Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!yale!cmcl2!kramden.acf.nyu.edu!brnstnd
From: brnstnd@kramden.acf.nyu.edu (Dan Bernstein)
Newsgroups: comp.lang.misc
Subject: Re: How to make a language downward-extensible?
Message-ID: <975:Oct800:52:3590@kramden.acf.nyu.edu>
Date: 8 Oct 90 00:52:35 GMT
References: <7950@scolex.sco.COM> <18718:Oct120:03:0090@kramden.acf.nyu.edu> <7976@scolex.sco.COM>
Organization: IR
Lines: 18

In article <7976@scolex.sco.COM> seanf (Sean Fagan) writes:
> You want things such that someone can describe an algorithm, in more or less
> generic terms, as well as a "better" replacement algorithm, and have the
> compiler automagically replace it?

Right. But both algorithms will be fully generic, and fully portable.
The compiler will make its best guess, or take the first choice if it
has no idea.

> I am not certain of this, but I think there has been some success in that
> area.  If this *is* what you mean, then comp.compilers would be a good place
> to (briefly) continue the discussion, since that's where those types hang
> out.

Well, it is a language issue, insofar as the language has to provide
quickpick (or something like it) in the first place.

---Dan