mwyrzykowski

Hi @Makio64 - I see this on iOS but with ToT WebKit so understandably that would be hard to verify without compiling WebKit as source into a 3rd party application:...

> The OOB behavior in the spec is currently per attribute and doesn't allow all attributes from the buffer to be considered OOB. My understanding from the specification as it...

@Makio64 it seems fine on ToT WebKit (I tried an iPhone 15 Pro): https://github.com/user-attachments/assets/a0e8735f-e7d1-4f9c-970c-8cdf67256907 Understandably it would be hard to verify, but if it looks correct on STP, I think...

This should be fixed by https://bugs.webkit.org/show_bug.cgi?id=277465 That fix did not make it into iOS 18.0 or 18.1, which is why you see the behavior on iOS but not Safari Technology...

Which version of STP @Makio64 ? I tried on an M2 Mac Studio with STP 204 and seems ok: https://github.com/user-attachments/assets/edb3cf31-1a0f-4dda-a44e-1ff9680fc745

I will attempt to reproduce on an M3 chip, maybe the flicker is due to a race condition

> I also don't love that, and am still a bit concerned that we need to commit to the granularity on a relatively short deadline. I'll put in a plug...

@SenorBlanco @kainino0x I suppose it is really necessary we know all the feature and limit names right now? We do have `readonly attribute DOMstring featureLevel;` being added to the GPUAdapter...

> Is an implementation is required to split command buffers? Implementation is not required to split command buffers, on Apple Silicon Macs and iOS devices WebKit never splits command buffers....

Following the WebGPU approach seems reasonable. Specifically, supporting the `viewFormats` approach would be great for srgb, as the native format for some devices is `bgra8unorm-srgb`: >NOTE: Canvas configuration cannot use...