Gnus package (was: xemacs vs emacs)

Gnus package (was: xemacs vs emacs)

[Reiner Steib]

On Sat, Apr 05 2008, Mike Kupfer wrote:

>>>>>> "Reiner" == Reiner Steib <reinersteib+gmane@imap.cc> writes:
>
> Reiner> But even the versions in "pre-release" are outdated.  Gnus
> Reiner> 5.10.8 has been released on 2006-04-11.  We will release 5.10.10
> Reiner> soon.  Emacs 22.2 already contains the code corresponding to
> Reiner> this version.
> ...
> Reiner> I don't know which versions are in the stable XEmacs packages.
>
> The stable gnus package is 5.10.8 plus a patch.  The pre-release version
> is the stable version plus an additional two patches.

Now that you mention it, I remember that 5.10.8 already was in the
stable XEmacs packages.  Reading "5.10.8" in Norbert's pre-release
announcement, I incorrectly assumed that the stable package must be
older (with regards to the upstream version).  Sorry for the confusion
and thanks for clarifying.  Please accept my apologies for getting it
all wrong.

Are the patches you mention already integrated upstream in the v5-10
branch?

> As for it being outdated, well, hey, 5.10.8 is the latest stable
> release, isn't it?  When 5.10.10 is released, I'll start work to update
> the XEmacs package.

Thank you.

If you sync from v5-10, we could add a tag (we already have
"xemacs-pkg-1_90") to Gnus repository so the Gnus maintainers can find
out which code corresponds to which XEmacs package version (when
cherry-picking certain patches, tagging probably doesn't make much
sense).  So the answer to your question [1] is "yes".  (I found [1] by
searching for this tag.  If you have questions/information to/for the
Gnus maintainers/users, please send them to ding@gnus.org as well (I
don't follow xemacs list regularly).  E.g. announcing the availability
of new Gnus XEmacs (pre-release) package on ding and/or gnu.emacs.gnus
is fine with me.

Bye, Reiner.

[1]
,----[ http://thread.gmane.org/gmane.emacs.xemacs.patches/9266/focus=9274 ]
| From: Mike Kupfer
| [...]
| By the way, the previous maintainer of the XEmacs Gnus package (Steve
| Youngs) mentioned to me that he would tag the Gnus CVS repository to
| indicate XEmacs packages.  For example, gnus-agent.el has a
| xemacs-pkg-1_90 tag.  Is that a practice you (the Gnus maintainers)
| would like me to continue?
`----
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Gnus package (was: xemacs vs emacs)

[Mike Kupfer]

>>>>> "Reiner" == Reiner Steib <reinersteib+gmane@imap.cc> writes:

Reiner> Please accept my apologies for getting it all wrong.

You're forgiven. :-)  Thanks for all your work on Gnus!

Reiner> Are the patches you mention already integrated upstream in the
Reiner> v5-10 branch?

Ah, good question.  It doesn't look like this one is:

2007-03-05  Mike Kupfer  <mike.kupfer@xemacs.org>

	* lisp/gnus-spec.el (gnus-parse-simple-format): Add required 2nd
	  argument to (setq dontinsert), using the fix from the HEAD
	  branch at gnus.org. 

I don't see this one, either.  I don't know if Aidan has approached
someone on the Gnus team about it yet.

2007-12-24  Aidan Kehoe  <kehoea@parhasard.net>

	* lisp/gnus-sum.el:
	* lisp/gnus-sum.el (put-display-table): New.
	* lisp/gnus-sum.el (get-display-table): New.
	Provide with #'defun-when-void, so as to not override the 21.5
	implementation. 
	* lisp/gnus-sum.el (gnus-summary-set-display-table):
	* lisp/gnus-xmas.el (gnus-xmas-summary-set-display-table):
	Use #'put-display-table, not #'aref, to deal with the case where
	the display table is a char table and not a vector. 

I didn't look in CVS, but Katsumi Yamaoka said the third patch (or
something equivalent) is in the v5-10 branch (see
http://calypso.tux.org/pipermail/xemacs-patches/2008-March/001531.html).

MikeK> When 5.10.10 is released, I'll start work to update the XEmacs
MikeK> package.

Reiner> Thank you.

My pleasure.  Though I keep forgetting about that pesky issue of
compatible GPL versions.  That could delay things. :-(

Reiner> So the answer to your question [1] is "yes".

Okay, great!  I guess this means I ought to get on the ding list.
http://gnus.org/distribution.html says I should mail Lars to get write
access to CVS; is that still the right thing for me to do?

Reiner> E.g. announcing the availability of new Gnus XEmacs
Reiner> (pre-release) package on ding and/or gnu.emacs.gnus is fine with
Reiner> me.

Good to know; thanks.

mike

Re: Gnus package

[Reiner Steib]

On Sat, Apr 05 2008, Mike Kupfer wrote:

>>>>>> "Reiner" == Reiner Steib <reinersteib+gmane@imap.cc> writes:
> Reiner> Are the patches you mention already integrated upstream in the
> Reiner> v5-10 branch?
>
> Ah, good question.  It doesn't look like this one is:
>
> 2007-03-05  Mike Kupfer  <mike.kupfer@xemacs.org>
>
> 	* lisp/gnus-spec.el (gnus-parse-simple-format): Add required 2nd
> 	  argument to (setq dontinsert), using the fix from the HEAD
> 	  branch at gnus.org. 

I suppose this is this patch?

2004-01-27  Jerry James  <james@xemacs.org>  (tiny change)

	* gnus-spec.el (gnus-parse-simple-format): Fix setq value
	omission.

--8<---------------cut here---------------start------------->8---
--- gnus-spec.el	5 Jan 2004 19:47:22 -0000	7.2
+++ gnus-spec.el	27 Jan 2004 05:55:53 -0000	7.3
@@ -665,7 +665,7 @@
        (list (car flist)))
       ;; A single number.
       ((string= fstring "%d")
-       (setq dontinsert)
+       (setq dontinsert t)
        (if insert
 	   (list `(princ ,(car flist)))
 	 (list `(int-to-string ,(car flist)))))
--8<---------------cut here---------------end--------------->8---

Hm, (setq dontinsert) is the same as (setq dontinsert nil) in Emacs.
OTOH, setting dontinsert to nil is a no-op here.

> I don't see this one, either.  I don't know if Aidan has approached
> someone on the Gnus team about it yet.

No.  In the form discussed in
<http://thread.gmane.org/gmane.emacs.xemacs.patches/8878> it's not
suitable for inclusion since it would break on Emacs.

> I didn't look in CVS, but Katsumi Yamaoka said the third patch (or
> something equivalent) is in the v5-10 branch (see
> http://calypso.tux.org/pipermail/xemacs-patches/2008-March/001531.html).

I don't know how you solved the problem in XEmacs, but I'd suggest to
check if you can use Katsumi Yamaoka's approach instead of a different
one.

> Though I keep forgetting about that pesky issue of compatible GPL
> versions.  That could delay things. :-(

Didn't you (XEmacs) decide to switch to GPLv3 as well?  I seem to
recall Steven wrote something like this recently.

> Reiner> So the answer to your question [1] is "yes".
>
> Okay, great!  I guess this means I ought to get on the ding list.

You may read the list via Gmane (gmane.emacs.gnus.general).
Non-subscriber postings are subject to moderation.

> http://gnus.org/distribution.html says I should mail Lars to get
> write access to CVS; is that still the right thing for me to do?

For getting write access, it's the right thing to write to Lars.  If
you just want to add a tag, it might not be worth the effort.  Someone
else (e.g. me) could do it on your request.  On a second thought, if
you only sync with upstream releases, we don't really need
xemacs-pkg-* tags because it would be the same as the v5-10-* release
tag.

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Gnus package

[Mike Kupfer]

>>>>> "Reiner" == Reiner Steib <reinersteib+gmane@imap.cc> writes:

Reiner> I suppose this is this patch?

Reiner> 2004-01-27 Jerry James <james@xemacs.org> (tiny change)

Reiner> * gnus-spec.el (gnus-parse-simple-format): Fix setq value
Reiner> 	omission.

Right (sorry for not providing a more precise reference).

Reiner> Hm, (setq dontinsert) is the same as (setq dontinsert nil) in
Reiner> Emacs.  OTOH, setting dontinsert to nil is a no-op here.

Hmm.  I didn't try to verify correctness, so I can't comment on what the
right value should be.

MikeK> I don't see this one, either.  I don't know if Aidan has approached
MikeK> someone on the Gnus team about it yet.

Reiner> No.  In the form discussed in
Reiner> <http://thread.gmane.org/gmane.emacs.xemacs.patches/8878> it's
Reiner> not suitable for inclusion since it would break on Emacs.

Okay.  I don't have time to pursue this right now, but I've added a note
to myself about it.  Or maybe Aidan will get to it first.

MikeK> I didn't look in CVS, but Katsumi Yamaoka said the third patch (or
MikeK> something equivalent) is in the v5-10 branch (see
MikeK> http://calypso.tux.org/pipermail/xemacs-patches/2008-March/001531.html).

Reiner> I don't know how you solved the problem in XEmacs, but I'd
Reiner> suggest to check if you can use Katsumi Yamaoka's approach
Reiner> instead of a different one.

Yes, I'm planning to look into that when I do update to 5.10.10.  (My
goal for all the packages I maintain is zero local patches.)

MikeK> Though I keep forgetting about that pesky issue of compatible GPL
MikeK> versions.  That could delay things. :-(

Reiner> Didn't you (XEmacs) decide to switch to GPLv3 as well?  I seem
Reiner> to recall Steven wrote something like this recently.

I've seen discussion of it, but I'll have to let Stephen explain what
the current status is.

Reiner> On a second thought, if you only sync with upstream releases, we
Reiner> don't really need xemacs-pkg-* tags because it would be the same
Reiner> as the v5-10-* release tag.

Okay.  If for some reason I need to sync with a non-release snapshot,
I'll mail ding to work out getting it tagged appropriately.

cheers,
mike