mrjo118

Might it be related to this? (line 135/258): https://github.com/gilesknap/gphotos-sync/blob/main/src/gphotos_sync/GooglePhotosDownload.py#L135 I didn't check where settings.case_insensitive_fs is derived, but I am using Windows

Having a deeper dig in the code I can see that when indexing the media it will take 2 different routes in the code depending on whether you specify --album-regex...

The inconsistancies are here: https://github.com/gilesknap/gphotos-sync/blob/main/src/gphotos_sync/GooglePhotosIndex.py#L161 And here: https://github.com/gilesknap/gphotos-sync/blob/main/src/gphotos_sync/GoogleAlbumsSync.py#L96

The thing with Windows is the default behaviour is you're not allowed to have files with the same name but in different case (therefore it fails this check https://github.com/gilesknap/gphotos-sync/blob/main/src/gphotos_sync/Checks.py#L108), yet...

I've managed to find out which caused this issue originally when merging the tesselation PR (PR 2534) to master. It turns the commit immediately before it on master (PR 2740)...

Could someone explain what incrementing the value of ShaderCodeGenVersion in Ryujinx.Graphics.Gpu/Shader/ShaderCache.cs actually does? I have commercial experience as a software developer, but graphics rendering and emulation is foreign to me....

Hi @gdkchan, thank you for your detailed analysis. I have retested my hypothesis and discovered that after deleting my existing shader cache and retesting my 'working' build which did not...

So I found out I can actually view the original commit history and retrieve the full source at those points in time from GitHub by clicking the links on a...

So with some investigation and experimentation with reverting commits which update ShaderCodeGenVersion between the relevant window of commits, I have managed to identify the cause of the regression in the...

Maybe the issue is relating to the circled line having a different behaviour due to the change in PR 2702? Possibly this line needs to call a slightly different leg...