Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!rutgers!cmcl2!kramden.acf.nyu.edu!brnstnd
From: brnstnd@kramden.acf.nyu.edu (Dan Bernstein)
Newsgroups: comp.lang.misc
Subject: Re: On whether C has first-class composable functions
Message-ID: <12547:Feb621:05:4491@kramden.acf.nyu.edu>
Date: 6 Feb 91 21:05:44 GMT
References: <1991Feb5.144815.23239@spool.cs.wisc.edu> <1234:Feb520:27:2391@kramden.acf.nyu.edu> <1991Feb6.161639.18311@spool.cs.wisc.edu>
Organization: IR
Lines: 36

In article <1991Feb6.161639.18311@spool.cs.wisc.edu> quale@picard.cs.wisc.edu (Douglas E. Quale) writes:
> In article <1234:Feb520:27:2391@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
> >As I said at the beginning of this discussion, you're wrong. The very
> >first implementation posted (mine) dynamically allocated the composed,
> >first-class functions.
> I'm quite certain that you are completely, absolutely and undisputedly wrong.
> Functions cannot be dynamically allocated in C.
> C does not have first-class functions.

You are making Jim Giles' mistake of confusing syntax and semantics.
Just because the C standard calls certain language objects ``functions''
doesn't mean that it's impossible to implement functions that don't have
the same restrictions as those objects. As a matter of fact, it's not
only possible, it's trivial.

The things that my implementation produced were not C functions, but
they were full-fledged functions in all respects.

We've hashed all this out before. (Everyone agrees with the above points
when I say ``implement''; the problem before was that some people didn't
like how I said ``C has first-class composable functions'' to mean
``first-class functions can be implemented in C.'') May I suggest that
you review the beginning of a discussion before jumping into the end of
it and throwing out ``you're wrong, wrong, wrong'' left and right?

May I also suggest that you learn to distinguish the syntax of a
language feature from its semantics?

> I have serious doubts about whether you would know what to do with
> dynamically allocated functions if you had them.

I've written a noticeable amount of Forth code, and I once found it
useful to implement dynamically allocated functions (in Pascal) for an
intepreter. Other than that I've never seen much use for the things.

---Dan