Michael Livshin usenet@cmm.kakpryg.net writes:
never had any perceivable performance problems because of this. interpreted elisp is just Fast Enough.
It is usually just Fast Enough, until you write loops with macros like loop or SLIME's destructure-case macro. SLIME already byte-compiles the most performance critical functions at load-time, so byte-compiling the entire file doesn't by very much.
Helmut.