button under the mouse, and so clicking the button failed without first
moving the mouse. Similar issues could happen with popups and switching
windows.
The problem is that a notifier was being used to deactivate possible other
activated buttons, but as notifiers are delayed it could undo the correct
activation too. If anyone notices hanging tooltips or other issues after this
fix, please report them.
When setting keymap properties to values equalling the RNA default, they will
get "unset" and automatic operator behavior is used. There is no way to
explicitly set the default value as a user.
1) To allow distinguishing uninitialized (not set) properties in the keymap
items, a few changes to the RNA struct comparison function are needed: Instead
of allowing only strict/non-strict comparison of 2 properties A and B in a
struct, this now has 3 modes:
* STRICT: compare only the actual property values (same as 'strict' before)
* UNSET_MATCH_ANY: if either A or B is unset, consider them a match (same as
non-strict before)
* UNSET_MATCH_NONE: if one property is set and the other not, consider them a
mismatch.
The new UNSET_MATCH_NONE mode is useful for keymaps, because it allows keeping
user-defined property values in the keymap even if they match the default
property value (see wm_keymap_diff function in wm_keymap.c)
2) A new operator is added for unsetting ID properties in the RMB context menu
and in user preferences next to keymap properties. This only works on ID
properties and deletes the ID property storage, so that the default value is used.
In the user preferences for keymaps the properties are shown in an inactive
layout to indicate that the default value is used (which some operators such as
the "select linked" op from the report use to trigger automatic behavior). When
the user sets a property it gets set and stays that way until explicitly "unset"
using the new operator.
When printing operator reports, ommit unset properties.
This is needed because in some cases operators check if a value is set or not, so filling in default arguments may change behavior.
curves and metaballs now behave the same as meshes wrt grid scaling.
remove WM_operator_view3d_distance_invoke(), and replace with a function called from exec which initializes defaults, this way operators can have their own invoke functions.
when running viewport operations with exec() rather then invoke(), perform the action immediately rather then using smoothview.
makes viewport operations usable from python scripts.
default settings in the startup.blend without having to actually save and embed
the file, which can be a tricky process and is problematic in branches and
patches.
This function can be emptied each time a new startup.blend is committed.
This allows to open windows stretched over monitors. Saving
stretched layouts open correctly as well.
Renamed screensize_all to desktopsize to better convey meaning.
- add missing headers from cmake (own omission)
- quiet rna_test.c unused define warnings.
- minor style edits
- spelling corrections and ignore all uppercase words with spell checking script.
rotation of the brush mask texture.
Unfortunately secondary path does not work here because we do not have a
permanent switch to choose between primary-secondary brush texture. Use
operator property instead.
This means calling invoke operators from python in states that dont support will give a meaningful error message.
also reduce context lookups when polling.
Instead of having ifdef __GNUC__ all over the headers
to use special compiler's hints use a special file where
all things like this are concentrated.
Makes code easier to follow and allows to manage special
attributes in more efficient way.
Thanks Campbell for review!
- Re-arrange locks, so no actual memory allocation
(which is relatively slow) happens from inside
the lock. operation system will take care of locks
which might be needed there on it's own.
- Use spin lock instead of mutex, since it's just
list operations happens from inside lock, no need
in mutex here.
- Use atomic operations for memory in use and total
used blocks counters.
This makes guarded allocator almost the same speed
as non-guarded one in files from Tube project.
There're still MemHead/MemTail overhead which might
be bad for CPU cache utilization
- reading bmp images was failing (needed to increase the size of the header to 64 bytes)
- the dnd image was being incorrectly checked (was always returning true even when none was used).
Now the viewport rendering thread will lock the main thread while it is exporting
objects to render data. This is not ideal if you have big scenes that might block
the UI, but Cycles does the same, and it's fairly quick because the same evaluated
mesh can be used as for viewport drawing. It's the only way to get things stable
until the thread safe dependency graph is here.
This adds a mechanism to the job system for jobs to lock the main thread, using a
new 'ticket mutex lock' which is a mutex lock that gives priority to the first
thread that tries to lock the mutex.
Still to solve: undo/redo crashes.
