Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!unmvax!ncar!tank!mimsy!cvl!haven!adm!cmcl2!lanl!jlg
From: jlg@lanl.gov (Jim Giles)
Newsgroups: comp.lang.misc
Subject: Other syntax issues (was: language commenting constructs)
Message-ID: <10582@lanl.gov>
Date: 17 Mar 89 01:53:12 GMT
References: <3452@ficc.uu.net>
Organization: Los Alamos National Laboratory
Lines: 25

From article <3452@ficc.uu.net>, by peter@ficc.uu.net (Peter da Silva):
> In article <39273@oliveb.olivetti.com>, chase@Ozona.orc.olivetti.com (David Chase) writes:
>> *  :=/= instead of =/== 
> 
> But this definitely was.
> 

I never liked ':=' for assignment.  It has been claimed that it was selected
because it _looked_like_ a left pointing arrow.  It doesn't, of course, look
anything like a left pointing arrow.  If I'm to be forced to use two char-
acters for the assignment operator (and one of them shifted no less) I 
would rather have '<-'.  This, at least, really does look like a left arrow.

Frankly, I prefer =/== to :=/=.  ':=' was only chosen in the early days
because some hardware didn't have '<' in the character set.

> Take your pick:
> 
> 	min2(a,b) { return (a < b) ? a : b; }
> 	#define min2(a,b) (((a) < (b)) ? (a) : (b))

Obviously, I pick the first one since the second is not even correct!
If either of the arguments to the macro have side-effects, the macro
may return the wrong answer - and it will certainly perform at least
one of the side-effects too many times.