Quote:
Originally posted by Quadraton
Yes indeed, it is a swapping algorithm that doesn't require a tertiary (temporary) variable. I believe it only works on integers (i.e. not floating points), and it should work for any combination of numbers (both positive and negative). As an additional benefit, the compiler can translate it into 3 simple XOR assembler statements, making it much more efficient to execute.
|
Well, it "only works" on integers 'cause the compiler won't let you do anything else. This method can be used on anything that has a binary representation. On a computer, that's everything. C let's you do anything you want (it's popularity is probably largely due to this), so you can simply cast your floats and do your thing...