From xemacs-m  Fri Dec  6 06:09:52 1996
Received: from gol1.gol.com (gol1.gol.com [202.243.48.4]) by xemacs.cs.uiuc.edu (8.8.3/8.8.3) with ESMTP id GAA09890 for <xemacs-beta@xemacs.org>; Fri, 6 Dec 1996 06:09:51 -0600 (CST)
Received: from pentagana (jhod@tokyo-04-148.gol.com [202.243.51.148]) by gol1.gol.com (8.8.3/8.8.3) with SMTP id VAA26043 for <xemacs-beta@xemacs.org>; Fri, 6 Dec 1996 21:09:50 +0900 (JST)
Sender: jhod@gol.com
Message-ID: <32A80E52.490EC67@po.iijnet.or.jp>
Date: Fri, 06 Dec 1996 21:15:14 +0900
From: Jareth Hein <jhod@po.iijnet.or.jp>
Organization: Sonic Software Planning, Tokyo
X-Mailer: Mozilla 3.01 (X11; I; Linux 2.1.13 i586)
MIME-Version: 1.0
To: XEmacs Beta Mailing List <xemacs-beta@xemacs.org>
Subject: b30 woes
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Well, I must have missed the announcement about b30, but I picked it up
this morn and started working away.

First off, my primary workstation at the moment is an Indigo2 running
IRIX 6.2. b30 compiles after editing the makefile (I'll create a patch
for the configure error tomorrow) and runs a grand total of once per
Xsession: the second invocation (even if non-simultanious) will crash
the X server with a stack to large error. I have since found that this
problem has existed since 19.13 (I don't have any earlier copies around)
and is most likely a problem with things getting included in the dumped
xemacs that shouldn't be. I'm sort of ashamed I didn't catch this
earlier, but as I usually fire up a copy of xemacs once and leave it
running all day, I hadn't caught this one yet. Has anyone else running
6.2 banged their head into this one yet? Or could it have something to
do with the mysterious memory glitches I'm getting with openGL programs
too (i.e. hardware oriented)?

Secondly, IRIX 6.2 defines library entry points for XmIm* functions, but
they seem to be stubs: in other words nothing happens when using
--with-xim=motif. The program runs, but no connection attempt is made to
the input server (in my case, xwnmo). Using --with-xim=xlib crashes just
as it has in b29 and in the same place, too, but as I'm fairly sure this
is a R6 thing, I'm chasing it from that end.

I'll build a Linux CDE version, a Linux Metro-Motif 2.0 and a HPUX 10
version tomorrow and let you know what the scoop is.

--Jareth Hein
Sonic Software Planning, Tokyo

