strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Juan LEON Lahoz Garcia]

Hi,

It seems that, (I supposse) due some error, a emacs 21.4 is in
ftp.gnu.org/pub/gnu/emacs/.

The file seems like a 21.3 version, but some sites (like distrowatch)
that uses script to check latest versions are reporting 21.4 as out.

Am I wrong?

Regards
juanleon

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Romain Francoise]

Juan LEON Lahoz Garcia <juanleon1@gmail.com> writes:

> It seems that, (I supposse) due some error, a emacs 21.4 is in
> ftp.gnu.org/pub/gnu/emacs/.

See the diff, Emacs 21.4 addresses a security vulnerability in movemail
(CAN-2005-0100) and is, apart from this fix, identical to version 21.3.

-- 
Romain Francoise <romain@orebokech.com> | How long will it be before he
it's a miracle -- http://orebokech.com/ | sees you own his legs but his
                                        | mind is free?

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[David Kastrup]

Juan LEON Lahoz Garcia <juanleon1@gmail.com> writes:

> It seems that, (I supposse) due some error, a emacs 21.4 is in
> ftp.gnu.org/pub/gnu/emacs/.
>
> The file seems like a 21.3 version, but some sites (like distrowatch)
> that uses script to check latest versions are reporting 21.4 as out.
>
> Am I wrong?

Yes, you are.  according to the diffs, we have

diff -rc2 emacs-21.3/README emacs-21.4/README
*** emacs-21.3/README	Tue Mar 18 10:20:35 2003
--- emacs-21.4/README	Wed Feb  2 02:34:12 2005
***************
*** 1,4 ****
! This directory tree holds version 21.3 of GNU Emacs, the extensible,
! customizable, self-documenting real-time display editor.
  
  You may encounter bugs in this release.  If you do, please report
--- 1,7 ----
! This directory tree holds version 21.4 of GNU Emacs, the extensible,
! customizable, self-documenting real-time display editor.  This release
! was made to fix a security bug, so most directories have no changes
! from 21.3.  The documentation still identifies the version as 21.3,
! since it has not been altered.
  
  You may encounter bugs in this release.  If you do, please report

  
diff -rc2 emacs-21.3/lib-src/ChangeLog emacs-21.4/lib-src/ChangeLog
*** emacs-21.3/lib-src/ChangeLog	Tue Mar 18 10:15:30 2003
--- emacs-21.4/lib-src/ChangeLog	Sun Feb  6 04:09:17 2005
***************
*** 1,2 ****
--- 1,8 ----
+ 2005-02-06  Richard M. Stallman  <rms@gnu.org>
+ 
+ 	* Version 21.4 released.
+ 
+ 	* movemail.c (popmail): Fix call to `error'.
+ 
  2003-03-18  Francesco Potortì  <pot@gnu.org>
  
diff -rc2 emacs-21.3/lib-src/movemail.c emacs-21.4/lib-src/movemail.c
*** emacs-21.3/lib-src/movemail.c	Fri Feb 15 07:07:14 2002
--- emacs-21.4/lib-src/movemail.c	Wed Feb  2 02:31:17 2005
***************
*** 788,792 ****
        if (pop_retr (server, i, mbf) != OK)
  	{
! 	  error (Errmsg, 0, 0);
  	  close (mbfi);
  	  return (1);
--- 788,792 ----
        if (pop_retr (server, i, mbf) != OK)
  	{
! 	  error ("%s", Errmsg, 0);
  	  close (mbfi);
  	  return (1);


That's pretty much it.  Richard, you should have asked anybody about
this.  After long and painful discussions on this list, we had agreed
that the new version of Emacs with the changes of years in it will be
called 21.4.  People have relied on it in announcements ("to be
expected in Emacs-21.4").  There was a long discussion about whether
the next big release should be called 22.0 and finally we agreed that
21.4 it would be.

So this bug fix release should have been called 21.3.1 or 21.3-b or
whatever.

I could just cry.  Now, AGAIN, I have a whole slew of documentation,
help files, announcements and stuff like that rendered false and junk
and misleading.  This was the _exact_ reason that it was proposed that
22.0 would become the next big version number, and it was turned down
finally by everybody agreeing that we'd make it 21.4 instead.

And now this.  I can't say how much hairtearing this will mean.  Is
there any chance that we get this changed to 21.3.1 or 21.3-b or
something like that _before_ the news spreads?

Please, Richard?

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Juan LEON Lahoz Garcia]

On Mon, 07 Feb 2005 12:59:15 +0100, David Kastrup <dak@gnu.org> wrote:
> Juan LEON Lahoz Garcia <juanleon1@gmail.com> writes:
> 
> > It seems that, (I supposse) due some error, a emacs 21.4 is in
> > ftp.gnu.org/pub/gnu/emacs/.
> >
> > The file seems like a 21.3 version, but some sites (like distrowatch)
> > that uses script to check latest versions are reporting 21.4 as out.
> >
> > Am I wrong?
> 
> Yes, you are.  according to the diffs, we have
> 

I am sorry. I was confused by the fact that NEWS file was unchanged
and because I thank 21.4 was meant to contain all new development done
last years.

[ Anyway, at least I was very happy between the 1.5 minute interval
between I read the announce in distrowatch and I downloaded the new
version and opened the NEWS file ;-) ]

Regards
juanleon

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Kim F. Storm]

David Kastrup <dak@gnu.org> writes:

> I could just cry.  Now, AGAIN, I have a whole slew of documentation,
> help files, announcements and stuff like that rendered false and junk
> and misleading.  This was the _exact_ reason that it was proposed that
> 22.0 would become the next big version number, and it was turned down
> finally by everybody agreeing that we'd make it 21.4 instead.

I don't think it was turned down as such -- everybody just seemed to
have their own idea exactly what the new scheme should be, so no
consensus was reached.

> And now this.  I can't say how much hairtearing this will mean.  Is
> there any chance that we get this changed to 21.3.1 or 21.3-b or
> something like that _before_ the news spreads?
>
> Please, Richard?

Sadly, I think it is too late.

So it's about time to reopen the "change the numbering to 22.x" story.
Sigh.

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

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Stefan Monnier]

Am I the only one who finds himself very much surprised that even though
he's a fairly active contributor and reads emacs-devel, emacs-pretest-bug,
gnu.emacs.help, gnus.emacs.bug, gnu.emacs.announce, and a few more places,
he has not seen any announcement (let alone discussion) of this new release?


        Stefan

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Jan D.]

Stefan Monnier wrote:

>Am I the only one who finds himself very much surprised that even though
>he's a fairly active contributor and reads emacs-devel, emacs-pretest-bug,
>gnu.emacs.help, gnus.emacs.bug, gnu.emacs.announce, and a few more places,
>he has not seen any announcement (let alone discussion) of this new release?
>

No.

Does this mean that customize variables marked 21.4 in CVS will have to 
be changed (I suspect that is the case)?

    Jan D.

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[David Kastrup]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> Am I the only one who finds himself very much surprised that even
> though he's a fairly active contributor and reads emacs-devel,
> emacs-pretest-bug, gnu.emacs.help, gnus.emacs.bug,
> gnu.emacs.announce, and a few more places, he has not seen any
> announcement (let alone discussion) of this new release?

This question is rhetorical.  The answer is obvious.  A slightly more
interesting question would be whether there is more than only one who
is _not_ surprised.

Answering the question is of secondary interest to answering the
question how to deal with it.  Personally, I'd like renaming this
thing to 21.3a even though Kim has already stated he considers it
unlikely that this might be a good idea/possible after the fact has
been established.

It would not appear, however, that either of our opinions would have
any relevance, anyway, so we might just stop musing about such things.

Given that, I recommend that we bother about what lies within our
power, namely editing Emacs, as releasing it is not under our control.
I propose that we now, where another renaming of everything has
seemingly become necessary, decide to call the next major release
22.0, and also fix all corresponding documentation and :version
strings (which are now completely fouled up) to refer to 22.0 instead.

This will prevent
a) us having to change all version strings again for the next major
release more than once,
b) us and third part package providers from looking like utter idiots
when talking about release numbers, in case another security fix
intervenes, invalidating all previously made statements and assurances
again.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Stefan Monnier]

>> Am I the only one who finds himself very much surprised that even
>> though he's a fairly active contributor and reads emacs-devel,
>> emacs-pretest-bug, gnu.emacs.help, gnus.emacs.bug,
>> gnu.emacs.announce, and a few more places, he has not seen any
>> announcement (let alone discussion) of this new release?

> This question is rhetorical.

Partly.  But maybe I had missed the announcement/discussion.

> It would not appear, however, that either of our opinions would have
> any relevance, anyway, so we might just stop musing about such things.
[...]
> I propose that we now, where another renaming of everything has
> seemingly become necessary, decide to call the next major release

Your first two lines above should make it clear to you that the second two
lines would be a waste of time.


        Stefan

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[David Kastrup]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> Am I the only one who finds himself very much surprised that even
>>> though he's a fairly active contributor and reads emacs-devel,
>>> emacs-pretest-bug, gnu.emacs.help, gnus.emacs.bug,
>>> gnu.emacs.announce, and a few more places, he has not seen any
>>> announcement (let alone discussion) of this new release?
>
>> This question is rhetorical.
>
> Partly.  But maybe I had missed the announcement/discussion.
>
>> It would not appear, however, that either of our opinions would have
>> any relevance, anyway, so we might just stop musing about such things.
> [...]
>> I propose that we now, where another renaming of everything has
>> seemingly become necessary, decide to call the next major release
>
> Your first two lines above should make it clear to you that the
> second two lines would be a waste of time.

The work to rename versions throughout a release to arrive at a
consistent set is nontrivial and nothing that I expect a single person
to do in the time frame for an emergency release such as the latest
one we have seen.  And an emergency release is unlikely to occur from
the trunk.

Under the premise that 21.4 as a release number is here to stay,
version numbers all over the trunk will have to be changed, again.
IIRC, it was a work of several days for a single volunteer last time.
This work is not immediately tied to a release.  If the release
system/scripts don't accommodate emergency releases like 21.3a, as
established facts make it appear, then it would seem that we better
keep 21.5 off from both announcements and code, as the number might be
needed sooner than anticipated.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Stefan Monnier]

>>> It would not appear, however, that either of our opinions would have
>>> any relevance, anyway, so we might just stop musing about such things.
>> [...]
>>> I propose that we now, where another renaming of everything has
>>> seemingly become necessary, decide to call the next major release
>> 
>> Your first two lines above should make it clear to you that the
>> second two lines would be a waste of time.

Just to make it clear, I'm referring to the "decide" part when I talk about
a waste of time.


        Stefan

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Jérôme Marant]

David Kastrup <dak@gnu.org> writes:

> Under the premise that 21.4 as a release number is here to stay,
> version numbers all over the trunk will have to be changed, again.
> IIRC, it was a work of several days for a single volunteer last time.
> This work is not immediately tied to a release.  If the release
> system/scripts don't accommodate emergency releases like 21.3a, as
> established facts make it appear, then it would seem that we better
> keep 21.5 off from both announcements and code, as the number might be
> needed sooner than anticipated.

Since it is too late, it would be better to take the opportunity to
adopt a sane numbering scheme right now which would take emacs-devel
away from such troubles in the future, like numbering the next release
as 22.1 instead of 21.5 (saner and more consistant).

Cheers,

-- 
Jérôme Marant

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[David Kastrup]

Jérôme Marant <jmarant@free.fr> writes:

> David Kastrup <dak@gnu.org> writes:
>
>> Under the premise that 21.4 as a release number is here to stay,
>> version numbers all over the trunk will have to be changed, again.
>> IIRC, it was a work of several days for a single volunteer last
>> time.  This work is not immediately tied to a release.  If the
>> release system/scripts don't accommodate emergency releases like
>> 21.3a, as established facts make it appear, then it would seem that
>> we better keep 21.5 off from both announcements and code, as the
>> number might be needed sooner than anticipated.
>
> Since it is too late, it would be better to take the opportunity to
> adopt a sane numbering scheme right now which would take emacs-devel
> away from such troubles in the future, like numbering the next
> release as 22.1 instead of 21.5 (saner and more consistant).

Actually, that was supposed to be my proposal, even though I managed
not actually mentioning it in so many words.  Sorry for that.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Jérôme Marant]

Quoting David Kastrup <dak@gnu.org>:

> > Since it is too late, it would be better to take the opportunity to
> > adopt a sane numbering scheme right now which would take emacs-devel
> > away from such troubles in the future, like numbering the next
> > release as 22.1 instead of 21.5 (saner and more consistant).
>
> Actually, that was supposed to be my proposal, even though I managed
> not actually mentioning it in so many words.  Sorry for that.

Actually, it wasn't. Stefan already proposed this months ago when
we discussed about bugfix releases.

Cheers,

--
Jérôme Marant

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[David Kastrup]

[Following up to myself]

David Kastrup <dak@gnu.org> writes:

> Answering the question is of secondary interest to answering the
> question how to deal with it.  Personally, I'd like renaming this
> thing to 21.3a even though Kim has already stated he considers it
> unlikely that this might be a good idea/possible after the fact has
> been established.

[...]

> Given that, I recommend that we bother about what lies within our
> power, namely editing Emacs, as releasing it is not under our control.
> I propose that we now, where another renaming of everything has
> seemingly become necessary, decide to call the next major release
> 22.0, and also fix all corresponding documentation and :version
> strings (which are now completely fouled up) to refer to 22.0 instead.
>
> This will prevent
> a) us having to change all version strings again for the next major
> release more than once,
> b) us and third part package providers from looking like utter idiots
> when talking about release numbers, in case another security fix
> intervenes, invalidating all previously made statements and assurances
> again.

To underline that we would want to avoid having to do such a step
again, try a google search for "Emacs-21.4 -X-Emacs-21.4".  That are
about 8430 pages/articles talking about Emacs-21.4.  It is a safe bet
that at this point of time, pretty much none of them is talking about
what has been released as 21.4 by now.

Anybody volunteering for writing to the maintainers of those web pages
and articles?  I am still not convinced that the 21.3a idea would not
be less painful in the long run, even though at the moment there exist
signed, though unannounced, 21.4 Emacs packages.

But again: it's not our call to make.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Kim F. Storm]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> Am I the only one who finds himself very much surprised that even though
> he's a fairly active contributor and reads emacs-devel, emacs-pretest-bug,
> gnu.emacs.help, gnus.emacs.bug, gnu.emacs.announce, and a few more places,
> he has not seen any announcement (let alone discussion) of this new release?

I am equally surprised!!

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

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Date: Mon, 07 Feb 2005 10:05:32 -0500
> Cc: Juan LEON Lahoz Garcia <juanleon1@gmail.com>, emacs-devel@gnu.org
> 
> Am I the only one who finds himself very much surprised that even though
> he's a fairly active contributor and reads emacs-devel, emacs-pretest-bug,
> gnu.emacs.help, gnus.emacs.bug, gnu.emacs.announce, and a few more places,
> he has not seen any announcement (let alone discussion) of this new release?

I'm not surprised: it takes time for the announcement to get from
Richard's mail to gnu.org servers and from there to propagate to the
news groups.  On top of that, gnu.emacs.announce was in the past known
to have problems with announcements.

By contrast, a file upload to ftp.gnu.org is instanteneous, and also
is generally done _before_ the announcement goes out.

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Stefan Monnier]

>> Am I the only one who finds himself very much surprised that even though
>> he's a fairly active contributor and reads emacs-devel,
>> emacs-pretest-bug, gnu.emacs.help, gnus.emacs.bug, gnu.emacs.announce,
>> and a few more places, he has not seen any announcement (let alone
>> discussion) of this new release?

> I'm not surprised: it takes time for the announcement to get from
> Richard's mail to gnu.org servers and from there to propagate to the
> news groups.  On top of that, gnu.emacs.announce was in the past known
> to have problems with announcements.

Well, I'm still surprised that the only mention of this decision was only
sent to something like gnu.emacs.announce which is known to be at
best flaky.

I would have appreciated some warning *beforehand* in emacs-devel.
Just to ackowledge that we actually exist.  But I guess the method used
reflects the fact that we simply have no say whatsoever in those matters.


        Stefan

Re: strange emacs 21.4 on ftp.gnu.org:/pub/gnu/emacs

[Richard Stallman]

The people who reported the bug asked me not to discuss this publicly
before the bug was announced, so I could not discuss it here until
afterwards.  Using a name such as 21.3.1 or 21.3-b would have caused
problems, so that was not a real option.  I had no real alternative
except what I did.

I am sorry if it made someone feel bad, but I can only ask you not
to take it personally.