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

Re: function signature type checking



In article <200111270444.XAA04121@life.ai.mit.edu>, Kim Barrett 
<kab@camellia.org> wrote:

> limited(<function>, ...) seems like the right approach.  However, I
> think there are people who would dispute that "quite easily".  Several
> people at Apple and Harlequin (and maybe CMU too, I don't remember
> off hand) spent some time working on this, because it seemed like a
> serious hole in the Dylan type system.  However, it seemed that every
> attempt at a specification got arbitrarily hairy around variadic
> parameter lists and return types. I have memories of a whiteboard
> densely covered with small print purporting to describe all the
> possible permutations...

The compiler *already* has a representation for this stuff -- it's more 
a question that the programmer can't at the moment communicate known 
information to it.  It already works pretty well within a compilation 
unit.

I don't think there is any shame in a 90% solution that occasionally 
throws up its hands and says "too hard -- just use the slow calling 
sequence".  After all, that's what happens now *all* the time.

 
> If someone is really interested in pursuing this, I could probably be
> persuaded to dredge up any discussions I have archived.

It would be excellent to have that, at least for future reference!  
Anything else you have, too.  I'm sure we can find a good home for any 
amount of stuff at www.gwydiondylan.org.

-- Bruce



References: