Regression: Trim curve make spline disapear. #101379

New Issue

Paul · 2022-09-26T12:47:51+02:00

Paul commented

2022-09-26 12:47:51 +02:00

System Information
Operating system: Windows-10-10.0.19044-SP0 64 Bits
Graphics card: AMD Radeon (TM) R9 390 Series ATI Technologies Inc. 4.5.14830 Core Profile Context 22.6.1 27.20.20912.1002

Blender Version
Broken: version: 3.4.0 Alpha, branch: master, commit date: 2022-09-26 00:52, hash: 3a7dc572dc
Worked: (13 Sep 22 before moving trim spline to another block)

Caused by eaf416693d

Short description of error
[Trim curve node make cyclic curve disappear when not trimmed]

Exact steps for others to reproduce the error
[Add circle node & trim spline node afterwords, this makes spline disappear. If you only trim to 0,999 instead of 1 (or from 0.0001 instead of 0.0) it comes back.
Trim spline bug.blend

{F13574735}]
[Based on the default startup or an attached .blend file (as simple as possible)]

**System Information** Operating system: Windows-10-10.0.19044-SP0 64 Bits Graphics card: AMD Radeon (TM) R9 390 Series ATI Technologies Inc. 4.5.14830 Core Profile Context 22.6.1 27.20.20912.1002 **Blender Version** Broken: version: 3.4.0 Alpha, branch: master, commit date: 2022-09-26 00:52, hash: `3a7dc572dc` Worked: (13 Sep 22 before moving trim spline to another block) Caused by eaf416693d **Short description of error** [Trim curve node make cyclic curve disappear when not trimmed] **Exact steps for others to reproduce the error** [Add circle node & trim spline node afterwords, this makes spline disappear. If you only trim to 0,999 instead of 1 (or from 0.0001 instead of 0.0) it comes back. [Trim spline bug.blend](https://archive.blender.org/developer/F13574727/Trim_spline_bug.blend) ![image.png](https://archive.blender.org/developer/F13574731/image.png) {[F13574735](https://archive.blender.org/developer/F13574735/image.png)}] [Based on the default startup or an attached .blend file (as simple as possible)]

Paul commented

2022-09-26 12:47:51 +02:00

Added subscriber: @genesis2303

blender-admin commented

2022-09-26 12:47:51 +02:00

#101837 was marked as duplicate of this issue

Philipp Oeser commented

2022-09-26 13:06:02 +02:00

Added subscriber: @lichtwerk

Philipp Oeser commented

2022-09-26 13:06:02 +02:00

Changed status from 'Needs Triage' to: 'Confirmed'

Philipp Oeser commented

2022-09-26 13:06:02 +02:00

Can confirm, will check

Philipp Oeser commented

2022-09-26 13:42:25 +02:00

Added subscribers: @HooglyBoogly, @Osares

Philipp Oeser commented

2022-09-26 13:42:25 +02:00

Caused by eaf416693d indeed

Only an issue for cyclc curves and Start exactly 0.0 and End exactly 1.0

CC @Osares
CC @HooglyBoogly

Caused by eaf416693d indeed Only an issue for cyclc curves and Start exactly 0.0 and End exactly 1.0 CC @Osares CC @HooglyBoogly

Philipp Oeser changed title from ~~Trim curve make spline disapear.~~ to Regression: Trim curve make spline disapear.

2022-09-26 13:43:41 +02:00

Mattias Fredriksson commented

2022-09-26 21:24:18 +02:00

Not sure how to remedy this as this is the current intended behavior as the start and endpoints are equivalent reducing the curve to a single point (see image below). While the behavior may be unintuitive when passing 0.0 and 1.0 as inputs these values do represent the same point on the cyclical curve. This behavior is consistent with any other point on the curve (since input is clamped to [0.0, 1.0]).

Originally I added a mode for cyclical splines which allowed the user to set the behavior to either split, copy or reduce the curve to a point when endpoints where equivalent. The decision was made to wait with introducing such a feature and instead go with the default behavior of reducing the curve to a point since this is the behavior of non-cyclic curves. Due to that decision this report was unavoidable.

The reason for why this 'works' in previous versions is also due to the patch adding support for trimming cyclical curves (previous behavior was just to copy the original curve).

Possible solutions

Solutions to this is more of a design choice as possible solutions could be to:

Option for taking the 'inverted' trim result (the part that is discarded/trimmed).
Add the trim mode back for cyclical splines (initially only supporting copy/reduce to point). This would make most sense if selection input was exposed at the same time (since it would be the simplest solution).
Actually show curves consisting of a single point in the viewport (plausible render similar to mesh vertices).

Not sure what @HooglyBoogly thinks though as it would be his call in the end.

As can be inspected in the spreadsheet editor the curve is reduced to a single point which is displayed by converting to mesh.

Not sure how to remedy this as this is the current intended behavior as the start and endpoints are equivalent reducing the curve to a single point (see image below). While the behavior may be unintuitive when passing 0.0 and 1.0 as inputs these values do represent the same point on the cyclical curve. This behavior is consistent with any other point on the curve (since input is clamped to [0.0, 1.0]). Originally I added a mode for cyclical splines which allowed the user to set the behavior to either split, copy or reduce the curve to a point when endpoints where equivalent. The decision was made to wait with introducing such a feature and instead go with the default behavior of reducing the curve to a point since this is the behavior of non-cyclic curves. Due to that decision this report was unavoidable. The reason for why this 'works' in previous versions is also due to the patch adding support for trimming cyclical curves (previous behavior was just to copy the original curve). **Possible solutions** Solutions to this is more of a design choice as possible solutions could be to: - Option for taking the 'inverted' trim result (the part that is discarded/trimmed). - Add the trim mode back for cyclical splines (initially only supporting copy/reduce to point). This would make most sense if selection input was exposed at the same time (since it would be the simplest solution). - Actually show curves consisting of a single point in the viewport (plausible render similar to mesh vertices). Not sure what @HooglyBoogly thinks though as it would be his call in the end. ![Screenshot (142).png](https://archive.blender.org/developer/F13576465/Screenshot__142_.png) As can be inspected in the spreadsheet editor the curve is reduced to a single point which is displayed by converting to mesh.

Iliya Katushenock commented

2022-09-26 21:33:29 +02:00

Added subscriber: @mod_moder

Iliya Katushenock commented

2022-09-26 21:33:29 +02:00

why between 0 and 1 we have length 0 ?

More precisely, how to determine which part of the curve will have length 0 and which part 1?

why between 0 and 1 we have length 0 ? More precisely, how to determine which part of the curve will have length 0 and which part 1?

Mattias Fredriksson commented

2022-09-26 22:07:37 +02:00

In #101379#1423040, @mod_moder wrote:
why between 0 and 1 we have length 0 ?

More precisely, how to determine which part of the curve will have length 0 and which part 1?

Not sure exactly what you are referring to, the values reference a point on a curve and 0 is a perfectly valid point. Values in this case are also factors multiplied with the total length of the curve (in which case the interval [0.0, 1.0] makes sense). But I might miss your point as you didn't include a quote indicating what you replied to.

> In #101379#1423040, @mod_moder wrote: > why between 0 and 1 we have length 0 ? > > More precisely, how to determine which part of the curve will have length 0 and which part 1? Not sure exactly what you are referring to, the values reference a point on a curve and 0 is a perfectly valid point. Values in this case are also factors multiplied with the total length of the curve (in which case the interval [0.0, 1.0] makes sense). But I might miss your point as you didn't include a quote indicating what you replied to.

Iliya Katushenock commented

2022-09-26 22:10:19 +02:00

If the cyclic curve is divided by 2 points into 2 segments, from with equal positions of the points, the segments will be of length 1 and 0

Mattias Fredriksson commented

2022-09-26 22:38:43 +02:00

Ah, not sure if you consider or hint a solution for adding a node option to return the second part. Then one just has to decide which one is the first part (assuming the one of length 1 to avoid the OP problem). Hadn't thought of it that way.

Mattias Fredriksson self-assigned this 2022-10-05 23:19:43 +02:00

Pratik Borhade commented

2022-10-15 06:19:17 +02:00

Added subscriber: @Leroy-Xie

blender-admin commented

2022-11-09 17:50:51 +01:00

This issue was referenced by 11f6c65e61

This issue was referenced by 11f6c65e61a22ea9423962d6165f54ec29e221a8

Hans Goudey commented

2022-11-09 19:00:34 +01:00

Changed status from 'Confirmed' to: 'Resolved'

Hans Goudey closed this issue

2022-11-09 19:00:34 +01:00

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Regression: Trim curve make spline disapear. #101379