|How to change the Java VM to support functional programming? email@example.com (1997-05-08)|
|Re: How to change the Java VM to support functional programming? firstname.lastname@example.org (1997-05-08)|
|Re: How to change the Java VM to support functional programming? eanders@u98.CS.Berkeley.EDU (1997-05-08)|
|Re: How to change the Java VM to support functional programming? email@example.com (William D Clinger) (1997-05-12)|
|From:||William D Clinger <firstname.lastname@example.org>|
|Date:||12 May 1997 00:23:39 -0400|
Dwight VandenBerghe <email@example.com> wrote:
> I'm thinking that what is needed for
> continuations is something like....
> ...allocating blocks on the heap of what are essentially little
> stacks, and then transferring them to the machine stack when needed.
> Am I on the right track?
Yes. First-class continuations aren't part of Standard ML, though,
and I don't think they're part of Caml. (SML/NJ supports them as an
extension.) First-class continuations also don't have much to do with
first-class functions unless you convert to continuation-passing
style, and even then you probably will want to distinguish
continuations from functions. See
Clinger, Hartheimer, and Ost. Implementation strategies
for continuations. 1988 ACM Conference on Lisp and
Functional Programming, pages 124-131.
Hieb, Dybvig, and Bruggeman. Representing control in
the presence of first-class continuations. 1990 ACM
SIGPLAN '90 PLDI, pages 66-77. Available via
The first paper is harder to find, but I like it better. :-)
Return to the
Search the comp.compilers archives again.