[raman@users.sf.net: url-http.el breaks Emacs/W3]

[raman@users.sf.net: url-http.el breaks Emacs/W3]

[Richard Stallman]

Would someone please deal with this, then ack?

------- Start of forwarded message -------
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Wed, 13 Sep 2006 20:59:45 -0700
To: emacs-devel@gnu.org
CC: 
From: "T. V. Raman" <raman@users.sf.net>
Subject: url-http.el breaks Emacs/W3
Reply-To: raman@users.sf.net
X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=failed 
	version=3.0.4

I finally  got a little more detail on where this breakage is
occuring.

Symptoms:

If you use Emacs/w3 and use the URL package out of Emacs CVS you
hit an error on sites that do a redirect e.g. http://sf.net

The error is very difficult to track down because it's being
thrown from the w3 site in w3-fetch-callback.

But I have verified that the problem appears to be  isolated in
url-http.el, i.e. if you use  package url out of Emacs 22 CVS,
and load url-http.el by hand from William Perry's original CVS
snapshot on savannah, then everything works correctly.

There are  a number of bug  fixes in url-http.el 
in the Emacs CVS repos -- so simply going back to Bill's version
is not a good idea either.

Could someone here help me chase this down further?

To reproduce the problem:

Pull w3 from savannah cvs, ./configure
- --with-emacs=<path-to-emacs-22> make 
then do 
M-x w3-fetch http://sf.net 
the above URL produces a 301 redirect to the actual SF site which
triggers the error
 Thanks, 
 --Raman

- -- 
Best Regards,
- --raman

      
Email:  raman@users.sf.net
WWW:    http://emacspeak.sf.net/raman/
AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
Google: tv+raman 
IRC:    irc://irc.freenode.net/#emacs


_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel
------- End of forwarded message -------

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[Andreas Schwab]

Richard Stallman <rms@gnu.org> writes:

> Would someone please deal with this, then ack?

It was broken by this change:

2005-07-15  Richard M. Stallman  <rms@gnu.org>

	* url-http.el (url-http-parse-headers): Add :redirect arg-pair
	when calling url-retrieve, to indicate a redirect.

	* url.el (url-retrieve): The callback function can get an additional
	keyword arg pair.

w3-fetch-callback does not know how to deal with that.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[T. V. Raman]

Thanks for tracking this down!

The problem is that w3 is moribund -- has no maintainer, so even
if I fix the problem locally in myinstall of emacs, it still
breaks for all users of emacspeak -- who I suspect ar epretty
much th only users of W3.

So while w3 waits for a maintainer, or while emacspeak users wait
for a browsing solution that works, could we work around the
problem in url-http.el?

>>>>> "Andreas" == Andreas Schwab <schwab@suse.de> writes:
    Andreas> Richard Stallman <rms@gnu.org> writes:
    >> Would someone please deal with this, then ack?
    Andreas> 
    Andreas> It was broken by this change:
    Andreas> 
    Andreas> 2005-07-15 Richard M. Stallman <rms@gnu.org>
    Andreas> 
    Andreas> 	* url-http.el (url-http-parse-headers): Add
    Andreas> :redirect arg-pair when calling url-retrieve, to
    Andreas> indicate a redirect.
    Andreas> 
    Andreas> 	* url.el (url-retrieve): The callback function
    Andreas> can get an additional keyword arg pair.
    Andreas> 
    Andreas> w3-fetch-callback does not know how to deal with
    Andreas> that.
    Andreas> 
    Andreas> Andreas.
    Andreas> 
    Andreas> -- Andreas Schwab, SuSE Labs, schwab@suse.de SuSE
    Andreas> Linux Products GmbH, Maxfeldstraße 5, 90409
    Andreas> Nürnberg, Germany PGP key fingerprint = 58CA 54C7
    Andreas> 6D53 942B 1756 01D3 44D5 214B 8276 4ED5 "And now for
    Andreas> something completely different."
    Andreas> 
    Andreas> 
    Andreas> _______________________________________________
    Andreas> Emacs-devel mailing list Emacs-devel@gnu.org
    Andreas> http://lists.gnu.org/mailman/listinfo/emacs-devel

 Thanks, 
 --Raman

--
Best Regards,
--raman


Email:  raman@users.sf.net
WWW:    http://emacspeak.sf.net/raman/
AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
Google: tv+raman
IRC:    irc://irc.freenode.net/#emacs

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[Richard Stallman]

    The problem is that w3 is moribund -- has no maintainer, so even
    if I fix the problem locally in myinstall of emacs, it still
    breaks for all users of emacspeak -- who I suspect ar epretty
    much th only users of W3.

How about if you distribute a fixed version of W3?

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[T. V. Raman]

I'm reluctant to take over maintainance of W3 because I already
have my hands full with emacspeak.

>>>>> "Richard" == Richard Stallman <rms@gnu.org> writes:
    Richard>     The problem is that w3 is moribund -- has no
    Richard> maintainer, so even if I fix the problem locally in
    Richard> myinstall of emacs, it still breaks for all users of
    Richard> emacspeak -- who I suspect ar epretty much th only
    Richard> users of W3.
    Richard> 
    Richard> How about if you distribute a fixed version of W3?

-- 
Best Regards,
--raman

      
Email:  raman@users.sf.net
WWW:    http://emacspeak.sf.net/raman/
AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
Google: tv+raman 
IRC:    irc://irc.freenode.net/#emacs

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[Richard Stallman]

    I'm reluctant to take over maintainance of W3 because I already
    have my hands full with emacspeak.

To distribute a version of W3 which fixes this bug
does not mean you have to make a substantial commitment
of time to W3 maintenance.

Anyway, it seems to be that or nothing.
We've concluded that the problem is in W3, so the fix
has to be in W3.

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[Richard Stallman]

    It was broken by this change:

    2005-07-15  Richard M. Stallman  <rms@gnu.org>

	    * url-http.el (url-http-parse-headers): Add :redirect arg-pair
	    when calling url-retrieve, to indicate a redirect.

	    * url.el (url-retrieve): The callback function can get an additional
	    keyword arg pair.

    w3-fetch-callback does not know how to deal with that.

The change I installed fixes a bug.  ISTR that my fix was the only
natural one that I saw.  It may be that someone will find another good
way to fix this which is 100% compatible.

Otherwise, I think w3 needs to be changed to handle this convention.

Re: [raman@users.sf.net: url-http.el breaks Emacs/W3]

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>     It was broken by this change:
>
>     2005-07-15  Richard M. Stallman  <rms@gnu.org>
>
> 	    * url-http.el (url-http-parse-headers): Add :redirect arg-pair
> 	    when calling url-retrieve, to indicate a redirect.
>
> 	    * url.el (url-retrieve): The callback function can get an additional
> 	    keyword arg pair.
>
>     w3-fetch-callback does not know how to deal with that.
>
> The change I installed fixes a bug.  ISTR that my fix was the only
> natural one that I saw.  It may be that someone will find another good
> way to fix this which is 100% compatible.
>
> Otherwise, I think w3 needs to be changed to handle this convention.


The w3 repository seems to have a fix for this problem now:

revision 1.36
date: 2006-11-01 00:21:50 +0000;  author: legoscia;  state: Exp;  lines: +24 -8
lisp/w3.el (w3-fetch-redirect-callback): Handle both old and new style
of `url-retrieve' callback.


So I have deleted the following item from FOR-RELEASE / BUGS:

** W3 status with latest url package.


-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk