It's actually a comparison of Smalltalk and Scheme, not Smalltalk and Lisp.
What's that a picture of?
Posted by: Kevin Kim on July 7, 2005 10:49 AMweird. I just googled for that yesterday, and now you post it today ...
Are you aware of how lame it is to force me to enter an email address?
Posted by: oddly on July 7, 2005 05:26 PMPascal, Lisp is a family of languages. Scheme is a Lisp dialect. Really.
oddly: It would be totally lame if it were true. But as far as I can tell I have "allow anonymous comments" enabled.
Posted by: John Wiseman on July 8, 2005 11:45 AMConcerning free smalltalk implementations, there is GNU Smalltalk, which does not look too appealing, because it lacks the visual component, which some consider as the most important thing in smalltalk.
Then there is squeak. This is an interactive graphical system, "just as smalltalk was supposed to be". I played around with an earlier version of it, and one of the first things I tried to do was to define a factorial function using only anonymous functions and the fixed-point principle.
Unfortunately - I do not know whether this is fixed in the standard distribution by now, and it seems as if there were a special patch for quite some time - at that time, the system had a badly broken implementation of closures. Effectively, every "function object" just had one single activation context tied to it, so if an anonymous function calls itself, things may work fine if you are exceptionally lucky (i.e. your function has more nice properties than what you tell the system). But normally, everything just goes boom.
Ugh. First thought that came to my mind was: "is this what you get when you let people that like to think in terms of objects implement an evaluator?"
Well...
(BTW: It seems as if GNU Smalltalk always did this the proper way.)