Thread with 3 posts
jump to expanded post
I like the way Apple APIs name things… for example, what GLSL would call a mat2x3, Core Graphics calls a CGAffineTransform. no thinking about whether that’s row-major or column-major, and it tells you why you want those numbers (most efficient representation of an affine transform)
meanwhile I use a 3-by-3 for affine transforms because of uhh… uhhhh… starts sweating
…
(the real reason is that it’s hard to remember which parameter of matrix multiplication is the column count, and which is the row count, so I make them the same value to shrink the space of possible wrong implementations,,,)
(can you believe that, in my most recent job, me knowing about matrix maths was absolutely critical? me neither. I got hired despite fumbling an interview question about it, hehe)