home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.aix
- Path: sparky!uunet!icmv!paris!mtranle
- From: mtranle@paris.IntelliCorp.COM (Minh Tran-Le)
- Subject: Re: More IBM "oversights"
- Message-ID: <mtranle.724400592@paris>
- Date: 15 Dec 92 06:23:12 GMT
- Reply-To: tranle@intellicorp.com
- References: <4348@unisql.UUCP> <Bz9A6n.6xq@constellation.ecn.uoknor.edu>
- Nntp-Posting-Host: paris
- Organization: IntelliCorp Inc.
- Lines: 31
-
- dwight@geohub.gcn.uoknor.edu (Dwight D. Moore) writes:
-
- >In article <4348@unisql.UUCP>, wrat@unisql.UUCP (wharfie) writes:
- >|
- >|>The xlc compiler is absolutely recalcitrant about accepting
- >|>anything like " Window foo = NULL". But the X11 headers
- >|>are full of stuff like "XtWindowOfObject(object) != NULL"
- >|>which of course won't compile with the native compiler. We
- >|>have to sprinkle our code with crap like "#ifdef IBM #undef XtIsRealized"
- >|>and fix it.
- >|>
- >|>If IBM has a better compiler (I can't believe they compiled X with
- >|>xlc) can a customer get it? Or does xlc have an undocumented
- >|>"Stop behaving like turbo pascal" mode?
-
- >A better compiler.? Don't blame
- >IBM and their compiler for doing its job; you need to understand the
- >nuances between K & R C, ANSI C, and the other mutations. xlc does
- >what it is supposed to do. ANSI C has strict type checking
- >if the compiler enforces it, which is the default for xlc.
- >Read the file /usr/lpp/xlc/bin/README.xlc for more information
- >on the compiler.
-
- If you really want to stick with the strict ansi C version xlc or c89, you
- need to compile with the following define:
-
- c89 -D_ALL_SOURCE -DNULL=0 ...
-
- The X11R5 headers should be more correct it its useage of NULL.
-
- Minh Tran-Le.
-