If we do have a strong reason, we should state it in the spec.I don't know. It doesn't make sense to my why this would be helpful, particularly with establishing interleaved attribs.If we don't have a strong reason, this requirement should be removed.On Thu, Mar 26, 2015 at 7:46 AM, Florian Bösch <pyalot@gmail.com> wrote:I was wondering if there's a reason that WebGL imposes multiple of type size requirements on offset and stride arguments to vertexAttribPointer in respect to the passed type argument.The invalid operation error is not defined for that case for either ES 2, 3 or any version of desktop GL.