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

Re: CPS in Parrot




On Wednesday, August 6, 2003, at 02:43 PM, Michael Vanier wrote:

>
>> Date: Wed, 6 Aug 2003 13:51:17 -0400
>> From: Matthias Felleisen <matthias@ccs.neu.edu>
>>
>> On Wednesday, August 6, 2003, at 02:29 AM, Michael Sperber wrote:
>>
>>>>>>>> "Michael" == Michael Vanier <mvanier@cs.caltech.edu> writes:
>>>
>>> Michael> "And, just to top it off, someone (no, I don't know who, so
>>> this might be
>>> Michael>  CS Legend) once proved that you can build any control
>>> structure you can
>>> Michael>  think of with continuations."
>>>
>>> @INPROCEEDINGS{Filinski1994-popl,
>>>         CROSSREF = {POPL1994},
>>>         AUTHOR = {Andrzej Filinski},
>>>         TITLE = {Representing Monads},  YEAR = 1994,
>>>         PAGES = {446-457}
>>> }
>>>
>>
>> It depends on what you mean with "build". It is certainly not true 
>> that
>> FunctionalScheme + callcc
>> can express F (my good old friend that produces composable
>> continuations, which eventually
>> made it into Mach 4). I also don't see how FunctionalScheme + callcc
>> can express prompt.
>
> Pardon my ignorance, but what are F and prompt?  Do you mean System F
> (polymorphic typed lambda calculus)?  Also, what does Mach 4 refer to?
> I'm lost :-(

No, F is one of my control operators. It is like callcc except that it 
turns the
continuation into a true function (a thing that returns to the place 
where it
is called). You thus get "composable" continuations, but they really are
just functions. See POPL 1988.

>> I am also willing to throw in set! and challenge you again to define
>> "build" in a way that excludes
>> at least thing.
>
> At least thing?

at least one thing. [Sorry I deleted too much from the message during 
editing]
>>


>> But Mike's reference to Clinger's paper is appropriate.
>
> I can't find it on citeseer, though.  Does anyone know if it's 
> available
> online?
> =

You will have to go to the library and check out HOSC.

-- Matthias