From xemacs-m  Fri Mar 28 18:25:38 1997
Received: from mailhost.lanl.gov (mailhost.lanl.gov [128.165.3.12])
	by xemacs.org (8.8.5/8.8.5) with ESMTP id SAA13320
	for <xemacs-beta@xemacs.org>; Fri, 28 Mar 1997 18:25:38 -0600 (CST)
Received: from branagh.ta52.lanl.gov (branagh.ta52.lanl.gov [128.165.144.9]) by mailhost.lanl.gov (8.8.5/8.8.3) with SMTP id RAA02847 for <xemacs-beta@xemacs.org>; Fri, 28 Mar 1997 17:25:39 -0700 (MST)
Received: by branagh.ta52.lanl.gov (SMI-8.6/SMI-SVR4)
	id RAA24043; Fri, 28 Mar 1997 17:20:46 -0700
Date: Fri, 28 Mar 1997 17:20:46 -0700
Message-Id: <199703290020.RAA24043@branagh.ta52.lanl.gov>
From: "John A. Turner" <turner@branagh.ta52.lanl.gov>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: xemacs-beta@xemacs.org
Subject: frame-icon problem (was Re: "live-icon" problem)
In-Reply-To: <199703281558.IAA21525@branagh.ta52.lanl.gov>
References: <199703281524.HAA08084@network-services.uoregon.edu>
	<199703281558.IAA21525@branagh.ta52.lanl.gov>
X-Mailer: VM 6.22 under 19.15 XEmacs Lucid
Reply-To: turner@lanl.gov

My environment is 19.15final with Solaris 2.5.1/CDE 1.0.2.  Previous
report from David Meyer (the live-icon report) was 20.1-beta10 and
Solaris 2.5.1.  I failed to mention this before.

I wrote:

 > Interesting.  I've seen that message as well, but it was related to
 > frame-icon.
 > 
 > It was strange, and it turned out I was doing something a bit wrong,
 > so I never reported it.  Maybe it's worth mentioning, though.
 > 
 > o In my .emacs, after (require 'frame-icon) I used to have code to add
 >   in the mappings for cvs-mode and f90-mode to icon-mode-alist, before
 >   they were added to frame-icon.el itself.
 > 
 > o I forgot about that code, so it remained in there even after they
 >   were added to frame-icon.el.
 > 
 > o It never presented a problem, until one day I did a cvs-update, and
 >   later switched to another virtual desktop (I'm using Solaris 2.5.1
 >   and CDE 1.0.2).
 > 
 > o Not long after that, the XEmacs frame in the original desktop pops
 >   up in the new desktop, with something similar to the above.
 >   Amazingly, I just switched to that desktop, and it turns out I had
 >   saved the message, and it was still there.  Here it is:
 > 
 > (1000) (specifier/warning) (error "Type of image instance not allowed here" #<image-instance (text) on #<x-device on ":0.0" 0x6f1> "/opt/local/xemacs-19.15-b104/etc/frame-icon/tree.xbm" 0x27c4>)
 > 
 >   Notice the (1000).  A new one of this would appear whenever I did
 >   anything (move the mouse, anything) and the number would increment.
 > 
 > Anyway, all this went away when I removed the code in my .emacs after
 > the (require 'frame-icon) mentioned above.
 > 
 > I don't know if this helps or not, but it was similar enough to the
 > live-icon thing mentioned above that I figured it might.

I knew the above was weird, and was a screwy report (that's why
I didn't send it in originally, but only after I saw something
similar).

Well, it just happened again, so removing the code that redundantly
added the maps for cvs-mode and f90-mode didn't make it go away after
all (I know, it never really made sense that it would in the first
place).

So then I thought it was something weird about virtual desktops,
etc. and CDE.  It still might be, but I just had it happen on the same
desktop under somewhat different circumstances.  Basically I closed
and opened and closed again a VM frame, and whamo, that message above
started spewing in the *Warnings* buffer.  Yes, it was the above
message, about tree.xbm (which is mapped to cvs-mode), and I had a
*cvs* buffer in another frame in which I was committing stuff.

Interestingly, quitting VM stopped the messages, and they haven't
started up again, even after restarting VM and opening/closing the
frame several times.

Is there anything I can do when this starts happening to try and see
what is going on?  I know this is a pretty sorry report.  Something
strange is going on, it's happened more than once, and I'm just
hoping that if I describe it enough, sooner or later there will be
enough clues that something will click with someone.

-- 
John Turner
http://www.lanl.gov/home/turner

