Good catch. Confirmed. Missing undo push when using the color picker?

I realize that it might be intentional and is not a deal-breaker.
Using duplicating props in the operator whose values are then set on execute to the object's prop seems to work fine.

@Salvatore De Paolis (iwkse) issues with PythonLibsUnix should be reported with details of why it fails exactly. See:

My CPU is an AMD A6-3400M with Radeon HD Graphics.

Thanks for the update Phillip.
I've posted a change proposal in Right Click Select.
https://blender.community/c/rightclickselect/g6cbbc/meta-object-families-refactor

I've been going over the output of make format, disabled formatting for a handful of areas where it's not helpful (mostly large definitions which read much better when hand aligned). See: D4185.

Sorry, but you are not finished with this bug hahaha

I had to toggle the viewport render mode from rendered to solid and back once (or change to shader values) to get this to trigger for the first time. After that I consistently got a black cube. It didn't matter what metallic was set to. And I also got black cube with the other shader node too.

Unless the command save file before crash it will not help.

Steps are much too vague.

Hi Clément, Sebastian and Philipp,

That is really strange. Try updating your graphic drivers first.

Well we could try to improve frustum plane extraction but that's not planned currently. So I would close the report.

You can make the object appear diffuse in the reflections if you want, by using the light path node and a mix shader.

On Windows Blender doesn't create crash report when a segfault occurs. You can start it from command line though and add the "--debug" flag, this will keep the console open with debug prints. More specific flags are available if detailed information for e.g. GPU rendering is required.

The bug is that Weight Tools do not work as expected, possibly because <bmesh>.verts.layers.deform.verify() is needed to fix it

I get an assert opening the file in a debug build.
This has to do with some armature dependency cycles and bendy bones (but I cannot find any armature in the file).
After simplifying the file (leaving only the 2 meshes) I get no error and no shading errors (I get no shading errors even before).

First I tried:
rpbcacueaiibh@Minisys:~/Blender 2.80.41$ ./blender --debug-gpu-force-workarounds
and I was getting a lot of X11 errors with the blender errors in the terminal so I thought I save the output into a file like this:
rpbcacueaiibh@Minisys:~/Blender 2.80.41$ ./blender --debug-gpu-force-workarounds > ~/Desktop/Output
but the file only contains the blender errors like so:

In T60526#600174, @Jacques Lucke (JacquesLucke) wrote:

I can't reproduce it. Please upload a .blend file that shows the problem.

Hi Brecht,

Cant the tabs get a seperate area they already seem sort of separate because ui is rotated there. Im not sure how this part work in GUI

I guess this is caused by accidentally zooming the buttons in the sidebar, as you reported in T60540: Middle Click navigation in 2.8 UV Editor randomly stops and starts working?

Yes. I was thinking of making the region dynamically sized based on the contents. But unfortunately it's more complicated than that because the tabs are part of the same region, so we have a kind of non-rectangular region.

So, I guess the solution is for the input to fall back to the underlying view when there is no panel under the cursor?