Great to see you all again. Here is the article I mentioned:
http://gmarceau.qc.ca/blog/2009/05/speed-size-and-dependability-of.html
-mark
A couple more links. For those interested in the original SICP lectures: http://groups.csail.mit.edu/mac/classes/6.001/abelson-sussman-lectures/
... and for Doug, there's some talk about "Continuations, closures, etc." on the LispWorks forum (which appears to be down at the moment): http://news.gmane.org/gmane.lisp.lispworks.general/
There's mention there of DLAMBDA:
"I would posit that Doug Hoyte's DLAMBDA and the use of it to produce little state machines embodies much of what I was aiming for. Rather than standing on our heads trying to make Lisp perform stack tricks to simulate a CALL/CC -- which we almost already have with CATCH/THROW, how about translating continuation directed computations toward something akin to DLAMBDA state machines -- which BTW are also macrology magic. And these state machines may be more efficient than attempts at duplicating continuations with indefinite extent."
BC
Brian - thanks for the link. Perhaps not surprisingly, I agree with the author of that quote. I don't think continuations are necessary for most use cases and instead open up whole new classes of problems. State machines are usually the way to go in my opinion.
Good meeting you all (again in some cases) and thanks for the interesting discussions.
Best,
Doug
On Wed, Jun 03, 2009 at 09:55:54AM -0400 or thereabouts, Brian Connoy wrote:
A couple more links. For those interested in the original SICP lectures: http://groups.csail.mit.edu/mac/classes/6.001/abelson-sussman-lectures/
... and for Doug, there's some talk about "Continuations, closures, etc." on the LispWorks forum (which appears to be down at the moment): http://news.gmane.org/gmane.lisp.lispworks.general/
There's mention there of DLAMBDA:
"I would posit that Doug Hoyte's DLAMBDA and the use of it to produce little state machines embodies much of what I was aiming for. Rather than standing on our heads trying to make Lisp perform stack tricks to simulate a CALL/CC -- which we almost already have with CATCH/THROW, how about translating continuation directed computations toward something akin to DLAMBDA state machines -- which BTW are also macrology magic. And these state machines may be more efficient than attempts at duplicating continuations with indefinite extent."
BC
toronto-lisp mailing list toronto-lisp@common-lisp.net http://common-lisp.net/cgi-bin/mailman/listinfo/toronto-lisp
-
Brian Connoy -
doug＠hcsw.org -
Mark Bolusmjak