Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site watarts.UUCP
Path: utzoo!watmath!watarts!upen
From: upen@watarts.UUCP (Ue-Li Pen)
Newsgroups: net.lang.c
Subject: Re: identifiers
Message-ID: <8196@watarts.UUCP>
Date: Sun, 6-Jan-85 19:42:01 EST
Article-I.D.: watarts.8196
Posted: Sun Jan  6 19:42:01 1985
Date-Received: Mon, 7-Jan-85 03:14:02 EST
References: <8195@watarts.UUCP> <4875@utzoo.UUCP>
Organization: U of Waterloo, Ontario
Lines: 14

> > If there is so much disagreement over how long the identifier is significant
> > to, why not say:
> > 
> > a) That the number of characters actually stored must be at least 7.
> > b) The length of the whole identifier is significant.
> 
> Not a bad suggestion, except that the recalcitrant object-module formats
> that cause the length-significance problems don't have a "real length"
> field available either.
>
You wouldn't have to change the object-module format, just
store the length as an ascii character at the end of the identifier..
This gets around modifying your linker and compiler - however it
would reduce the size of your "significant characters" stored..