Hrvoje> When using a subset of a standardized language, you can still write Hrvoje> conformant programs. Sure, but it still means that those programs will have to be written consist of adding the missing features. Hrvoje> The reason to use the subset instead of the Hrvoje> unabridged experience is, of course, efficiency. Common Lisp implementations exist which produce code just about as efficient as anything else on the planet, even if they implement a large subset of the language. I think you're referring to memory footprint which indeed weighs in heavy, because the CL committee forgot to organize the language into layers and/or packages, and you basically always have to have the whole thing in-core. -- Cheers =8-} Chipsy Friede, Völkerverständigung und überhaupt blabla