Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site cheviot.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!genrad!teddy!panda!talcott!harvard!seismo!mcvax!ukc!cheviot!lindsay
From: lindsay@cheviot.UUCP (Lindsay F. Marshall)
Newsgroups: net.lang.c
Subject: pointers to structs
Message-ID: <208@cheviot.UUCP>
Date: Wed, 9-Jan-85 04:29:39 EST
Article-I.D.: cheviot.208
Posted: Wed Jan  9 04:29:39 1985
Date-Received: Sat, 12-Jan-85 07:35:21 EST
Reply-To: lindsay@cheviot.UUCP (Lindsay F. Marshall)
Organization: U. of Newcastle upon Tyne, U.K.
Lines: 19

<>
    What is the position as regards a) portability, and b) the "standard"
on using the following feature of C :

    .
    .
    struct STYPE *t;
    .
    .

where the size and shape of STYPE have not been defined and the elements
of t are NEVER accessed - the normal use is that t is assigned a value
from a function call and is then passed to another function, the contents
of the structure are not releveant. This certainly works
on the compilers I have tried it on, but is it "correct"?? (If it is, it's a
tremendous way of reducing the number of include files one needs!!)
Please mail replies if possible, as not all the news has the strength to reach
this far! I will summarise if it is worth it.

Lindsay F. Marshall - Computing Lab., U of Newcastle upon Tyne
  ARPA : lindsay%cheviot%newcastle.mailnet@MIT-MULTICS.ARPA
  UUCP : !ukc!cheviot!lindsay