double vs. float

OpenGL Development JOGL Development
#1

I’ve been working away through the various NeHe JOGL ports and noticed that a number of times the vertex variable types were ported as float from c/c++ to float in Java. Nothing seem odd…? Reading an earlier post it was said that LWJGL hides the OGL double commands.

But to write a float we need to either write 1.0F or cast it where as doubles need no special treatment. Or does this increase the computation needed? ???

Thanks.

#2

With current computers, floats are faster than doubles… Floats are 32 bits digits whereas Doubles are 64 bits one’s.

If you’re using a 64bits processor, you won’t see any speed difference. Else, use float. It’s alwais better.

Chman