emacspeak and emacs 22 GTK bug?

[Tim Cross <tcross@xxxxxxxxxxx>]

thanks Raman. However, thats pretty much the first test I tried. I've not had
time to look at this over the last week, but hope to investigate further. I've
found I can build the GTK version from source on Debian testing, but it fails
on Debian unstable. I've managed to repeat this on 5 different systems. Running
testing, all appears fine, but on Unstable it shows the problems I've listed in
a consistent manner. It appears that whatever the problem is, its quite subtle
and difficult to nail down. I'm still attempting to develop a clear test case.
will keep plugging away at it in my spare time.

Tim



T. V. Raman writes:
 > 
 > Here is a simple test case for you to run against your failing
 > emacs-gtk
 > 
 > In the scracth buffer, evaluate these lines in sequence, and run
 > M-x foo each time. Notice that the advice gets evaluated and
 > defined correctly.
 > 
 > (defun foo ()
 > (interactive)
 > (message "I'm foo."))
 > (defadvice foo (after first pre act comp)(message "After foo"))
 > 
 > (defadvice foo (after second pre act comp)(message "son of foo."))
 > 
 > 
 > >>>>> "Tim" == Tim Cross <tcross@xxxxxxxxxxx> writes:
 >     Tim> Hi Raman, I think you misunderstood me. I don't want you
 >     Tim> to grab a pre-built version. I wanted to ask if you
 >     Tim> wouldn't mind checking the describe function information
 >     Tim> with the version you have already built from CVS. I'm
 >     Tim> already working with the package maintainer to see if
 >     Tim> the problem lies there.
 >     Tim> 
 >     Tim> The issue I'm encountering is subtle and is only
 >     Tim> revealed iff you have extra advice defined for view-file
 >     Tim> other than emacspeak. Just running emacspeak without any
 >     Tim> non-emacspeak advice defined does not reveal the problem
 >     Tim> and therefore would not necessarily show up in a simple
 >     Tim> test.
 >     Tim> 
 >     Tim> All I wanted was some confirmation that the
 >     Tim> describe-function operation run on view-file in your CVS
 >     Tim> version of GTK emacs is correctly reporting that the
 >     Tim> function has been advised. The version I am running is
 >     Tim> not reporting this correctly on my system and I wanted
 >     Tim> to verify that yours is correct as this would indicate I
 >     Tim> have some other issue on my system not related to the
 >     Tim> CVS sources or the packaged version.
 >     Tim> 
 >     Tim> To summerise
 >     Tim> 
 >     Tim>  - After starting emacs 22 GTK version and loading
 >     Tim> emacspeak, do a C-h f view-file. Scroll to the bottom of
 >     Tim> the output where it should say "This function is advised
 >     Tim> etc". If it does not, try an apropos on view-file and
 >     Tim> see if you have an entry for ad-Orig-view-file. Note
 >     Tim> that at this point, view-file appears to work perfectly
 >     Tim> fine. If you do see this inconsistency, you will then
 >     Tim> find if you load my txutils.el utility and repeat the
 >     Tim> process, both the emacspeak and txutils advice are
 >     Tim> reported, but now view-file is broken and will generate
 >     Tim> a backtrace if debug on error is enabled.
 >     Tim> 
 >     Tim> If you can do this check, it would be appreciated. If
 >     Tim> you can't, then don't worry. I will have my download
 >     Tim> quota renewed in a couple of weeks and will grab a fresh
 >     Tim> copy of the CVS and build my self to check.
 >     Tim> 
 >     Tim> BTW Note that none of this inconsistency in reporting of
 >     Tim> advice or any of the other issues exist in the non-GTK
 >     Tim> version, which is built from exactly the same sources.
 >     Tim> 
 >     Tim> Tim
 >     Tim> 
 >     Tim> T. V. Raman writes:
 >     >> 
 >     >> Tim,
 >     >> 
 >     >> I dont trust pre-built versions of Emacs, so I'm not going
 >     >> to pull a particular build of Emacs that you might have
 >     >> gotten.
 >     >> 
 >     >> It looks like you have a bad build as best as I can tell.
 >     >> 
 >     >> The trust-worthy means of verifying if there is a problem
 >     >> with Emacs built against GTK is to:
 >     >> 
 >     >> A) Pull the CVS version of Emacs 22 B) ./configure
 >     >> --with-toolkit=gtk
 >     >> 
 >     >> I've done this last week and as you saw, it did not
 >     >> reproduce the bizarre behavior you're reporting.  I
 >     >> suggest you write to the person who built the emacs-gtk
 >     >> that you pulled.
 >     >> 
 >     >> 
 >     >> 
 >     >> >>>>> "Tim" == Tim Cross <tcross@xxxxxxxxxxx> writes:
 >     Tim> Hi Raman, I've updated to a later version of emacs 22
 >     Tim> GTK version (from 1st May) and I'm still getting the
 >     Tim> same problem. However, I have noticed something else a
 >     Tim> little odd. I was wondering if you could check the
 >     Tim> version you have to see if it is the same.
 >     Tim> 
 >     Tim> After starting emacs 22 GTK version with emacspeak
 >     Tim> loaded, if I do a describe function on view-file, it
 >     Tim> does not mention anything about the function being
 >     Tim> advised. However, if I do an apropos for view-file,
 >     Tim> there is a ad-Orig-view-file entry.
 >     Tim> 
 >     Tim> If I do the same thing with the non-GTK version, the
 >     Tim> describe function returns the expected note at the end
 >     Tim> about the function being advised by emacspeak. So, it
 >     Tim> looks like something is not right as they should both
 >     Tim> report exactly the same thing.
 >     Tim> 
 >     Tim> Note that view-file works fine until you load another
 >     Tim> piece of advice on that function. Then you get the
 >     Tim> max-lisp-eval-depth exceeded error. This does not occur
 >     Tim> with the non-GTK version.
 >     Tim> 
 >     Tim> In my case, this problem arises with my txutils package
 >     Tim> as it defines some advice on view-file. Note that there
 >     Tim> are no issues if I load my package without emacspeak. It
 >     Tim> is only when the two are loaded the error occurs.
 >     Tim> 
 >     Tim> Could you start emacs 22 GTK version and just have a
 >     Tim> look at the describe function for view-file and see if
 >     Tim> it reports that the function has been advised. If it
 >     Tim> doesn't, then we either have an issue with emacspeak or
 >     Tim> with the GTK version of emacs.
 >     Tim> 
 >     Tim> thanks,
 >     Tim> 
 >     Tim> Tim
 >     Tim> 
 >     Tim> P.S. I have also setup an old box and am using this for
 >     Tim> testing. I'm not loading any other packages apart from
 >     Tim> emacspeak and my txutils utility. This is to eliminate
 >     Tim> the chance of the problem being due to some unexpected
 >     Tim> interaction with another package that is being
 >     Tim> loaded. I'm trying to get things down to a bare minimum.
 >     Tim> -- Tim Cross tcross@xxxxxxxxxxx
 >     Tim> 
 >     Tim> There are two types of people in IT - those who do not
 >     Tim> manage what they understand and those who do not
 >     Tim> understand what they manage.
 >     Tim> 
 >     Tim> -----------------------------------------------------------------------------
 >     Tim> To unsubscribe from the emacspeak list or change your
 >     Tim> address on the emacspeak list send mail to
 >     Tim> "emacspeak-request@xxxxxxxxxxx" with a subject of
 >     Tim> "unsubscribe" or "help"
 >     >> 
 >     >> -- 
 >     >> Best Regards, --raman
 >     >> 
 >     >> 
 >     >> Email: raman@xxxxxxxxxxx WWW:
 >     >> http://emacspeak.sf.net/raman/ AIM: emacspeak GTalk:
 >     >> tv.raman.tv@xxxxxxxxxxx PGP:
 >     >> http://emacspeak.sf.net/raman/raman-almaden.asc Google:
 >     >> tv+raman IRC: irc://irc.freenode.net/#emacs
 >     Tim> 
 >     Tim> -- Tim Cross tcross@xxxxxxxxxxx
 >     Tim> 
 >     Tim> There are two types of people in IT - those who do not
 >     Tim> manage what they understand and those who do not
 >     Tim> understand what they manage.
 > 
 > -- 
 > Best Regards,
 > --raman
 > 
 >       
 > Email:  raman@xxxxxxxxxxx
 > WWW:    http://emacspeak.sf.net/raman/
 > AIM:    emacspeak       GTalk: tv.raman.tv@xxxxxxxxxxx
 > PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
 > Google: tv+raman 
 > IRC:    irc://irc.freenode.net/#emacs
 > 
 > -----------------------------------------------------------------------------
 > To unsubscribe from the emacspeak list or change your address on the
 > emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
 > subject of "unsubscribe" or "help"
 > 

-- 
Tim Cross
tcross@xxxxxxxxxxx

There are two types of people in IT - those who do not manage what they 
understand and those who do not understand what they manage.

-----------------------------------------------------------------------------
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
subject of "unsubscribe" or "help"