(Previous discussion continued)
Re: continuations and threads Matthias Felleisen 20 Mar 2000 00:38 UTC

Re: continuations and threads Matthias Felleisen 20 Mar 2000 00:38 UTC

In the past I considered all of your uses of call/cc as legit. I guess Mike
 hit the nail on the head when he said "perhaps your call/cc is just too
 slow." But at the same time, I have to say that I considered exceptions
 and pure threading the true domain of call/cc and since exceptions and
 threads are a part of MzScheme, I guess it takes away the motivation for
 many call/cc's.

;; ---

The monograph is based on my dissertation from 1986, and Dan and I worked
out the presentation in "Art of" thru a dialog.

Yes, the semantics in there (and used elsewhere) shows how global call/cc
is. I tried to tame it with proposals for restricted continuation grabs,
but the proposal never caught on. Appel actually put them into SML/NJ, but
until Gunter et alii came along and revised it nobody cared. I am not sure
anybody would care anyway.

Using prompt and control, many programs were easier to write than with
call/cc and set! and easier to understand, too. Why people didn't pick up
on it can only be explained with premature standardization on Scheme's
side.

;; ---

Thanks for the feedback to everyone.

-- Matthias