Page MenuHome

Cloth physics cache gets reset by changing object materials
Confirmed, LowPublicKNOWN ISSUE


System Information
Operating system: Darwin-19.4.0-x86_64-i386-64bit 64 Bits
Graphics card: AMD Radeon Pro 5500M OpenGL Engine ATI Technologies Inc. 4.1 ATI-3.8.24

Blender Version
Broken: version: 2.83 (sub 11), branch: master, commit date: 2020-03-26 23:49, hash: rBf1fb3eb97540
Worked: unknown

Short description of error
When using a cloth physics simulation, if you add or remove a material from an object, the cloth system cache is reset ("cache is outdated"). I don't think the material affects the cloth physics in any way (as it seems you can change any parameter within the material without the cache resetting), and therefore this should probably not happen. Please correct me if I am wrong and the material does affect the physics somehow.

Exact steps for others to reproduce the error
In the default scene, enable cloth physics for the cube. Let the playback run for a few frames, building up the cache. Add a material to the cube; now the cache gets reset.

Event Timeline

It seems there are also quite a few other options that do this. For example changing the viewport display options of the object, or even just changing whether it is selectable show this behaviour.

Richard Antalik (ISS) changed the task status from Needs Triage to Confirmed.Apr 1 2020, 12:58 PM
Jacques Lucke (JacquesLucke) triaged this task as Low priority.May 19 2020, 3:42 PM
Jacques Lucke (JacquesLucke) changed the subtype of this task from "Report" to "Bug".

It seems the cache is triggered by a call to WM_event_add_notifier(C, NC_OBJECT, ob);, so it might be a more general issue indeed.
As long as it does not invalidate the cache once you have baked it, I do not consider this bug to be very important, even though it can be annoying sometimes.

Maybe this should be a known issue, not sure.

Jacques Lucke (JacquesLucke) changed the subtype of this task from "Bug" to "Known Issue".Mon, Sep 14, 2:51 PM

It's unlikely that this will be worked on soon, so I'll reclassify this as known issue.