A few remarks:
These should be fixed to use the Windows print menu on Windows. I
guess an alternative would be to install an lpr program from Cygwin
I have cygwin,
but tried the XEmacs version tailored for M$ Windows, to
be able to also work in a pure Windows environment. Maybe I should try
to get a cygwin print program working, as you said.
It appears that your OS is Windows XP (not Vista, which apparently
brings a whole collection of new problems). Is that correct?
Yes, it is XP.
Would it be possible for you to run xemacs under a debugger and
reproduce the crash, then send us a stack trace?
Cannot promise anything. Will see
if that error will surface again, and
if so I will try to make it repeatable. Possibly there was something
unique in that specific system state that caused the crash.
What did you expect to recover? "Recover session" here
(AFAIK) simply
looks for autosave files, it does not try to put your xemacs back into
the state at which xemacs last checkpointed.
Well, the system told me (before
dying) that I could use recover-session
to get back whatever-it-was, and I sort of expected that the system on
restart at least would be able to know what buffers/files I might have
had open. Actually, I do not know if I had any buffer with 'autosave'.
XEmacs (stable versions) have been so reliable that I have had no reason
to keep a mental image of the state of the system, so I have not been
able to remember what buffers I had at the time of the crash. And
neither could Xemacs after restart.
Actually, we'd prefer you keep reporting bugs. :-) In general,
XEmacs
21.5 is quite stable on Windows AFAIK (I don't use Windows at all).
OK, I will
stay with a bleeding-edge version ;-)
/olle
/olle
Stephen J. Turnbull wrote:
> Olle Olsson writes:
>
> > Background: the three ways of printing a buffer (well, there are more
> > ways, but
> > lets be simple-minded first):
> > - M-x print-buffer: FAILS ... cannot find program "pr"
> > - the print icon on the top menu: FAILS ... cannot find program ""
>
> These should be fixed to use the Windows print menu on Windows. I
> guess an alternative would be to install an lpr program from Cygwin or
> something (I'm just assuming that they exist, so don't kill yourself
> looking for one).
>
> > - the File > Print menu item: sometimes works, but not this time ;-(
>
> Your version seems to be a vanilla 21.5.28 from the last beta release.
> If you're not using the latest sources from the Mercurial repository,
> you may wish to try those. (Mercurial is a Python program, and
> probably should work without trouble for this purpose on Windows. I
> recommend doing this since you have some interest in up-to-date
> XEmacsen.) See
http://www.xemacs.org/Develop/hgaccess.html for more
> info.
>
It appears that your OS is Windows XP (not Vista, which apparently
brings a whole collection of new problems). Is that correct?
>
> Regarding the actual crash, this
>
> > BEGIN DOS BOX =================================>>>
> >
> > Fatal error: assertion failed, file
> > c:\XEmacsBuild\xemacs-21.5-2008-10-12\src\device-impl.h, line 166, EQ
> > (DEVICE_TYPE (d), sym)
>
> is somewhat helpful, but to get any further we'll need somebody to
> reproduce the crash under the debugger. It's probably the case that
> it crashed in a dialog because XEmacs handed Windows some parameter it
> didn't like, but until we can limit the possibilities via a stack
> trace, it's like searching for a needle in a haystack.
>
> Would it be possible for you to run xemacs under a debugger and
> reproduce the crash, then send us a stack trace? (The whole thing,
> please, not just the last screenful of frames!)
>
> > By the way ...
> > - `M-x recover-session' did not work .... "nothing to recover"
>
> What did you expect to recover? "Recover session" here (AFAIK) simply
> looks for autosave files, it does not try to put your xemacs back into
> the state at which xemacs last checkpointed. I normally have 50-100
> buffers in my process, with probably 25-50 files open, but none of my
> ~/.saves-* files has more than one pair of file-autosave in it!
>
> If you want to actually recover the state of xemacs as of a
> checkpoint, use desktop.el or session.el.
>
> I admit this is a poorly named command, but its history goes back
> into the 1980s at least so I think that we can't complain too much.
>
> > - maybe I should go back to a more stable version of XEmacs-for-MSWindows.
>
Actually, we'd prefer you keep reporting bugs. :-) In general,
XEmacs
21.5 is quite stable on Windows AFAIK (I don't use Windows at all).
>
>
--
------------------------------------------------------------------
Olle Olsson olleo(a)sics.se Tel: +46 8 633 15 19 Fax: +46 8 751 72 30
[Svenska W3C-kontoret: olleo(a)w3.org]
SICS [Swedish Institute of Computer Science]
Box 1263
SE - 164 29 Kista
Sweden
------------------------------------------------------------------
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta