[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Coroutines




I remember reading that Strachey had an epiphany when developing
Denotational Semantics--namely that continuations were the right way
to model control flow.  Prior to having that insight, he had been
unable to come up with a suitable semantics for the goto statement.

If this is true, maybe he got the idea from Church?

++ Ken

Matthias Felleisen writes:
 > 
 > Correct. 
 > 
 > Scheme had first-class continuations from the very beginning. When I use
 > "call/cc", that's what I mean. Guy's Scheme had catch, which was a binding
 > construct. 
 > 
 > Clinger, Friedman, Wand introduced call/cc so that they didn't have to cope
 > with another binding construct around 1978. That trick was introduced by
 > Church in 1948; they just applied it to catch.
 > 
 > J is basically like call/cc and catch. I don't know whether Guy and Gerry
 > knew about this or whether their discovery was independent. Guy? 
 > 
 > Reynolds had first-class labels in Gedanken. He had escape in a 1972 paper,
 > which is Guy's catch. 
 > 
 > -- Matthias