Single page description of Smalltalk
dwighth at ipa.net
Tue Mar 7 05:59:37 UTC 2000
Alan Kay wrote:
> Of course, one is tempted to simply write the page (Dan can do this better
> than anyone).
> The only syntax that counts is
> <receiver> <message>
> and this could be expanded a little to deal with unaries, binaries and
> keywords. The semantics is all about "biology": sending, receiving, and
> remembering with a little about inheritance thrown in.
> Tempting as a challenge?
Yes, it's very tempting. Should we try to capture the little
idiosyncrasies in Squeak (relative to an ideal, not other Smalltalks) --
like the constraints on how "-" and "|" can be used in a binary
selector, and the way blocks won't actually allow no arguments and a
single bar - [ | 2 + 3 ] is not allowed but [ | | 2 + 3 ] is -- or
should we try for the cleanest, most general description?
(And yes, I've already started on it.)
More information about the Squeak-dev