1 and 5, feel free to do. They are quickies that I just never remembered to do. 4… well, I’m going to just have to push on Sun again.
1 is done. I have a question about 5 though.
The standard behavior for jME’s math classes is that
Vector3f a;
a.multiply(b);
returns a "new" vector that is a * b just like
Matrix3f a;
a.multiply(b);
returns a "new" matrix that is a*b.
If I want to keep this pattern, I would need to change the behavior of some functions. For example,
Matrix3f a;
a.add(5);
Acording to standard behavior for most jME math functions, this shouldn't change matrix a at all. It should simply return a new matrix3f that is a+5, and if I wanted to change a i would use
a.addLocal(5);
The problem is that a.add(5) actually changes a and returns void. To 'standardize' usage of math classes I'ld have to change it's behavior.
agreed.
This is an interesting thread on the topoc of shader/material systems. The good stuff is by Yann L
http://www.gamedev.net/community/forums/topic.asp?topic_id=169710
Not too sure how this would sit with the current rendering system, but shader-centric engines seem to be the way every thing is going.
http://triplebuffer.devmaster.net/misc/yannl.php covers loads of other stuff that is worth looking at.