Re: Coroutines

[Matthias Felleisen <address@hidden>]

As I indicated, these things are very academic ideas that are probably
unintersting to most people on this list, but sine a number askd off the
list, too: 

     >    It's also true that you do NOT need continuations to model
     >    catch/call-cc. You can do it with a very different semantics with
     >    pleasing properties. Sorry.

     This sounds interesting -- could you please go into more detail or
     give some references?

R, Cartwright and M. Felleisen.  Extensible denotational language
specifications.  In _Theoretical Aspects of Computer Software (TACS)_,
Lecture Notes in Computer Science. Springer Verlag, Berlin, 1994, 244--272.
 % invited talk at TACS 

;; --- 

     >    working with me, some logicians did. They embedded a part of classical
     >    logic into constructive logic and *that* turned out to be like cps-ing=
     .=20

     I am vaguely aware of this, but a canonical reference would be greatly
     appreciated.

Here is the original reference on that: 

Griffin, T.  A formulae-as-types notion of control.  In _Proc. 17th
ACM Symposium on Principles of Programming Languages_, 1990, 47--58.

The French school has a whole bunch of other things to say on this topic
(now that they accepted Griffin's result).

-- Matthias