Sorry but being able to reproduce problem is mandatory. Can't you strip down scene that exhibits this issue as much as possible?

I solved the issue, I had to reinstall Intel Graphics driver and then reinstall Blender.

@Troy Sobotka (sobotka) I do understand, and totally get the argument, but for my two pennies worth-

I can't reproduce this issue.

What are steps to reproduce? do I have to render or do I see this in preview? Asking in advance because I get crashes in this file so I can't check now

This issue is on TODO list T69819, so I will close this. Thanks for reporting anyway.

Oops did not want to confirm yet.

In case anyone is curious about the difference the fix makes, here's a 0.1° sun at very high focal length:

I have a similar problem with T77977 (both in 2.82 and 2.83)

Thanks for the report. I guess I missed the build window with that fix. But it works fine now.

definitely looks like intentionally removed - see T75721

This is not bug, but rather limitation of current design. Reason you get nothing is selected is that selected vertex does not belong to active object.

I have noticed this in last editing session. Thanks for narrowing this down, I haven't made connection with deleting strips.

In T78605#974892, @Robert Guetzkow (rjg) wrote:

@Moustafa (mnhht) The working version also use software rendering and llvmpipe is generally not considered to be supported by Blender. Unless the GPU team change their mind and plan to work on llvmpipe support, graphics glitches can happen on unsupported setups.

As person unfamiliar with this feature I really can't tell what is it's intended funcionality and purpose from documentation, so I guess I will go ahead and conirm this.

I think this is because grabbing canvas will invalidate dynamic paint cache. Since this is working in 2.79 I will confirm, but this may be technical limitation of cache system so it may not be a bug.

Slight;;y the same issue here also. After windows 20.04 update it started to happen to me also, no matter if i use 2.83.1 or 2.90 on a small default cube scene viewport i get an OpenCL error

@Moustafa (mnhht) The previous versions also use software rendering and llvmpipe is generally not considered to be supported by Blender. Unless the GPU team changes their mind and plans to work on llvmpipe support, graphics glitches can happen on unsupported setups.

I think the problem is, if it needs to recompile the shaders after a user change to the node tree, then the viewport isnt refreshed until a value is changed, the viewport is moved, or the frame is changed.

In T77854#974826, @Andrey Sokolov (so_records) wrote:

Still having this issue In the Blender 2.83.1, Windows, branch: master, commit date: 2020-06-25 09:47. One more step to reproduce it is to animate the active camera: in this version Blender almost doesn't crash with VDB if nothing is animated. Also I've found out that it is possible to render without crashing from time to time if all 3 following checkboxes are checked in the Render Settings -> Performance: Use Spatial Splits, Save Buffers, Persistent Images. Unchecking any of them leads to crash. In the latest daily build of Blender 2.9 everything seems to work just fine, I can't reproduce this issue in it.

In T78605#974789, @Robert Guetzkow (rjg) wrote:

@Moustafa (mnhht) Could you please upload your system information (Help > Save System Info in Blender) from one of the versions that work properly.

@Richard Antalik (ISS) okay, got it!

1. Alpha represents degree of occlusion.
2. Emissions can happen independent of occlusion.

I imagine that the original code was not made to support this case.
But I will confirm for now.

Confirmed the same problem on my MacOSX

To me it seems to be a bug.
I will wait for a third opinion.

I've just compiled and everything is running nice.
Thank you Clément and Rahul

version: 2.90.0 Alpha, branch: master, commit date: 2020-07-07 16:23, hash: 4990e4dd01f2, type: Release
build date: 2020-07-07, 16:27:25
Mac mini OS X Catalina 10.15.5
Blender crashes when trying to open UV editing workspace.

In T78690#974846, @Jesse Y (deadpin) wrote:

This was probably fixed just after that build there -- looks like the build you tried was dated July 3rd. It was fixed later that day as part of rB4a48939f0

Can you try again with a more recent linux build.

No activity for more than a week. As per the tracker policy we assume the issue is gone and can be closed.