Skip to content

Conversation

@Pessimistress
Copy link
Collaborator

For #9504

Change List

  • Attribute.getBufferLayout() skip sub-attributes with invalid vertex formats
  • IconLayer merge some attributes into an interleaved buffer (bindings exceeds maxVertexBuffers)
  • Plumbing

@Pessimistress Pessimistress requested a review from ibgreen October 11, 2025 04:21
@Pessimistress
Copy link
Collaborator Author

Not ready to merge - TextLayer is broken because it is not packing the correct instanceOffsets.

Copy link
Collaborator

@ibgreen ibgreen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  • I see you are running into the limited number of vertex buffers - it is quite inconvenient, stricter than WebGL.
  • 8 vertex buffers + 10 storage buffers seems to be the limit even on recent macbook https://webgpureport.org/
  • We can of course do something on the CPU side to interleave, but it really gets in the way of a simple GPU-buffer-based layer API.

@coveralls
Copy link

coveralls commented Dec 17, 2025

Coverage Status

coverage: 91.072% (+0.02%) from 91.055%
when pulling 90e2f8b on x/icon-layer-webgpu
into b0134f0 on master.

@Pessimistress Pessimistress marked this pull request as ready for review December 17, 2025 03:28
Copy link
Collaborator

@ibgreen ibgreen left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice work!

Nit: At some point I would like to change shaderAttributes to interleavedAttributes or something and maybe surface it a bit more in docs. That name always throw me off.

@Pessimistress Pessimistress merged commit a15c8ce into master Dec 19, 2025
5 checks passed
@Pessimistress Pessimistress deleted the x/icon-layer-webgpu branch December 19, 2025 02:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants