Andre Weissflog

I think it would require pretty deep changes in the public APIs (basically adding the concept of a 'computer-shader-stage', and on the buffer size a new buffer type 'storage-buffer').

...as a first step, readonly storage buffer support is now implemented, and compute shader support is planned for the "not too distant future".

Hmm, interesting. The CI test builds also compile with -Werror, my guess is that this problem only exists on some Linux distros. It definitely makes sense to add the define...

Ok thanks for the investigation, mentioning c99 vs gnu99 somehow rings a bell, I might have stumbled over the issue a long time ago, maybe in the context of Oryol,...

> Watch out! You have to define this before you include any of the libc headers! Hmm ok, the declaration section of the sokol headers are supposed to only include...

I prefer the _GNU_SOURCE version. Do you want to provide a PR for sokol_app.h and sokol_time.h which adds those blocks? (PS: not sure what happens on macOS, technically it's a...

Thanks for the headsup. I'll need to set aside a bit of time to play around and get familiar with your fork and my year-end vacation is coming up. It'll...

So from my side, I abandonded the old multiwindow branch I had started because it became too gnarly (and also required too many changes in sokol_gfx.h). sokol_gfx.h at least will...

I'll need to check what the current behaviour is. IIRC the problem might be that DXGI holds onto (discarded) swapchain surfaces for a couple of frames which then causes massive...

> Unless there's another way to achieve this perhaps? Not at the moment. The reason why you see a difference is the `_sapp_win32_update_dimensions()` call, this updates the values returned by...