Yes, likely this test just needs to be skipped because of a Mac bug. The above CL added this new test.

Note that we don't hope to pass WebGL 2.0 conformance on NVIDIA GPUs on macOS; there are far too many failures. It's likely there is a driver bug in handling of uniform buffers. Mo, did you add a suppression for a WebGL conformance test failure yesterday in this area?

Jamie, if it's not too hard for you to add a driver bug workaround that continues to dirty all the uniform buffer state on macOS with NVIDIA GPUs, that might be the best way to get this to pass.

Well, I think this is an order of operations thing. We should definitely add the workaround to ship MANGLE on OSX, but I don't think it's a high priority to just get a new test to pass when there are already likely existing issues filed for this. Is that ok to defer this work to someone who is going to be working on the OSX port?

Yes, that sounds ok. I thought that perhaps adding the workaround would be simpler.