[comp.sys.sgi] Response from MIT X Consortium concerning XGetDefault and GNU Emacs

BRUC@DINO.SQUIBB.COM ("Bob Bruccoleri 683-6165", 609) (01/23/91)

From:	IN%"gildea@expo.lcs.mit.edu"  "Stephen Gildea" 22-JAN-1991 18:00
To:	rms@ai.mit.edu
Subj:	RE: [BRUC@dino.squibb.com: GNU Emacs, Xdefaults, and version 3.3.1 of Irix]

Received: from EXPO.LCS.MIT.EDU by dino.squibb.com; Tue, 22 Jan 91 17:59 EST
Received: from excalibur.lcs.mit.edu by expo.lcs.mit.edu; Tue, 22 Jan 91
 18:01:20 EST
Received: by excalibur.lcs.mit.edu; Tue, 22 Jan 91 18:01:17 EST
Date: Tue, 22 Jan 91 18:01:11 EST
From: Stephen Gildea <gildea@expo.lcs.mit.edu>
Subject: RE: [BRUC@dino.squibb.com: GNU Emacs, Xdefaults, and version 3.3.1 of
 Irix]
To: rms@ai.mit.edu
Cc: bruc@dino.squibb.com
Message-id: <9101222301.AA23845@excalibur.lcs.mit.edu>
In-Reply-To: Message from rms@ai.mit.edu of 15 Jan 91 1:24:02 EST
X-Envelope-to: bruc

In R4, on some systems, the last two arguments of XGetDefault are
indeed reversed.  This bug did not exist in R3 and will not exist
in R5.  (The doc, spec, and header file are all correct.)

As you can see from the code fragment Bob enclosed, the function
header appears twice in the source code, once with ANSI-C-style
prototypes, and once with K&R style.  The ANSI-C is broken in R4.

Don't break Emacs to compensate; have the affected people edit
their Xlib and re-build.

 < Stephen
   MIT X Consortium