VertexFormat:
| "float32"
| "float32x2"
| "float32x3"
| "float32x4"
| "uint8"
| "uint8x2"
| "uint8x4"
| "sint8"
| "sint8x2"
| "sint8x4"
| "unorm8"
| "unorm8x2"
| "unorm8x4"
| "snorm8"
| "snorm8x2"
| "snorm8x4"
| "uint16"
| "uint16x2"
| "uint16x4"
| "sint16"
| "sint16x2"
| "sint16x4"
| "unorm16"
| "unorm16x2"
| "unorm16x4"
| "snorm16"
| "snorm16x2"
| "snorm16x4"
| "uint32"
| "uint32x2"
| "uint32x3"
| "uint32x4"
| "sint32"
| "sint32x2"
| "sint32x3"
| "sint32x4"
A vertex attribute format: component type and count.
The set is deliberately narrower than the full GPU vocabulary — see the module notes for what is excluded and why. Layouts that cannot be spelled here (three-component 8- and 16-bit types, for instance) remain expressible with the Batcher#addAttribute GL-enum form, which is supported indefinitely.