Ben Clayton

> Should we give a harder look (relook?) at `.length` and deprecate `arrayLength` for runtime-arrays? Then it would be a consistent `foo.length` call in both cases. I would be happy...

I'm not strongly opinionated on whether this should be accepted or not, but if it is, `var` is a far better name than introducing another keyword to the language.

+1. I've also found myself reaching for this. This is supported in HLSL.

@stegu - would the addition of a new `modulo` builtin, while preserving the existing `modf` alleviate your concerns?

Chrome is branching for WebGPU V1 in a matter of days, so changing the definition of `modf` at this point (whether it is poorly named or not) would cause a...

> The thing that I'd be more worried about the idea of calling an inverse in a shader is whether this is really a useful way of using your computing...

> Option 3: `identity4x4f32()` Given the existing [single letter suffixes for type aliases](https://www.w3.org/TR/WGSL/#matrix-types), wouldn't it me more consistent to be `identity4x4f()` ? I'll throw another (likely controversial) proposal into the...

The step to fetch the dependency is listed in the root README: https://github.com/google/shaderc?tab=readme-ov-file#getting-and-building-shaderc --- 1. Check out the source code: ``` git clone https://github.com/google/shaderc $SOURCE_DIR cd $SOURCE_DIR ./utils/git-sync-deps ``` ---...

You can download `clang-format` as part of the LLVM binaries at https://github.com/llvm/llvm-project/releases/tag/llvmorg-17.0.1

There are many validation tests that do this sort of pattern: ```ts t.expectCompileResult( t.params.arg === 'good' || t.params.arg === 'good_vec2_abstract_float', `const c = ${kFn}${kArgCases[t.params.arg]};` ); ``` Given that implementations may...