Path: utzoo!utgpu!water!watmath!clyde!rutgers!mailrus!nrl-cmf!cmcl2!brl-adm!brl-smoke!gwyn
From: gwyn@brl-smoke.ARPA (Doug Gwyn )
Newsgroups: comp.lang.c
Subject: Re: Ackermann's Function
Message-ID: <7219@brl-smoke.ARPA>
Date: 8 Feb 88 12:24:12 GMT
References: <4034@aw.sei.cmu.edu>
Reply-To: gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 16

In article <4034@aw.sei.cmu.edu> firth@sei.cmu.edu (Robert Firth) writes:
>The proper C translation is: ...

Actually, I liked John Gilmore's better!

>Arguing about a side effect that shouldn't even be there is a pretty
>pointless exercise, so I won't.

The issue is not tied to Ackermann's function, but is a more general
question about order of evaluation.  (My guess, since I don't have
the draft proposed standard at hand, is that Stallman is right --
it is not until the inner A() is evaluated that issues of interleaving
of ITS arguments are relevant; whether or not the outer A()'s other
argument is evaluated before or after this is implementation-dependent.
I do recall that the wording about sequence points changed somewhat
since the old draft, so it may be clearer in the current one.)