Matrix4x4 * AbstractVector{Vector4} about oldimmutablearrays.jl

[twadleigh]

Hmm. I'd be happy to entertain a pull request, but I'm not sure this really belongs in the library. You can accomplish this with an explicit loop or comprehension (or map, as you note), in which case it seems superfluous to add to the library (which is mostly about code generation), or you can convert it to generic matrix multiplication, in which case it wouldn't really belong in this library, either.

[babaq]

Your argument makes a lot of sense. the proposed functionality actually means to transform or multiply each of Vector4 in an array with a Matrix4x4. Adding to * function may pollute the meaning of matrix multiplication, but it does fit into .* which is simpler than the equivalent for loop or map. I was just thinking from a user perspective hoping those things automagically work:)

one can always convert Vector{Vector4} to generic matrix, that make me thinking is it right to support operation between ImmutableArrays and Arrays

tm=Matrix4x4(1.0)
a=Vector4(2.0)
tm*a # result in Vector4
[tm]*[a] # result in 4-element Vector
tm*[a] # doesn't work
[tm]*a # doesn't work

i am skeptical of any complication of support this, maybe any operation between ImmutableArrays and generic Arrays should return generic Arrays.

[twadleigh]

The "it should just work" argument does make sense. And, yes, I'd agree that any mixing of ImmutableArray and other arrays should make vanilla arrays.

Could you spec out some definitions for the additional functionality you'd like to see in a PR or gist?

Also, just a heads-up, but it's been a looong time since I've had my head in this code, so it might take me a while to spin up on thinking about the best way to integrate it.

Thanks for making the effort to contribute.

[babaq]

Sure, i'll try to make it happen:)