Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA
Path: utzoo!watmath!clyde!cbosgd!cbdkc1!desoto!packard!edsel!bentley!hoxna!houxm!ihnp4!mhuxn!mhuxj!mhuxr!ulysses!allegra!mit-eddie!godot!harvard!seismo!brl-tgr!tgr!olympus!bruce@su-shasta.arpa
From: bruce@su-shasta.arpa
Newsgroups: net.lang.c
Subject: Re: Alignment problems in C
Message-ID: <6893@brl-tgr.ARPA>
Date: Wed, 2-Jan-85 02:01:45 EST
Article-I.D.: brl-tgr.6893
Posted: Wed Jan  2 02:01:45 1985
Date-Received: Thu, 3-Jan-85 04:08:53 EST
Sender: news@brl-tgr.ARPA
Organization: Ballistic Research Lab
Lines: 14

Having ``packed'' structures in C would help tremendously in writing
protocol code.  There really are times that the data is defined by
someone else, and having a clear (structured) way of describing it
is MUCH better than packing and unpacking inside of the receiving/
generating code.  It may also be more efficient.

I recommend some mechanism for forcing ``packed'' structures.  If
a particular compiler cannot handle some alignments, it can complain
when it sees them, e.g., { char, short, long } can complain about the
short and/or long on an odd boundary.

This would be a useful ``standard extension'' to C.

/Bruce Borden