on 3/30/01 3:35 AM, Alan Kay at Alan.Kay@disney.com wrote:
Most importantly, class Class was an instance of itself, and class Object was a superclass of itself. The "magic of microcode" was what allowed those two relations to be circular. The explicit metaclasses of Smalltalk-80 really don't do enough work to justify their complexity IMNSHO. But, wait until after this summer .....
What will come after this summer;) Nathanael plans to work on that? This would be really great.
By the way, metaclasses are implicit in Smalltalk-80 not explicit ;)
I have the impression that a lot of other approaches could be experimented like the idea of pluggable metaclass that would decouple different meta aspect from the class. (I do not have a paper on that just discussed with Gilad Bracha from Sun).
Cheers,
Alan
At 9:08 PM -0500 3/29/01, Jecel Assumpcao Jr wrote:
On Thursday 29 March 2001 18:42, Stephen Pair wrote:
ÝAre there any papers (or code) for Smalltalk-76 on the web? ÝI'm sure Ýthey've been announced before (and wasn't there an implementation of Ýit for Squeak?), but I've forgotten.
This one is great:
http://users.ipa.net/~dwighth/smalltalk/St76/Smalltalk76ProgrammingSystem.ht m>> l
but doesn't mention meta-classes. On page 34 of Alan Kay's "The Early History of Smalltalk" paper there is a drawing labeled "Smalltalk-76 Metaphysics". All classes were instances of the class Class object, which was a subclass of the class Object. Little Smalltalk initially had the same design (which makes all classes have the exact same set of class methods) but has moved to a more Smalltalk-80-like system in version 4.
There is a simulation in Squeak of Smalltalk-72, not 76.
-- Jecel