[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

mICQ packaging (was: Re: On the matter of Qt packaging)



[copy&paste...copying message IDs from followups...]

>--[Jordi Mallach]--<jordi@debian.org>
# On Thu, Feb 06, 2003 at 03:38:26PM +0100, Alexander Neumann wrote:

# > "You're using the mICQ package provided by Debian. Since the Debian
# > maintainer is extremely uncooperative, you're adviced to use the better
# > quality package from micq.org. Simply add the following line to your
# > /etc/apt/sources.list to track stable versions of mICQ:
# > deb http://www.micq.org/debian stable main
# I can also say that Micq upstream is not being nice either. I have a few
# nice mails from Rüdiger from the time when Micq was mostly unmaintained
# upstream and it didn't work at all due to the ICQ protocol change.

mICQ was never "unmaintained" upstream. Development was a bit slow, but
direct connections had been added recently, so while it still didn't do the
new protocol, communication was still possible, so it wasn't "unusable"
either. Also note that at the time mICQ was removed, Sander hadn't even
updated his Debian package to the newest version since quite a while,
and also note that licq wasn't usable as well (and was not removed), and
that there's still an ICQ client in Debian that doesn't work with the
current protocol (everybuddy #80361, everybuddy #86838, #84323, ...)
and hasn't been removed.

# Sander decided to remove the package from Debian entirely, and Rüdiger
# basically insulted us all for having done so.

There was no reason to do so, because it still worked somewhat, and I wasn't
even told by the maintainer (who was on the mICQ mailing list) and only
found out accidently when the bug reports mysteriously disappeared. Good
reasons to be pissed.

# On Mon, 4 Mar 2002 13:11:03 +0100, he wrote

"he" means me... nice precise attribution line.

# > Actually this is the worst mismaintainment I've seen so far
# Heh, so [...] Sander is the worst maintainer in the world [...]

In opposite to you, I can differentiate between doer and deed.

# > after having several bugs closed that weren't fixed (like KDE still
# > having trouble with the € - the maintainer didn't see the light and just
(fixing your quote)
# > closed it (#122533)
# [...] and DanielS suck too.

The shame here is that on a virgin Debian woody KDE install, the €uro
won't work out-of-the box, you have to fiddle with a configuration file
that will constantly be overwritten (against policy) by locales with the
wrong values, ...

# > while the locales maintainer considered a simplistic fix (well,
# > workaround) for not feasible (#130259)).
# [...] and BenC [...] suck too. We're surrounded by incompetents!

... though making locales set the value that would make KDE2 do the right
thing was first accepted in 2002-01, but later removed silently without
reopening the bug report. About your sarcasm - "BenC", as you call him,
showed indeed some incompetence about locale issues...

# [1] In a mail from Rüdiger to Madkiss, "Jordi Mallach war nie Maintainer
# von mICQ. Auch nicht Sponsor, oder weiß der Geier was. Also erfind nicht
(fixing your quote again)
# irgendwelchen Schwachsinn.", which appears to translate to "Jordi
# Mallach was never Maintainer of mICQ. Nor he was Sponsor or whatever. So
# please don't invent stories like that."

To put this mail into perspective: First, it was written when I had given up
hope to have intelligent live at the other end; second, you need to know
that there was a time when I received every two weeks a mail from the Debian
mICQ maintainer of the day that he was the new maintainer, but none of them
except "madkiss" uploaded a package at all -- but Jordi was none of them;
and finally, this was an answer of his "you could have brought your
packaging into Debian with Jordi" after a lengthy discussion in which he
didn't understand the difference between upstream and Debian packaging, but
ignored all critiques to his packaging.

>--[Eduard Bloch]--<edi@gmx.de>
# Rico -mc- Gloeckner schrieb am Thursday, den 06. February 2003:
 
# > Ruediger knows exactly what a Maintainer is, despite your wrong (or
# > misunderstandable) Translation.
# Maybe, and he tries to become a developer.

With packaging mICQ being the most important reason.

# > When a DD experiences Problems with an Upstream Author, then he should
# > probably ask more experienced DDs for Support, not ignore the Upstream.
# When a wanna-be-DD sees an concurent wanna-be-DD and throws stones in
# his way using the upstream package, I cannot see anything good in this
# behaviour.

Actually I couldn't care less whether he is already a DD or not. I do care
about the package he creates. Considering that I sent that application after
the differences between me and "madkiss" became unsolvable, this is somewhat
far fetched. Just remember that I'm the one who still sometimes gets complains
about #167610.

# And he can complain loudly about bad packaging as Ralf done,
# there is no real reason for such childish private wars.

From those 3 of the more important stuff that I eventually reported, two are
still unresolved in woody. But I can flood it with minor bug reports, if you
want. It's just that a private mail to the maintainer _should_ be less work
if it doesn't need to be fixed immediately.

# > I dont know if Ruediger agress with this, but maybe someone could check
# > Ruedigers debs and if they are approriate be sponsor for Ruedigers micq
# > debs - this seems to be a good solution to me.
# ACK.

Actually I tried to sync my packaging with the Debian package as much as
possible by adopting all sensible changes made (so I don't claim it to be
genuinly mine packaging). Having a look at the diff that "madkiss" generated
is rather telling:

(http://ftp.debian.org/debian/pool/main/m/micq/micq_0.4.10.1-1.diff.gz)
* dropping FAQ, AUTHORS and README from the package (his debian/copyright even
  refers to the README).
* (unnecessary reordering of debian/docs)
* dropping of mICQ's logo menu entry
* (rewriting it's description; I don't care about that one as I'll some day write
  a better description anyway)
* (90% unnecessary changes in white spaces in debian /rules)
* the removal of the -O4 option - if you really want to triple its size...
* [a change in DEB_BUILD_OPTIONS, which is probably correct]
* the removal of the "Debian" extra version marker -- I FREAKIN' NEED AS MUCH
  EXTRA VERSION AS POSSIBLE IF A USER COMES UP WITH A PROBLEM, SET IT TO
  "Debian" BUT JUST DON'T REMOVE IT, THANK YOU. Mandrake did manage to get
  that right, so Debian should be, too.
* [replacing of two upstream changelog entries with his own, informationless
  ones (they could be better, but ok)]
* (adding lots of trailing spaces to the changelog - we don't have anything
  better to do, do we?)
* (some changes in debian/copyright - we just have to do it differently.
  I'm not sure what the policy currently says about encoding, so transcoding
  it from UTF8 for the official package is probably okay)

So his package has three things left out, a huge pile of totally useless
changes, and a few changes that might actually be usefull. Add to that a
trivial to fix bug in woody that's so annoying that you don't want to use it
and a copyright violation as well in woody, and, just for the fun of it, a
Debian security alert that's so insulting to call me "an upstream developer"
when I wrote 90% of current mICQ, then you know why I am less than happy
about Debian.

Yours, Rüdiger (mICQ author, former hoster of an m68k buildd).

-- 
         100 DM =  51  € 13 ¢.
         100  € = 195 DM 58 pf.
  mailto:ruediger@ruediger-kuhlmann.de
    http://www.ruediger-kuhlmann.de/

Attachment: pgpTeQm4qy2Fp.pgp
Description: PGP signature


Reply to: