x-face problems

x-face problems

[John H Palmieri]

GNU Emacs 20.2.1 (sparc-sun-solaris2.5.1, X toolkit), qgnus-0.32

I just realized that I haven't been seeing any x-faces today.  There
are several possibilities: I've been changing to the newest versions
of Quassia Gnus as they appear, so I suppose something there could
have changed (although I've looked at the ChangeLog and haven't
spotted any likely candidates); alternatively, my system administrator
could have done something which had the side effect of screwing things
up.  (I suppose I could have done something, too, but I haven't
changed my .emacs, .gnus, or .cshrc in any relevant ways.)

My questions:
  Has anything related to x-face handling changed in the recent
versions of gnus?
  If not, is there any way to monitor what gnus does when it tries to
display an x-face?  I have gnus-article-display-x-face as part of my
gnus-article-display-hook, and I've tried running the commands in
gnus-article-x-face-command from an xterm one at a time, and succeeded
(i.e., uncompface, icontopbm, and xv all work individually).  Any
ideas about how can I find out what's going wrong?

-- 
John H. Palmieri, Visiting Assistant Professor
Department of Mathematics, University of Notre Dame
Email: palmieri@member.ams.org
URL: http://www.nd.edu/~jpalmier/

Re: x-face problems

[Francois Felix Ingrand]

John H Palmieri <John.H.Palmieri.2@nd.edu> writes:

> GNU Emacs 20.2.1 (sparc-sun-solaris2.5.1, X toolkit), qgnus-0.32
> 
> I just realized that I haven't been seeing any x-faces today.  There
> are several possibilities: I've been changing to the newest versions
> of Quassia Gnus as they appear, so I suppose something there could
> have changed (although I've looked at the ChangeLog and haven't
> spotted any likely candidates); alternatively, my system administrator
> could have done something which had the side effect of screwing things
> up.  (I suppose I could have done something, too, but I haven't
> changed my .emacs, .gnus, or .cshrc in any relevant ways.)
> 
> My questions:
>   Has anything related to x-face handling changed in the recent
> versions of gnus?

I had a smimilar pb a while ago. It appeared that an hook was cleaning up
headers before X-Face where display. Prepending instead of appending fixed the
pb. 

(add-hook 'gnus-article-display-hook 'gnus-article-display-x-face t)

Note the t...
-- 
-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Francois Felix INGRAND     "La guerre, qu'on voit danser le long des golfes
Professional: felix@laas.fr       clairs, a des reflets d'argent" (Anonyme)
Personal: felix@dial.oleane.com            WWW: <http://www.laas.fr/~felix>

Re: x-face problems

[John H Palmieri]

Francois Felix Ingrand <felix@laas.fr> writes:

> John H Palmieri <John.H.Palmieri.2@nd.edu> writes:
> 
> > GNU Emacs 20.2.1 (sparc-sun-solaris2.5.1, X toolkit), qgnus-0.32
> > 
> > I just realized that I haven't been seeing any x-faces today.
> > 
> > My questions:
> >   Has anything related to x-face handling changed in the recent
> > versions of gnus?
> 
> I had a smimilar pb a while ago. It appeared that an hook was cleaning up
> headers before X-Face where display. Prepending instead of appending fixed the
> pb. 
> 
> (add-hook 'gnus-article-display-hook 'gnus-article-display-x-face t)
> 
> Note the t...

I tried adding the t, but it didn't help.  Maybe I should back up:
when I first set up gnus, I tried having x-faces displayed, but had
problems until I put in the lines:

      (make-variable-buffer-local 'shell-file-name)))
      (add-hook 'gnus-article-display-hook 
		'(lambda ()
		   (setq shell-file-name "/bin/sh")))))

Now, for instance, if I comment these out, when I read a message with
an X-Face header, xv starts up and tells me

  File '/tmp/xva003Gl' contains no data.  (Zero length file.)

As of a day or two ago, if I don't comment these lines out, then
nothing seems to happen when I read a message with an X-Face header.
So I ask again, how can I figure out what's going wrong?

By the way, I do have the same problem as David Goldberg (Re: [PB]
window configs / new frames): 

> On another note, does anyone else have trouble with Didier's messages
> having two x-face headers.  They caused emacs to core dump on me.
> Next time I'm going to run without a coredump size limit and submit a
> bug report, but I'm curious if I'm the only one.

-- 
John H. Palmieri, Visiting Assistant Professor
Department of Mathematics, University of Notre Dame
Email: palmieri@member.ams.org
URL: http://www.nd.edu/~jpalmier/

Re: x-face problems

[Edward J. Sabol]

Excerpts from mail: (25-Feb-98) Re: x-face problems by John H Palmieri

John H Palmieri <John.H.Palmieri.2@nd.edu> writes:
John> I just realized that I haven't been seeing any x-faces today.

Francois Felix Ingrand <felix@laas.fr> writes:
Felix> I had a smimilar pb a while ago. It appeared that an hook was cleaning
Felix> up headers before X-Face where display. Prepending instead of
Felix> appending fixed the pb.
Felix> 
Felix> (add-hook 'gnus-article-display-hook 'gnus-article-display-x-face t)
Felix> 
Felix> Note the t...

John> I tried adding the t, but it didn't help.  Maybe I should back up:
John> when I first set up gnus, I tried having x-faces displayed, but had
John> problems until I put in the lines:
John>
John> (make-variable-buffer-local 'shell-file-name)))
John> (add-hook 'gnus-article-display-hook 
John>           '(lambda ()
John>              (setq shell-file-name "/bin/sh")))))
John>
John> Now, for instance, if I comment these out, when I read a message with
John> an X-Face header, xv starts up and tells me
John>
John> File '/tmp/xva003Gl' contains no data.  (Zero length file.)
John>
John> As of a day or two ago, if I don't comment these lines out, then
John> nothing seems to happen when I read a message with an X-Face header.
John> So I ask again, how can I figure out what's going wrong?

Sorry, I don't know how to figure out what's going wrong in your situation.
(Did you change your shell at all in the past few days?) But maybe it will be
of some help to you if I tell you what I use and know to work in my
configuration. My shell is tcsh, and the default gnus-article-x-face-command
is simply incompatible with csh-like shells, as you probably know since you
set shell-file-name to "/bin/sh" in gnus-article-display-hook. But instead of
doing what you do and changing shell-file-name, I instead change
gnus-article-x-face-command to be csh-compatible (in addition to specifying
the geometry for xv). Here's the elisp lines that I use:

(add-hook 'gnus-article-display-hook 'gnus-article-display-x-face t)
(setq gnus-article-x-face-command "( echo '/* Width=48, Height=48 */'; uncompface; ) | icontopbm | xv -geometry +0+195 -quit -")

Later,
Ed

Re: x-face problems

[John H Palmieri]

"Edward J. Sabol" <sabol@alderaan.gsfc.nasa.gov> writes:

> Excerpts from mail: (25-Feb-98) Re: x-face problems by John H Palmieri
> 
> John H Palmieri <John.H.Palmieri.2@nd.edu> writes:
> John> I just realized that I haven't been seeing any x-faces today.
> 
>     But instead of
> doing what you do and changing shell-file-name, I instead change
> gnus-article-x-face-command to be csh-compatible (in addition to specifying
> the geometry for xv). Here's the elisp lines that I use:
> 
> (add-hook 'gnus-article-display-hook 'gnus-article-display-x-face t)
> (setq gnus-article-x-face-command "( echo '/* Width=48, Height=48 */'; uncompface; ) | icontopbm | xv -geometry +0+195 -quit -")
> 
> Later,
> Ed

Thanks, this works.  (I still wonder why it stopped working a few days
ago.  I didn't change my shell or anything like that.  But the issue
is not so crucial now that the x-faces are back.)

-- 
John H. Palmieri, Visiting Assistant Professor
Department of Mathematics, University of Notre Dame
Email: palmieri@member.ams.org
URL: http://www.nd.edu/~jpalmier/

Re: x-face problems

[Dave Love]

>>>>> "John" == John H Palmieri <John.H.Palmieri.2@nd.edu> writes:

 John> Thanks, this works.  (I still wonder why it stopped working a few days
 John> ago.  I didn't change my shell or anything like that.  But the issue
 John> is not so crucial now that the x-faces are back.)

Because of this.  I think the hook order is a red herring.  [Cue
interesting discussion from around the herring-pond?]

1998-02-25  Dave Love  <d.love@dl.ac.uk>

	* gnus-art.el (gnus-article-x-face-command): Replace leading `{'.

*** gnus-art.el	1998/02/25 20:07:05	1.1
--- gnus-art.el	1998/02/25 20:07:25
***************
*** 167,173 ****
    :group 'gnus-article-hiding)
  
  (defcustom gnus-article-x-face-command
!   " echo '/* Width=48, Height=48 */'; uncompface; } | icontopbm | xv -quit -"
    "*String or function to be executed to display an X-Face header.
  If it is a string, the command will be executed in a sub-shell
  asynchronously.	 The compressed face will be piped to this command."
--- 167,173 ----
    :group 'gnus-article-hiding)
  
  (defcustom gnus-article-x-face-command
!   "{ echo '/* Width=48, Height=48 */'; uncompface; } | icontopbm | xv -quit -"
    "*String or function to be executed to display an X-Face header.
  If it is a string, the command will be executed in a sub-shell
  asynchronously.	 The compressed face will be piped to this command."

Re: x-face problems

[Lars Magne Ingebrigtsen]

John H Palmieri <John.H.Palmieri.2@nd.edu> writes:

> 		   (setq shell-file-name "/bin/sh")))))

Well, in general you shouldn't ever have this variable set to anything
other than sh or bash.  Setting it to something derived from csh is
like asking Emacs to kill you.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen