Matrix operations under BufferData

[Checkmate50]

The following code snippet is a bit annoying, in that I need to declare an identity matrix before assigning it to two separate pieces of bufferData.

let imat = generate_identity_matrix();
let model_mat = BufferData::new(imat); 
let model_mat2 = BufferData::new(translate(imat, 2.0, 0.0, 0.0));

Perhaps it's an unavoidable part of the current BufferData paradigm, but it would be nice if I didn't need "in-between" matrices for each operation. Also, how does this work with a loop? Would I need to make a new piece of BufferData for each iteration of the loop?

[Checkmate50]

Concretely, this can be seen in the multicube example. What I would what instead is something like

let model_mat = BufferData::new(generate_identity_matrix()); 
let model_mat2 = BufferData::new(translate(model_mat2, 2.0, 0.0, 0.0));

[Pat-Lafon]

I'm assuming you meant to use model_mat to create model_mat2. Since BufferData::new is just a wrapper around the matrix and doesn't do anything with it, it seems pretty safe to just make it's data field public as I've done in 7a0125095916afa513f7b2acb0a4939303b2469f . This would allow you to write

let model_mat = BufferData::new(generate_identity_matrix()); 
let model_mat2 = BufferData::new(translate(model_mat.data, 2.0, 0.0, 0.0));

There's probably an alternative where I implement From/Into for BufferData back into the inner data, something like

let model_mat = BufferData::new(generate_identity_matrix()); 
let model_mat2 = BufferData::new(translate(model_mat.into(), 2.0, 0.0, 0.0));

but I think just being able to access the field is fine.

[Pat-Lafon]

Closed by #21