I just ran into something interesting. It was John Foust's review
(semi-review) of Aztec C, in the June Amazing Computing. One of the things
he said about the smaller code size was that you could with Lattice get the
code about as small as Aztec, it was just that Aztec did it automaticly.
(oops, it wasn't june, it was May, sorry) He did some things, like linking
with Amiga.lib before lc.lib, and replacing standard c calls with AmigaDOS
calls. It seems that most of the optimizing that Aztec does is simple
trimming overhead off of the standard c calls, rather than improving the
assembler itself (Which I'm sure it does some of too) He didn't do speed
comparasons for running code, but Aztec did compile faster. You might want
to read the article. Its the C tutorial part four, not a separite review.
Later, <MOWGLI>