Dzmitry Malyshau

See also - https://github.com/gpuweb/gpuweb/pull/1121#discussion_r497710116 from a year ago :)

Thank you for the investigation! Push constants look like an easy feature to have for us, since all the backends can efficiently support it. I don't see much controversy of...

@Kangz why 128 *bits*? @msiglreith suggested 128 bytes, and it seems reasonable, given the example Root Signature layout they provided, and also the fact `maxPushConstantsSize >= 128` on Vulkan. If...

What's conceptually a problem with the user using push constants for matrices if they need to be updated every draw call and they fit within limits? > Also leaves space...

@kainino0x that position is understandable. But how do you come up with a "very small" metric that is safe w.r.t. our future requirements?

Overall, we have the following options: 1. expose /do not expose / hint 2. soft limit hard-coded in API / hard limit in API (e.g. 128 bytes) / hard limit...

@litherum thank you for performing the benchmarks! Can you make sense of the data? It's hard to see the difference from the graphs, although there does appear to be one....

You've almost lost me at the first sentence :) > Generating mipmaps is something that most 3D applications want to do. Most 3D applications that are performance-sensitive wouldn't want to...

@deltakosh interesting that you mention this, because Three.js and Babylon.js are in a good position to implement it once and for all in their own codebase, using the most efficient...

@deltakosh is your use of WebGL justified by the inability to generate mipmaps with WebGPU directly? I mean, just having a shader that samples an image, and then running it...