LOGO-L> Re: ToonTalk versus MicroWorlds

[bh@anarres.CS.Berkeley.EDU (Brian Harvey)]

"Ken Kahn" <KenKahn@ToonTalk.com> writes:
>I woke up this morning thinking I should implement a tiny Lisp interpreter
>in ToonTalk to illustrate its universality. A friend suggested Logo,
>instead. My guess is that the implementing a parser for Logo is more work
>than implementing "eval" and "apply" in Lisp. Is that right?

Absolutely.  Berkeley Logo's evaluator started out as one of the Scheme
interpreters in SICP (the explicit control evaluator in chapter 5) and
then we had to double the size of eval as well as adding a parser!
I suppose it would be less of a job if you didn't care about tail call
elimination, but even then, the tokenization is tricky, and you have to
re-parse procedures if some procedure is redefined to take a different
number of arguments.



---------------------------------------------------------------
Please post messages to the Logo forum to logo-l@gsn.org.  Mail
questions about the list administration to logofdn@gsn.org.  To
unsubscribe send    unsubscribe logo-l    to majordomo@gsn.org.