Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!pwa-b!mmintl!franka
From: franka@mmintl.UUCP (Frank Adams)
Newsgroups: net.lang.c
Subject: = and :=
Message-ID: <1568@mmintl.UUCP>
Date: Mon, 23-Jun-86 00:44:21 EDT
Article-I.D.: mmintl.1568
Posted: Mon Jun 23 00:44:21 1986
Date-Received: Thu, 19-Jun-86 03:38:18 EDT
References: <5565@alice.UUCP> <8200005@ztivax.UUCP>
Reply-To: franka@mmintl.UUCP (Frank Adams)
Organization: Multimate International, E. Hartford, CT
Lines: 20


The more I think about it, the more I like the idea of having "=" be an
assignment which does not return a value, and ":=" be an assignment which
returns its left side after the assignment (i.e., the current C definition
of "=").  I am not sure whether I like this well enough to actually support
putting it into the language, however.  Certainly, it would have to be
possible to set a switch so as to not get an error for the old use of "="
in any compiler written in this century, at the least.

An extension would be to use some variant (perhaps "$=" or "=:") to mean
that the result of the expression is the lvalue *before* the assignment.
This would enable, e.g., an exchange to be written as "x = y $= x".

One drawback is that the idea does not generalize in an immediate and
obvious way to the increment and decrement operators.  Presumably, "++x"
should mean "x+:=1" and "x++" should mean "x+$=1"; but there should be some
such notation which means "x+=1".

Frank Adams                           ihnp4!philabs!pwa-b!mmintl!franka
Multimate International    52 Oakland Ave North    E. Hartford, CT 06108