From xemacs-m  Sun Jun 22 23:40:50 1997
Received: from crystal.WonderWorks.COM (crystal.WonderWorks.com [192.203.206.1])
	by xemacs.org (8.8.5/8.8.5) with ESMTP id XAA25673
	for <xemacs-beta@xemacs.org>; Sun, 22 Jun 1997 23:40:48 -0500 (CDT)
Received: by crystal.WonderWorks.COM 
	id QQcvda02767; Mon, 23 Jun 1997 00:40:49 -0400 (EDT)
Date: Mon, 23 Jun 1997 00:40:49 -0400 (EDT)
Message-Id: <QQcvda02767.199706230440@crystal.WonderWorks.COM>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Kyle Jones <kyle_jones@wonderworks.com>
To: xemacs-beta@xemacs.org
Subject: Re: `skip_debugger'
In-Reply-To: <m2en9une8d.fsf@altair.xemacs.org>
References: <kigu3iqpqr4.fsf@jagor.srce.hr>
	<QQcvcp25247.199706230156@crystal.WonderWorks.COM>
	<m2k9jmnj4l.fsf@altair.xemacs.org>
	<QQcvct01182.199706230257@crystal.WonderWorks.COM>
	<m2hgeqngqk.fsf@altair.xemacs.org>
	<QQcvcx02113.199706230352@crystal.WonderWorks.COM>
	<m2en9une8d.fsf@altair.xemacs.org>
X-Mailer: VM 6.33 under 20.3 "Copenhagen" XEmacs Lucid (beta8)
X-Face: /cA45WHG7jWq>(O3&Z57Y<"WsX5ddc,4c#w0F*zrV#=M
        0@~@,s;b,aMtR5Sqs"+nU.z^CSFQ9t`z2>W,S,]:[+2^
        Nbf6v4g>!&,7R4Ot4Wg{&tm=WX7P["9%a)_da48-^tGy
        ,qz]Z,Zz\{E.,]'EO+F)@$KtF&V

Steven L Baur writes:
 > Kyle Jones <kyle_jones@wonderworks.com> writes:
 > 
 > > Steven L Baur writes:
 > 
 > >> Yes.  Build with Hrvoje's patch calling Ferror_message_string and Lisp 
 > >> union types.  Then force an error during loadup.  The easiest way to
 > >> do it is to touch a dumped lisp file (like simple.el) which will cause 
 > >> update-elc.el to load all lisp and start bytecompiling when next doing 
 > >> `make all-elc'.
 > 
 > > Done.  No crash.
 > 
 > (Sorry, with the running-temacs-p check disabled).

Yeah, I did yank that stuff.

 > Actually I think I've gotten it figured out, though the manifestation
 > of the problem mystifies me.  Ferror_message_string() isn't declared.
 > I've declared it, GCPRO'ed data and I don't get a crash either.  I'm
 > going to try it next without the GCPRO on data.
 > 
 > [...]
 > 
 > It worked.

The question is: why?

