generic functions
Rainer Blome
rainer@physik3.gwdg.de
Fri, 21 Jul 1995 14:51:10 +0200
> > > And chaining single arguement dispatches is a mess.
> > no need to when "generic methods" exist.
by generic method i mean a method (belonging to some object) that behaves
differently for different argument types. in other words, a generic
function that can be invoked by sending the appropriate message to the
object that owns it. this degenerates to just a gf when it is owned by
some global namespace.
point being, methods may behave like the oldfashioned gfs, but usually
shouldn't. clos, dylan and stklos and so on have made gfs the default, EVERY
function is potentially generic. that
a) introduces a performance penalty, which, i admit, vanishes when a proper
compiler is used, and
b) is poison for modularity.
rainer