wheel is not the way... ;) )";"None";"None";"None" 11742;1;"Open";3;7246;"Markus Ilmola";100;"Nobody";"2008-05-18 17:36";"1970-01-01 01:00";"2008-09-22 18:06";"Skandinavian characters ä, ö and å don't display correctly in the filebrowser";"A file or a folder name with ä, ö or å in it doesn't display correctly in the filebrowser. For example a folder named ""Työpöytä"" displays as ""Työpöytä"". Blender version: the latest svn version (Rev: 14883) OS: Ubuntu Linux 7.10 ";"None";"None";"None" 11923;1;"Open";3;4430;"Pawel Kowal";100;"Nobody";"2008-05-20 15:33";"1970-01-01 01:00";"2008-05-22 18:15";"Changing Use blender units doesn't change start and end limits";"When you change 'Use blender units' it doesn't change the limits for the strand start and end. So when you set this option you can't lower start value below 0.25 but you can set it more than 2.0. You have to reopen Strand popup for getting proper limits. Checked on two WinXP installations.";"None";"None";"None" 12808;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-05-26 14:29";"1970-01-01 01:00";"2008-09-09 17:51";"task manager and render";"WinXP, Blender 2.46 1. start a render 2. press Ctrl+shift+esc (or shift+ctrl+esc) to start Windows Task Maganer - sometimes this stop the render (like when I press esc to stop render)";"None";"None";"None" 13022;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2008-05-27 12:00";"1970-01-01 01:00";"2008-07-20 13:38";"""Only shadow"" option broken for shadow buffer lamps";"See attached file. As soon as you set ""Soft"" (shadow softness) to anything other than 0, the outline of the light cone on the surface will generate a shadow pattern. Makes the ""Only Shadow"" option pretty much useless.";"None";"None";"None" 13127;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-05-28 00:22";"1970-01-01 01:00";"2008-07-24 19:52";"object as texture coordinate bug";"1. load the attached scene and start render: texture will not move 2. now turn on layer 2 and start render: texture will move So the texture-mover-object on layer 2. When we turn off this layer, the material cant reach the texture-coordinate-object. I dont know this is a bug or a feature... This happens when using local view of the plane.";"None";"None";"None" 13286;1;"Open";3;4754;"Allan Liddle";100;"Nobody";"2008-05-28 22:29";"1970-01-01 01:00";"2008-05-28 22:29";"Particle system ""corrupts"" when changing material colour and used in conjunction with softbodies";"This happens in Blender 2.46 release version on Windows XP on a 3GHz dual core Intel processor with 2GB of RAM. The best way to explain this is to do the following: 1. Using the attached .blend file, go to Soft Body -> Collision tab. Bake the soft body (I set it to 50 frames to save time). 2. Animate it to verify that the baking worked. 3. Go to the particle system, Visualization tab. Go to the ""Render"" heading and change the Material to ""1"". 4. Look what happens to the particles - it incorrectly groups together in one ""blob"" and the animation of the particle system does not work anymore. 5. The SAME HAPPENS when you repeat steps 1 and 2 (skip 3) and you change the amount of children (in particles window, Children tab, change ""Amount"" value). I THINK this should not happen, because only the particles are baked and not the children, correct? Thanks, Allan Liddle (AnyMation) (PS: Again - sorry, but I don't know how to verify the issue - as requested above. Please let me know how to do it)";"Tools";"None";"New" 13593;1;"Open";3;8648;"Andrew Hunter";100;"Nobody";"2008-06-01 18:23";"1970-01-01 01:00";"2008-06-04 02:49";"External Asset Management";"Summary: When using Blender with external assets, such as external video files, there is no facility to keep track of the imported files or the metadata that goes with it. A system through which assets could be referenced in the project but not included in the sequencer. Scenario: Sally has found that Blender meets her needs as a non-linear editor for a short video project but has multiple takes of each shot. She is frustrated that she has to Add > Video + Audio for each take directly onto the timeline, making switching between takes cumbersome. Solution: She would prefer that there be a file browser like window that she could drag the clips directly into the timeline from. In addition, it would be nice to keep track of the metadata, such as the scene, shot and take number, in the asset window for easy reference. ";"None";"None";"None" 13594;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2008-06-01 23:00";"1970-01-01 01:00";"2008-09-09 14:59";"Broken libdata handling";"When you link an object from a file, then delete the object, then append it, it is broken. If you assign a material (which doesn't get imported automatically anymore), you can't deassign it. If you try to edit it, it will say ""Can't edit libdata"", even though it is no libdata (and it is even drawn pink in the view). ";"None";"None";"None" 13636;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-05 19:00";"1970-01-01 01:00";"2008-09-09 14:38";"border render and defocus node";"1. make a scene with defocus node 2. set border render: set a small part of the screen 3. press render: the render will very slow because of the border render: on the part of the screen that out of the border will very big blur";"None";"None";"None" 13638;1;"Open";3;4405;"Killian De Volder";100;"Nobody";"2008-06-05 20:32";"1970-01-01 01:00";"2009-04-20 15:32";"Blender has trouble receiving input.";"Depending on the window manager blender accepts or doesn't accept any input: No windowmanager (launched from a xterm): Input doesn't work. Launched from wmii: Input works if you are lucky and switch deskops enough. Once the render window has been infront, you risk loosing control again. Using TWM: seems to work. Bug occured after updating to 2.46";"None";"None";"Fixed" 13647;1;"Open";3;7677;"Yves Poissant";100;"Nobody";"2008-06-06 07:30";"1970-01-01 01:00";"2008-09-19 18:25";"SAH-BVH acceleration structure temporary code";"This is the code I've been working on for a while. It works and is rather well debuged but it is incomplete. I still need to add some more optimization tricks. Heuristics to help intersecting the bounding boxes in the BVH and help selecting which child to try first. I saved the patch there to protect it. I start a new job next monday, developing real-time ray tracing renderer and I will have to sigh a NDA. There is a possibility that this might prevent me from working on Blender code. In this event, the code will be there for someone else to pursue. I will try to negociate being able to continue working on that code though. The patch uses the rt value to enable different versions of the acceleration structures. I needed to find which one was more efficient. The most important codes are: 330: acceleration structure is skipped. 0: original octree 111: original BIH as supplied by Brecht. I added a couple optimization. 333: the BIH with cached triangle data. Caching triangle improved render time by about 10% on average on my test scenes. 444: the straight BVH with cached data. The spliting is done on mean position only. 450: SAH BVH with cached data. At tree build time, a search is made for 16 positions on each axis fo find the partition that will result in the least costly traversal. 460: SAH BVH with cached data but the ray-triangle intersection uses the 3D barycentric method. This is the fastest so far and can be further optimized. This method requires that the data be cached because it must be transformed at tree build time. So I cached the data for the other methos to be able to compare the render times. The SAH BVH is the most stable structure for acceleration. It is usually faster than the octree but there are a few exceptional scenes for which it is slower while some other scenes are spectacularly faster. The BIH is usually a tiny bit faster than the SAH BVH but when it fails, it fails spectacularly too. SAH BVH never failed spectacularly. Beside, there are a few more heuristics I need to implement that should make BVH faster than BIH in the end. There is an issue that is still pending and is currently partially solved though. It is related to re-testing neighbor triangles on edges. The code currently works because it relies on a new jittering code that was added in v2.46. But that jittering causes render artifacts so it may be removed eventually or it will certainly be modified and this could break the current implementation of the BVH because it does not do the re-testing stuff. That is it for now.";"None";"None";"None" 13648;1;"Open";3;4679;"Juho Vepsalainen";100;"Nobody";"2008-06-06 15:20";"1970-01-01 01:00";"2008-09-09 14:15";"UV coordinate weirdness";"Check out the file provided. If you open up the UV editor and see what the UVs look like for the selected object, you should see that's definitely something wrong. I don't know what I did to corrupt it unfortunately. I did try to unwrap it in 2.42a, 2.46 and current SVN (rev. 15148) and the result was always the same so I suppose the issue is not there. I suspect the UVs got messed out by some mesh editing tool. That's just one possibility though. I eventually solved the issue by removing one of the offending vertices and by creating a new to its place. I restored the faces of course. After I did this for the two offending places, it unwrapped correctly.";"None";"None";"None" 13650;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-06 21:27";"1970-01-01 01:00";"2008-09-09 14:04";"UV Test grid deletes my drawed texture";"1. open the uv/image window and make a new texture 2. turn on texture paint and draw something 3. press n to open image props dialog and on it press UV Test grid button: it deletes your drawed texture without ask... and no undo (try undo: strange!)";"None";"None";"None" 13656;1;"Open";3;4902;"Fabrice Tiercelin";100;"Nobody";"2008-06-07 20:11";"2008-06-09 18:55";"2008-06-22 22:44";"Bad rotation computing for manual rotation higher than 180 degrees";"This has been a wanted feature for a long time as you can see :) (en) http://blenderartists.org/forum/showthread.php?t=112477 (fr) http://blenderclan.tuxfamily.org/html/modules/newbb/viewtopic.php?viewmode=flat&topic_id=13950&forum=1 (and other links I just can't remember) The issue : When you make a manual rotation, you want the rotation to be as higher as you draw it (let's say with several revolutions). However, Blender will compute the shorter path (sometimes will rotate on the other side). This patch takes into account all the revolutions the user has made. Regards, Fabrice TIERCELIN";"None";"None";"None" 13669;1;"Open";3;7666;"giuseppe canino";100;"Nobody";"2008-06-08 15:52";"1970-01-01 01:00";"2008-06-29 22:25";"Redraw of bones transformation doesn't update on ESC key in complex setup";"Hi,after some thinking I have decided to submit to the bug tracker some wrong behaviours of the constraints system,blender armatures system has always had these problems since years,I hope that at some points these can be fixed. I was trying to do a good Ikspline setup for spine riggings and I found an annoying problem. When you start a transformation,moving a bone in pose mode and after pressing ESC key(to abort) without finishing the transformation the transformation of driven elements(bones,empty,and so on)remain fixed to the previous position just before pressing ESC,not where they should be(they follow the movement correctly but when you press ESC they do not return to the original position). pressing G,R or Alt+A after this updates the redrawing correctly,bones and control objects go in the right position. Other than in this case,there is the same problem when you open a scene with this kind of setup,only pressing G,R,Alt+A(forcing redrawing)update the bones positions correctly. The same problems can be found on some scripts that transform bones or other elements related to bones,even if you do a script link with redraw as option when pressing ESC elements don't update correctly(but after you force redraw all goes fine). For me the problem is that Blender doesn't update the dependency graphs for bones in pose mode when the user press ESC key,but I'm not sure. Maybe the person that we'll answer to this will point out that these kind of problems are due to cyclic dependencies problems,but in MHO : IF A drives B AND B tries to drive A THEN CYCLIC DEPENDENCIES=TRUE but my setup are: A drives B AND B drives C THEN CYCLIC DEPENDENDENCIES=FALSE Problem here is that A object and C object are bones,so in Blender become: A=armature.bone1 B=object outside armature(empty hook,curve control points,and so on) C=armature.bone2 SO A drives B AND B drives C is like A drives B AND B drives A !(because bones are INSIDE armatures object) IMHO this needs to be solved at some point,the constraint system is powerfull,but if the system doesn't update correcty working become difficoult(I have rigging experience on some others programs,like maya and xsi,and there the system is able to update correctly the transformation,I understand blender is another story,but if you can do these setups,the setup should be updated correctly. I have attached 2 blend files to explain better,below some more explanations of the setups. I hope it's useful. Simple ikspline setup(IKSplineRigTest.Zip). For trying move the red bones that control the curve and rotate local y axis for the white bones for twisting(local y axis rotations is interpolated across the lenght of the curve,for now I have used constraints for this,maybe is better doing a pyconstraint using slerp,but it's a prototype,first the system should work good before trying more automation). Explaining a bit,I have used stretch to constraint and point constraint to lock bone position to fixed point on curve length(with the new option for using vertex group I was able to use a mesh as fake curve,I deform with a curve modifier a mesh,that mesh has vertex groups with only 1 vertex that drive the individual position of a bones on curve) I also used a little script to do strechy ikspline,I use 2 curve deform on 2 different meshes and I use a slider(a bone)to decide if and how much the spline is streched. ) Simple Hook deformation from bones to curves(BoneHooks.zip) Another case,bone hooks(a little script that binds control points curve position where bones are in space,for now in object space to avoid some computations,I always center my object at 0 0 0 to optimize a bit). It's really simple,and here I don't see cyclic dependencies at all,as bones drives the curve control points that deform(curve deform)a mesh(which here has nothing to do with the armature,if it worked without redraw problems, I would use that mesh as driver for spline bones(and here maybe there are cyclic dependencies in Blender jargon) . The problem is the same on ESC key,missing some view redraw(try to move bone in pose mode and press esc to stop the trasformation,redraw here is wrong). BYE RENDERDEMON";"None";"None";"None" 13670;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-08 19:22";"1970-01-01 01:00";"2008-09-09 13:43";"render preview and ambient occl.";"1. make a scene and press shift+p to open render preview window 2. now turn on or off the ambient occlusion, or set something in it (type, energy etc) - preview render dont change 3. move or resize preview window - preview render dont change only if you pan/zoom/turn the 3d view the preview render window updates the ao settings";"None";"None";"None" 13682;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-09 16:38";"1970-01-01 01:00";"2008-06-28 04:57";"edge panel right side cutted";"I think I have found the smallest bug in Blender! :) See attached image: right side of the edge set panel seems cutted. :)";"None";"None";"None" 13761;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-11 22:07";"1970-01-01 01:00";"2008-09-09 12:32";"ugly envmap filtering";"see attached blend it contains an cubemap envmap that created with blender I set its filter to maximum (25) you will see the strange filtering and the edges of the cubemap note: if you enable the subsurf modifier on the model, the filter looks better this is strange because the model has smoothed normals and has fine geometry without subsurf";"None";"None";"None" 13986;1;"Open";3;4889;"rupert plumridge";100;"Nobody";"2008-06-13 09:23";"1970-01-01 01:00";"2008-09-09 12:03";"opening network path in Vista fails";"Not sure if this is a Vista issue or a Blender issue. I am using Vista Home Premium and Blender 2.46. I didn't change any settings when installing Blender, just used default paths. I have copied all of the Peach DVD production contents to a networked drive. When I browse to this folder using Vista's Explorer and try to open a .blend file, Blender opens, but then I get an error from Blender saying it can't open the path, but the path it shows is not the one I selected. The file path I browsed to: \\XPC\Media\blender\peach\chars\bird.blend Path that Blender shows as giving error: C:\XPC\Media\blender\peach\chars\XPC\Media\blender\peach\chars\bird.blend Clearly the path is being passed to Blender incorrectly, though I don't know if this is a Vista issue or a Blender issue.";"None";"None";"None" 14389;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-16 19:13";"1970-01-01 01:00";"2008-09-09 11:00";"strange rendering of nonplanar quad";"load attached blend and render it it will render two frames you will see a quad with strange render, flipping triangles";"None";"None";"None" 14395;1;"Open";3;4679;"Juho Vepsalainen";100;"Nobody";"2008-06-17 12:05";"1970-01-01 01:00";"2008-06-17 18:02";"Snap, Cursor -> Selection doesn't work if ""applied"" array modifier is used";"See provided .blend. Open the file and go to the edit mode. Select all vertices and hit shift-s. Select ""Cursor -> Selection"". Note that it doesn't take all vertices (ones applied with modifier) into count. Tested this on 2.46 on SuSE 10.2.";"None";"None";"None" 14417;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2008-06-19 03:29";"1970-01-01 01:00";"2008-09-24 02:47";"""Skin faces/Edge loops"" uses wrong face flags";"FKEY menu -> ""Skin faces/Edge loops"" always creates smooth faces regardless of the original state of the source faces. Also, it uses the wrong material index (always the first one) regardless of current / faces' material index.";"None";"None";"None" 14424;1;"Open";3;4857;"Dalai Felinto";7473;"Benoit Bolsee";"2008-06-20 03:27";"1970-01-01 01:00";"2008-08-05 18:49";"Linked objects doesn't share properties in the Game Engine";"1)I linked one object (i.e. an empty) to a new scene (Ctrl+L). 2)In Blender, when I change a property in one of the objects, the other one changes automatically. 3) In Game Engine, however, if I change a property of the original object and overlay the scene with the linked object, the property didn't change automatically. They work as if I have two different copies of the same object, but not as if I have the same object linked in two different scenes. Is it by design, or can be considered a bug? Have a nice day!";"None";"None";"None" 14626;1;"Open";3;980;"Alexander Ewering ";100;"Nobody";"2008-06-22 23:08";"1970-01-01 01:00";"2008-09-08 19:51";"Groups do not appear in Add -> Group menu";"In the attached file, there's a group ""Blumenkasten"" (created the usual way, not via some satanic python script or similar), and it is proven to work and exist by assigning it to an empty, instancing it correctly. However, it does not ever appear in the ""Add -> Group ->"" menu...";"None";"None";"None" 14659;1;"Open";3;4446;"samran abed alsalam";100;"Nobody";"2008-06-23 00:41";"1970-01-01 01:00";"2008-07-13 15:23";"insert key Problem";"Maby not bug When I het I for insert key “ in material or any defiant edit “ I don’t Have full animation list but I have in ipo ok ‘ in ipo when I het I didn’t create animation curve ! . And Particle did not have insert key list when I het I in Particle window I have fields animation key list not Particle And all this didn’t work with auto key! thanks";"None";"None";"None" 14668;1;"Open";3;5075;"Reuben Martin";4246;"Peter Schlaile";"2008-06-23 01:02";"1970-01-01 01:00";"2008-08-10 17:46";"FUSE can't mount the frameserver";"I'm wanting to mount the http based frameserver as a FUSE based filesystem. However, in order to do that, the frameserver needs to generate a file index when accessing http://localhost:8080/images/ppm/ Without this there is no way to populate the mount point with files. (also, if other types of images are to be supported besides ppm in the future, then http://localhost:8080/images/ needs to generate and index as well) ";"None";"None";"None" 14797;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-23 17:44";"1970-01-01 01:00";"2008-09-08 19:43";"setting the render frames";"1. set start frame to 50 2. set end frame to 100 3. now enter start frame field and type 1 then press tab to jump end frame field and type 10 - after pressing enter the 10 will change to 50 I think there is a check of the start>end at a wrong position in the code.";"None";"None";"None" 16052;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2008-06-27 22:11";"1970-01-01 01:00";"2008-09-08 19:48";"Removing an Empty messes up the depgraph";"In this example, an armature has an empty parent. When the empty is removed, the mesh does not deform with the armature. This only happens when the empty is still in a second scene. In the attached scene remove the empty and play the animation,";"None";"None";"None" 16372;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-06-28 14:12";"1970-01-01 01:00";"2008-06-29 20:23";"movie or sequence texture dont update on viewport";"1. make a plane object 2. load a sequence or a movie texture and turn on ""auto refresh"" in image props 3. enter edit mode and append this texture to the plane 4. switch to textured mode on the viewport 5. now change frames: the texture of the plane on the viewport dont updated The actual frame of the texture will update only when you press ""Reload"" button in the Image menu (or press alt+r).";"None";"None";"None" 16662;1;"Open";3;4002;"john herreño";100;"Nobody";"2008-06-29 20:43";"2008-07-20 18:06";"2008-09-19 23:55";"Curve modifier is ignored on lattice when using OpenGL Rendering Preview";"(Tested on 2.46 official and graphicall.org 15379M) Steps: - Create a Bezier Curve (make it 3D and tweak if wanted) - Create a Lattice (add resolution in W value). Scale up on Z direction. - Add a curve modifier to the lattice, use the curve as the guide. Select ""Z"" direction in the modifier. Tweak lattice Z position to better accomodate in the curve. - (Optionally the direction of the curve can be switched, for easeness; ""Switch Direction"" in Edit Mode, I mean) - Click on button to do a OpenGL Render Preview (rightmost landscape button, header of 3DView) - When closing the render window, the lattice will be dettached from the curve, as if the modifier didn't exist. To get it applied, it's necessary to ""shake"" the lattice (starting any transformation and cancelling) The bug also happens when doing a OpenGL Render Preview for the whole animation. Any mesh modified by the lattice will be affected also, including a particle system attached to the mesh. On file loading it manifests also, so I think it's not possible to render from the command-line when using such a setup. Demo blend file attached. Question: I would like to know if this bug is DAG-related. System Info (from 2.46 release): ===================================== = Blender 2.46 System Information = ===================================== Platform: win32 ======== Python: ====== - Version: 2.5 (r25:51908, Sep 19 2006, 09:52:17) [MSC v.1310 32 bit (Intel)] - Paths: G:\Progs\Blender Foundation\ShellBlender G:\Progs\Blender Foundation\ShellBlender\python25.zip E:\Progs\Python25\Lib E:\Progs\Python25\DLLs E:\Progs\Python25\Lib\lib-tk G:\Progs\Blender Foundation\ShellBlender E:\Progs\Python25 E:\Progs\Python25\lib\site-packages E:\Progs\Python25\lib\site-packages\win32 E:\Progs\Python25\lib\site-packages\win32\lib E:\Progs\Python25\lib\site-packages\Pythonwin G:\Progs\BLENDE~1\SHELLB~1 G:\Progs\Blender Foundation\ShellBlender\.blender\scripts G:\Progs\Blender Foundation\ShellBlender\.blender\scripts\bpymodules - Directories: Blender home dir: G:\Progs\Blender Foundation\ShellBlender\.blender Default dir for scripts: G:\Progs\Blender Foundation\ShellBlender\.blender\scripts Default ""bpydata/"" data dir for scripts: G:\Progs\Blender Foundation\ShellBlender\.blender\scripts\bpydata User defined dir for scripts: -- not found Data dir ""bpydata/"" inside user defined dir: -- not found Default config data ""bpydata/config/"" dir: G:\Progs\Blender Foundation\ShellBlender\.blender\scripts\bpydata\config - Modules: all basic ones were found. OpenGL: ====== - Renderer: GeForce4 MX 4000/AGP/SSE/3DNOW! - Vendor: NVIDIA Corporation - Version: 1.5.4 - Extensions: GL_ARB_imaging GL_ARB_multitexture GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_shader_objects GL_ARB_shading_language_100 GL_ARB_texture_compression GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_texture_rectangle GL_ARB_transpose_matrix GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_window_pos GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_clip_volume_hint GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_multi_draw_arrays GL_EXT_packed_pixels GL_EXT_paletted_texture GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shared_texture_palette GL_EXT_stencil_wrap GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_object GL_EXT_vertex_array GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KTX_buffer_region GL_NV_blend_square GL_NV_fence GL_NV_fog_distance GL_NV_light_max_exponent GL_NV_packed_depth_stencil GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_register_combiners GL_NV_texgen_reflection GL_NV_texture_env_combine4 GL_NV_texture_rectangle GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_program GL_NV_vertex_program1_1 GL_SGIS_generate_mipmap GL_SGIS_multitexture GL_SGIS_texture_lod GL_SUN_slice_accum GL_WIN_swap_hint WGL_EXT_swap_control - Simplistic almost useless benchmark: Redrawing all areas 10 times took 0.2152 seconds.";"None";"None";"None" 16674;1;"Open";3;4679;"Juho Vepsalainen";100;"Nobody";"2008-07-01 13:11";"1970-01-01 01:00";"2008-07-06 14:53";"Render preview doesn't work properly after a new material is added to an object";"Open provided .blend. Add a new material to the cube selected. Change the color of the material. None of the preview panels except for the once in Buttons Window get updated after this here. Rotating a view and moving the preview panel in the biggest viewport makes it work again in that particular view. I tested this using 2.46 on SuSE 10.2.";"None";"None";"None" 16734;1;"Open";3;5860;"filip mond";100;"Nobody";"2008-07-02 13:33";"2008-07-06 14:29";"2008-07-07 14:45";"Text deform by Curve Modifier";"I looked forward to TEXT object deform by Curve Modifier. But it doesnt work correctly. It just smash all letters to begining of a curve. Do you think are there possibilities to fix it soon? Or did I something wrong? If its not working for now it would be better dont show the modifier panel when Text object is selected (or give some note about non function now) I think a lot of people spend a lot of time by changing setup (x.y.z) to fix it without success result. Thanks for help. vklidu run on iMac 24, intel, OS X 10.4.11, Blender 2.46";"None";"None";"None" 17020;1;"Open";3;4170;"Patrick Ye";100;"Nobody";"2008-07-04 07:00";"1970-01-01 01:00";"2008-09-14 22:37";"Blender 2.46 crashes due to unmap_chunk error";"My system: Ubuntu 8.04 AMD64 Blender: latest build from SVN, version 2.46 GTK: version 2.0 The following python code will cause Blender to commit an unmap_chunk error if run in the script window of Blender: [code] import gtk textBuffer = gtk.TextBuffer() textView = gtk.TextView(textBuffer) textView.set_wrap_mode(gtk.WRAP_WORD) scrollWindow = gtk.ScrolledWindow() scrollWindow.add(textView) [/code] The exact error message is: [error] *** glibc detected *** /home/jingy/install/svn/blender/install/linux2/blender: m unmap_chunk(): invalid pointer: 0x0000000001f9e6c0 *** ======= Backtrace: ========= /lib/libc.so.6(cfree+0x1b6)[0x2b754477bd46] /usr/lib/libgobject-2.0.so.0(g_object_newv+0x23f)[0x2aaaabc2f54f] /var/lib/python-support/python2.5/gtk-2.0/gobject/_gobject.so[0x2aaaab9dfcf1] /var/lib/python-support/python2.5/gtk-2.0/gtk/_gtk.so[0x2aaaaccc5aba] /usr/lib/libpython2.5.so.1.0[0x2b754234fdd8] /usr/lib/libpython2.5.so.1.0(PyObject_Call+0x13)[0x2b7542306f53] /usr/lib/libpython2.5.so.1.0(PyEval_EvalFrameEx+0x25b0)[0x2b75423849f0] /usr/lib/libpython2.5.so.1.0(PyEval_EvalCodeEx+0x796)[0x2b75423888c6] /usr/lib/libpython2.5.so.1.0(PyEval_EvalCode+0x32)[0x2b75423889e2] /home/jingy/install/svn/blender/install/linux2/blender(BPY_txt_do_python_Text+0x 1f9)[0x85d4e9] /home/jingy/install/svn/blender/install/linux2/blender(run_python_script+0x22)[0 x6522b2] /home/jingy/install/svn/blender/install/linux2/blender(winqreadtextspace+0x53a)[ 0x65301a] /home/jingy/install/svn/blender/install/linux2/blender(screenmain+0x43b)[0x6ab10 b] /home/jingy/install/svn/blender/install/linux2/blender(main+0x5b1)[0x5815e1] /lib/libc.so.6(__libc_start_main+0xf4)[0x2b75447221c4] /home/jingy/install/svn/blender/install/linux2/blender(atan+0x321)[0x580a99] ======= Memory map: ======== 00400000-0130e000 r-xp 00000000 08:11 37257265 /home/j ingy/install/svn/blender/install/linux2/blender 0150e000-01698000 rw-p 00f0e000 08:11 37257265 /home/j ingy/install/svn/blender/install/linux2/blender 01698000-01fad000 rw-p 01698000 00:00 0 [heap] 40eb9000-40f3e000 rw-p 00000000 00:0e 1055 /dev/ze ro 41627000-41629000 rwxp 00000000 00:0e 1055 /dev/ze ro 2aaaaaaab000-2aaaaabb3000 rw-p 2aaaaaaab000 00:00 0 2aaaaabb3000-2aaaaabb4000 rw-s fd060000 00:0e 14745 /dev/nv idia0 2aaaaabe5000-2aaaaac24000 r--p 00000000 08:07 4039488 /usr/li b/locale/en_AU.utf8/LC_CTYPE 2aaaaac24000-2aaaaac2b000 r--s 00000000 08:07 4015800 /usr/li jingy@odin:~$ blender Compiled with Python version 2.5.2. Checking for installed Python... got it! *** glibc detected *** /home/jingy/install/svn/blender/install/linux2/blender: munmap_chunk(): invalid pointer: 0x0000000001fa0080 *** ======= Backtrace: ========= /lib/libc.so.6(cfree+0x1b6)[0x2b7b78e49d46] /usr/lib/libgobject-2.0.so.0(g_object_newv+0x23f)[0x2aaaac88954f] /var/lib/python-support/python2.5/gtk-2.0/gobject/_gobject.so[0x2aaaac639cf1] /var/lib/python-support/python2.5/gtk-2.0/gtk/_gtk.so[0x2aaaad91faba] /usr/lib/libpython2.5.so.1.0[0x2b7b76a1ddd8] /usr/lib/libpython2.5.so.1.0(PyObject_Call+0x13)[0x2b7b769d4f53] /usr/lib/libpython2.5.so.1.0(PyEval_EvalFrameEx+0x25b0)[0x2b7b76a529f0] /usr/lib/libpython2.5.so.1.0(PyEval_EvalCodeEx+0x796)[0x2b7b76a568c6] /usr/lib/libpython2.5.so.1.0(PyEval_EvalCode+0x32)[0x2b7b76a569e2] /home/jingy/install/svn/blender/install/linux2/blender(BPY_txt_do_python_Text+0x1f9)[0x85d4e9] /home/jingy/install/svn/blender/install/linux2/blender(run_python_script+0x22)[0x6522b2] /home/jingy/install/svn/blender/install/linux2/blender(winqreadtextspace+0x53a)[0x65301a] /home/jingy/install/svn/blender/install/linux2/blender(screenmain+0x43b)[0x6ab10b] /home/jingy/install/svn/blender/install/linux2/blender(main+0x5b1)[0x5815e1] /lib/libc.so.6(__libc_start_main+0xf4)[0x2b7b78df01c4] /home/jingy/install/svn/blender/install/linux2/blender(atan+0x321)[0x580a99] ======= Memory map: ======== 00400000-0130e000 r-xp 00000000 08:11 37257265 /home/jingy/install/svn/blender/install/linux2/blender 0150e000-01698000 rw-p 00f0e000 08:11 37257265 /home/jingy/install/svn/blender/install/linux2/blender 01698000-01fb1000 rw-p 01698000 00:00 0 [heap] 412f0000-41375000 rw-p 00000000 00:0e 1055 /dev/zero 41e44000-41e46000 rwxp 00000000 00:0e 1055 /dev/zero 2aaaaaaab000-2aaaaabb3000 rw-p 2aaaaaaab000 00:00 0 2aaaaabb3000-2aaaaabb4000 rw-s fd060000 00:0e 14745 /dev/nvidia0 2aaaaabe5000-2aaaaac24000 r--p 00000000 08:07 4039488 /usr/lib/locale/en_AU.utf8/LC_CTYPE 2aaaaac24000-2aaaaac2b000 r--s 00000000 08:07 4015800 /usr/lib/gconv/gconv-modules.cache 2aaaaac2b000-2aaaaac2c000 rw-s 11a7a9000 00:0e 14745 /dev/nvidia0 2aaaaac2c000-2aaaaac2d000 rw-s fd641000 00:0e 14745 /dev/nvidia0 2aaaaac2d000-2aaaaac2e000 rw-s cf9ac000 00:0e 14745 /dev/nvidia0 2aaaaac2e000-2aaaaac2f000 rw-s cf837000 00:0e 14745 /dev/nvidia0 2aaaaac2f000-2aaaaac30000 rw-s e0900000 00:0e 14745 /dev/nvidia0 2aaaaac30000-2aaaaac34000 rw-s b95ef000 00:0e 14745 /dev/nvidia0 2aaaaac4c000-2aaaaac55000 r-xp 00000000 08:07 4014300 /usr/lib/libXcursor.so.1.0.2 2aaaaac55000-2aaaaae55000 ---p 00009000 08:07 4014300 /usr/lib/libXcursor.so.1.0.2 2aaaaae55000-2aaaaae56000 rw-p 00009000 08:07 4014300 /usr/lib/libXcursor.so.1.0.2 2aaaaae56000-2aaaaae5f000 r-xp 00000000 08:07 4014330 /usr/lib/libXrender.so.1.3.0 2aaaaae5f000-2aaaab05e000 ---p 00009000 08:07 4014330 /usr/lib/libXrender.so.1.3.0 2aaaab05e000-2aaaab05f000 rw-p 00008000 08:07 4014330 /usr/lib/libXrender.so.1.3.0 2aaaab05f000-2aaaab064000 r-xp 00000000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab064000-2aaaab263000 ---p 00005000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab263000-2aaaab264000 rw-p 00004000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab05f000-2aaaab064000 r-xp 00000000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab064000-2aaaab263000 ---p 00005000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab263000-2aaaab264000 rw-p 00004000 08:07 4014310 /usr/lib/libXfixes.so.3.1.0 2aaaab264000-2aaaab286000 rw-s 00000000 00:09 0 /SYSV00000000 (deleted) 2aaaab2a6000-2aaaab2e6000 rw-s d3fa6000 00:0e 14745 /dev/nvidia0 2aaaab2e6000-2aaaab2e7000 rw-s fdc06000 00:0e 14745 /dev/nvidia0 2aaaab2e7000-2aaaab3e7000 rw-s d3e64000 00:0e 14745 /dev/nvidia0 2aaaab3e7000-2aaaab3eb000 rw-s df790000 00:0e 14745 /dev/nvidia0 2aaaab3eb000-2aaaab5eb000 rw-s df7d5000 00:0e 14745 /dev/nvidia0 2aaaab5eb000-2aaaab5ec000 rw-s 00000000 00:09 655367 /SYSV00000000 (deleted) 2aaaab5ec000-2aaaab5ed000 rw-s 00000000 00:09 688136 /SYSV00000000 (deleted) 2aaaab5ed000-2aaaab5fd000 rw-s e09ec000 00:0e 14745 /dev/nvidia0 2aaaab612000-2aaaab615000 rw-s 00000000 00:09 156094 /dev/zero (deleted) 2aaaab615000-2aaaab618000 rw-s 00000000 00:09 156095 /dev/zero (deleted) 2aaaab618000-2aaaab61b000 rw-s 00000000 00:09 156096 /dev/zero (deleted) 2aaaab61b000-2aaaab61e000 rw-s 00000000 00:09 156097 /dev/zero (deleted) 2aaaab61e000-2aaaab621000 rw-s 00000000 00:09 156098 /dev/zero (deleted) 2aaaab621000-2aaaab624000 rw-s 00000000 00:09 156099 /dev/zero (deleted) 2aaaab624000-2aaaab627000 rw-s 00000000 00:09 156100 /dev/zero (deleted) 2aaaab627000-2aaaab62a000 rw-s 00000000 00:09 156101 /dev/zero (deleted) 2aaaab62a000-2aaaab62d000 rw-s 00000000 00:09 156102 /dev/zero (deleted) 2aaaab62d000-2aaaab630000 rw-s 00000000 00:09 156103 /dev/zero (deleted) 2aaaab630000-2aaaab687000 r-xp 00000000 08:07 4015079 /usr/lib/libtiff.so.4.2.1 2aaaab687000-2aaaab886000 ---p 00057000 08:07 4015079 /usr/lib/libtiff.so.4.2.1 2aaaab886000-2aaaab889000 rw-p 00056000 08:07 4015079 /usr/lib/libtiff.so.4.2.1 2aaaab889000-2aaaab909000 rw-s d3f48000 00:0e 14745 /dev/nvidia0 2aaaab92a000-2aaaab932000 r-xp 00000000 08:06 1052724 /lib/libnss_compat-2.7.so 2aaaab932000-2aaaabb31000 ---p 00008000 08:06 1052724 /lib/libnss_compat-2.7.so 2aaaabb31000-2aaaabb33000 rw-p 00007000 08:06 1052724 /lib/libnss_compat-2.7.so 2aaaabb33000-2aaaabb49000 r-xp 00000000 08:06 1052722 /lib/libnsl-2.7.so 2aaaabb49000-2aaaabd48000 ---p 00016000 08:06 1052722 /lib/libnsl-2.7.so 2aaaabd48000-2aaaabd4a000 rw-p 00015000 08:06 1052722 /lib/libnsl-2.7.so 2aaaabd4a000-2aaaabd4c000 rw-p 2aaaabd4a000 00:00 0 2aaaabd4c000-2aaaabd56000 r-xp 00000000 08:06 1052738 /lib/libnss_nis-2.7.so 2aaaabd56000-2aaaabf55000 ---p 0000a000 08:06 1052738 /lib/libnss_nis-2.7.so 2aaaabf55000-2aaaabf57000 rw-p 00009000 08:06 1052738 /lib/libnss_nis-2.7.so 2aaaabf57000-2aaaabf61000 r-xp 00000000 08:06 1052728 /lib/libnss_files-2.7.so 2aaaabf61000-2aaaac161000 ---p 0000a000 08:06 1052728 /lib/libnss_files-2.7.so 2aaaac161000-2aaaac163000 rw-p 0000a000 08:06 1052728 /lib/libnss_files-2.7.so 2aaaac163000-2aaaac363000 rw-s 11a98c000 00:0e 14745 /dev/nvidia0 2aaaac363000-2aaaac563000 rw-s cfbcb000 00:0e 14745 /dev/nvidia0 2aaaac564000-2aaaac625000 rw-p 2aaaac564000 00:00 0 2aaaac625000-2aaaac651000 r-xp 00000000 08:07 4080951 /usr/lib/python-support/python-gobject/python2.5/gtk-2.0/gobject/_gobject.so 2aaaac651000-2aaaac850000 ---p 0002c000 08:07 4080951 /usr/lib/python-support/python-gobject/python2.5/gtk-2.0/gobject/_gobject.so 2aaaac850000-2aaaac856000 rw-p 0002b000 08:07 4080951 /usr/lib/python-support/python-gobject/python2.5/gtk-2.0/gobject/_gobject.so 2aaaac877000-2aaaac8ba000 r-xp 00000000 08:07 4016934 /usr/lib/libgobject-2.0.so.0.1600.3 2aaaac8ba000-2aaaacaba000 ---p 00043000 08:07 4016934 /usr/lib/libgobject-2.0.so.0.1600.3 2aaaacaba000-2aaaacabc000 rw-p 00043000 08:07 4016934 /usr/lib/libgobject-2.0.so.0.1600.3 2aaaacabc000-2aaaacac0000 r-xp 00000000 08:07 4017044 /usr/lib/libgthread-2.0.so.0.1600.3 2aaaacac0000-2aaaaccbf000 ---p 00004000 08:07 4017044 /usr/lib/libgthread-2.0.so.0.1600.3 2aaaaccbf000-2aaaaccc0000 rw-p 00003000 08:07 4017044 /usr/lib/libgthread-2.0.so.0.1600.3 2aaaaccc0000-2aaaaccc8000 r-xp 00000000 08:06 1052762 /lib/librt-2.7.so 2aaaaccc8000-2aaaacec7000 ---p 00008000 08:06 1052762 /lib/librt-2.7.so 2aaaacec7000-2aaaacec9000 rw-p 00007000 08:06 1052762 /lib/librt-2.7.so 2aaaacec9000-2aaaacf88000 r-xp 00000000 08:07 4016909 /usr/lib/libglib-2.0.so.0.1600.3 2aaaacf88000-2aaaad187000 ---p 000bf000 08:07 4016909 /usr/lib/libglib-2.0.so.0.1600.3 2aaaad187000-2aaaad188000 rw-p 000be000 08:07 4016909 /usr/lib/libglib-2.0.so.0.1600.3 2aaaad188000-2aaaad189000 rw-p 2aaaad188000 00:00 0 2aaaad189000-2aaaad190000 r-xp 00000000 08:07 4014536 /usr/lib/libffi.so.4.0.1 2aaaad190000-2aaaad390000 ---p 00007000 08:07 4014536 /usr/lib/libffi.so.4.0.1 2aaaad390000-2aaaad391000 rw-p 00007000 08:07 4014536 /usr/lib/libffi.so.4.0.1 2aaaad391000-2aaaad3aa000 r-xp 00000000 08:06 1052764 /lib/libselinux.so.1 2aaaad3aa000-2aaaad5aa000 ---p 00019000 08:06 1052764 /lib/libselinux.so.1 2aaaad5aa000-2aaaad5ac000 rw-p 00019000 08:06 1052764 /lib/libselinux.so.1 2aaaad5ac000-2aaaad5ad000 rw-p 2aaaad5ac000 00:00 0 2aaaad5ad000-2aaaad5d2000 r-xp 00000000 08:07 4014952 /usr/lib/libpcre.so.3.12.1 2aaaad5d2000-2aaaad7d2000 ---p 00025000 08:07 4014952 /usr/lib/libpcre.so.3.12.1 2aaaad7d2000-2aaaad7d3000 rw-p 00025000 08:07 4014952 /usr/lib/libpcre.so.3.12.1 2aaaad7d3000-2aaaada01000 r-xp 00000000 08:07 4087815 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/gtk/_gtk.so 2aaaada01000-2aaaadc00000 ---p 0022e000 08:07 4087815 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/gtk/_gtk.so 2aaaadc00000-2aaaadc4d000 rw-p 0022d000 08:07 4087815 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/gtk/_gtk.so 2aaaadc6e000-2aaaae031000 r-xp 00000000 08:07 4017053 /usr/lib/libgtk-x11-2.0.so.0.1200.9 2aaaae031000-2aaaae230000 ---p 003c3000 08:07 4017053 /usr/lib/libgtk-x11-2.0.so.0.1200.9 2aaaae230000-2aaaae23b000 rw-p 003c2000 08:07 4017053 /usr/lib/libgtk-x11-2.0.so.0.1200.9 2aaaae23b000-2aaaae23d000 rw-p 2aaaae23b000 00:00 0 2aaaae23d000-2aaaae2d2000 r-xp 00000000 08:07 4017050 /usr/lib/libgdk-x11-2.0.so.0.1200.9 2aaaae2d2000-2aaaae4d2000 ---p 00095000 08:07 4017050 /usr/lib/libgdk-x11-2.0.so.0.1200.9 2aaaae4d2000-2aaaae4d7000 rw-p 00095000 08:07 4017050 /usr/lib/libgdk-x11-2.0.so.0.1200.9 2aaaae4d7000-2aaaae4f5000 r-xp 00000000 08:07 4014371 /usr/lib/libatk-1.0.so.0.2209.1 2aaaae4f5000-2aaaae6f4000 ---p 0001e000 08:07 4014371 /usr/lib/libatk-1.0.so.0.2209.1 2aaaae6f4000-2aaaae6f7000 rw-p 0001d000 08:07 4014371 /usr/lib/libatk-1.0.so.0.2209.1 2aaaae6f7000-2aaaae710000 r-xp 00000000 08:07 4017051 /usr/lib/libgdk_pixbuf-2.0.so.0.1200.9 2aaaae710000-2aaaae910000 ---p 00019000 08:07 4017051 /usr/lib/libgdk_pixbuf-2.0.so.0.1200.9 2aaaae910000-2aaaae911000 rw-p 00019000 08:07 4017051 /usr/lib/libgdk_pixbuf-2.0.so.0.1200.9 2aaaae911000-2aaaae91b000 r-xp 00000000 08:07 4014936 /usr/lib/libpangocairo-1.0.so.0.2000.1 2aaaae91b000-2aaaaeb1b000 ---p 0000a000 08:07 4014936 /usr/lib/libpangocairo-1.0.so.0.2000.1 2aaaaeb1b000-2aaaaeb1c000 rw-p 0000a000 08:07 4014936 /usr/lib/libpangocairo-1.0.so.0.2000.1 2aaaaeb1c000-2aaaaeb5e000 r-xp 00000000 08:07 4014934 /usr/lib/libpango-1.0.so.0.2000.1 2aaaaeb5e000-2aaaaed5d000 ---p 00042000 08:07 4014934 /usr/lib/libpango-1.0.so.0.2000.1 2aaaaed5d000-2aaaaed60000 rw-p 00041000 08:07 4014934 /usr/lib/libpango-1.0.so.0.2000.1 2aaaaed60000-2aaaaedc9000 r-xp 00000000 08:07 4014406 /usr/lib/libcairo.so.2.17.3 2aaaaedc9000-2aaaaefc8000 ---p 00069000 08:07 4014406 /usr/lib/libcairo.so.2.17.3 2aaaaefc8000-2aaaaefcb000 rw-p 00068000 08:07 4014406 /usr/lib/libcairo.so.2.17.3 2aaaaefcb000-2aaaaefce000 r-xp 00000000 08:07 4016933 /usr/lib/libgmodule-2.0.so.0.1600.3 2aaaaefce000-2aaaaf1cd000 ---p 00003000 08:07 4016933 /usr/lib/libgmodule-2.0.so.0.1600.3 2aaaaf1cd000-2aaaaf1ce000 rw-p 00002000 08:07 4016933 /usr/lib/libgmodule-2.0.so.0.1600.3 2aaaaf1ce000-2aaaaf1d0000 r-xp 00000000 08:07 4014298 /usr/lib/libXcomposite.so.1.0.0 2aaaaf1d0000-2aaaaf3cf000 ---p 00002000 08:07 4014298 /usr/lib/libXcomposite.so.1.0.0 2aaaaf3cf000-2aaaaf3d0000 rw-p 00001000 08:07 4014298 /usr/lib/libXcomposite.so.1.0.0 2aaaaf3d0000-2aaaaf3d2000 r-xp 00000000 08:07 4014302 /usr/lib/libXdamage.so.1.1.0 2aaaaf3d2000-2aaaaf5d1000 ---p 00002000 08:07 4014302 /usr/lib/libXdamage.so.1.1.0 2aaaaf5d1000-2aaaaf5d2000 rw-p 00001000 08:07 4014302 /usr/lib/libXdamage.so.1.1.0 2aaaaf5d2000-2aaaaf601000 r-xp 00000000 08:07 4014538 /usr/lib/libfontconfig.so.1.3.0 2aaaaf601000-2aaaaf801000 ---p 0002f000 08:07 4014538 /usr/lib/libfontconfig.so.1.3.0 2aaaaf801000-2aaaaf803000 rw-p 0002f000 08:07 4014538 /usr/lib/libfontconfig.so.1.3.0 2aaaaf803000-2aaaaf805000 r-xp 00000000 08:07 4014318 /usr/lib/libXinerama.so.1.0.0 2aaaaf805000-2aaaafa04000 ---p 00002000 08:07 4014318 /usr/lib/libXinerama.so.1.0.0 2aaaafa04000-2aaaafa05000 rw-p 00001000 08:07 4014318 /usr/lib/libXinerama.so.1.0.0 2aaaafa05000-2aaaafa0c000 r-xp 00000000 08:07 4014328 /usr/lib/libXrandr.so.2.1.0 2aaaafa0c000-2aaaafc0b000 ---p 00007000 08:07 4014328 /usr/lib/libXrandr.so.2.1.0 2aaaafc0b000-2aaaafc0c000 rw-p 00006000 08:07 4014328 /usr/lib/libXrandr.so.2.1.0 2aaaafc0c000-2aaaafc37000 r-xp 00000000 08:07 4014938 /usr/lib/libpangoft2-1.0.so.0.2000.1 2aaaafc37000-2aaaafe37000 ---p 0002b000 08:07 4014938 /usr/lib/libpangoft2-1.0.so.0.2000.1 2aaaafe37000-2aaaafe38000 rw-p 0002b000 08:07 4014938 /usr/lib/libpangoft2-1.0.so.0.2000.1 2aaaafe38000-2aaaafe64000 r-xp 00000000 08:07 4014962 /usr/lib/libpixman-1.so.0.10.0 2aaaafe64000-2aaab0064000 ---p 0002c000 08:07 4014962 /usr/lib/libpixman-1.so.0.10.0 2aaab0064000-2aaab0066000 rw-p 0002c000 08:07 4014962 /usr/lib/libpixman-1.so.0.10.0 2aaab0066000-2aaab0088000 r-xp 00000000 08:07 4014530 /usr/lib/libexpat.so.1.5.2 2aaab0088000-2aaab0288000 ---p 00022000 08:07 4014530 /usr/lib/libexpat.so.1.5.2 2aaab0288000-2aaab028a000 rw-p 00022000 08:07 4014530 /usr/lib/libexpat.so.1.5.2 2aaab028a000-2aaab029d000 r-xp 00000000 08:07 4089204 /usr/lib/python2.5/site-packages/cairo/_cairo.so 2aaab029d000-2aaab049d000 ---p 00013000 08:07 4089204 /usr/lib/python2.5/site-packages/cairo/_cairo.so 2aaab049d000-2aaab04a1000 rw-p 00013000 08:07 4089204 /usr/lib/python2.5/site-packages/cairo/_cairo.so 2aaab04a1000-2aaab04c4000 r-xp 00000000 08:07 4087813 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pango.so 2aaab04c4000-2aaab06c4000 ---p 00023000 08:07 4087813 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pango.so 2aaab06c4000-2aaab06cb000 rw-p 00023000 08:07 4087813 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pango.so 2aaab06cb000-2aaab070a000 r-xp 00000000 08:07 4087811 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/atk.so 2aaab070a000-2aaab0909000 ---p 0003f000 08:07 4087811 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/atk.so 2aaab0909000-2aaab0911000 rw-p 0003e000 08:07 4087811 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/atk.so 2aaab0911000-2aaab0915000 r-xp 00000000 08:07 4087814 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pangocairo.so 2aaab0915000-2aaab0b15000 ---p 00004000 08:07 4087814 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pangocairo.so 2aaab0b15000-2aaab0b16000 rw-p 00004000 08:07 4087814 /usr/lib/python-support/python-gtk2/python2.5/gtk-2.0/pangocairo.so 2aaab0b16000-2aaab0b17000 r--p 00000000 08:07 1624345 /usr/lib/locale/en_AU.utf8/LC_IDENTIFICATION 2aaab0b17000-2aaab0b18000 r--p 00000000 08:07 4039490 /usr/lib/locale/en_AU.utf8/LC_MEASUREMENT 2aaab0b18000-2aaab0b19000 r--p 00000000 08:07 1624344 /usr/lib/locale/en_AU.utf8/LC_TELEPHONE 2aaab0b19000-2aaab0b1a000 r--p 00000000 08:07 1624343 /usr/lib/locale/en_AU.utf8/LC_ADDRESS 2aaab0b1a000-2aaab0b1b000 r--p 00000000 08:07 4039561 /usr/lib/locale/en_AU.utf8/LC_NAME 2aaab0b1b000-2aaab0b1c000 r--p 00000000 08:07 4039494 /usr/lib/locale/en_AU.utf8/LC_PAPER 2aaab0b1c000-2aaab0b1d000 r--p 00000000 08:07 4039510 /usr/lib/locale/en_AU.utf8/LC_MESSAGES/SYS_LC_MESSAGES 2aaab0b1d000-2aaab0b1e000 r--p 00000000 08:07 1624342 /usr/lib/locale/en_AU.utf8/LC_MONETARY 2aaab0b1e000-2aaab0bff000 r--p 00000000 08:07 4039487 /usr/lib/locale/en_AU.utf8/LC_COLLATE 2aaab0bff000-2aaab0c00000 r--p 00000000 08:07 4039509 /usr/lib/locale/en_AU.utf8/LC_TIME 2aaab0c00000-2aaab0c01000 r--p 00000000 08:07 4039562 /usr/lib/locale/en_AU.utf8/LC_NUMERIC 2aaab0c01000-2aaab0c06000 r--p 00000000 08:07 1558330 /usr/share/locale-langpack/en_AU/LC_MESSAGES/gtk20.mo 2aaab0c06000-2aaab0c07000 r--p 00000000 08:07 1558331 /usr/share/locale-langpack/en_AU/LC_MESSAGES/gtk20-properties.mo 2b7b7611e000-2b7b7613b000 r-xp 00000000 08:06 1052658 /lib/ld-2.7.so 2b7b7613b000-2b7b7613e000 rw-p 2b7b7613b000 00:00 0 2b7b7633b000-2b7b7633d000 rw-p 0001d000 08:06 1052658 /lib/ld-2.7.so 2b7b7633d000-2b7b7635f000 r-xp 00000000 08:07 4014823 /usr/lib/libjpeg.so.62.0.0 2b7b7635f000-2b7b7655f000 ---p 00022000 08:07 4014823 /usr/lib/libjpeg.so.62.0.0 2b7b7655f000-2b7b76560000 rw-p 00022000 08:07 4014823 /usr/lib/libjpeg.so.62.0.0 2b7b76560000-2b7b76584000 r-xp 00000000 08:07 4014966 /usr/lib/libpng12.so.0.15.0 2b7b76584000-2b7b76784000 ---p 00024000 08:07 4014966 /usr/lib/libpng12.so.0.15.0 2b7b76784000-2b7b76785000 rw-p 00024000 08:07 4014966 /usr/lib/libpng12.so.0.15.0 2b7b76785000-2b7b7679b000 r-xp 00000000 08:07 4015144 /usr/lib/libz.so.1.2.3.3 2b7b7679b000-2b7b7699b000 ---p 00016000 08:07 4015144 /usr/lib/libz.so.1.2.3.3 2b7b7699b000-2b7b7699c000 rw-p 00016000 08:07 4015144 /usr/lib/libz.so.1.2.3.3 2b7b7699c000-2b7b7699d000 rw-p 2b7b7699c000 00:00 0 2b7b7699d000-2b7b76ad2000 r-xp 00000000 08:07 4015000 /usr/lib/libpython2.5.so.1.0 2b7b76ad2000-2b7b76cd2000 ---p 00135000 08:07 4015000 /usr/lib/libpython2.5.so.1.0 2b7b76cd2000-2b7b76d05000 rw-p 00135000 08:07 4015000 /usr/lib/libpython2.5.so.1.0 2b7b76d05000-2b7b76d0d000 rw-p 2b7b76d05000 00:00 0 2b7b76d0d000-2b7b76d87000 r-xp 00000000 08:07 4014546 /usr/lib/libfreetype.so.6.3.16 2b7b76d87000-2b7b76f87000 ---p 0007a000 08:07 4014546 /usr/lib/libfreetype.so.6.3.16 2b7b76f87000-2b7b76f8c000 rw-p 0007a000 08:07 4014546 /usr/lib/libfreetype.so.6.3.16 2b7b76f8c000-2b7b770a5000 r-xp 00000000 08:07 4017591 /usr/lib/libgettextlib-0.17.so 2b7b770a5000-2b7b772a4000 ---p 00119000 08:07 4017591 /usr/lib/libgettextlib-0.17.so 2b7b772a4000-2b7b772aa000 rw-p 00118000 08:07 4017591 /usr/lib/libgettextlib-0.17.so 2b7b772aa000-2b7b772ad000 rw-p 2b7b772aa000 00:00 0 2b7b772ad000-2b7b772f4000 r-xp 00000000 08:07 4014920 /usr/lib/libopenal.so.0.0.0 2b7b772f4000-2b7b774f4000 ---p 00047000 08:07 4014920 /usr/lib/libopenal.so.0.0.0 2b7b774f4000-2b7b774f6000 rw-p 00047000 08:07 4014920 /usr/lib/libopenal.so.0.0.0 2b7b774f6000-2b7b774fb000 rw-p 2b7b774f6000 00:00 0 2b7b774fb000-2b7b77501000 r-xp 00000000 08:07 4014350 /usr/lib/libalut.so.0.1.0 2b7b77501000-2b7b77700000 ---p 00006000 08:07 4014350 /usr/lib/libalut.so.0.1.0 2b7b77700000-2b7b77703000 rw-p 00005000 08:07 4014350 /usr/lib/libalut.so.0.1.0 2b7b77703000-2b7b77745000 r-xp 00000000 08:07 4014252 /usr/lib/libHalf.so.2.0.2 2b7b77745000-2b7b77944000 ---p 00042000 08:07 4014252 /usr/lib/libHalf.so.2.0.2 2b7b77944000-2b7b77945000 rw-p 00041000 08:07 4014252 /usr/lib/libHalf.so.2.0.2 2b7b77945000-2b7b77946000 rw-p 2b7b77945000 00:00 0 2b7b77946000-2b7b779b4000 r-xp 00000000 08:07 4014263 /usr/lib/libIlmImf.so.2.0.2 2b7b779b4000-2b7b77bb3000 ---p 0006e000 08:07 4014263 /usr/lib/libIlmImf.so.2.0.2 2b7b77bb3000-2b7b77bb6000 rw-p 0006d000 08:07 4014263 /usr/lib/libIlmImf.so.2.0.2 2b7b77bb6000-2b7b77bd0000 r-xp 00000000 08:07 4014261 /usr/lib/libIex.so.2.0.2 2b7b77bd0000-2b7b77dd0000 ---p 0001a000 08:07 4014261 /usr/lib/libIex.so.2.0.2 2b7b77dd0000-2b7b77dd4000 rw-p 0001a000 08:07 4014261 /usr/lib/libIex.so.2.0.2 2b7b77dd4000-2b7b77dd8000 r-xp 00000000 08:07 4014265 /usr/lib/libImath.so.2.0.2 2b7b77dd8000-2b7b77fd8000 ---p 00004000 08:07 4014265 /usr/lib/libImath.so.2.0.2 2b7b77fd8000-2b7b77fd9000 rw-p 00004000 08:07 4014265 /usr/lib/libImath.so.2.0.2 2b7b77fd9000-2b7b77fda000 rw-p 2b7b77fd9000 00:00 0 2b7b77fda000-2b7b78041000 r-xp 00000000 08:07 4014279 /usr/lib/libSDL-1.2.so.0.11.1 2b7b78041000-2b7b78241000 ---p 00067000 08:07 4014279 /usr/lib/libSDL-1.2.so.0.11.1 2b7b78241000-2b7b78243000 rw-p 00067000 08:07 4014279 /usr/lib/libSDL-1.2.so.0.11.1 2b7b78243000-2b7b7826f000 rw-p 2b7b78243000 00:00 0 2b7b7826f000-2b7b78308000 r-xp 00000000 08:07 4015550 /usr/lib/libGL.so.169.12 2b7b78308000-2b7b78407000 ---p 00099000 08:07 4015550 /usr/lib/libGL.so.169.12 2b7b78407000-2b7b7843b000 rwxp 00098000 08:07 4015550 /usr/lib/libGL.so.169.12 2b7b7843b000-2b7b7843d000 rwxp 2b7b7843b000 00:00 0 2b7b7843d000-2b7b784c0000 r-xp 00000000 08:07 4014250 /usr/lib/libGLU.so.1.3.070002 2b7b784c0000-2b7b786bf000 ---p 00083000 08:07 4014250 /usr/lib/libGLU.so.1.3.070002 2b7b786bf000-2b7b786c1000 rw-p 00082000 08:07 4014250 /usr/lib/libGLU.so.1.3.070002 2b7b786c1000-2b7b786c2000 rw-p 2b7b786c1000 00:00 0 2b7b786c2000-2b7b787c1000 r-xp 00000000 08:07 4014287 /usr/lib/libX11.so.6.2.0 2b7b787c1000-2b7b789c0000 ---p 000ff000 08:07 4014287 /usr/lib/libX11.so.6.2.0 2b7b789c0000-2b7b789c5000 rw-p 000fe000 08:07 4014287 /usr/lib/libX11.so.6.2.0 2b7b789c5000-2b7b789ce000 r-xp 00000000 08:07 4014316 /usr/lib/libXi.so.6.0.0 2b7b789ce000-2b7b78bcd000 ---p 00009000 08:07 4014316 /usr/lib/libXi.so.6.0.0 2b7b78bcd000-2b7b78bce000 rw-p 00008000 08:07 4014316 /usr/lib/libXi.so.6.0.0 2b7b78bce000-2b7b78bd0000 r-xp 00000000 08:06 1052783 /lib/libutil-2.7.so 2b7b78bd0000-2b7b78dcf000 ---p 00002000 08:06 1052783 /lib/libutil-2.7.so 2b7b78dcf000-2b7b78dd1000 rw-p 00001000 08:06 1052783 /lib/libutil-2.7.so 2b7b78dd1000-2b7b78dd2000 rw-p 2b7b78dd1000 00:00 0 2b7b78dd2000-2b7b78f2a000 r-xp 00000000 08:06 1052678 /lib/libc-2.7.so 2b7b78f2a000-2b7b7912a000 ---p 00158000 08:06 1052678 /lib/libc-2.7.so 2b7b7912a000-2b7b7912d000 r--p 00158000 08:06 1052678 /lib/libc-2.7.so 2b7b7912d000-2b7b7912f000 rw-p 0015b000 08:06 1052678 /lib/libc-2.7.so 2b7b7912f000-2b7b79134000 rw-p 2b7b7912f000 00:00 0 2b7b79134000-2b7b791b4000 r-xp 00000000 08:06 1052715 /lib/libm-2.7.so 2b7b791b4000-2b7b793b3000 ---p 00080000 08:06 1052715 /lib/libm-2.7.so 2b7b793b3000-2b7b793b5000 rw-p 0007f000 08:06 1052715 /lib/libm-2.7.so 2b7b793b5000-2b7b793b7000 r-xp 00000000 08:06 1052695 /lib/libdl-2.7.so 2b7b793b7000-2b7b795b7000 ---p 00002000 08:06 1052695 /lib/libdl-2.7.so 2b7b795b7000-2b7b795b9000 rw-p 00002000 08:06 1052695 /lib/libdl-2.7.so 2b7b795b9000-2b7b795ba000 rw-p 2b7b795b9000 00:00 0 2b7b795ba000-2b7b795d0000 r-xp 00000000 08:06 1052756 /lib/libpthread-2.7.so 2b7b795d0000-2b7b797d0000 ---p 00016000 08:06 1052756 /lib/libpthread-2.7.so 2b7b797d0000-2b7b797d2000 rw-p 00016000 08:06 1052756 /lib/libpthread-2.7.so 2b7b797d2000-2b7b797d6000 rw-p 2b7b797d2000 00:00 0 2b7b797d6000-2b7b798c5000 r-xp 00000000 08:07 4015069 /usr/lib/libstdc++.so.6.0.9 2b7b798c5000-2b7b79ac5000 ---p 000ef000 08:07 4015069 /usr/lib/libstdc++.so.6.0.9 2b7b79ac5000-2b7b79acb000 r--p 000ef000 08:07 4015069 /usr/lib/libstdc++.so.6.0.9 2b7b79acb000-2b7b79ace000 rw-p 000f5000 08:07 4015069 /usr/lib/libstdc++.so.6.0.9 2b7b79ace000-2b7b79ae1000 rw-p 2b7b79ace000 00:00 0 2b7b79ae1000-2b7b79aee000 r-xp 00000000 08:06 1052703 /lib/libgcc_s.so.1 2b7b79aee000-2b7b79cee000 ---p 0000d000 08:06 1052703 /lib/libgcc_s.so.1 2b7b79cee000-2b7b79cef000 rw-p 0000d000 08:06 1052703 /lib/libgcc_s.so.1 2b7b79cef000-2b7b79cf0000 rw-p 2b7b79cef000 00:00 0 2b7b79cf0000-2b7b79dc7000 r-xp 00000000 08:07 4014365 /usr/lib/libasound.so.2.0.0 2b7b79dc7000-2b7b79fc7000 ---p 000d7000 08:07 4014365 /usr/lib/libasound.so.2.0.0 2b7b79fc7000-2b7b79fce000 rw-p 000d7000 08:07 4014365 /usr/lib/libasound.so.2.0.0 2b7b79fce000-2b7b7a039000 r-xp 00000000 08:07 4014480 /usr/lib/libdirectfb-1.0.so.0.1.0 2b7b7a039000-2b7b7a239000 ---p 0006b000 08:07 4014480 /usr/lib/libdirectfb-1.0.so.0.1.0 2b7b7a239000-2b7b7a23c000 rw-p 0006b000 08:07 4014480 /usr/lib/libdirectfb-1.0.so.0.1.0 2b7b7a23c000-2b7b7a23d000 rw-p 2b7b7a23c000 00:00 0 2b7b7a23d000-2b7b7a245000 r-xp 00000000 08:07 4014550 /usr/lib/libfusion-1.0.so.0.1.0 2b7b7a245000-2b7b7a444000 ---p 00008000 08:07 4014550 /usr/lib/libfusion-1.0.so.0.1.0 2b7b7a444000-2b7b7a445000 rw-p 00007000 08:07 4014550 /usr/lib/libfusion-1.0.so.0.1.0 2b7b7a445000-2b7b7a459000 r-xp 00000000 08:07 4014478 /usr/lib/libdirect-1.0.so.0.1.0 2b7b7a459000-2b7b7a658000 ---p 00014000 08:07 4014478 /usr/lib/libdirect-1.0.so.0.1.0 2b7b7a658000-2b7b7a65a000 rw-p 00013000 08:07 4014478 /usr/lib/libdirect-1.0.so.0.1.0 2b7b7a65a000-2b7b7af97000 r-xp 00000000 08:07 4015551 /usr/lib/libGLcore.so.169.12 2b7b7af97000-2b7b7b097000 ---p 0093d000 08:07 4015551 /usr/lib/libGLcore.so.169.12 2b7b7b097000-2b7b7b20a000 rwxp 0093d000 08:07 4015551 /usr/lib/libGLcore.so.169.12 2b7b7b20a000-2b7b7b211000 rwxp 2b7b7b20a000 00:00 0 2b7b7b211000-2b7b7b212000 rw-p 2b7b7b211000 00:00 0 2b7b7b212000-2b7b7b213000 r-xp 00000000 08:07 4015553 /usr/lib/libnvidia-tls.so.169.12 2b7b7b213000-2b7b7b312000 ---p 00001000 08:07 4015553 /usr/lib/libnvidia-tls.so.169.12 2b7b7b312000-2b7b7b313000 rw-p 00000000 08:07 4015553 /usr/lib/libnvidia-tls.so.169.12 2b7b7b313000-2b7b7b324000 r-xp 00000000 08:07 4014308 /usr/lib/libXext.so.6.4.0 2b7b7b324000-2b7b7b523000 ---p 00011000 08:07 4014308 /usr/lib/libXext.so.6.4.0 2b7b7b523000-2b7b7b524000 rw-p 00010000 08:07 4014308 /usr/lib/libXext.so.6.4.0 2b7b7b524000-2b7b7b525000 r-xp 00000000 08:07 4015130 /usr/lib/libxcb-xlib.so.0.0.0 2b7b7b525000-2b7b7b724000 ---p 00001000 08:07 4015130 /usr/lib/libxcb-xlib.so.0.0.0 2b7b7b724000-2b7b7b725000 rw-p 00000000 08:07 4015130 /usr/lib/libxcb-xlib.so.0.0.0 2b7b7b725000-2b7b7b726000 rw-p 2b7b7b725000 00:00 0 2b7b7b726000-2b7b7b741000 r-xp 00000000 08:07 4015132 /usr/lib/libxcb.so.1.0.0 2b7b7b741000-2b7b7b940000 ---p 0001b000 08:07 4015132 /usr/lib/libxcb.so.1.0.0 2b7b7b940000-2b7b7b941000 rw-p 0001a000 08:07 4015132 /usr/lib/libxcb.so.1.0.0 2b7b7b941000-2b7b7b943000 r-xp 00000000 08:07 4014293 /usr/lib/libXau.so.6.0.0 2b7b7b943000-2b7b7bb42000 ---p 00002000 08:07 4014293 /usr/lib/libXau.so.6.0.0 2b7b7bb42000-2b7b7bb43000 rw-p 00001000 08:07 4014293 /usr/lib/libXau.so.6.0.0 2b7b7bb43000-2b7b7bb44000 rw-p 2b7b7bb43000 00:00 0 2b7b7bb44000-2b7b7bb49000 r-xp 00000000 08:07 4014304 /usr/lib/libXdmcp.so.6.0.0 2b7b7bb49000-2b7b7bd48000 ---p 00005000 08:07 4014304 /usr/lib/libXdmcp.so.6.0.0 2b7b7bd48000-2b7b7bd49000 rw-p 00004000 08:07 4014304 /usr/lib/libXdmcp.so.6.0.0 2b7b7bd49000-2b7b7bd4c000 rw-p 2b7b7bd49000 00:00 0 7fff34975000-7fff34989000 rwxp 7ffffffe9000 00:00 0 [stack] 7fff34989000-7fff3498b000 rw-p 7fffffffd000 00:00 0 7fff349fd000-7fff349ff000 r-xp 7fff349fd000 00:00 0 [vdso] ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall] Aborted [/error] However, using the Blender (2.45) that comes with Ubuntu 8.04, this error does not occur. Furthermore, the exact same build on my Unbutu 32 bit x86 OS/machine also works without a problem. Thanks a lot. Patrick";"None";"None";"None" 17270;1;"Open";3;7656;"Gary Tolley";100;"Nobody";"2008-07-05 08:00";"1970-01-01 01:00";"2008-07-06 11:05";"Bezier points don't match frame";"I know the summary isn't much to go on, so here is what is happening I'm working on a script to convert a bezier curve to a set of IPO curves - In the attached Blend file, select the curve - make sure its in object mode - right click in the script - and select execute script - switch to the console - you should see every bezier point and handle as its float number - for this example look at the ""Current iteration at:"" number - look for 41 - take note of Px, Py and Pz values - go back to the editor - the frame number should be at 41 - in the 3D window press tab - you should now see that the transform manipulator is on a bezier point - also the ""Transform Properties"" window will change to show the current values of where the transform manipulator is - at this stage switch back to the console and move it so its next to the Transform Properties"" window - looking back and forth at each of the values for Px, Py and Pz Its quite obvious that the values in the ""Transform Properties"" window has clipped the entire float value down to 4 significant digits - Look also at the ""Empty"" which is already tracked to frame 41 you can easily see that the Emptys' Object Centre is not zero'd onto the bezier point in 3D space which it should, and I believe its due to the clipping of the float value as its is being applied to the IPO graph ";"None";"None";"None" 17293;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2008-07-08 20:39";"1970-01-01 01:00";"2008-10-08 18:19";"Crop Node problems: changes values when I set the render size for 50% or lower, and don't recover the values after I restore the 100% size";"The blender file attached illustrates it better then my words :) In the Render Nodes window: 1) Create a Crop (Add->Distort->Crop) node and link it to the Render Layer. 2) Set the Crop values X2 and Y2 to maximum values 3) Set the Crop values X1 and Y1 to medium values 4) Render the image (with Do Composite activated) in 100% size mode. Here you can see that the RenderLayer maximum size will resize the Crop maximum value. Bug 1 (small bug): shouldn't this adjust be made even before I render it? If I load normal images (Add->Input->Image) the adjustment is automatic. 5) Now render the image to 50% size mode. Bug 2 (medium bug): the X2 and Y2 value are adjusted to the new size, but not the X1 and Y1 (eventually they are clamped to fit the new maximum value). 6) Now render the image to 100% size again. Bug 3 (super bug): the X2 and Y2 value are still the adjusted values, as are X1 and Y1. 7) Now forget about the RenderLayer and add an image and a procedural texture ass input. (check the file, is the nodes below the RenderLayer. 8) Link the texture to the Image with a Color->Mix Node. 9) Now Link the texture to a Crop node and then link it to the image using the same mix node. Bug 4: Something weird happens. Maybe it is because the Texture is not an image, but the crop screw with it. The Result is completely different. Maybe this is related to this bug: https://projects.blender.org/tracker/index.php?func=detail&aid=13607&group_id=9&atid=125 Tested in Blender 2.46 and current BlenderSVN";"None";"None";"None" 17332;1;"Open";3;9343;"Andre Alves";100;"Nobody";"2008-07-15 03:53";"1970-01-01 01:00";"2008-09-08 14:27";"Render with composite nodes that blurry the image, like blur or vector blur, blurry the fields too.";"I need to do some animations for tv and for this the rendered images must have fields (lower case in my country, but if I use some composite nodes that blutty the image like, blur or vector blur, it blurry the fields lines too, making it unusable for tv, it that a way to make this images correctly with that nodes? Or make the fields processing go to the end of the render process? Or if it's not other way, how I can use the sequencer to calculate fields on an image sequence? ";"None";"None";"None" 17338;1;"Open";3;5609;"Tobias Minich";100;"Nobody";"2008-07-16 06:54";"1970-01-01 01:00";"2008-09-07 20:34";"Curve modifier is not applied correctly on curves";"OS: Windows XP Blender: SVN 15551 from GraphicAll, also occurred on earlier SVN version I want to apply a curve modifier on a Bezier curve. The deforming curve is also a Bezier curve with some tilt. The 3D view shows the modified curve correctly, but when I apply the modifier, the curve start and stops too early (its stretched by the deforming curve). The curious thing is that the part that sits correctly is also transformed correctly. The issue is probably best described by an example. The blend file contains the deforming curve (_deform), a deformed mesh object (Cube, with array modifier) and two deformed Bezier curves (strL/R_spline). As you will see, the two curves ""run on"" the mesh object (as they are supposed to do), but as soon as you apply the curve modifier on one of them, you'll observe the effect described above.";"None";"None";"None" 17343;1;"Open";3;177;"Matt Ebb ";100;"Nobody";"2008-07-17 08:20";"1970-01-01 01:00";"2008-09-22 14:24";"AO pass doesn't factor in material 'amb' value";"In shots I'm working on now, I'm rendering several characters holding glass objects. I'm using AAO, at render time, and again as a render pass, tinted and multiplied on in post, for some extra enhancement and control. The glass objects look very strange, and overly dark and opaque with the AO effect on full strength, so I've lowered the material's 'amb' value down to 0.1 or so, so it still gets some shadowing from the characters' hands, but not overpowering it. This works very nicely in the render itself, however it doesn't work at all in the AO pass - when I multiply it back on top, it's still at the full strength. The AO pass seems to be storing the total occlusion for that face, but it's also inconsistent since objects with 'amb' = 0.0 don't get any occlusion at all (even though eg. amb = 0.01 get occluded as usual). As a side issue, the AO pass also seems to store 0.0 in areas where there are no faces (sky). This is incorrect too, since the definition of the ambient occlusion pass is 1.0 = no occlusion, 0.0 = fully occluded. Obviously the sky is not fully occluded, so it should be 1.0. This means you have to do stupid tricks like adding the inverted alpha channel to it in every comp, which is annoying, resource consuming, and shouldn't be necessary in the first place. Anyway, fixing the original problem is a bit tricky, in the sense of maintaining compatibility on existing render setups. I want the amb slider to have an effect, but I also don't want to change all my existing materials that were made with a default amb 0.5 (this is an odd default to start with). Currently I've patched in the code below, which only uses the amb range 0.0-0.5 and scales it to 1.0-0.0 in the AO pass. It works well for my purposes, but it's a bit icky... Any thoughts? Index: blender/render/intern/source/shadeoutput.c =================================================================== --- blender/render/intern/source/shadeoutput.c (revision 15608) +++ blender/render/intern/source/shadeoutput.c (working copy) @@ -1600,10 +1600,14 @@ if(R.wrld.mode & WO_AMB_OCC) { if(((passflag & SCE_PASS_COMBINED) && (shi->combinedflag & SCE_PASS_AO)) || (passflag & SCE_PASS_AO)) { + float ambfac; /* AO was calculated for scanline already */ if(shi->depth) ambient_occlusion(shi); VECCOPY(shr->ao, shi->ao); + ambfac = 2.0f*(1.0f-shi->amb); + CLAMP(ambfac, 0.0f, 1.0f); + VecMulf(shr->ao, ambfac); } } ";"None";"None";"None" 17352;1;"Open";3;2630;"Eckhard M. Jäger";100;"Nobody";"2008-07-17 23:46";"1970-01-01 01:00";"2008-07-20 19:38";"No access to World gravity";"For various reasons (supporting realtime physics, particles) it would nice to have an access to the gravity slider of the World panel ""Mist Stars Pysics"". ";"None";"None";"None" 17358;1;"Open";3;9472;"josh rung";100;"Nobody";"2008-07-21 00:36";"1970-01-01 01:00";"2008-09-07 20:27";"lwo export does not export any vertex painting";"steps to reproduce: make a plane (num7, then space add -> mesh -> plane) vertex paint it so theres a clear pattern UV map it, (go to top view then in edit mode with mesh selected press u and select project from view from dropdown texture it on uv map, ( but also be sure to put entries in the F5 menu's ""MA"" field and F6's ""TE"" field) export to lwo format To observe the lack of vertex painting I was using a game engine, one similar to doom3 would work (quake4 or doom3) following the guide detailed here: http://www.katsbits.com/htm/tutorials/doom_3_texture_blending.htm The vertex painting is not visible upon importing the exported lwo file either, but I am not sure if it is supported ";"None";"None";"None" 17361;1;"Open";3;9475;"Mike Spreitzer";100;"Nobody";"2008-07-21 04:51";"1970-01-01 01:00";"2008-07-27 04:57";"ipo unable to rotate around X after object is rotated 90 degrees around Y";"Steps to reproduce the bug: Start with the standard scene (File - New). This has just one object, a cube that is 2 units long on each side. View is from above (along Z axis). The 3D view window is in Object mode. Rotate the cube +90 degrees around the Y axis. NUM3 in the 3D view window to switch to viewing along the X axis. Object - Insert Keyframe - Rot. Open an ipo curve editor window somewhere. Observe you have three channels selected: RotX, RotY, and RotZ. Select just the RotX channel. Ctrl-LMB to add one control point, e.g. at about frame=50, value=50deg. Use LMB to shift the view around. The attached file x-equals-z.blend shows this state, at frame 15. Observe that the cube has been rotated around Z, not X. For comparison, do the same procedure but use the RotZ channel instead of RotX; observe that you get the same result --- a cube rotating around Z. For another point of comparison: instead of using an ipo, just take the cube, after rotating 90 degrees around Y, and R - X to rotate around the X axis --- observe this does indeed rotate around X, not Z. This is the first release of blender that I have used, so I can not compare with another release. Hardware: MacBook Pro, with main memory increased to a total of 4 GB. Blender version 2.46 Built on 2008-05-18 12:57:56, Rev-14890 version darwin dynamic system-info.txt contains the following: ===================================== = Blender 2.46 System Information = ===================================== Platform: darwin ======== Python: ====== - Version: 2.3.5 (#1, Dec 7 2006, 14:50:51) [GCC 4.0.1 (Apple Computer, Inc. build 5363) (+4864187)] - Paths: /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python23.zip /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3 /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/plat-darwin /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/plat-mac /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/plat-mac/lib-scriptpackages /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/lib-tk /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/lib-dynload /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site-packages /System/Library/Frameworks/Python.framework/Versions/2.3/Extras/lib/python /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender/scripts /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender/scripts/bpymodules - Directories: Blender home dir: /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender Default dir for scripts: /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender/scripts Default ""bpydata/"" data dir for scripts: /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender/scripts/bpydata User defined dir for scripts: -- not found Data dir ""bpydata/"" inside user defined dir: -- not found Default config data ""bpydata/config/"" dir: /Applications/blender-2.46-OSX-py2.3-intel/blender.app/Contents/MacOS/.blender/scripts/bpydata/config - Modules: all basic ones were found. OpenGL: ====== - Renderer: NVIDIA GeForce 8600M GT OpenGL Engine - Vendor: NVIDIA Corporation - Version: 2.0 NVIDIA-1.5.8 - Extensions: GL_ARB_transpose_matrix GL_ARB_vertex_program GL_ARB_vertex_blend GL_ARB_window_pos GL_ARB_shader_objects GL_ARB_vertex_shader GL_EXT_multi_draw_arrays GL_EXT_clip_volume_hint GL_EXT_rescale_normal GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_gpu_program_parameters GL_APPLE_client_storage GL_APPLE_specular_vector GL_APPLE_transform_hint GL_APPLE_packed_pixels GL_APPLE_fence GL_APPLE_vertex_array_object GL_APPLE_vertex_program_evaluators GL_APPLE_element_array GL_APPLE_flush_render GL_APPLE_aux_depth_stencil GL_NV_texgen_reflection GL_NV_light_max_exponent GL_IBM_rasterpos_clip GL_SGIS_generate_mipmap GL_ARB_shading_language_100 GL_ARB_imaging GL_ARB_point_parameters GL_ARB_texture_env_crossbar GL_ARB_texture_border_clamp GL_ARB_multitexture GL_ARB_texture_env_add GL_ARB_texture_cube_map GL_ARB_texture_env_dot3 GL_ARB_multisample GL_ARB_texture_env_combine GL_ARB_texture_compression GL_ARB_texture_mirrored_repeat GL_ARB_shadow GL_ARB_depth_texture GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_occlusion_query GL_ARB_point_sprite GL_ARB_texture_non_power_of_two GL_ARB_vertex_buffer_object GL_ARB_pixel_buffer_object GL_ARB_draw_buffers GL_ARB_shader_texture_lod GL_EXT_compiled_vertex_array GL_EXT_framebuffer_object GL_EXT_texture_rectangle GL_ARB_texture_rectangle GL_EXT_texture_env_add GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_texture_lod_bias GL_EXT_abgr GL_EXT_bgra GL_EXT_stencil_wrap GL_EXT_texture_filter_anisotropic GL_EXT_secondary_color GL_EXT_blend_func_separate GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_depth_bounds_test GL_EXT_texture_compression_s3tc GL_EXT_texture_compression_dxt1 GL_EXT_texture_sRGB GL_EXT_blend_equation_separate GL_EXT_texture_mirror_clamp GL_EXT_packed_depth_stencil GL_EXT_geometry_shader4 GL_EXT_transform_feedback GL_EXT_gpu_shader4 GL_APPLE_flush_buffer_range GL_APPLE_ycbcr_422 GL_APPLE_vertex_array_range GL_APPLE_texture_range GL_APPLE_float_pixels GL_ATI_texture_float GL_ARB_texture_float GL_ARB_half_float_pixel GL_APPLE_pixel_buffer GL_APPLE_object_purgeable GL_NV_point_sprite GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_blend_square GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_fog_distance GL_NV_depth_clamp GL_NV_multisample_filter_hint GL_NV_fragment_program_option GL_NV_fragment_program2 GL_NV_vertex_program2_option GL_NV_vertex_program3 GL_ATI_texture_mirror_once GL_ATI_texture_env_combine3 GL_ATI_separate_stencil GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod - Simplistic almost useless benchmark: Redrawing all areas 10 times took 0.2073 seconds. .. (*) Found 2 notices, documented in the text above. ";"None";"None";"None" 17364;1;"Open";3;9502;"Blake Ketchum";100;"Nobody";"2008-07-22 05:17";"1970-01-01 01:00";"2009-04-20 15:39";"shift+scroll does not work with Blender, wacom Intuos3, and Mac OS X";"I have a Mac and use OS X. Using the Wacom Intuos3 tablet driver in two recent versions, I have my intuos3 tablet, which has two touch strips, set so that one does horizontal scroll (ctrl+scroll) and the other set to (shift + scroll). (see image on the settings) The tablet will perform the horizontal scroll on blender, but not the vertical scroll. The tablet works fine on my PC with Blender and performs the proper scroll functions, and so I do not think it is a hardware issue. On the mac, however, it neither shift nor shift+ scroll will work with the Wacom mouse, buttons, pen nor tablet in Blender. Using my keyboard and my logitech mouse, scroll+shift generates a vertical scroll in blender on my mac, as it should. Shift functions DO work in other programs, like word and firefox, on the Mac with the tablet. I have done this several times, but only have access to one mac, so I have not tested it on another OS X machine. Because the tablet has two touch strip areas, it would be incredibly useful if the vertical/horizontal scroll capacity was operational. I presented this on the Blender Underground and it was suggested that I post a potential bug here.The bug may reside in the tablet software, too. I can't tell. I did not find reference to it in other posts. Thanks for your work on Blender. It's just about the coolest thing since sliced bread.";"None";"None";"None" 17369;1;"Open";3;4679;"Juho Vepsalainen";100;"Nobody";"2008-07-22 14:19";"1970-01-01 01:00";"2008-07-31 16:11";"Group linking weirdness";"1. Download provided .blend. 2. Link the lamp group found in that .blend to your scene. For some reason nothing appears to the outliner view. OOPS view shows the linked blocks. In this case they seem to be separate and have red squares on them (probably they are invalid somehow). 3. Append the same lamp group. Blender prints out ""append: already linked"" message to the console. Expected behavior would be to give clear notice about why linking groups does not work and not to link at all instead of leaving them this way.";"None";"None";"None" 17387;1;"Open";3;2960;"Petri Rantanen";100;"Nobody";"2008-07-24 16:36";"1970-01-01 01:00";"2008-07-31 15:19";"Halo materials do not get ""overridden"" when using a replacement material (render layers)";"Hi, I noticed this when I tried to get a ""pass"" out of the edge effect of the renderer (you know, the toony edges). So, I created a shadeless white material and a white world background. Then I assigned the shadeless white material to be the replacing material, in the rendering buttons, in the Render Layers panel, in the Mat: field. But, it is not replacing the halo materials, instead they remain as they are and do not use the shadeless white material. It's not bad that the halos do not change into ""non-halos"", but they *should* change their color. Let's say you've made clouds or mist with halos, and you wish it to remain, but want the ""edge pass"" out of it. So you still want the halos to cover stuff, but don't want them to be of any other color but white. This way the edge effect doesn't show under the halos (which is good). So, what is the edge pass? I combine the shadeless white render, which only has the edges as black, with the original render in the nodes, using a mix node with the multiply option. This way the black edges are added to the original render, offering more flexibility. Of course... and actual edge effect pass in the nodes would be nice ;). Anyway, the problem / bug is that the halo materials do not get overriden. I think it'd be better that they only changed the colors to whatever color the overriding material is, and not the type of the material (so no change from halo to a normal material). It'd just take the colors (diffuse, spec and mir) and replace those... Hmm... that might get a bit complicated. Actually, sometimes it is good that the halo's don't get replaced. So in conclusion, I don't think the Mat: field should override halo materials, unless you wanted to. So, I guess there should be a button next to the Mat: field that enables halo materials to be overridden. Maybe same for the strand materials. Sorry about the long report, this was kind of a thinking process. -Petri Rantanen aka Falgor";"None";"None";"None" 17391;1;"Open";3;6601;"James Monteath";100;"Nobody";"2008-07-25 16:41";"1970-01-01 01:00";"2008-07-31 15:18";"Lib Proxy: When converting the armature to a proxy the original Display Type info not preserved";" I posted a bug about this on #17316, however i did not realize, at that time, that it actually resets the original values from the linked in file. This is confusing behavior for users starting out with blender. You think that the Display Type properties of the linked object will be preserve and it is not. Also see 17390 as it may be related. To fix this i must set it manually myself in my new file. What other data might not be preserved ? ";"None";"None";"None" 17392;1;"Open";3;6261;"Filipe Soares Dilly";100;"Nobody";"2008-07-25 16:58";"1970-01-01 01:00";"2008-07-31 15:15";"Recursive Dupli Groups don't appear in Add menu";"Recursive Dupli Groups don't appear on the list of ""Shortcut-SPACE"" ADD > Groups. Example: Create a ""Group_01""; Create a ""Group_02""; Add these two groups to the scene; Then Group the two created groups (""Group_03""); The Created Group_03 appear on the list of ""Shortcut-SPACE"" ADD > Groups, but the others (01 and 02) don't. They all appear on the ""Add menu"" (the one on the User Preferences header). So its not a really bad bug. But its annoying. :P File attached.";"None";"None";"None" 17416;1;"Open";3;9668;"Earnest Blimp";100;"Nobody";"2008-07-31 16:36";"1970-01-01 01:00";"2008-07-31 17:44";"Images with Alpha lose Alpha on File Save";"Hi, Open the file alphabug.blend In the UV Editor, click Image->New and set Alpha to 0.0, click OK Set Image->Texture Painting to True, and draw around a little bit in the texture Click the 'Draws only Alpha' button in the UV Editor header You should see the alpha representation of whatever you drew Now just save the scene, and reload it The image is now white, and the alpha appears to be gone It's like internally, alpha is maintained and all works, but as soon as the file(s) are saved to disk and/or read back, the alpha channel is gone. Kubuntu Hardy Heron NVIDIA 8800GT 512MB Blender 2.46 Thanks!";"None";"None";"None" 17425;1;"Open";3;8948;"Aldrik Dunbar";100;"Nobody";"2008-08-02 21:48";"1970-01-01 01:00";"2008-08-03 13:45";"3D view + image sequence";"The 3D view dose not update when you change the image sequence frame via Auto Refresh or the image properties (with Textured Draw type, in Texture Paint mode or with the Solid Tex option in View Properties). In Texture Paint mode edits do affect the correct frame (and are visible in the image viewer) but are also not displayed correctly in the 3D view. OS: Archlinux (64bit) Gfx: nvidia 6800GT (using proprietary drivers)";"None";"None";"None" 17469;1;"Open";3;7357;"Jean Montambeault";100;"Nobody";"2008-08-11 00:00";"1970-01-01 01:00";"2008-09-07 13:20";".B.blend crash in versions 2.45 and older caused by 2.46 et 2.47rc1";"If the .B.blend used by all versions of Blender installed was created by version 2.46 or newer it becomes impossible to use versions 2.45 and older as they crash on starting with a segmentation fault. Procedure to verify: - delete or rename the .B.blend in the home directory. - start any version of Blender (I own 2.44 up to 2.47rc1 and several build of branches) and there is no problem. - now start any version 2.46 or newer. - do nothing but save the default settings (Ctrl-U) - now any version 2.45 or older crashes. My system : Ubuntu Hardy Bi-processor Intel Xeon 2.4 GHz 512 Mb RAM ATI 9600 Pro 128 Mb DDR2 ram proprietary driver Jean";"None";"None";"None" 17477;1;"Open";3;9914;"Chris Folea";100;"Nobody";"2008-08-13 05:01";"1970-01-01 01:00";"2008-09-09 21:20";"Adding loops to unrelated vgroups makes edited hair particle system act up";"When I'm working on my character, I need to be able to add loops while playing with the awesome new cloth sim for more reliability on collisions. However, when I add loops to unrelated vertex groups(AKA not the hair Vgroup) the hair system goes completely berzerk. To repeat this glitch(it does the same thing on both of my pc's and even in a new file) 1. Have an object with a dedicated hair vertex group: http://www.badluckbob.com/BUG/b1.jpg 2. Add a loop to a completely unrelated vertex group using CTRL+R. In this case...the arm. http://www.badluckbob.com/BUG/b2.jpg 3. The hair system goes completely crazy. http://www.badluckbob.com/BUG/b3.jpg Thank you so much for your hard work developers! -Chris";"None";"None";"None" 17494;1;"Open";3;4962;"Stanley Svensson";100;"Nobody";"2008-08-19 01:37";"1970-01-01 01:00";"2008-09-07 12:30";"Material nodes give wrong colors in comp passes (2.47 showstopper?)";"When rendering with material nodes, only the last selected material node diffuse color is passed to the color and diffuse channels in the render passes. The shadow pass doesn't look right either? This shows up in 2.46 and 2.47RC1 for windows. To repeat the bug: 1. Open the provided blend file. 2. Click on the material node named 'base' 3. Press F12 4. Update the comp nodes - the 'Color' and 'Diffuse' should now have an orange color and the 'Shadow' pass blue channel is for some reason gone. 5. Click on the material node named 'refl' 6. Press F12 7. Update the comp nodes - 'Color' and 'Diffuse' is suddenly black and the 'Shadow' pass is now also lacking data for all rgb channels. Windows XP Pro SP3 Athlon 3200+ 1.5 GB ram GeForce 7600 GS ";"None";"None";"None" 17496;1;"Open";3;2960;"Petri Rantanen";100;"Nobody";"2008-08-19 18:27";"1970-01-01 01:00";"2008-09-07 13:01";"Depsgraph updating (displace) excludes image sequence textures";"Hi, Here's how it goes down (same text in the blend-file) I add a mesh (16x16) grid (on layer1). Then I add an image sequence as a displacement texture for the mesh (40 frames png, RGB). I use this texture (named ""disp"") in the displace modifier which I give to this object. Then I render an animation. Everything works. BUT. If I add dupliverts to the object (say a cube as the duplivert object, layer2), the displacement map/modifier is no longer calculated, except to one frame, the one you start to render from. Also, it's not previewable. And if you want to change that frame, you have to go to the frame and enter/exit the edit mode - no automation whatsoever. So, there should be previewable (in 3d-view) animated displacement maps, and they should also work with objects that have dupliverts. This should work according to all logic even now, but it just doesn't. The dupliverts are what make it not to work. Nor do any animated textures (like procedural textures which have their settings animated). Think about how cool stuff you could make, oceans of cubes or spheres, animated with animated displacement maps. (*in case the sequence frames are not packed, just use some image sequence of you own instead*) -Falgor (Petri Rantanen)";"None";"None";"None" 17512;1;"Open";3;107;"Chris Want ";100;"Nobody";"2008-08-22 16:56";"2008-10-09 19:20";"2008-10-09 19:35";"Bad Beveling";"Attached is a relatively simple blend file for which the bevel modifier has done a poor job. Chris ";"None";"None";"None" 17521;1;"Open";3;9208;"Rian Tut";4042;"Janne Karhu";"2008-08-25 02:42";"1970-01-01 01:00";"2008-10-06 00:10";"Incorrect UV cords for texture controling new particle system density and lightyear render time caused by free particle controll points.";" OS: Windows XP pro service pack2 VidioCard: Nvidia 7600 Blender 2.46 or 2.47RC2 I'm not sure how the attach file works so if its not there leave a comment. Texture controlling particle density with 'DENS' reads UV's incorrectly if particle system is below a subsurf modifier. Start the Test particle UV's.blend make sure the subsurf modifier is above the particle system. The uv mapped texure on the cube should be set to multiply. In the new textured view watch the cube as you change the (texture particle setting) between normal and inverse 'DENSITY'. Either particles will not be emitted at all, or they will be emitted where the texture is not white! Now move the particle system below the subsurf modifier and disable the subsurf from being visible. Notice the particles are emitted correctly - only where the texture is white. Second problem is render's that take lightyears due to free particle controll points. Open same file, delete subsurf modifier on 'Cube', and make sure the texture is controlling the density of the particles. Set particle system to editable and enter particle mode; make sure selection mode is in point select mode so that all the control points for paticles are visible; select 'Comb' brush and comb across the cube where there dosn't appear to be particle strands but surface has root controll points. You should see free points appear around the cube as you begin to brush over the surface; if you attempt to render at this point render should initialize and appear to freeze. If you acctually let it set for about thirty minutes and then pres 'esc' you would find that it managed to render a couple of lines!!! Last note: if free points are manually deleted render usually works normally but particle are emitted in all the wrong places. ";"None";"None";"None" 17541;1;"Open";3;10312;"Someone Someone";100;"Nobody";"2008-08-28 19:19";"1970-01-01 01:00";"2008-09-07 19:41";"Blender dosent display key alpha properly";"Blender doesn't display key alpha properly: In render results it doesn't apply the alpha channel it just fills areas with no color with black which results in OSA renders looking like OSA did nothing (In reality it did if you open the result in an external image viewer/editor (GIMP, or windows picture/fax viewer as an example) it will have the alpha channel applied correctly and the OSA showing like it should) This doesn't happen with premultiplied alpha has the alpha channel seems to be applied properly then. There is also another problem with key alpha in the compositor when using key alpha I'm getting white pixels around the image. i added screen shots of this. This could be tested using the default scene by just changing the sky,premul,key settings under the render panel. and rendering, looking in the compositor render preview. I'm using windows xp sp3 Blender 2.47 my 3d card is ati radeon 9600 with driver 8.7";"None";"None";"None" 17548;1;"Open";3;6441;"luke etyrnal";100;"Nobody";"2008-08-29 21:29";"1970-01-01 01:00";"2008-09-07 20:32";"Displace Modifier not following frame changes in image-sequenc-texture";"When using a texture that is based on an image-sequence (not a video, but the correctly named frame-series i have ouput) , to drive the Displacement modifier, the Displacement modifier doesn't seem to update it's displacement of the affected vertices when the frame is advanced -- unless i turn the displace modifier off and back on... the if i move the timeline around to different frames, the texture is loading the correct frame, but the modifier is NOT. it's as though, once the process of rendering a range of frames is begun, the displacement modifier reads an image from the image-sequence, then it keeps it for all following frames... never updating to make sure it's got the right image for the curremt frame. help. OSX 10.4.11 / Blender PPC 2.46";"None";"None";"None" 17558;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-09-01 15:27";"1970-01-01 01:00";"2008-09-08 16:30";"grid problem";"WinXP+Geforce7800 use top/front/side views to see the grid 1. go to the center of the global coordinate system 2. zoom in and look the grid, pan the view and look the grid: all okay 3. now pan from center to X=1000 4. zoom in and look the grid, pan the view and look the grid: grid drawing is gone mad! strange, but the snapping is works good everywhere, so maybe this is an opengl related bug";"None";"None";"None" 17563;1;"Open";3;10261;"Trygve Wastvedt";100;"Nobody";"2008-09-03 02:21";"1970-01-01 01:00";"2008-09-18 13:39";"Particles in DupliGroups don't change with group.";"1. Add a cube, make it part of a group, and add a particle system to it. 2. Add an empty and use DupliGroup to instance the cube. 3. Move the empty to a new location. 4. Run the animation so that there are particles on both the cube and its instance. 5. Preview render both cubes - while particles show up in the 3d window on the instance, none are rendered. My computer: Dell Inspiron 6400 with ATI Mobility Radeon X1400 using fglrx driver Ubuntu 8.04.1 Blender 2.47 Thanks, -Trygve";"None";"None";"None" 17607;1;"Open";3;10378;"David Silverman";100;"Nobody";"2008-09-12 04:26";"1970-01-01 01:00";"2008-09-21 18:57";"UV editor (painting) does not update other views or texture users";"when painting in the UV editor the lock icon says: ""updates other affected window panes automatically to reflect changes in real time"" This does not happen. Tried with displace modifier and color on material. Must jump into edit mode and out to see texture and displace update in 3d view. Would be nice for the tool tip to reflect whether the feature is on. maybe a lock isn't an appropriate icon. this happened on an XP machine";"None";"None";"None" 17612;1;"Open";3;10378;"David Silverman";100;"Nobody";"2008-09-13 02:24";"1970-01-01 01:00";"2008-09-14 10:56";"brush ""stamp"" datablock selector needs tooltip explaining what it is";"I spent a lot of time thinking that the ""dropup"" datablock selector chose what image you were painting onto. It doesn't have an appropriate tool-tip or label explaining it is the image-stamp for the brush. The brush chooser should probably say brushes instead of choices.";"None";"None";"None" 17614;1;"Open";3;4862;"Matjaž Lamut";100;"Nobody";"2008-09-13 15:36";"2008-09-15 10:43";"2008-09-23 00:18";"Normal Map bug with mirrored UVs";"There appears to be a bug with the normal maps, when the UVs are mirrored. The original side shows properly, while the mirrored side has some weird artifacts. This is independent of the mirroring method, be it the modifier, the tool or anything else. The normals were re-calculated with no proper result. There are also no double vertices or anything. All in all it appears to be an issue with blender not properly applying the normal map to a mirrored UV. This problem was first observed by a friend of mine and it's the same on my computer. I used Blender 2.47, one of the Apricot builds and the SVN 16477 version. The problem appears in the GLSL mode in the 3d window, in the rendering preview, as well as when actually rendering the object. I'm using Windows XP with SP3 Athlon 1,9 ghz 1GM of RAM Gefore 7600GT with semi-latest drivers I'm attaching a test blend file and a render result. Left side is the mirrored one and the right side is the original. One of the comments my friend had is that it appears the red channel of the normal map is flipped on the mirrored part of the model. When he flipped the red channel in the actual texture, the result was just the opposite - the mirrored side was showing correctly and the original side had these artifacts. ";"None";"None";"None" 17624;1;"Open";3;6623;"Aaron Digulla";100;"Nobody";"2008-09-15 08:52";"1970-01-01 01:00";"2008-09-15 08:52";"Limit sculpt to selected vertices";"For more precision during sculpting, add a ""select"" mode (using the sculpt tool on a mesh will select vertices instead of modifying them) and an option ""limit to selected vertices""";"Tools";"None";"None" 17625;1;"Open";3;6623;"Aaron Digulla";100;"Nobody";"2008-09-15 09:00";"1970-01-01 01:00";"2008-09-15 09:00";"New sculpt tools: Decimate, subsurf, destress";"I'm unhappy with multires mesh and the way subsurf and decimate works. Often, I only need more detail in a specific place of my mesh. Therefore, I'd like three new sculpt tools: - Decimate: Reduce the vertices under the cursor - Subsurf: Subdivide the vertices under the cursor - Destress: Move vertices to destress the mesh, i.e. take the strain away. This tool should try to move vertices to create faces which have equal sides and equal angels while preserving the overall geometry.";"Tools";"None";"None" 17644;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2008-09-17 22:03";"2008-09-18 11:44";"2008-09-18 17:07";"Fluid Simulation doesn't report FrameChanged (script link) to the last frame";"Hi, i don't know if this is a bug. The file attached showcase the problem I'm having. I'm using a python scriptLink set as FrameChanged. If i bake a particle simulation from frame 1 to 5, the script is called in the following frames: 1, 1 (twice), 2, 3, 4, 5, 1 BUT if I bake a fluid simulation from frame 1 to 5 the script is called in the following frames: 1, 1 (twice again), 2, 3, 4, 1 There is no frame 5 (the baking works, but the frame is not change to activate the script). I understand that it can be by design. However the differences between these two baking outputs make scripting a little more hacky. checked in Blender 247 and SVN";"None";"None";"None" 17661;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2008-09-21 19:48";"1970-01-01 01:00";"2009-04-23 23:33";"BGE : Object children wrong transformation";"This is a model from yo-frankies menu, when the game runs youll see the text planes rotate in a way that is unlike the 3d view. ";"None";"None";"None" 17700;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2008-09-28 19:17";"1970-01-01 01:00";"2008-09-29 12:40";"pyNode update and disconnect";"Open the file, click update in the pyNode window. The two other nodes will be disconnected to the pyNode. Bug present in Blender 246, 247 and SVN 16791 Maybe is the file (I got this example from the PyNode Cookbook)";"None";"None";"None" 17743;1;"Open";3;4188;"Grzegorz Rakoczy";100;"Nobody";"2008-10-03 16:17";"1970-01-01 01:00";"2008-10-03 16:45";"Can't append or link group with fluid outside source dir";"Hi, I've found strange bug: when I have some object with fluid sim baked and try to link whole group into another blend file, I can do this only inside the same dir. When I try to link from another directory then fluid sim data isn't loaded. I've tried absolute and relative paths and still the same. Object is loaded but I see cube instead of baked fluid sim. ";"None";"None";"None" 17749;1;"Open";3;3284;"ys vry";100;"Nobody";"2008-10-04 00:19";"1970-01-01 01:00";"2008-10-04 15:08";"Particle Group alignment fixed to X-axis only...";"2.48rc1, group vizualization doesnt work correct, object in diffrent direction and doesnt follow the hair.";"None";"None";"None" 17751;1;"Open";3;177;"Matt Ebb ";100;"Nobody";"2008-10-04 08:26";"1970-01-01 01:00";"2008-10-04 15:52";"Atmosphere rendering adds colour incorrectly over transparent objects";"Hi Ton, I've assigned this to you since you've been working on this lately - hope it's ok. There's a pretty major flaw in the way that the 'atmosphere' rendering works - it's completely non-physical, just adding colour to the surface, no matter what the surface properties are. I noticed this while testing it on volumetrics in the sim_physics branch, but after more testing and investigation, I've noticed that it gives wrong results in other cases too, like for raytraced transparency. Check the attached image samples to see - obviously the glass sphere in the atmosphere image should not look so dark. I looked inside and I can see what the problem is, the atmosphere is implemented in quite a strange way, as an additional render pass, doing its own tile processing in a low level of the renderer. This means it's oblivious to surface properties, just adding colours on top of other colours. I suspect zaghaghi may have just followed the same process as was used for the sky shading itself, but I don't think it's the best way of doing it. It seems to me like unnecessary effort - all it needs as input is a view vector and distance, and that it would be much more simple and suitable to do these calculations within the solid shading process itself. In fact it could be implemented as a node, too.";"None";"None";"None" 17754;1;"Open";3;7357;"Jean Montambeault";100;"Nobody";"2008-10-04 14:03";"1970-01-01 01:00";"2008-10-04 17:20";"In the F9 button window, when a panel is moved while in edit mode it won't stay put once we go to object mode and back";"OK, this may not be a bug per se but it sure is a bugger, In the factory settings scene, with the default cube in edit mode, move the ""Mesh Tools More"" panel just to the right of the ""Mesh Tools"" panel. Then Tab-Tab to go back and forth between object mode and edit mode and you'll see that ""Mesh Tools More"" got pushed on position to the right. Tab-Tab again and that panel has reached the end of the line. Make UV Calculation an independent panel and position it just to the right of the ""Mesh Tools"" panel. Tab-Tab, Tab-Tab ... same dance. Thanks for your time guys. :) Jean";"None";"None";"Fixed" 17763;1;"Open";3;4948;"Thomas Brandt";100;"Nobody";"2008-10-05 12:42";"1970-01-01 01:00";"2008-10-05 13:36";"Cloth does not render with a subsurface in front of the modifier stack";"Hello, i want to message that the cloth modifier does not show in the final render, when a subsurface modifier is above in the modifier stack. The opengl view is OK. 1. Bake some frames 2. try to render the last of the frames -> the render shows the first frame with no cloth influence. To fix this move the subsurface modifier under the cloth, bake some frames and the final render is as desired.";"None";"None";"None" 17788;1;"Open";3;177;"Matt Ebb ";100;"Nobody";"2008-10-09 00:42";"1970-01-01 01:00";"2008-10-09 12:22";"File output nodes are not antialiased when using FSA";"I'm trying to save out multiple passes using file output nodes in the compositor, however when I use FSA, although the main render result looks fine, all the other images saved out with file out nodes don't get antialiased. This makes it a bit impossible to comp back together afterwards, especially with ID masks etc. Demo file is attached, with example render cheers";"None";"None";"None" 17824;1;"Open";3;8596;"David Sheftick";100;"Nobody";"2008-10-12 04:10";"1970-01-01 01:00";"2008-10-12 13:17";"Original lamp uses old default values which can cause blender to crash";"This is not really a bug but it causes blender to crash so I thought I should report it. When I start blender using the factory settings, setting the lamp to spotlight then setting it to use shadow buffer, while in textured GLSL mode, causes blender to crash. This is because the original lamp uses the old default values for the shadow buffer. The old shadow buffer size is 2880 instead of the new setting of 512. My computer can only handle a maximum size of 2048. Not a bug, but it took me a while to figure out what was wrong, so I imagine this is giving some other people problems as well. By the way the shadow buffer size can not be increased using the arrow buttons in the Shadow buffer size box, but it can be decreased.";"None";"None";"None" 17826;1;"Open";3;8462;"Reece Robinson";100;"Nobody";"2008-10-12 05:42";"1970-01-01 01:00";"2008-10-12 13:21";"The sound does not play when using the Intel Mac version of blender but does using PowerPC version.";"Hi, This report relates to a 24"" Intel iMac. 1) Press 'P' to start the game engine. 2) Press the 'space' bar to hear the helloworld wav file play. The sound does not play when I use the Intel Mac version of blender in the 'sound block buttons' sub context of the scene (F10) menu or in the game engine. BUT if I run the PowerPC version on the same Intel Mac I do hear the sound. This problem appears on 2.47 and 2.46 official releases as well. I have also confirmed that the game sound output does not work on an Intel MacBook I own. ===================================== = Blender 2.47 System Information = ===================================== Built on 2008-10-02 21:21:58, Rev-16882 Version darwin dynamic Platform: darwin ======== Python: ====== - Version: 2.5.1 (r251:54863, Jan 17 2008, 19:35:17) [GCC 4.0.1 (Apple Inc. build 5465)] - Paths: /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python25.zip /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5 /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/plat-darwin /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/plat-mac /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/plat-mac/lib-scriptpackages /System/Library/Frameworks/Python.framework/Versions/2.5/Extras/lib/python /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib-tk /System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/lib-dynload /Library/Python/2.5/site-packages /System/Library/Frameworks/Python.framework/Versions/2.5/Extras/lib/python/PyObjC /Users/reece/Documents/blender/scripts /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender/scripts /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender/scripts/bpymodules - Directories: Blender home dir: /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender Default dir for scripts: /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender/scripts Default ""bpydata/"" data dir for scripts: /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender/scripts/bpydata User defined dir for scripts: /Users/reece/Documents/blender/scripts/ Data dir ""bpydata/"" inside user defined dir: -- not found Default config data ""bpydata/config/"" dir: /Applications/blender-2.48-RC1-osx-intel-py-2.5.zip/blender.app/Contents/MacOS/.blender/scripts/bpydata/config - Modules: all basic ones were found. OpenGL: ====== - Renderer: ATI Radeon HD 2600 PRO OpenGL Engine - Vendor: ATI Technologies Inc. - Version: 2.0 ATI-1.5.30 - Extensions: GL_ARB_transpose_matrix GL_ARB_vertex_program GL_ARB_vertex_blend GL_ARB_window_pos GL_ARB_shader_objects GL_ARB_vertex_shader GL_ARB_shading_language_100 GL_EXT_multi_draw_arrays GL_EXT_clip_volume_hint GL_EXT_rescale_normal GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_gpu_program_parameters GL_EXT_geometry_shader4 GL_EXT_transform_feedback GL_APPLE_client_storage GL_APPLE_specular_vector GL_APPLE_transform_hint GL_APPLE_packed_pixels GL_APPLE_fence GL_APPLE_vertex_array_object GL_APPLE_vertex_program_evaluators GL_APPLE_element_array GL_APPLE_flush_render GL_APPLE_aux_depth_stencil GL_NV_texgen_reflection GL_NV_light_max_exponent GL_IBM_rasterpos_clip GL_SGIS_generate_mipmap GL_ARB_imaging GL_ARB_point_parameters GL_ARB_texture_env_crossbar GL_ARB_texture_border_clamp GL_ARB_multitexture GL_ARB_texture_env_add GL_ARB_texture_cube_map GL_ARB_texture_env_dot3 GL_ARB_multisample GL_ARB_texture_env_combine GL_ARB_texture_compression GL_ARB_texture_mirrored_repeat GL_ARB_shadow GL_ARB_depth_texture GL_ARB_shadow_ambient GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_occlusion_query GL_ARB_point_sprite GL_ARB_texture_non_power_of_two GL_ARB_vertex_buffer_object GL_ARB_pixel_buffer_object GL_ARB_draw_buffers GL_ARB_shader_texture_lod GL_EXT_compiled_vertex_array GL_EXT_framebuffer_object GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXT_texture_rectangle GL_ARB_texture_rectangle GL_EXT_texture_env_add GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_texture_lod_bias GL_EXT_abgr GL_EXT_bgra GL_EXT_stencil_wrap GL_EXT_texture_filter_anisotropic GL_EXT_separate_specular_color GL_EXT_secondary_color GL_EXT_blend_func_separate GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_texture_compression_s3tc GL_EXT_texture_compression_dxt1 GL_EXT_texture_sRGB GL_EXT_blend_equation_separate GL_EXT_texture_mirror_clamp GL_EXT_packed_depth_stencil GL_EXT_draw_buffers2 GL_APPLE_flush_buffer_range GL_APPLE_ycbcr_422 GL_APPLE_vertex_array_range GL_APPLE_texture_range GL_APPLE_float_pixels GL_ATI_texture_float GL_ARB_texture_float GL_ARB_half_float_pixel GL_APPLE_pixel_buffer GL_APPLE_object_purgeable GL_NV_blend_square GL_NV_fog_distance GL_ATI_texture_mirror_once GL_ATI_blend_equation_separate GL_ATI_blend_weighted_minmax GL_ATI_texture_env_combine3 GL_ATI_separate_stencil GL_ATI_texture_compression_3dc GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod GL_SGI_color_matrix - Simplistic almost useless benchmark: Redrawing all areas 10 times took 0.1769 seconds. . (*) Found 1 notice, documented in the text above. ";"None";"None";"None" 17835;1;"Open";3;11442;"Jordan Kohler";100;"Nobody";"2008-10-14 16:21";"1970-01-01 01:00";"2009-04-23 20:44";"setPosition() and the constraint actuator do not work for soft bodies";"Soft bodies do not seem to be affected by setPosition() or the constraint actuator.";"None";"None";"None" 17862;1;"Open";3;3132;"Daniel Salazar";100;"Nobody";"2008-10-19 01:44";"1970-01-01 01:00";"2008-10-19 01:44";"Colored Shadows not suported in GLSL";"Harkyman's colored shadows are not working in GLSL shadow rendering cheers";"None";"None";"None" 17884;1;"Open";3;139;"Martin Poirier";100;"Nobody";"2008-10-22 00:58";"1970-01-01 01:00";"2009-02-23 00:04";"Scons: Freetype linked to WITH_BF_INTERNATIONAL";"Freetype is used only when WITH_BF_INTERNATIONAL is true. It should be independent (with its own WITH_blah_blah) as it is used for text objects too.";"None";"None";"None" 17885;1;"Open";3;7111;"Domino Marama";100;"Nobody";"2008-10-22 10:17";"1970-01-01 01:00";"2009-04-20 16:27";"Subdivide Tool gives inconsistant results on UV map";"The mesh shape seems to affect the subdivide results on the UV map, where I'd expect the edge positions for the new vertices on the UV map to be calculated purely from the UV map, the same as subsurf and multires do. In the attached blend there are 4 objects: Sphere Sphere.Subdivide Plane Plane.Subdivide The Sphere and Plane are the starting mesh that W-Subdivide was applied to. The *.Subdivide versions are the results. On the Sphere.Subdivide the added UV points are not always at the midpoints of the edges as expected. The Plane.Subdivide shows the expected results. Adding a multires level, or a subsurf modifier and applying gives the correct results on all shapes. It's just the Subdivide tool that is showing the problem.";"None";"None";"None" 17886;1;"Open";3;8140;"Ralf L.";100;"Nobody";"2008-10-22 12:58";"1970-01-01 01:00";"2009-02-25 10:11";"No audio in Gameengine when build with mingw";"I have done a build from source with MSVC9 and sound in the Gameengine works. I used the YoFrankie Demo When I build with Scons and mingw there is no sound. Have seen this with other builds from www.graphicall.org too. See also here for a thread at BlenderArtists: http://blenderartists.org/forum/showthread.php?t=138764 I don't now (yet) if this is also valid for Linux. I used the source from Blender-2.48-release, and also SVN 17082 - 17088 and latest from yesterday - 17149 Tested on AMD64 3000+ with 2GB Ram - WindowsXP Sp3 and DELL Laptop Core Duo T2400 with 1GB RAM WindowsXP SP3";"None";"None";"None" 17892;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-10-23 00:14";"1970-01-01 01:00";"2008-10-24 04:03";"fly mode ignores protected rotation axis.";"1. make a camera and press NumPad 0 to viewing from this camera 2. press N to view Transform props 3. lock all transform props 4. press shift+F to fly mode: you can rotate the locked rotations winXP";"None";"None";"None" 17898;1;"Open";3;11140;"ronan ducluzeau";4042;"Janne Karhu";"2008-10-24 17:23";"1970-01-01 01:00";"2008-11-01 19:38";"Friction does not have any effect when Dynamic button is enabled";"Release: 2.48a OS: WinXP François Grassard (aka <>) asked me to report the bug. Plane's friction setting does not have any effect when Dynamic button is enabled. It used to work well with Blender 2.47. It was useful to simulate explosion. Fragments's rotation could naturally stop after several rebounds. Here is the .blend. Play the animation. Then, enable Dynamic button and replay the animation. ";"None";"None";"None" 17899;1;"Open";3;2241;"Tommy Helgevold";100;"Nobody";"2008-10-24 19:34";"1970-01-01 01:00";"2008-11-01 15:53";"Tangent shader - shadow biases badly";"Machine: Hewlett packard Pavilion 6133sc Processor: Q6600 - 2.4 Ghz Core 2 Duo Quad Ram: 8 GB OS: Ubuntu Linux 8.04 - latest updates - 64 bit. Blender: 2.48a SVN Problem: Severe render artifacts using the TANGENT Shading option, especially with rounded, smoothed bevels. Creates some sort of zig-zag bounce pattern that is impossible to ""bias"" or ""ajust"" away, no matter what shader used. Please see included example on image (top of pot) Also - please render closeups of the top corner of the pot (furthest away, just zoom in on the included blend file - and render) and the bug becomes even more apparent. Problem 2: There are also some ray shadow biasing issues (look at the blocky shadows in the image example), Biassing has been used - and turned on but to no effect. ";"None";"None";"None" 17903;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2008-10-25 17:33";"1970-01-01 01:00";"2008-11-01 16:31";"When I join more then 16 objects I don't receive a warm about the lost-to-be materials";"Blender has a material number limitation (16 I think). When I join more then 16 objects with different materials, the mesh is joined perfectly. However the materials get lost. I don't mind the 16 materials limitation, however I would love to see a pop-up alerting that some materials would be lost. PS.: This is what messy up with my BConf presentation slideshow :/ I joined my 50 slides together to make it easier for the BGE to handle it. However I forgot about this limitation and I realized that too late. (stupid me :p)";"None";"None";"None" 17905;1;"Open";3;7724;"endi endike";100;"Nobody";"2008-10-25 22:56";"1970-01-01 01:00";"2008-11-01 16:35";"camera can be move but it has locked transformations";"camera can be move but it has locked transformations 1. make a camera and make it active with ctrl+NumPad0 2. press n to see transformation props, and lock all 3. move editor view somewhere 4. press ctrl+alt+NumPad 0 : camera will move to editor view WinXP";"None";"None";"None" 17907;1;"Open";3;11751;"Christoph Trassl";100;"Nobody";"2008-10-26 02:00";"1970-01-01 01:00";"2008-11-01 15:36";"CVE-2008-1103: Insecure usage of temporary files";"Various vendors reported security issues with the way blender makes use of temporary files in /tmp. This issue seems to be unresolved in Blender 2.48a. This problems applies to: * autosaved files * quit.blend * rendered pictures The ""quit.blend"" issue is sometimes patched by writing /tmp/quit.blend with O_EXCL (by distros, eg. Fedora Blender 2.47 in Fedora 9), but as stated in [3] the solution should be using HOME/.blender/ or HOME/.blender/tmp as default temporary directory instead. Fedora has a patch for O_EXCL solution (see [11], but as stated in [3] should not be applied, see above) and one that uses the home directory instead of /tmp (see [12]). Debian has a patch using the home directory (see [13]) but does not use HOME/.blender as far as i can see. Both patches seem to be outdated as of blender-2.48a release and will not apply cleanly. I guess best solution would be to default the temporary directory to HOME/.blender/ or HOME/.blender/tmp and therefore the solution should be a mix of [12] and [13], but implemented according to blender coding guidelines. This solution should mitigate future bugs relating to issues with /tmp usage. I suggest mentioning CVE-2008-1103 either in Svn log or in the next release notes clarifying blender.org's position regarding CVE-2008-1103 and the status of the fix. Then packagers hopefully don't use custom patches anymore. Tia. References: [1] http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-1103 [2] https://bugzilla.redhat.com/show_bug.cgi?id=CVE-2008-1103 [3] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=298167 [4] https://bugs.launchpad.net/ubuntu/+source/blender/+bug/6671 [11] http://cvs.fedora.redhat.com/viewvc//rpms/blender/F-9/blender-2.45-cve-2008-1103-1.patch?view=markup [12] http://cvs.fedora.redhat.com/viewvc//rpms/blender/F-9/blender-2.45-cve-2008-1103-2.patch?view=markup [13] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=298167#17 Gentoo, Mandrake, SuSE and possibly others have also created bugreports - these are not mentioned here. ";"None";"None";"None" 17928;1;"Open";3;10327;"Jose Ricarte";100;"Nobody";"2008-10-29 11:35";"1970-01-01 01:00";"2008-10-30 13:17";"Bezier curve rendering/converting to mesh";"Working with curves, material zoffs and Shrinkwrap Can be a powerful tool for toon style but I think there is an error rendering ";"None";"None";"None" 17930;1;"Open";3;11833;"Sundial Services";100;"Nobody";"2008-10-29 19:50";"1970-01-01 01:00";"2008-11-01 17:40";"Can't ""make proxy"" from linked-object because current-obj can't be DEselected";"(1) Object of interest is contained in a ""linked scene."" (2) From another scene, with anything (e.g. a camera) selected, it is not possible to (reliably) do... (a) Shift+F4 ... select the (linked) armature you want to proxy. (b) Ctrl+Alt+P ... Error is: ""Can only make a proxy for a referenced object or group."" And the reason... as evidenced (e.g.) by pressing ""Shift+F4"" again... is that the (say...) camera that was previously selected is still selected. (The object-name in the lower left corner of the screen also continues to reflect ""that camera,"" rather than the object that you attempted to select.) In ""Shift+F4,"" both the camera AND the object that you wish to proxy are both highlighted... but it doesn't matter. You can't tell Blender to stop ""zeroing its binary attention upon that dammed camera."" If in ""Shift+F4"" I explicitly de-select the camera and explicitly select the object that I want to make a proxy of, then ""it appears to work"" but ""it doesn't 'stick.'"" ";"None";"None";"None" 17932;1;"Open";3;2380;"Vilem Novak";100;"Nobody";"2008-10-30 13:53";"2009-04-20 16:42";"2009-04-21 10:05";"wacom pressure sensitivity not working on ubuntu 8.10 32bit";"Actually, as the title says. In inkscape and gimp everything works as supposed to. using the automatic configuration of the new X version used in interpid. Wacom used is intuos 3 PTZ-631W";"None";"None";"None" 17934;1;"Open";3;11885;"Mike S.";100;"Nobody";"2008-10-31 23:20";"1970-01-01 01:00";"2009-02-17 04:03";"PLY import/export V texture coordinate flipping";"Blender 2.48 Built on 2008-10-14 22:38:17, Rev-17080M Version win32 dynamic Python: - Version: 2.5.2 (r252:60911, Oct 13 2008, 20:50:37) [MSC v.1500 32 bit (Intel)] - Renderer: RADEON X300/X550 Series x86/SSE2 - Vendor: ATI Technologies Inc. - Version: 2.0.5885 WinXP Release The ply exporter exports the uv coordinates directly (no flipping). ply_export.py line 138: if faceUV: uv = f.uv While the importer flips the v. ply_import.py line 219: face.uv.append((vertex[uvindices[0]], 1.0 - vertex[uvindices[1]])) I'm not sure which way is correct, but I assume they should be consistent. How to reproduce: Create a new blank project with a cube. Unwrap the cube uv coords. Move the points in such a way that there is a clear top and bottom to the points (e.g. a pyramid) Export the model as PLY. Import the PLY exported model. The UV coords will be upside down.";"None";"None";"None" 17945;1;"Open";3;6675;"Yuri Shchedov";100;"Nobody";"2008-11-03 15:09";"1970-01-01 01:00";"2008-11-03 15:09";"Bezier curves: Taper/Bevel objects suggesstions (to make them really helpful)";"Hello, I was trying to model with Bezier and taper/bevel options and seems Blender needs some small improvements to bring it to the next level as model application Here is what I'm suggesting. I think it would be extremely helpful is user could add 2 TaperOb for one main curve. This will help to align curve form in second dimention. Also it would be awesome to have two Stick options for Side 1 and Side 2 of the main curve for each TaperOb. This will allow user to stick one of the side of the main curve and keep it plain. Having multiple BevelOb for one curve will be super helpful too. I'm sure this will bring Blender to the next level for architect, furniture and other designers Attaching a picture I tried to draw showing my thoughts. Combining these features and allowing to align main curve under bezier toolset will save a lot of modeller's time Thank, Yuri ";"Tools";"None";"New" 17968;1;"Open";3;5660;"Andrew Davies";100;"Nobody";"2008-11-09 01:03";"1970-01-01 01:00";"2008-11-10 19:08";"Fill tool only fill's half the selection and only with tri's";"Hey all, I think i may have found a bug with the fill tool. When the selection in the attached file is filled using Shift+F, it only fills half the selection and only with tri's. As a point of reference, when in edge mode and the 2 end edges are deselected the Skin Face/Edge tools works though with the same method (without the 2 ends deselected) blender says ""no vertloops found"". Im using the 64bit windows version of 2.48a with Vista64, AMD 5600+ and 6GB DDR2 Ram. Its also been tested and confirmed with 2.47/WinXP and ubuntu ibex/current svn. Any ideas? Thanks for looking :) ";"None";"None";"None" 17980;1;"Open";3;9596;"Maurice Raybaud";100;"Nobody";"2008-11-13 19:46";"1970-01-01 01:00";"2008-11-14 11:25";"skined faces are smooth shaded on solid mesh";"This might be a feature request you can close immediately, BUT the skin tool (hit F with two edge loops selected>skin) creates smooth shaded geometry by defaults, Okay that's not bad for the organic modeling blender is more oriented to, but it does so even if the mesh we're working on is set to solid shading. I would expect it to contextually switch. Sorry to be picky, it's really to be constructive. :)";"None";"None";"None" 17981;1;"Open";3;9596;"Maurice Raybaud";100;"Nobody";"2008-11-13 19:52";"1970-01-01 01:00";"2008-11-14 11:11";"Extrude has to be validated before scale";"It seems to me that the behaviour of the Extrude tool used to be more flexible: one should be able to hit e key, then immediately hit S end scale the extrusion. But now, we have to right or left click before we can scale. If maintained as is, this would create the need for a (useless) inset tool... :) ";"None";"None";"None" 17985;1;"Open";3;12208;"Nicholas Porcaro";100;"Nobody";"2008-11-15 08:01";"1970-01-01 01:00";"2009-04-22 23:34";"Renamed - Wrong alpha sorting problem in the BGE";"In the attached file (and also here: http://www.blendernyc.org/twoside+alpha-problems.blend ) I sometimes get sort of ""tearing"" or ""flickering"" when two objects are moving together and are butted up against each other. I have Two-Side, Alpha, and Sort selected in the Texture Face tab of the Editing Panel for each of the meshes. Here is what you do to illustrate (and work around) the problem in the .blend above: - Press P to start the game engine - Press A to start the IPO curves for each mesh - Notice the flickering when the meshes get close to each other. - ""fix"" the flickering by removing Two-Side from the Texture Face. Do this for each mesh: - Go into Edit Mode - Go to the Editing Panel (F9) - Select the Texture Face tab - Select a Face - Leave only the following buttons on: Tex, Two-Side, Opaque - Press A to select all faces - Press Control-C then pick Active Mode - Press Control-C then pick Active Transp Also note this problem does not occur when you don't start the game engine, In otherwords, just doing Option-A in the 3D view will not produce the flickering. ";"None";"None";"None" 17993;1;"Open";3;12238;"Carsten Thiele";100;"Nobody";"2008-11-16 10:31";"1970-01-01 01:00";"2009-04-20 16:53";"Keyboard Layout with german keyboard not working except in Textinput-fields";"When running blender unter Linux (tested on Ubuntu 8.10 amd64) the keyboardlayout is kind of mixed up. In Textfields the Layout is perfectly ok, y, z, ö, ä, ß and all other german-keyboard specific Keys work ok. BUT the shortcuts dont use this layout. e.g. when i want to constraint the scale-command the Z-Axis i have to Press Y and vice versa. When i Want to use undo i have to press Ctrl-y instead of Ctrl-z, Wireframe/solid change is with y. Tested with blender 2.46, 2.48 and 2.48a, always the same behavior. I deleted the .blend directory completely but nothing changed. Because in text-input fields all Keys work as expected, and all other applications on the system have the right layout, i think this must be a bug in the application. On Windows XP all works very good. I am Using the python 2.5-version from the blender-download page. ";"None";"None";"None" 18008;1;"Open";3;10371;"Thomas Dinges";100;"Nobody";"2008-11-17 19:09";"1970-01-01 01:00";"2009-02-08 17:28";"Full Sample lose OSA";"Open the attached Blend-File. Render it. The render and also the backdrop image in the compositor is antialised. Change some values in the gamma node...it stays antialiased. Now enable in the Render Panel ""Save Buffers"" and ""FullSample"". Render the image again. The render is fine but the backdrop has no OSA now. Also when you change some settings in the Nodes for example the gamma, the render is without OSA too now. Blender: 2.48a Windows 32 Vista (official blender.org build) GPU: nVidia Geforce 8600GT ";"None";"None";"None" 18020;1;"Open";3;5447;"Atomic ";100;"Nobody";"2008-11-21 21:21";"1970-01-01 01:00";"2009-04-20 18:25";"DupliGroup Scale Zero Render Bug";"Hi All, Blender windows 2.48a XP SP3. I wanted to use empties to dupligroup a set of cubes. I want to scale each empty in the Z-Axis to have different heights for my cubes. For an audio meter. To re-create the bug. Open the attached BLEND file. Press Render and we see all three cubes. Even though one is scaled in the Z-Axis by zero. Typically a scale of 0 should not render anything. Delete the tall cube ""Empty.001"". Press Render and everything still looks fine. Add a new empty and press dupligroup and type in ""Group"". Move it away from the origin and press Render. Blink-The original source disappears. Something flaky going on indeed. It seems that when a dupligrouped object is scaled to zero, that scale information is getting rippled back to the original source causing the source to disappear.";"None";"None";"None" 18047;1;"Open";3;4547;"Matthias Schleinkofer";100;"Nobody";"2008-12-02 09:12";"1970-01-01 01:00";"2009-04-20 18:46";"Constraint ""Follow Path"" produces barging";"The path was produced by a polygon and converted to a Bezier curve. The ""empty"" object is linked with a constraint to that path. In the segment which forms the loop the empty does not keep the ""up"" direction (time from 450 to 550). To see it, ""play back animation"" (Alt+a). The tilt is set correctly (btw: is there a way to see the actual set value for tilt?). Regards Matthias";"None";"None";"None" 18052;1;"Open";3;4547;"Matthias Schleinkofer";100;"Nobody";"2008-12-04 12:06";"1970-01-01 01:00";"2009-04-20 19:15";"Children with DupliFaces don't respond to track/up settings";"It is the task to place an object at several positions by DupliFaces. 1. Although the faces are orientated in the same way, the axis and so in consequence the objects are rotated. 2. And there is no chance to have any influence by using Track(x,y,z,-x,-y,-z) and Up(x,y,z).";"None";"None";"None" 18053;1;"Open";3;5677;"squareline ";100;"Nobody";"2008-12-04 19:37";"1970-01-01 01:00";"2009-02-15 17:05";" import module bug";"please try this file, and say this is a bug? http://www.mediafire.com/?sharekey=4...db6fb9a8902bda sometime works if i reload blender and sometime not, press p and check the console/terminal window";"None";"None";"None" 18056;1;"Open";3;3487;"Campbell Barton";100;"Nobody";"2008-12-06 02:56";"1970-01-01 01:00";"2009-04-20 19:24";"Quad directly infront of the camera gives artifacts";"rendering a quad thats directly in front of the camera can give artifacts. (diagonal black area) I have had this trouble more then once on 32/64 bit systems. Attached a file with the artifact packed into an image.";"None";"None";"None" 18059;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2008-12-07 12:34";"1970-01-01 01:00";"2009-04-20 19:30";"Frame Remapping: 1) Particles don't playback (alt+a) or render properly 2) IPO frame selection";"The two map parameters found in the [Scene Panel]::[Anim/Playback buttons] are Map Old and Map New. They allow you to render animations in Blender files simulating different fps. (eg if you map 25 to 30 you don't need to rescale your IPOs to resync the animations). However I found two problems using this feature. 1) (check example file particlesRemap.blend there you can [bake], [play], then [free bake], change [Map New], [bake], and [play] to see the problem) Particles always use the original map (never the new one) to bake the particles. If you bake a particle using MapOld/MapNew = 1.25, while you are baking you can see the particles as if you were using MapOld/MapNew = 1.0. For when you play/render the animation it is not displayed/rendered correct. In fact if you bake the particles with MapOld/MapNew = 1.0, change the MapOld/MapNew = 1.25 and play (using the same cache) you see the same problematic result. The issue maybe be related with the life duration of the particles. 2) IPO issues: change the Old Map to 125 and try to change the current frame clicking (RMB) in the screen. When you are in the IPO window, when you click in a frame (eg 100) the current render frame became this frame (eg 100). However the current frame displayed in IPO window (green vertical line) is the transformed one (eg 125). My suggestion here a) change the displayed frame in the bottom of the screen according to the ratio MapOld/MapNew (G.scene->r.framelen) so you can see the right frame corresponding for each Control Vert. b)once we have this (a) we make the current frame to change (RMB in the screen) to use the new remapped frame instead of the original one. Blender 2.48a/SVN windows vista 64bits, nvidia 9600, ... ";"None";"None";"None" 18071;1;"Open";3;4682;"Werner Höhrer";100;"Nobody";"2008-12-10 15:01";"1970-01-01 01:00";"2009-04-14 19:33";"Linked blocks vanish if library file is not found";"I have some animated materials and some groups linked from other files. Now I did a quick test to see what happens if the library file was renamed/moved (or if I simply want to use it on a different computer with a different directory layout). As soon as I load the .blend file the linked stuff is completely gone (no materials left and all groups deleted). I can change the file path/name in the outliner, but that does not bring the materials nor the groups back. Also ""File->External Data -> Find Missing File"" will replace the path correctly with the new one (as long as the file name was not changed), but the linked data is still lost. Reproducible with Blender 2.46 (Ubuntu) and Blender 2.48a (Windows XP). ----- I've attached a small test case (also available at http://mattn.ninex.info/files/blender_linking_problem.zip) for the problem: * lib.blend Defines two materials named ""red"" and ""blue"". * obj1/1.blend Here I use the two linked materials from the library. Simply move/rename the lib.blend file or something similar to see the problem. (Chance has it that you can't even open the 1.blend as-is without loosing the (linked) materials on the cubes. Although it should work fine - I used a relative path for the lib.) ----- I've already asked in the blenderartists forum if there is a way to restore the links, but it doesn't look like there is one. I don't seem to be the only one having this problem though. Forum thread is here: http://blenderartists.org/forum/showthread.php?p=1274750 Werner";"None";"None";"None" 18076;1;"Open";3;4857;"Dalai Felinto";4042;"Janne Karhu";"2008-12-11 11:33";"1970-01-01 01:00";"2009-04-04 15:50";"Particle system cache problems";"1) Make an emitter particle and bake it. 2) Save the file As a new file Bug 01: the cached particles (eg blendcache_file01) are not copied to the new file cache folder (eg blendcache_file02) 3) Now try to render the animation Bug 02: When Blender doesn't find the cache file, instead of re-baked it, it ignore it completely without showing any particle and with no output error (typical:p) 4) If you go to the particle bake tab you can see that Blender still think that particles are still stored and you need to manually free the memory. Example file (cached in my computer) attached. Bug 02.5: According to Matt Ebb after you delete the file, the pointer to the cache is not deleted. 5) Now free the cache and change the initial render frame (not the bake one) to 5. Press ANIM Bug 03: Blender will not calculate the particles unless you start the rendering from frame 01 or bake it. Please don't tell me that this is by design :) Thanks a lot, Dalai";"None";"None";"None" 18100;1;"Open";3;13003;"Andrey Dubravin";100;"Nobody";"2008-12-18 15:57";"1970-01-01 01:00";"2009-04-14 19:16";"Internationalisation tooltip problem";"When I set russian language and set translate only tooltips (not buttons and toolbox) I get wrong behaviour. The length of tooltip window hase wrong width. It seems that width is calculating using english text. When I set to translate all strings (buttons and toolbox) - all works. I check this on Windows XP with 2.47 release only, but this may be a bug for 2.48 blender also.";"None";"None";"None" 18129;1;"Open";3;6601;"James Monteath";100;"Nobody";"2008-12-29 02:40";"1970-01-01 01:00";"2009-01-27 17:26";"Render: SSS turned on while render region will have incorrect results";" Tried to render region for a scene that has a model with SSS turned on. I have seen this problem since 2.45ish. I stopped doing this then, however I really need this to work better. You can create any scene with SSS turned on the model. I tried rendering with xParts at 1,2 and then 32 all the same result, on the same machine and on different ones too. Then render region using the blender command line. here is the command line: ( this is not the problem ) '/opt/blender/blender', '-d', '-b', '/media/Renders/Jobs/isabella_bust-2008_12_28-19_37_08/model/challenge/isabella/isabella_bust.blend', '-S', 'Scene', '-o', '/media/Renders/BlenderTest/Frames/models/isabella_bust/0000/0003', '-P', '/home/jmonteath/workspace/Tools/bin/blender-region.py', '--', '32', '30', '0003', '50', '2' Python script is attached. I don't think the script is a problem either, but just in case. Here is the link to the image: http://i305.photobucket.com/albums/nn213/ponderz/blender/0001.png Once all the regions are rendered you need to assemble them, i use imagemagik to do that. PS: This works fine with no SSS turned on objects. ============================================================================= Blender 2.48 (sub 1) Build Linux ponder-04 2.6.24-22-generic #1 SMP Mon Nov 24 19:35:06 UTC 2008 x86_64 GNU/Linux NVRM version: NVIDIA UNIX x86_64 Kernel Module 169.12 Thu Feb 14 17:51:09 PST 2008 GCC version: gcc version 4.2.3 (Ubuntu 4.2.3-2ubuntu7) model name : Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz model name : Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz model name : Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz model name : Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz ============================================================================= ";"None";"None";"None" 18136;1;"Open";3;7656;"Gary Tolley";100;"Nobody";"2008-12-31 01:26";"1970-01-01 01:00";"2008-12-31 01:26";"General rendering feature";"As GPU's get more powerful, it would really speed up rendering movies if the re rendering is done of the GPU, for both the internal renderer, and with YafRay At the same time making sure that YafRay also works as an alternative (broken as of 2.48a)";"Render (internal)";"None";"New" 18144;1;"Open";3;13254;"Ruslan L.";100;"Nobody";"2009-01-03 21:27";"1970-01-01 01:00";"2009-01-29 02:20";"fbx exporter - normals";"At first, my english isn't perfect. I export models to .fbx format file. When import models into my program; it shows them with smooth faces, even if i did them solid (or sharp, i don't sure how correct). For example, in attached screenshot (fbx_cube.jpg) you can see how it looks with non-modified cube model, which is present in every new blender scene by default. I found only one way to resolve this problem - to apply EdgeSplit modifier - the result you can see in another screenshot (fbx_cube_EdgeSplit.jpg). Tried [Editing]->[Link and Materials]->[Set Solid] and [Editing]->[Mesh]->[AutoSmooth], that didn't made model's faces solid/sharp in my program. All problems is to display models faces in my program; in Blender they displayed correct. Thanks. RAM: 1024 MB Graphic card: Radeon x1300 OS: Windows familly Blender: 2.47,2.48a";"None";"None";"None" 18162;1;"Open";3;13382;"Sergey Stoma";100;"Nobody";"2009-01-10 23:25";"1970-01-01 01:00";"2009-04-13 19:24";"Collada armature export doesn't handle bones with . (dot) in the name";"Bones with dots in their name are ignored during the export, and don't generate any entries. Looks like it is caused by translator.py, line 1037: jointSourceArray.data.append(adjustedName) Everywhere else it uses adjustedName.replace('.','_') instead of the object name. Changing that line generated to match the rest seems to be fixing things. Blender 2.48a, Windows Vista, iMac/ATI 2600HD Pro.";"None";"None";"None" 18172;1;"Open";3;10283;"Simon Rainerson";100;"Nobody";"2009-01-13 11:21";"1970-01-01 01:00";"2009-04-29 16:28";"Copying objects doesn't copy logic brick links";"When copying objects with a controller, link between sensors from other objects and controller of copied does not copy. When copying an object with an actuator connected another objects controller the link does not copy. What does work: When copying objects with a sensor connected to an other objects controller. The controller-actuator link when copying the controller object. Attached a file with three objects, ""sensor"", ""controller"" and ""actuator"", each with a brick (of that kind) connected to each other. Copy each object and watch the links between the bricks.";"None";"None";"None" 18174;1;"Open";3;13469;"Stefan Lesicnik";100;"Nobody";"2009-01-15 17:41";"1970-01-01 01:00";"2009-04-08 20:09";"Insecure creation of files in /tmp when rendering. (Linked to CVE-2008-1102)";"Hi, There is currently an Ubuntu Bug open that describes this behaviour. https://bugs.edge.launchpad.net/blender/+bug/6671 From the bug Reproduced in versions: 2.37a-1ubuntu1.1 (breezy?) 2.41-1ubuntu4 (dapper) 2.42a-linux-glibc232-py24-i386-static (blender.org binary) 2.42a-1ubuntu1.1 (edgy) 2.44-2ubuntu2 (gutsy) 2.45-4ubuntu1 (hardy) 2.46+dfsg-4 (intrepid) Blender writes to files in /tmp/ in an insecure fashion. For example, launching blender and then selecting ""Render > Render Animation"", writes to the file /tmp/0001.jpg. This can be exploited by a malicious user to overwrite arbitrary files of another user using blender: mallory@myhost$ ln -s /home/bob/thesis.tex /tmp/0001.jpg -- I have also reproduced this in 2.48a. I know you can change the render location with -o, but would think it would be better for blender to write these to possibly to the home directory somewhere (.blender/render) ? or the like by default. Thanks!";"None";"None";"None" 18178;1;"Open";3;13478;"Augustin Ma";100;"Nobody";"2009-01-16 05:14";"1970-01-01 01:00";"2009-04-08 20:06";"Dead keys do not work to input text";"Hello, Dead keys do not work. In some languages and depending on the keyboard layout, some accented characters need to be inputed with the help of dead keys. http://en.wikipedia.org/wiki/Dead_key The attached blender file has the following content both in a text object and in a text file (this example assumes input using a French keyboard layout): ========= Accented letters without dead keys: éçàè with dead keys ¨e^e¨i^i ========== The last line with dead keys is supposed to be ëêïî. I am not sure any of the info below is relevant, but just in case it is... = Blender 2.48 System Information = ===================================== Blender REV-17181M linux2 dynamic. Platform: linux2 ======== Python: ====== - Version: 2.5.2 (r252:60911, Jul 31 2008, 17:44:49) [GCC 4.2.3 (Ubuntu 4.2.3-2ubuntu7)] OpenGL: ====== - Renderer: GeForce 6200/AGP/SSE/3DNOW! - Vendor: NVIDIA Corporation - Version: 2.1.2 NVIDIA 169.12 ";"None";"None";"None" 18234;1;"Open";3;13662;"Martin Lubich";100;"Nobody";"2009-01-27 21:03";"1970-01-01 01:00";"2009-04-22 13:22";"Making external objects/data local looses references if paths are relative in the library";"Lets start with the following directory structure project-- library-- textures Assume you have a library library.blend in project/library which references through a texture to an image file checker.png in projects/library/textures. The option is chosen to make all paths relative, so the internal path to the image is set to //textures/checker.png (texture files). We create a main project file main.blend in projects. We link the object from library.blend which in turn references checker.png. Lets load main.blend again and decide to make the linked object including its data local. Save main.blend and reload it. Now you end up with a reference to checker.png as //textures/checker.png which of course is invalid as now we are based against main.blend. So it seems that upon localizing external data eiterh no adjustment for relative paths is done or the adjustment is done against the wrong base directory. This behaviour is reproducible and verified with blender 2.48a and 2.47. Other versions weren't tested. Tests done on the following system: - blender 2.48 64bit linux static version - ubuntu 8.04 hardy heron with actual updates as of 27.01.2009 - AMD Phenom X4 9750 @ 4x2.4 Ghz - 8Gb RAM - NVidia 8500 GT w 256MB See attached a zip file containing the problem described above.";"None";"None";"None" 18255;1;"Open";3;11509;"futar ";100;"Nobody";"2009-02-05 00:54";"1970-01-01 01:00";"2009-04-06 11:07";"Node Material passes don't work in the Compositor";"When you create a Node Material using a mix of two different materials to achieve a combined effect, only the information from the material highlighted in the node tree gets sent to the compositor (e.g. colour pass, reflection pass, specular pass) instead of a mixture of the two materials passes. This makes compositing with node materials a nightmare. In the example .blend try selecting a different material in the node tree each time you render to see that the output is dependent on the material selected. If this is deliberate then I apologize... but I can't see why you would want this to occur. Mac OSX 10.4.11 PPC ATI Radeon 9600 Pro Blender 2.48a";"None";"None";"None" 18267;1;"Open";3;6874;"Chris Cherrett";100;"Nobody";"2009-02-09 04:53";"1970-01-01 01:00";"2009-04-22 19:44";"Defocus Node Gives Poor Results While Animating";"When pulling a rack focus with the defocus node, you get strange artifacts on the edges of foreground objects. It looks like flickering. In the video I posted you can see as the man in the foreground comes into focus the edges of his head flicker in and out of focus. I have tried every setting I can think of and have tried this many different times in many different scenes and always get the same type of problem. Thanks!";"None";"None";"None" 18270;1;"Open";3;13809;"Moris Molino";100;"Nobody";"2009-02-09 15:51";"1970-01-01 01:00";"2009-04-05 20:24";"Outliner Alphabetical Disorder";"I've been using the sequence editor quite a bit at work for some fast re-editing of episodes. I recently needed to use the outliner to keep track of clip versions. It was then to my surprise that I discovered something that I had come across many years back in Houdini. It was a bug to them, so I figured... this might be a bug in Blender. You can decide. I'm assuming that the out liner displays the order of things by default in alphabetical order. That order either does not consider numerical order or the sorting only goes so far. I've added a screen shot to show what I mean. I was able to re-create this for the screen grab. You can see that for the most part it is alphabetical and case sensitive, it's once you get into the numbers that things go crazy. Is there a sort option that I'm missing? Anyway... I didn't do much as far as the steps I used. I basically tried to emulate the way things have been working on this project. I load the clips in the order that they should appear in the show. That isn't necessarily the order that they are numbered. Anyway.. what ever order I load them up in shouldn't matter in the out liner, as far as I'm concerned. Anyway... take a look. Let me know what you think. Thanks M";"None";"None";"None" 18280;1;"Open";3;6689;"Ed Britton";100;"Nobody";"2009-02-13 19:58";"1970-01-01 01:00";"2009-02-22 23:03";"Inconsistent L Hotkey";"When using the L hotkey to select linked islands of vertices from the cursor position: If done in the UV/Image view, you must hold Shift+L to add or remove from the selection, and pressing L without shift will deselect all but the currently highlighted island. This is inconsistent to the 3D view, where pressing L without shift will add to the selection and Shift+L will only allow you to remove a selection. Blender SVN 248.2, Windows XP, Nvidia GeForce 8800";"None";"None";"None" 18281;1;"Open";3;5756;"Andrew Price";100;"Nobody";"2009-02-14 00:36";"1970-01-01 01:00";"2009-04-05 19:51";"Particles do not reverse when using time IPO";"Step one: Add a plane. Step two: Create a particle system for that plane. Step three: Change the particle settings so that it emits something. I used AccZ: 1.00 and Random: 0.400. Step four: Go to the IPO curve. Step five: Select the 'Time' ipo and create a hill type shape so that the particles are supposed to launch outwards then stop and come back in again. Step six: Hit Alt-A to see the results. As soon as the time ipo curve starts to go backwards the particles suddenly stick to the emitter then slowly die. System info: ===================================== = Blender 2.48 System Information = ===================================== Platform: win32 ======== Python: ====== - Version: 2.5.2 (r252:60911, Oct 13 2008, 20:50:37) [MSC v.1500 32 bit (Intel)] OpenGL: ====== - Renderer: GeForce 8800 GTS 512/PCI/SSE2 - Vendor: NVIDIA Corporation - Version: 2.1.2 ------------------------------------------------- I have run this on a linux system with the same results. Apparently 2.47 has the same issue, but in 2.45 it works fine. There are threads about the issue here: http://blenderartists.org/forum/showthread.php?t=144940 and http://blenderartists.org/forum/showthread.php?t=148109 Thank you for your troubles!!";"None";"None";"None" 18292;1;"Open";3;7536;"Uwe Girlich";100;"Nobody";"2009-02-18 23:53";"1970-01-01 01:00";"2009-04-05 19:33";"softbody sometimes does not interact with same layer collision objects";"I use a self-compiled SVN 19035 on a Linux 32 bit x86 system. Softbodies should interact with collision objects on the same layer. Unfortunately, this is true only, if the layer of the collision object is ""right"" just at the beginning of the softbody simulation. It looks like the softbody simulator collects the list of potential collision objects at the first frame and does not take into account layer switch IPOs, which can bring in collision objects from other layers at any later time. I have a big softbody, which interacts with objects on the same layer. Later on, another collision object comes in from another layer but this new object is ignored. When I only simulate the softbody around the new collision, everything works because now the new object is at the right layer already at the beginning of the simulation. It took me almost a day to find the reason: the layer switch. Completely stupid workaround: remove the layer change IPO, bake the softbody and then re-add the layer change IPO. The attached blend file contains two situations: * Left: a collision object interacts correctly with a softbody on the same layer. * Right: the collision object starts on layer 2 and comes to the layer 1 only shortly before the actual collision would happen. The right softbody does not react to the collision sphere. I also set an object group but this does not help. ";"None";"None";"None" 18308;1;"Open";3;2970;"trip fbi";100;"Nobody";"2009-02-21 20:03";"1970-01-01 01:00";"2009-04-05 19:19";"Keyboard layout Dvorak partially works on OSX 10.5.6";"On OSX 10.5.6 the universal keyboard support of Dvorak does not work in the view and defaults back to QWERTY when trying to use hotkeys. However in the property controls, sliders, and text editor it keeps the Dvorak awareness of the keyboard layout. ";"None";"None";"None" 18333;1;"Open";3;14412;"Larry Phillips";4290;"Joshua Leung";"2009-02-27 13:00";"1970-01-01 01:00";"2009-02-28 11:51";"Action editor / grease pencil mode - no jump to prev / next keyframe";"Jump to next or previous keyframe (menu or hotkey) does not work in Action Editor Grease Pencil mode. Jump to next or previous marker hotkey doesn't either (menu does). Workflow hit for animation previs but there is a workaround (key a mesh as you do each drawing) Thanks ";"None";"None";"None" 18349;1;"Open";3;6183;"Troy Sobotka";100;"Nobody";"2009-02-28 18:21";"1970-01-01 01:00";"2009-04-05 20:16";"NODE PREVIEW THUMBNAILS: All thumbnails are corrupted with PNGs.";"Summary: When using PNG images created in other applications, the alpha channels appear inverted in all node thumbnail windows. This results in translucent / transparent areas appearing opaque white while the _actual_ opaque areas reveal the cross hatch transparent background. Attachment reveals how this bug manifests itself in context.";"None";"None";"None" 18364;1;"Open";3;11108;"bert baum";100;"Nobody";"2009-03-03 15:39";"1970-01-01 01:00";"2009-04-06 18:56";"loop cut: uv coordinates are always in the middle, ignoring the actual loop position";"Hi! assuming you have a uv-unwrapped model and you want to add a loop cut 1/3rd between two existing loops. What happens is that in the uv layout the loop is ALWAYS set right in the middle of the two existing loops, ignoring the actual position of the new loop. It would be much better if the new loop inside the uv layout would also be set in the same position realtive to the two existing loops. Then, the texture doesn't get distored by adding a loop that is not located right in the middle. Have a look at the example pics and the sample .blend Cheers, zapman";"None";"None";"None" 18367;1;"Open";3;12201;"Istvan Burbank";4042;"Janne Karhu";"2009-03-05 00:05";"1970-01-01 01:00";"2009-04-04 17:01";"boids die on hit and sticky";"While the die on hit and sticky features work for the default particles, they seem to do nothing for boids. Please see the .blend. The boids particles pass right through the mesh that has collision enabled.";"None";"None";"None" 18397;1;"Open";3;5447;"Atomic ";100;"Nobody";"2009-03-10 20:25";"1970-01-01 01:00";"2009-04-05 20:38";"Text window does not correctly interpret CR character code embdded in text.";"Hi All, I have a text file that uses only CR to terminate a line. The Blender text window fails to correctly interpret this kind of document, thus all code ends up on a single line. Quite frustrating really. I am attaching a text file that demonstrates this flaw. Open the file in the Blender text window and observe the bug.";"None";"None";"None" 18409;1;"Open";3;5365;"Roger Wickes";4246;"Peter Schlaile";"2009-03-14 13:49";"1970-01-01 01:00";"2009-05-21 12:31";"Strip Multiply alpha lost";"open attached file. first pic (balloon) alphaover color strip works. Second (girl circle) does not. Girl circle is a multiply blending mode of base image and mask. Mask is formed either by multiplying scene strip, or by multiplying the converted scene strip; either way, somehow it's lost. If you mute the alphaover and color, you will see that the render result of girl circle is correct; it is alpha 1 where the mask is, and 0 elsewhere. However, when it is alphaovered a color, the alpha for the intermediate is somehow lost. ";"Interface";"None";"None" 18415;1;"Open";3;14836;"jason mclaughlin";100;"Nobody";"2009-03-16 16:32";"1970-01-01 01:00";"2009-03-16 16:32";"connected wireframe movement";"what about a feature where joints in a 3d wireframe can be connected to move at the same time with handles? given a handle that's connected to many joints. it would be to say that the move of the handle is a movement of joints at the same time, with saying for the handle how much to move the angle of each joint at the same time. so one handle to move some is for one joint to move a few degrees one way and another to move a few degrees another way. so one handle to move many joints at the same time can be for making a whole hand move like open or closed. many handles for the same hand maybe for different ways a hand can move. then coulnd't a 3d model be made to move all at once into different character poses and such? like.. it could be that given one joint and another, and a handle between both, that an algebraic-like expression of how much the handle moves is for how much each joint moves at the same time... so the handle to move up and down is for both joints to move at the same time different. so a handle to open and close a hand maybe.. with fingers in or out maybe? ";"None";"None";"None" 18433;1;"Open";3;3191;"Charles Wardlaw";100;"Nobody";"2009-03-20 15:05";"1970-01-01 01:00";"2009-04-08 19:52";"undo behaves differently between keyed and unkeyed objects";"Hi folks, This is pretty major for me, a character animator. I never noticed it before because I always used AutoKey. Essentially, undo pushes for movement on objects with set keys are ignored. I have a simple demo you can do to see the difference: 1) How it should work. Take the default cube. Move it three times, then hit undo three times. Each time it will undo back to the previous position, until it's back where it started. 2) How it works with keys Set a LocRot key on the cube, then try the above. When you hit undo it moves the cube back to the last keyed position. This is bad. It really needs to behave like #1 in both situations. As someone who's animated a lot in over the last year, I can say with confidence that #2's workflow is just not as nice as #1's. Each motion should have its own undo push, and the setting of the key should also have an undo push. And before someone replies, ""This is a feature, not a bug -- use AutoKey"": It's not a feature. ;) Thanks so much for all your tireless work, guys.";"None";"None";"None" 18442;1;"Open";3;5422;"James Crosby";100;"Nobody";"2009-03-22 19:44";"1970-01-01 01:00";"2009-04-08 20:15";"Icon for OS X isn't big enough.";"Icons for leopard are meant to contain sizes up to 512x512, the current icon only contains up to 128x128. I believe that in including these sizes doesn't render the icon incompatible with older versions. I used the icon svg from svn (release/freedesktop/icons/) to re-create the sizes, however the results aren't quite identical to the old icon. It may be better to contact the original artist. Please find a new .icns file, and the svg files for large and small sizes. The smallest size in the .icns file remains unchanged. ";"None";"None";"None" 18444;1;"Open";3;15037;"Francis Engels";100;"Nobody";"2009-03-23 22:57";"1970-01-01 01:00";"2009-04-13 18:31";"problem with obstacles using subsurf modifier in the fluidsystem";"When using an object with the subsurf modifier in its stack, the fluid (domain) does not collide with the obstacle properly after baking. The fluidmesh takes the form of the original non-subsurfed object, as if the fluidsystem doesn't take the subsurf modifier into account. According to this post on the community forum there are at least two cases known.(http://www.blender.org/forum/viewtopic.php?t=14546) Also, in previous versions this behavior was not happening. Workaround is to 'apply' the subsurf modifier on the obstacle object, removing it from the stack, thus later adjustments aren't possible anymore. My specs ------------------------------------------------------------------------ Processor: Intel(R) Core(TM)2 CPU 6600 @ 2.40GHz (2400 MHz) Operating System: Microsoft Windows XP, 32-bit (Service Pack 3) DirectX version: 10.0 GPU processor: GeForce 9800 GTX/9800 GTX+ Driver version: 182.08 Stream processors: 128 Core clock: 675 MHz Shader clock: 1688 MHz Memory clock: 1100 MHz (2200 MHz data rate) Memory interface: 256-bit Memory: 512 MB Video BIOS version: 62.92.3B.00.05 IRQ: 16 Bus: PCI Express x16 ------------------------------------------------------------------------- Blender 2.48a : Built on 2008-10-22 21:29:17, Rev-17183M ";"None";"None";"None" 18475;1;"Open";3;15172;"Yap Chun Fei";100;"Nobody";"2009-03-30 07:33";"1970-01-01 01:00";"2009-04-13 19:18";"Baking textures accuracy bug from one low poly model to another same low poly model with different UV ";"Ok I've been having this problem whereby I have 2 similar low poly models with different UV layouts. Trying to bake textures (color, normals..etc...) from low poly model A to low poly model B however there are some accuracy problems as some pixels are distorted or out of place. original texture on low poly model A image: http://img403.imageshack.us/img403/7619/originalr.jpg baked texture on low poly model B image: http://img4.imageshack.us/img4/5426/problemq.jpg This distortion does not happen with 3ds max. To ""fix"" this problem I've had to increase mesh tesselation for low poly model B, bake it and re optimize it back to its original mesh shape because of this baking inaccuracy which is time consuming. ""fix"" image: http://img403.imageshack.us/img403/2589/fixe.jpg Also I've noticed that this distortion gets more apparent towards the center of a mesh face hence the need for mesh face tesselation. center problem image: http://img5.imageshack.us/img5/8228/meshface.jpg I have included a zip file bug.zip which includes a blend file with textures old.jpg and new.jpg which contains the distortions. Within the blend file are 2 similar models objects (called old and new) with different UV layouts to bake textures from old onto the new model with different UV layout. Please help. This feature is important for game artists.";"None";"None";"None" 18483;1;"Open";3;8753;"Tenth Planet";100;"Nobody";"2009-04-03 23:39";"1970-01-01 01:00";"2009-04-03 23:39";"Background picture - separate aspect ratio from context.aspX/Y";"2 3DViews with different cameras/bgpics may need different aspect ratios when modeling from historical photos / book pictures which have been cropped. Current workarounds: 1. The last bgpic loaded gets its aspect ratio in context.AspY OR 2. view.c hack so images are rendered with natural aspect ratio imageY/imageX 3. artists suggestions: use textured plane instead of bgpic but then its in textured mode and everthing in front needs to be in wireframe - awkward Related problem: Bgpic not associated with camera. When cameras are changed, it would be nice to have the bgpic associated with the camera and have it changed too. Ideal for stills: On Camera Editing panel, a PerCameraBgpic toggle, which when depressed would bring up another panel, and it would have an aspect ratio as well as other bgpic parameters AND be exposed to the python API so can be exported as a photo tour to GoogleEarth KML. ";"Interface";"None";"New" 18498;1;"Open";3;8997;"Keith Boshoff";100;"Nobody";"2009-04-08 20:45";"1970-01-01 01:00";"2009-04-25 19:22";"Projection paint bug";"If uv's overlap on an object that's being projection painted artefacts appear. Video evidence: http://www.wahooney.net/files/ProjectionPaintBug.ogm";"None";"None";"None" 18512;1;"Open";3;7038;"Constantin Rahn";100;"Nobody";"2009-04-11 14:15";"1970-01-01 01:00";"2009-04-21 15:19";"Preferences ""Aligned to View"" doesn't work in ""Edit Mode""";"The option ""Preferences"" -> ""Edit Modes"" -> ""Aligned to View"" only works in ""Object Mode"". In ""Edit Mode"" the option ""Aligned to View"" is ignored. You can add a new meshes to an existing mesh in Edit Mode with ""Add"", too. Not only in Object Mode. A .blend makes no sence here. Simply add a cube or plane, to the existing cube in the default scene, in Edit Mode. Make sure you have switched off the ""Aligned to view"" preference.";"None";"None";"None" 18513;1;"Open";3;6265;"Hannu Hoffrén";4042;"Janne Karhu";"2009-04-11 16:26";"1970-01-01 01:00";"2009-05-02 01:55";"Boids with crowd value of zero are emitted in groups when emitter mesh is small";"This one is for jahka. When boids have crowd value of zero they are still emitted in groups when the emitter mesh is small. Attached blend file has same particle system in both large plane and small plane. The effect can be seen at around frame 30, Additional question: Why boids are drawn to spherical field when the newtonian particles are repelled. This effect can be seen in the same file.";"None";"None";"None" 18526;1;"Open";3;7384;"Lance Flavell";100;"Nobody";"2009-04-15 07:21";"1970-01-01 01:00";"2009-04-21 15:37";"Image node does not recognise valid channels for OpenEXR format.";"According to the page at http://www.blender.org/development/release-logs/blender-243/imaging/ Blender Node Editor is able to import OpenEXR images showing all the channels on the node. This seems true when the .exr has been made in Blender. However, if using multichannel EXR as part of a production pipeline where the images were not originally made in Blender, then the only options available are ""Image"", Alpha"" and ""Z"". Blender image node does not automatically pick up on the other tags like it is supposed to for this format. I have tried samples from... * http://www.openexr.com/downloads.html (direct link to EXR files: http://download.savannah.nongnu.org/releases/openexr/openexr-images-1.5.0.tar.gz ) * http://www.fnordware.com/OpenEXR/ (direct link: http://www.fnordware.com/OpenEXR/x35_3a.exr ) ...with Blender not being able to recognise the available pictures in these images. I've just had a production team put off using Blender as their compositing tool as a result of this fault. Forum link for further discussion: http://blenderartists.org/forum/showthread.php?t=153389 Cheers - Lancer.";"None";"None";"None" 18551;1;"Open";3;15626;"Cameron Rapp";100;"Nobody";"2009-04-16 21:52";"1970-01-01 01:00";"2009-04-21 22:26";"Stereo Rendering no functioning properly in game engine 2.48a";"The Stereo renderer appears to have some issues, with anaglyph rendering the separation is the same regardless of distance, producing a flat image even when viewed through glasses. The hues might also be a bit incorrect for standard glasses. Also in Split screen mode, Images from left and right cameras look completely different, Using yofrankie as an example, screenshots attached.";"None";"None";"None" 18580;1;"Open";3;15731;"Brad Gonsalves";100;"Nobody";"2009-04-21 10:42";"1970-01-01 01:00";"2009-04-21 17:54";"mirror messes up colorband + nor input";"I discussed this problem here..... http://blenderartists.org/forum/showthread.php?p=1363278&posted=1#post1363278 I was following the following tutorial on how to achieve backface culling- (because single sided planes are only transparent in opengl and not the renderer, I had to find an alternative solution) http://wiki.blender.org/index.php/Doc:Tutorials/Textures/Map_Input_Techniques see section right at the bottom labeled single sided face if you look at the third picture I have submitted of the bathroom, as per the above mentioned tutorial, I made the outer wall simulate backface culling, so standing out side of the bathroom looking in, the wall will appear transparent, but standing inside the bathroom looking out the wall is completely opaque.... this works perfectly until I introduce a mirror reflecting the wall in question, then, as you can see, the wall is invisible, (only the walls reflectivity is still present) to reproduce this I have included a .blend file and 2 images- one of a box with blue assigned to its +z on the normal and red to the -z of the normal. when the box is filmed through the opening you can see that the inside of the box is blue, where as the outside is red- ie the normals are facing inward. (see image 1) which is working as it should as soon as you look through the opening with the use of a mirror, the far end of the box reflects red now instead of blue (see image 2) I believe that a mirror is reflecting the wrong side of the plane";"None";"None";"None" 18584;1;"Open";3;4544;"Kai Kostack";100;"Nobody";"2009-04-21 23:28";"1970-01-01 01:00";"2009-04-22 13:58";"GLSL: strange alpha problems with an array of planes";"i have uploaded a blend file with packed textures. in there you can see an array of textured planes with alpha maps. from the starting position it looks correct but if you rotate around the array you will notice that from one side the planes will become invisible except the one in front of all. it seems not to have anything to do with the face normals because flipping of those doesn't change anything. ";"None";"None";"None" 18621;1;"Open";3;15882;"Fabio Cavicchio";100;"Nobody";"2009-04-26 15:39";"1970-01-01 01:00";"2009-04-29 20:11";"Textures with mipmap option in spherical projection: artifacts at poles.";"Equirectangular textures, when applied to spheres with option ""Mipmap"" give artifacts near the north and south pole of the sphere. See the attached blend file: if you disable the Mipmap option then the result is OK. A sample texture is included. It seems that with a 2000x1000 equirectangular image the top and bottom 7-8 rows are ignored by the mapping, and the texture is evaluated as 2000x985. This cannot be seen on simple Earth textures since the poles are all white, but if you use a high quality image with some details on the polar region then the problem is evident. ";"None";"None";"None" 18625;1;"Open";3;15871;"David Lindsley";100;"Nobody";"2009-04-27 01:29";"1970-01-01 01:00";"2009-05-14 14:06";"Linked particle mesh loses its particles";"This is hard to describe, so I have attached 2 blend files. The first blend file, DadTest.blend has a character with particle hair on a separate mesh from the head - like a wig. The second blend file, LinkedDadTest.blend links to the first. When that character is linked into the second file, all is well, except the linked particles always reverts back to its initial position, rather than where the mesh is. The mesh is in the right spot, but the particles are not. If I select the mesh, then go to Particle Mode, the particles will snap into their correct placement. So while, this is an okay work-around, there's no way I can send this file to a render farm, because every time the file is opened, the hair is in the wrong spot until I manually fix it via Particle Mode and then Object Mode. This is the same on my WinXP 64 machine and my WinXp 32 bit and a Ubuntu 32 machine. ";"None";"None";"None" 18641;1;"Open";3;14430;"Blendiac the Third";100;"Nobody";"2009-04-28 23:11";"1970-01-01 01:00";"2009-04-28 23:11";"Cameras in Linked Groups Can't be Viewed (Append Works Fine)";"This one's simple. If you link a group with a camera in it into a scene, you can't view that camera! Strangely enough it *does* work if you append, just not with link. According to someone on the forum, you can see it if you use a camera actuator, but then that forces you to use that actuators camera movement. I just need a camera parented to my character which can be linked in to any scene file. You can see this by linking and appending the group ""cam"" from the attached file.";"None";"None";"None" 18646;1;"Open";3;15874;"Fredrik Hansson";100;"Nobody";"2009-04-29 04:45";"1970-01-01 01:00";"2009-05-02 23:22";"Projection paint brush texture not stuck to the pen";"not exactly sure if this is intended as it could be useful in a few rare cases however its inconstant with how the painting in the uv/image editor and can't be turned off. anyway when i paint in the 2d view i get the brush stuck nicely on the pen tip but if i paint on the mesh i get a tiling pattern with no control over how its going to tile or if i even want it to tile instead of just work like the 2d version does.";"None";"None";"None" 18658;1;"Open";3;1964;"Mort ";100;"Nobody";"2009-04-30 22:17";"1970-01-01 01:00";"2009-05-03 13:10";"| character in textbuttons";"the | (pipe) character behaves as a row of white space filling the row, new characters appear on the far right (cursor location also incorrect)";"None";"None";"None" 18663;1;"Open";3;14430;"Blendiac the Third";7473;"Benoit Bolsee";"2009-05-01 07:37";"1970-01-01 01:00";"2009-09-12 00:28";"Soft bodies ignore constraint actuators";"If you take a look at the file attached (created on 2.48a), you'll notice there is a constraint that should keep the ball object on the 2d plane perpendicular to the camera. It works fine when the object is set to dynamic or rigid body, but seems to stop doing anything when the object is set to soft body type. Use arrow keys to navigate left and right.";"None";"None";"None" 18688;1;"Open";3;15874;"Fredrik Hansson";4827;"Nicholas Bishop";"2009-05-05 03:11";"1970-01-01 01:00";"2009-09-12 00:27";"Projection painting on mirrored uv edge gives artifacts";"painting along the center line of a mirrored object seems like it can't make up its mind about what fade strength to use and also it never uses the bleed value to fill up outside the edge leaving a line along the center. and undoing this also gives trouble as i get all sorts of patterns from doing that. i get the worst results with this if i draw quickly back and forth over the seam. system: windows vista x64 intel core2 quad q6600 GeForce GTX 260 8gb ram svn version: Built on 2009-05-04 16:55:22";"None";"None";"None" 18719;1;"Open";3;15871;"David Lindsley";146;"Brecht Van Lommel ";"2009-05-11 03:12";"1970-01-01 01:00";"2009-09-12 00:27";"Softbody in linked/proxied objects or groups fails.";"Here is something that is either very broken, or I am just not doing something right. This works very well in 2.45. The attached blend file, softcube.blend is the basic cube as a soft body, and a simple armature in it. Here are the steps to reproduce the problem. Open up softtest245.blend in Blender 2.45. You will see the linked object from softcube.blend in it. It is a very simple animation. Run it and you will see the cube rock from one side to the other, bouncing around a bit, because it has a jello-like softbody applied to it. Okay so far so good. Now open up softtest249.blend using Blender 2.49. It links to the same softcube.blend file, and does the same thing as before. EXCEPT that the softbody-ness of the cube is GONE! It's just rigid. What happened here? I tried this in 2.46 and 2.48 with slightly different results, but none of them got it right as in the 2.45 test. This is either a major bug, or there has been a change in the way things happen with linking in softbody objects starting with version 2.46. (I tried to see if 2.49 RC2 is any better, but 2.49RC2 doesn't work at all. Whenever I try to run it, it says, ""The application failed to initialize properly (Oxc0150002). Click on OK to terminate the application"") ";"None";"None";"None" 18725;1;"Open";3;11085;"Nathan Braswell";1013;"Jiri Hnidek";"2009-05-11 15:42";"1970-01-01 01:00";"2009-09-12 00:23";"Particle group duplication cannot include metaballs, recursion happens";"I am using svn revision 20122M when I found the bug. I have confirmed it with versions 2.49 RC 2 and 2.47. It dosn't work in 2.45 because partical system disipears. (before partical system rewrite?) SETUP: I was trying to create a energy blast thing for my new movie when I notised this bug. I was using a UV sphere with a partical system using a group of a meta ball and a lamp with a low power blue light. Then I use a wind empty to move them. ERROR: When I try to render it renders the full frame, then quits or crashes. When run from the command line no error is reported, I just get the regular prompt again. No quit.blend is created. When appending the objects into a different blend file it crashes immedently.(before i render as it is appending) Is reproducable, I have created a similar setup from scratch, same error results. * CPU: Mobile INtel(R) Celeron(R) CPU 2.00GHz * RAM: 512 MB * GFX: NVIDIA GeForce4 420 Go (Toshiba) * OS: Windows XP Home, service pack 3 * Blender: 2.48.4 2009-05-09 17:27:18 Rev-20122M";"None";"None";"None" 18789;1;"Open";3;8241;"Bo Johansen";4246;"Peter Schlaile";"2009-05-19 14:38";"1970-01-01 01:00";"2009-09-12 00:22";"AVI files containing alpha doesn't work";"Using an AVI file in textures/background/composite nodes etc. doesn't work if the AVI file contains an alpha channel. In earlier versions 2.4x it did work. Not sure when it broke, but it doesn't work in 2.48/2.49. I'm using Blender in Windows Vista32 Home Premium (Q6600/4GB/GeForce 8800GT) ";"None";"None";"None" 18806;1;"Open";3;4638;"Simon Nielsen";7473;"Benoit Bolsee";"2009-05-23 06:49";"1970-01-01 01:00";"2009-09-12 00:20";"Dynamically added and dupligroup properties";"When adding objects dynamically (scene.addObject), or using dupligroups, their state and property debug output don't behave the same as those of directly instantiated objects. In the case of states, the line is on screen, but no value shown. In the case of properties, they show their initial value even though the copy's value changes. In the test .blend, the debug states and properties should behave the same as the directly added object's, as the exact same thing is going on. Note that I get both the dupli and added objects from scene.objects by name - which seems to yield the added/dupligrouped copy just fine. Yet, I'm guessing, the originals' properties are the ones showing up in debug. IOW, the debug flag isn't inherited (or the list is compiled at engine init), and objects on hidden layers show up with no useful info. I might very well be the one doing things wrong by getting the copies by object name - in that case, an alternate method (storing object IDs in GameLogic.globalDict?) should be documented somewhere. I'll gladly help with this. The .blend is a bit complex, but should be self-explanatory.";"None";"None";"None" 18807;1;"Open";3;16503;"Yuval Levy";4246;"Peter Schlaile";"2009-05-23 07:46";"1970-01-01 01:00";"2009-09-12 00:20";"AVCHD files (24Mb/S) ";"These files are straight out of a Canon Vixia HF11. Only very recently VLC Media Player has been able to play them, and I understand Blender, like VLC, uses ffmpeg to decode them. I tried them on Blender 2.48a in both Windows XP (official download from blender.org) and Ubuntu 9.04 (via apt-get). The decoding is not right. They appear distorted, with block-like artefacts. Attached is one file for you to test. Thank you.";"None";"None";"None" 18825;1;"Open";3;9125;"Alexandros Sigalas";100;"Nobody";"2009-05-25 20:14";"1970-01-01 01:00";"2009-05-26 19:45";"Separate by material creates empty meshes for material IDs that aren't used by some face";"make a cube with 6 materials assigned each to each face, delete some faces, seperate by material. There are 6 objects created, even though less faces actually existed. All surplus objects are empty meshes.";"None";"None";"None" 18834;1;"Open";3;6730;"Remigiusz Fiedler";100;"Nobody";"2009-05-27 07:47";"1970-01-01 01:00";"2009-05-27 16:58";"Apply Scale/Rotation to ObData for mirrored objects produces negative normals";"working on output for mirrored geometry in dxf-exporter.py i found this bug: CubeA is negative scaled (mirrored) mesh-object CubeA2 is the same object, after ""Apply Scale/Rotation to ObData"": all normals are flipped inside, but should not! *)tested on 2.49rc3 winXP32 **)attached blend file Patch idea: before start ""Apply..."": 1. check if obj.matrix is a left-hand-system 2. if true: flip all face-normals (=set opposite vertex order) A prototype in python works well: def isLeftHand(matrix): #Is the matrix a left-hand-system, or not? ma = matrix.rotationPart() crossXY = Mathutils.CrossVecs(ma[0], ma[1]) check = Mathutils.DotVecs(ma[2], crossXY) if check < 0.00001: return 1 return 0 if isLeftHand(obj.matrix): # then change vertex-order in every face for f in faces: f.reverse() #f = [f[-1]] + f[:-1] #TODO: maybe needed to preserve UV etc... i can try to write C code, but would prefer someone else do this. migius";"None";"None";"None" 18861;1;"Open";3;16671;"matthew lagoe";100;"Nobody";"2009-06-01 14:52";"1970-01-01 01:00";"2009-06-12 18:00";"-02 flag renders white background and frame 70 wont render";"Frame 70 will not render correctly there are 3 things it sometimes does 1) doesnt render at all and gives boolean error 2) renders white background 3) renders the wrong frame Also this scene will only render white background when -02 is enabled";"None";"None";"None" 18904;1;"Open";3;5768;"Gwenael Le Page";100;"Nobody";"2009-06-08 19:08";"1970-01-01 01:00";"2012-10-29 12:35";"Tex Node Editor: group doesn't take changes into account";"This seems to be a bug to me: in the tex node editor, do the following: click ""use nodes"" unlink checker and output add mix and rotate like following: -----Rotate-- / \ / MIX ------------- Output / / Checker --------------- Group Rotate and Mix Now if you are in compact viewing mode, the changes you make are updated in the output node and are effective. But if you switch to extended view: _ the values haven't been changed. _ if you change them now, the output won't update _ if you then go back to compact view, the value changes won't be carried over. In fact it is as if the input fields were numb and disconnected while in extended view. I join a .blend file";"None";"None";"None" 18910;1;"Open";3;6764;"Juan Pablo Bouza";146;"Brecht Van Lommel ";"2009-06-09 15:00";"1970-01-01 01:00";"2009-09-12 00:15";"Two IK constraints problem";"I´ve noticed that when you have two IK Constraints in the same bone, and the influence of the top IK is IPO driven, when the influence starts going down, the bone starts returning to the rest position instead of pointing to the bone targeted in the second IK constraint. Then, when the driver bone reaches the position where the influence of the first constraint is 0, the affected bone jumps to the bone targeted in the second IK constraint all of a sudden. So, the result is that the change in the influences of the constraints is jumpy, there is no smooth transition, regardless of the IPO curve. I´ve attached a blend example.";"None";"None";"None" 19006;1;"Open";3;17380;"Hans Wurst";100;"Nobody";"2009-07-11 15:50";"1970-01-01 01:00";"2009-07-11 15:50";"change the order of properties";"interface improvment in the logic panel (F4): allow to move single property entries up and down in the list.";"Interface";"None";"New" 19069;1;"Open";3;13885;"Kevin Delbrayelle";100;"Nobody";"2009-07-29 19:01";"1970-01-01 01:00";"2009-08-31 00:46";"No warning about unsaved file when opening new file/exiting.";"Sorry for posting this bug which is much a feature request than a real bug. It is understood the document states this warning : ""Blender expects that you know what you are doing! When you load a file, you are not asked to save unsaved changes to the scene you were previously working on, completing the file load dialog is regarded as being enough confirmation that you didn't do this by accident. Make sure that you save your files."" I'm a heavy blender user and don't get much problem with this, but some of my colleges do sometime lose hard hours of work. Let me explain while a warning prompt saying the current file has not been saved yet would be welcome by several blender users: 1. In a development environment, especially when working in team in the same office, people can be disturbed/asked by other and make mistakes, such as forgetting to save a file. 2. This is especially true when people do work a lot and get tired. Mistake are more and more common. 3. It is steal possible to accidentally hit ""F1"" instead of ""Shitf-F1"", and then load a file rather than append/linking data. Thank you very much for work done on blender, and thank you for reading :) Kévin";"None";"None";"None" 19770;1;"Open";3;8509;"Matt Runion";100;"Nobody";"2009-10-31 04:34";"1970-01-01 01:00";"2009-12-05 00:59";"Higher Number of Hair Particles Crash Blender";"When working with hair particles in Blender 2.5 (latest build on my machine, built from SVN on 10/29), increasing the particle count to 4000 and using child particles (using 100 per parent) will cause Blender to crash when rendering and the ""ray tree"" step is processing. Quick overview: 4000 hair particles, 100 child particles each Ray Tracing enabled, Approximate AO used. Using Blender 2.49 (latest released version), the crash does not occur. I have attached a sample file (hopefully I've done it correctly!). Thanks! ";"None";"None";"None" 19885;1;"Open";3;18626;"Bjørnar Hansen";100;"Nobody";"2009-11-14 11:57";"1970-01-01 01:00";"2009-11-14 11:57";"Render to image and video ""in parallel""";"Blender 2.5: When I click on animate I expect Blender to render a video. But by default it renders a series of JPEGs instead. Please make this behave more clever: Have two separate ""output"" settings boxes, one for images and one for video. Expected behavior: When you click on ""Render"" button - render an image. When you click on ""Animate"" button - by default render video. Have an option to render a series of images instead by selecting this (disabled by default). Current behavior: When you click on ""Render"" button - render an image. When you click on ""Animate"" button - render a series of images ";"Interface";"None";"None" 19898;1;"Open";3;3480;"Jorge Bernal";100;"Nobody";"2009-11-15 20:59";"1970-01-01 01:00";"2010-01-31 02:16";"Fullscreen toggle button doesn't restore window mode";"Hi, When you push toggle fullscreen button, blender goes to fullscreen correctly but when you press it again to restore the normal window mode blender looses the previous dimensions and the title toolbar. I attach a screenshot in the restored window mode to show the problem. My system is: linux Mandriva 2009.1 Gfx card: Nvidia 8800GT with latest nvidia drivers. regards, lordloki";"None";"None";"None" 20211;1;"Open";3;4857;"Dalai Felinto";100;"Nobody";"2009-12-02 17:12";"1970-01-01 01:00";"2009-12-19 10:06";"Instanced objects (Alt+D) with materials per object are loaded in BGE with one single material";"We have here 4 instances of the same object (red) each one with one material assigned to the object datablock. However when you launch the game (P) they are sharing the same material. It happens in all modes (Multitexture, GLSL, ...). It happens in Blender 2.49b and Blender 2.5";"None";"None";"None" 20300;1;"Open";3;8622;"Torsten Glose";100;"Nobody";"2009-12-08 23:09";"1970-01-01 01:00";"2010-01-30 16:02";"unicode char in object name breaks python";"Blender2.50Alpha0 As soon as a unicode char(I've used an ""ä"", unicode ""\u00e4"") appear in object names python throws an ""UnicodeDecodeError"". This affect scripts depending on object names. In Blender 2.49b an ""ä"" is printed out as \xe4 rather than throw an exception. The first screenshot shows the exception thrown as I try to print out the name of the default cube after adding an ""ä"" to its name. The other both screenshots shows the difference in UI if an ""ä"" is present or not in a bone name. If the unicode char is present the console shows the following errors: Traceback (most recent call last): File ""D:\BLENDE~1.5-A\.blender\scripts\ui\properties_data_bone.py"", line 81, in draw pchan = ob.pose.bones[context.bone.name] UnicodeDecodeError: 'utf8' codec can't decode byte 0xe4 in position 4: unexpected end of data Traceback (most recent call last): File ""D:\BLENDE~1.5-A\.blender\scripts\ui\properties_data_bone.py"", line 165, in draw pchan = ob.pose.bones[context.bone.name] UnicodeDecodeError: 'utf8' codec can't decode byte 0xe4 in position 4: unexpected end of data Traceback (most recent call last): File ""D:\BLENDE~1.5-A\.blender\scripts\ui\properties_data_bone.py"", line 211, in draw pchan = ob.pose.bones[context.bone.name] UnicodeDecodeError: 'utf8' codec can't decode byte 0xe4 in position 4: unexpected end of data";"None";"None";"None" 20314;1;"Open";3;6183;"Troy Sobotka";100;"Nobody";"2009-12-10 03:28";"1970-01-01 01:00";"2009-12-10 03:28";"SEQUENCER: Setting strobe to any value breaks view output";"Summary: Within the sequencer, setting strobe to any value results in the sequencer preview viewport turning all black. This permanently loses the viewport and requires a restart often requiring a force close for Blender. Version: SVN 25262 Reproduction: 1) Switch to a sequencer view and load a strip. 2) Select the strip and change the Filter->Strobe value to anything other than 1.0. 3) Watch viewport turn black. ";"Interface";"None";"None" 20371;1;"Open";3;6256;"Dietrich Bollmann";100;"Nobody";"2009-12-14 04:33";"1970-01-01 01:00";"2009-12-25 23:34";"Operators fail in background mode: SystemError when executing mesh commands in scripts started from the command line.";"The blender context is not set correctly when executing operators in background mode - for example when executing simple mesh commands as 'bpy.ops.mesh.primitive_cube_add()' from scripts executed via the -P