Debian Developers - 185 Packages that look orphaned

This is Interesting: Free IT Magazines  
Home > Archive > Debian Developers > August 2004 > 185 Packages that look orphaned





You are viewing an archived Text-only version of the thread. To view this thread in it's original format and/or if you want to reply to this thread please [click here]

Author 185 Packages that look orphaned
Goswin von Brederlow

2004-01-26, 11:34 am

Hi,

I looked through the differences between testing and unstable and
picked out everything older than 100 days. Reasons why those packages
are not in testing are:

- non-free / contrib packages nobody tried to compile
- FTBFS or RC bugs
- possibly failure of the testing script to detect it
- other packages hold you back (get involved in those other packages)


Noone has cared enough about these packages to get them compiled,
fixed or pushed into sarge so I am assuming the packages don't have a
caring maintainer or fan community. Ergo they should be orphaned.

If you maintain one of thses packages then tell me (including the
names of packages you maintain) during the next week. If you are using
one of these packages and could maintain (or NMU some fixes) you
should contact the maintainer and me to work things out. If I hear
nothing about a package soon I will start with the oldest and do a few
packages every day.

Package list follows.

MfG
Goswin

PS: Dear debain QA group, you own a few packages in the list and at
least one isn't shown as orphaned.

----------------------------------------------------------------------
3270

agrep allegro4 amiwm aplus-fsf argouml avifile axe

babel bass bluez-pan brahms bterm-unifont btools bulkmail

checkinstall cl-ftp cl-imho cl-odcl cl-uncommonsql clustalw
commons-beanutils

dchub devel-protocols dip distributed-net distributed-net-pproxy
drscheme

eglade eshell

free-java-sdk

ghemical gnews gnome-pim gnu-smalltalk gnumach gnumach1 gnuyahoo gpcl
gspy gtkam guile-oops gwydion-dylan

hunt

ibm-jdk1.1-installer inform

jasper jetty jikes-contrib

kernel-image-2.2.25-mac-udeb kernel-image-2.4.22-alpha-udeb
kernel-patch-2.4-kgdb kernel-patch-lpp kernel-patch-mppe kmusicdb
ksetisaver kvim

l2tpd lgeneral lgrind libcrypto++ libdbmusic libembperl-perl
libfilesystem-ruby libgd-perl libgnomeuimm1.3 libjcommon-java
libjfreechart-java libjfreereport-java libmail-cclient-perl libpcd
libpixie-java libsdl-erlang libstruts1.1-java limewire
linuxvideostudio lmbench logjam lsbappchk lsbdev lush lyx-cjk

mffm-fftw mga-vidsnacc mico microwindows mig mondo motv mozart
mozart-gtk mozart-stdlib mozilla-locale-zh-hk mozilla-snapshot mueller
muffin muse mush mysql++

netjuke netsaint-plugins nhc98 ntop

ohphone onshore-timesheet openh323 openmcu openmotif opie-applets-fb
opie-apps-fb opie-base-fb opie-comnet-fb opie-games-fb opie-i18n
opie-inputmethods-fb opie-multimedia-fb opie-pim-fb opie-settings-fb
opie-themeing-fb opie-tools-fb

palo-installer peacock pgi pgplot5 php4-interbase php4-vpopmail pja
plib1.7 pose-skins povray pwlib pydb python-fixedpoint

qmail-pop3-sv qmail-qfilter qmail-svqmail-smtp-sv qps qvplay

s390-dasd s390-netdevice satan saxon-catalog scalemail sip-qt3
smarteiffel soap4r spim spip-eva statcvs strategoxt svtools
swh-plugins swig sxid syscalltrack sysklogd

tempest thoughttracker trafstats tse3 tuxracer

units-filter

vcr vdkbuilder vflib3 vic vipec vlad vmailmgr vpopmail

w3-el wims wims-modules-fr wmsensors

xdelta2 xdrawchem xephem xfce4 xfce4-themes xffm4-icons
xfonts-greek-ph xshodo

yank

zangband zblast


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Brett Cundal

2004-01-26, 12:34 pm

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:

[snip]
quote:

> Noone has cared enough about these packages to get them compiled,
> fixed or pushed into sarge so I am assuming the packages don't have a
> caring maintainer or fan community. Ergo they should be orphaned.



[snip]
quote:

> ghemical gnews gnome-pim gnu-smalltalk gnumach gnumach1 gnuyahoo gpcl
> gspy gtkam guile-oops gwydion-dylan



I maintain gnu-smalltalk, but noone seems interested in sponsoring an
upload for me. My regular sponsor has not been able to do an upload in
ages. Requests for a new sponsor have gone unanswered.

There is interest in the package... I've been contacted by several
people regarding the status of this package, but unfortunately none of
them were DDs.

If anyone *is* interested in doing an upload, please contact me for
information.

-- Brett

Chuan-kai Lin

2004-01-26, 12:34 pm

Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> wrote:
quote:

> nhc98



Well, the only reason why nhc98 has FTBFS bugs is that it does not
support 64-bit architectures yet: this is a known limitation [1],
and the offending ia64 and alpha architectures have been dropped from
the control file.

I would say that nhc98 is in good hands (and in good shape too).

[1] http://www.cs.york.ac.uk/fp/nhc98/install.html

--
Chuan-kai Lin
http://www.cse.ogi.edu/~linchuan/


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Wouter Verhelst

2004-01-26, 12:34 pm

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> Hi,
>
> I looked through the differences between testing and unstable and
> picked out everything older than 100 days. Reasons why those packages
> are not in testing are:
>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)



You should really have looked through the list before spamming such a
huge amount of people.
quote:

> kernel-image-2.2.25-mac-udeb



This one is made to make d-i possible for m68k. However, as m68k doesn't
have a usable debian-installer yet (which I'm sure you know), this can't
go in yet.

--
Wouter Verhelst
Debian GNU/Linux -- http://www.debian.org
Nederlandstalige Linux-documentatie -- http://nl.linux.org
"Stop breathing down my neck." "My breathing is merely a simulation."
"So is my neck, stop it anyway!"
-- Voyager's EMH versus the Prometheus' EMH, stardate 51462.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Ben Pfaff

2004-01-26, 2:33 pm

Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> writes:
quote:

> I looked through the differences between testing and unstable and
> picked out everything older than 100 days. Reasons why those packages
> are not in testing are:
>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)



[...]
quote:

> w3-el



w3-el was supposed to be removed from unstable. Anyone know why
it's still showing up in the Sources.gz for unstable?
--
Ben Pfaff
email: blp@cs.stanford.edu
web: http://benpfaff.org


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Martin Michlmayr

2004-01-26, 2:33 pm

reassign 227738 ftp.debian.org
thanks

* Ben Pfaff <blp@cs.stanford.edu> [2004-01-26 19:44]:
quote:

>
> w3-el was supposed to be removed from unstable. Anyone know why
> it's still showing up in the Sources.gz for unstable?



It hasn't been removed from the archive yet. I was waiting for the
maintainer of eshell, which is in the same situation as w3-el (depends
on emacs20), to confirm. However, since it's a fairly obvious removal
request, let's go ahead now.
--
Martin Michlmayr
tbm@cyrius.com


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-26, 3:34 pm

Wouter Verhelst <wouter@grep.be> writes:

You didn't have to CC all 150 DDs for this.
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
>
> You should really have looked through the list before spamming such a
> huge amount of people.
>
>
> This one is made to make d-i possible for m68k. However, as m68k doesn't
> have a usable debian-installer yet (which I'm sure you know), this can't
> go in yet.



Why can't that go into sarge? Ok it wouldn't be much use there now but
nothing prevents it from being there. And when d-i support for macs is
added it doesn't have to be pushed in then.

Last I heard kernel 2.6 support for mac was being worked on and 2.4 is
getting skipped alltogether. I don't expect any newer udeb to replace
this any time soon.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-26, 3:34 pm

Brett Cundal <bcundal@cundal.net> writes:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
>
> [snip]
>
>
> [snip]
>
>
> I maintain gnu-smalltalk, but noone seems interested in sponsoring an
> upload for me. My regular sponsor has not been able to do an upload in
> ages. Requests for a new sponsor have gone unanswered.
>
> There is interest in the package... I've been contacted by several
> people regarding the status of this package, but unfortunately none of
> them were DDs.
>
> If anyone *is* interested in doing an upload, please contact me for
> information.
>
> -- Brett



Thanks for the quick reply.

On the off chance I come accross any DD with smalltalk knowledge I
will point him your way.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-26, 3:34 pm

Thanks for your quick reply.

Anand Kumria <wildfire@progsoc.org> writes:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
>
> Sigh. No wonder people dislike you.
>
>
> No, if you going to send out this kind of email you need to look more
> deeply at the problem. swh-plugins requires fftw to work. That doesn't
> because GCC 3.3.3 ICEs on m68k. The fix is in gcc 3.3.4 but you, no
> doubt, already did the work to discover this rather than wasting my
> time.



Actually I took the time to compile swh-plugins a few times on m68k
myself a while back. Sorry I overlooked that in the generated list, I
caught a few others.
quote:

> Feel free to orphan/nmu gcc, orphan/nmu fftw3 and then nmu (but not
> orphan) swh-plugins. I mean, you _did_ check to see that the package in



Knowing your still around and following the problem is enough. No
orphaning needed.
quote:

> testing versus the package in unstable has significant functionality
> differences to warrant all this, right?



Never saw that stated as a requirement for sarge.
quote:

> It is nice that you did take the time to look at this problem but I
> think, in future, you'll have a greater degree of success if you email
> people seperately and then do things in public if they don't respond.



Then I couldn't look for the maintainer and for other DDs using the
package at the same time. Twice the waiting period.
quote:

> Anand



MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-26, 3:34 pm

Nathanael Nerode <> writes:
quote:

> Don't be too hard on these; getting them compiled on all arches has
> been next to impossible for quite a while. :-/ Frankly, the
> practical advice I'd give is to stop making these Architecture: all.
> :-(



Maybe we could make a list of packages that should be autobuild in
contrib/non-free and distribute that to the buildds.

Or rather make a list of packages not to be build and allow
contrib/non-free in general, since the buildds have that mechanism
already.
quote:

> Don't include packages with "Keep this out of testing" RC bugs.



Too lazy to look at every package, sorry. I don't expect experimental
packages to be >100 days old without a new upload. I will detect them
before acting on orphaning them.
quote:

> Don't be too hard on these; people are often unaware of their dependency
> chains; and also some packages have been stuck in four or more different
> dependency chains, which can be a real pain in the neck.



Well, now they are aware that its a problem.
quote:

> Minus the situations I commented on above.
>
> <snip>
> I strongly advise starting by only orphaning those packages with RC bugs open
> longer than a week (and without "keep this out of testing" bugs). Please
> give the others a break for now, as some of them may actually be maintained,
> and at any rate they deserve closer scrutiny.



All packages listed are between 1000 and 100 days old. I suspect any
bugs on those are roughly the same age. A fresher bug means there is
at least one person out there using the package, don't worry there.
quote:

> You'll have *plenty* to orphan for quite a while if you just do these, which
> are more definitely unmaintained than the others.
>
> --Nathanael



Plenty of packages to pick from.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Nano Nano

2004-01-26, 4:34 pm

On Tue, Jan 27, 2004 at 04:24:23AM +0100, Goswin von Brederlow wrote:[QUOTE][color=darkred]
> Anand Kumria <wildfire@progsoc.org> writes:

How much faster could Debian releases occur if fewer archs. were
supported? Has a cost/benefit ratio been established? (Bug counts v.
installed base).


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Anthony Towns

2004-01-26, 5:34 pm

On Mon, Jan 26, 2004 at 08:38:59PM -0800, Nano Nano wrote:
quote:

> On Tue, Jan 27, 2004 at 04:24:23AM +0100, Goswin von Brederlow wrote:
> How much faster could Debian releases occur if fewer archs. were
> supported?



Until we have a d-i that works well on at least i386 and powerpc, the
answer is "negligibly". (Probably after then too, IME)

Cheers,
aj

--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.

Linux.conf.au 2004 -- Because we could.
http://conf.linux.org.au/ -- Jan 12-17, 2004

Andreas Tille

2004-01-26, 6:34 pm

On Tue, 27 Jan 2004, Goswin von Brederlow wrote:
quote:

> Maybe we could make a list of packages that should be autobuild in
> contrib/non-free and distribute that to the buildds.


This reasonable approach was discussed / suggested several times
(just search the archive) and ended always in: It's free. Would you
volunteer to do it? I would love if someone would grab this hut.
quote:

> Or rather make a list of packages not to be build and allow
> contrib/non-free in general, since the buildds have that mechanism
> already.


This would also work but this is kind of politics which concerns non-free.
So if we try to go with the positive list approach above some people might
be more happy ...

Kind regards

Andreas.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Stefan Frank

2004-01-26, 6:34 pm

Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> writes:
quote:

> If you maintain one of thses packages then tell me (including the
> names of packages you maintain) during the next week. If you are using
> one of these packages and could maintain (or NMU some fixes) you
> should contact the maintainer and me to work things out. If I hear
> nothing about a package soon I will start with the oldest and do a few
> packages every day.
>



quote:

> netjuke



On 12/21 i sent an RFS for netjuke to debian-mentors without success
after i contacted the Maintainer on 11/29.

The Maintainer then replied to me on 12/27 telling me he's waiting for a
his new key to be added to the keyring. As i didn't heard anything from
him again, i mailed him again on 01/24 asking about his plans about his
package. So it seems he's not MIA completely. I hope he'll incorporate
some of the changes i made into his package and upload a new
version 'soon'.

For anyone interested, an updated package is available from:
http://home.arcor.de/sfr/debian/

This package fixes most of the bugs filed in the bts afaict.
But as i'm not a developer there are probably still packaging issues
to be resolved.

ciao, Stefan


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Wouter Verhelst

2004-01-26, 6:34 pm

On Tue, Jan 27, 2004 at 05:01:12AM +0100, Goswin von Brederlow wrote:
quote:

> Wouter Verhelst <wouter@grep.be> writes:
>
> You didn't have to CC all 150 DDs for this.



Uh. Whoops.

[...]
quote:

>
> Why can't that go into sarge? Ok it wouldn't be much use there now but
> nothing prevents it from being there. And when d-i support for macs is
> added it doesn't have to be pushed in then.
>
> Last I heard kernel 2.6 support for mac was being worked on and 2.4 is
> getting skipped alltogether. I don't expect any newer udeb to replace
> this any time soon.



No, but for the time being they're required as neither 2.6 nor 2.4 works
on macs currently. Yes, it's being worked on, but we can't guarantee
it'll be available in time for sarge. As long as that isn't the case, we
still need 2.2

--
Wouter Verhelst
Debian GNU/Linux -- http://www.debian.org
Nederlandstalige Linux-documentatie -- http://nl.linux.org
"Stop breathing down my neck." "My breathing is merely a simulation."
"So is my neck, stop it anyway!"
-- Voyager's EMH versus the Prometheus' EMH, stardate 51462.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Petter Reinholdtsen

2004-01-26, 8:33 pm

[Goswin von Brederlow]
quote:

> Why can't that go into sarge?



The scripts that update the package list in testing ignore udebs. So
a package which only produces udebs, do not propagate into testing.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Emanuele Rocca

2004-01-26, 9:33 pm

* On 27-01-04 - 00:58, Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> wrote:
quote:

> I looked through the differences between testing and unstable and
> picked out everything older than 100 days.


[...]
quote:

> PS: Dear debain QA group, you own a few packages in the list and at
> least one isn't shown as orphaned.


[...]
quote:

> qmail-pop3-sv qmail-qfilter qmail-svqmail-smtp-sv


I am working on a QA upload for qmail-qfilter.
Actually qmail-sv, qmail-smtp-sv and qmail-pop3-sv are not orphaned.

The problem here is that all these packages depends by qmail which
isn't present because it's not allowed to distribute it in compiled
form.

What about making these packages Recommend qmail-src, perhaps giving
additional informations in README.Debian?

This is what I thought to do with qmail-qfilter.

Cheers,
ema
--
Emanuele Rocca | Lui che offrì la faccia al vento, la gola al vino e mai
1024D/EAF19B60 | un pensiero, non al denaro non all'amore nè al cielo...

Ian Lynagh

2004-01-26, 10:33 pm


I'm sure I've seen this said WRT similar lists, but listing maintainers
next to packages would make things an awful lot easier.

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

>
> nhc98



Was waiting for bug 226508 (on ftp.debian.org) to be acted on after
which it should have sailed into testing. Unfortunately, from the bug
close message, it looks like there was some confusion and it has been
completely removed rather than just the IA64 deb. I'll do another upload
if it disappears/has disappeared completely.


Thanks
Ian


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Nathanael Nerode

2004-01-26, 10:33 pm

>I looked through the differences between testing and unstable and
quote:

>picked out everything older than 100 days. Reasons why those packages
>are not in testing are:
>
>- non-free / contrib packages nobody tried to compile


Don't be too hard on these; getting them compiled on all arches has been
next to impossible for quite a while. :-/ Frankly, the practical advice I'd
give is to stop making these Architecture: all. :-(
quote:

>- FTBFS or RC bugs


Don't include packages with "Keep this out of testing" RC bugs.
quote:

>- possibly failure of the testing script to detect it
>- other packages hold you back (get involved in those other packages)


Don't be too hard on these; people are often unaware of their dependency
chains; and also some packages have been stuck in four or more different
dependency chains, which can be a real pain in the neck.
quote:

>Noone has cared enough about these packages to get them compiled,
>fixed or pushed into sarge so I am assuming the packages don't have a
>caring maintainer or fan community. Ergo they should be orphaned.


Minus the situations I commented on above.

<snip>
quote:

>If I hear nothing about a package soon I will start with the oldest and do a
>few packages every day.


I strongly advise starting by only orphaning those packages with RC bugs open
longer than a week (and without "keep this out of testing" bugs). Please
give the others a break for now, as some of them may actually be maintained,
and at any rate they deserve closer scrutiny.

You'll have *plenty* to orphan for quite a while if you just do these, which
are more definitely unmaintained than the others.

--Nathanael


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Herbert Xu

2004-01-26, 10:33 pm

Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> wrote:
quote:

>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)



You better leave my packages alone.
--
Debian GNU/Linux 3.0 is out! ( http://www.debian.org/ )
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Anand Kumria

2004-01-26, 10:33 pm

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> Hi,



Sigh. No wonder people dislike you.
quote:

> I looked through the differences between testing and unstable and
> picked out everything older than 100 days. Reasons why those packages
> are not in testing are:
>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)
>
>
> Noone has cared enough about these packages to get them compiled,
> fixed or pushed into sarge so I am assuming the packages don't have a
> caring maintainer or fan community. Ergo they should be orphaned.



No, if you going to send out this kind of email you need to look more
deeply at the problem. swh-plugins requires fftw to work. That doesn't
because GCC 3.3.3 ICEs on m68k. The fix is in gcc 3.3.4 but you, no
doubt, already did the work to discover this rather than wasting my
time.

Feel free to orphan/nmu gcc, orphan/nmu fftw3 and then nmu (but not
orphan) swh-plugins. I mean, you _did_ check to see that the package in
testing versus the package in unstable has significant functionality
differences to warrant all this, right?

It is nice that you did take the time to look at this problem but I
think, in future, you'll have a greater degree of success if you email
people seperately and then do things in public if they don't respond.

Anand

--
`` We are shaped by our thoughts, we become what we think.
When the mind is pure, joy follows like a shadow that never
leaves. '' -- Buddha, The Dhammapada


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Petter Reinholdtsen

2004-01-26, 10:33 pm

[Herbert Xu]
quote:

> You better leave my packages alone.



I suspect that the proper way to make sure your packages are left
alone, is for you to make sure they propagate into testing as they
should. If they don't, the responsibility is yours if the packages
are orphaned.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Colin Watson

2004-01-26, 10:33 pm

On Tue, Jan 27, 2004 at 12:14:39PM +0100, Petter Reinholdtsen wrote:
quote:

> [Herbert Xu]
>
> I suspect that the proper way to make sure your packages are left
> alone, is for you to make sure they propagate into testing as they
> should. If they don't, the responsibility is yours if the packages
> are orphaned.



The responsibility is also on people orphaning packages to actually
check why they aren't propagating into testing, rather than operating
scattergun.

Cheers,

--
Colin Watson [cjwatson@flatline.org.uk]


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
H. S. Teoh

2004-01-26, 10:33 pm

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> Hi,
>
> I looked through the differences between testing and unstable and
> picked out everything older than 100 days. Reasons why those packages
> are not in testing are:
>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)


[snip]
quote:

> If you maintain one of thses packages then tell me (including the
> names of packages you maintain) during the next week. If you are using
> one of these packages and could maintain (or NMU some fixes) you
> should contact the maintainer and me to work things out. If I hear
> nothing about a package soon I will start with the oldest and do a few
> packages every day.



Please leave axe alone. I am trying to get it built for the architectures
it is out-of-date on.


T

--
What's a "hot crossed bun"? An angry rabbit.

Jeremy T. Bouse

2004-01-27, 12:34 am

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> Hi,
>
> I looked through the differences between testing and unstable and
> picked out everything older than 100 days. Reasons why those packages
> are not in testing are:
>
> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)
>
>
> Noone has cared enough about these packages to get them compiled,
> fixed or pushed into sarge so I am assuming the packages don't have a
> caring maintainer or fan community. Ergo they should be orphaned.
>
> If you maintain one of thses packages then tell me (including the
> names of packages you maintain) during the next week. If you are using
> one of these packages and could maintain (or NMU some fixes) you
> should contact the maintainer and me to work things out. If I hear
> nothing about a package soon I will start with the oldest and do a few
> packages every day.
>
> Package list follows.
>
> MfG
> Goswin
>
> PS: Dear debain QA group, you own a few packages in the list and at
> least one isn't shown as orphaned.
>
> gspy



This one is a case of being only for i386 so I don't even believe it
has a chance of making it into stable. Furthermore the release of 0.1.7
packaging is due to trying to find time to maintain the software itself
and try to fix the problems reported as upstream has not made any
release in over a year now (0.1.7 was released Nov 20, 2002) so I'm
trying to make the fixes myself along with other contributed patches. If
you have more time than me to deal with it, have at it as it'd give me
more time to prepare the settlement of my divorce. Otherwise I don't
feel this is a major hurdle for releasing Testing as stable, just remove
it and leave it in Unstable only.

Regards,
Jeremy T. Bouse

Wouter Verhelst

2004-01-27, 2:34 am

Op di 27-01-2004, om 04:08 schreef Goswin von Brederlow:
quote:

> Maybe we could make a list of packages that should be autobuild in
> contrib/non-free and distribute that to the buildds.



No, we couldn't. The only requirement for a package to go into non-free
is "Debian must be allowed to distribute it". That does not include
automated compilation (or even compilation, for that matter) or porting
the thing to a different architecture. Apart from that, many autobuilder
admins (including me) will simply not want to run an autobuilder
building non-free packages.

(No, that doesn't mean I won't build a non-free package if the
maintainer asks for it, which I have done in the past, but that's a
different question)
quote:

>
> Too lazy to look at every package, sorry.



That's OK, but then don't start pointing fingers. Thanks.

--
Wouter Verhelst
Debian GNU/Linux -- http://www.debian.org
Nederlandstalige Linux-documentatie -- http://nl.linux.org
Most people have two reasons for doing anything -- a good reason, and
the real reason

Bastian Blank

2004-01-27, 2:34 am

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> - non-free / contrib packages nobody tried to compile
> - FTBFS or RC bugs
> - possibly failure of the testing script to detect it
> - other packages hold you back (get involved in those other packages)


- they only builds udebs for non-releasable archs
quote:

> 3270



non-free but neccesary for any people which work on s/390.
quote:

> palo-installer
> s390-dasd s390-netdevice



udeb only

bastian

--
Emotions are alien to me. I'm a scientist.
-- Spock, "This Side of Paradise", stardate 3417.3

Andreas Barth

2004-01-27, 2:34 am

* Brett Cundal (bcundal@cundal.net) [040127 02:25]:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:


quote:

>
> [snip]
>
[QUOTE][color=darkred]
> I maintain gnu-smalltalk, but noone seems interested in sponsoring an
> upload for me. My regular sponsor has not been able to do an upload in
> ages. Requests for a new sponsor have gone unanswered.



I'd sponsor it, if you create an up-to-date version of this package
(that also closes all RC-bugs on it).


Cheers,
Andi
--
http://home.arcor.de/andreas-barth/
PGP 1024/89FB5CE5 DC F1 85 6D A6 45 9C 0F 3B BE F1 D0 C5 D1 D9 0C


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Andreas Tille

2004-01-27, 2:34 am

On Tue, 27 Jan 2004, Wouter Verhelst wrote:
quote:

> Op di 27-01-2004, om 04:08 schreef Goswin von Brederlow:
>
> No, we couldn't. The only requirement for a package to go into non-free
> is "Debian must be allowed to distribute it".


Well this is required. Since when does this requirement exclude the
following?
quote:

> That does not include
> automated compilation (or even compilation, for that matter) or porting
> the thing to a different architecture.


As I said in this thread there are packages where license is just
unclear like: "This is free software."
Why should this stuff not be allowed to be compiled on different
architectures?

Goswin was talking about a list of packages which can be compiled
manually on all architectures without licensing problems. Why should
occure licensing problems if this is done automatically?

Kind regards

Andreas.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Javier =?iso-8859-15?Q?Fern=E1ndez-Sanguino_Pe

2004-01-27, 5:34 am

On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
quote:

> Hi,
>
> I looked through the differences between testing and unstable and


(...)
quote:

> Noone has cared enough about these packages to get them compiled,
> fixed or pushed into sarge so I am assuming the packages don't have a
> caring maintainer or fan community. Ergo they should be orphaned.



Not having a caring maintainer is not equal to not being autobuilt. Given
the fact that for those non-free packages I maintain, some have been
autobuilt at _some_ point
quote:

> bass


Don't touch this unless you are willing to work on it, thanks
quote:

> lmbench


Ditto.
quote:

> satan


Ditto.

Now, I really don't remember seeing a mail on d-announce saying that
autobuilders would not build non-free packages, but maybe I've missed it.
Since I don't see in the Policy or Devel's reference [1] why/when this
changed I don't intend to play the game and start begging autobuilders to
do what they have previously done before. Now, this was discussed in -devel
[2], with no consensus AFAIK.

If you think that QA would do a better job at mantaining these packages
(which I don't) please do explain yourself. I find it funny that some think
that we should not waste inexpensive (autobuilders=machine) resources in
non-free packages but we are willing to waste expensive (QA=people)
resources in them.

Regards

Javi


[1]
http://www.debian.org/doc/developer...rter-automation
[2]
http://lists.debian.org/debian-deve...1/msg00270.html
And please don't tell me it shouldn't be documented because both the Policy
and Reference talk about non-free.

Andreas Barth

2004-01-27, 6:34 am

* [040127 12:10]:
quote:

> Don't include packages with "Keep this out of testing" RC bugs.



For this (and other reasons) I'd like some parsable way to detect
that, e.g. Subject starting with "keep out of testing".


Cheers,
Andi
--
http://home.arcor.de/andreas-barth/
PGP 1024/89FB5CE5 DC F1 85 6D A6 45 9C 0F 3B BE F1 D0 C5 D1 D9 0C


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Arnaud Vandyck

2004-01-27, 9:34 am

Goswin von Brederlow

2004-01-27, 10:34 am

Thanks for the quick reply.

Stefan Frank <sfr@gmx.net> writes:
quote:

> Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> writes:
>
> On 12/21 i sent an RFS for netjuke to debian-mentors without success
> after i contacted the Maintainer on 11/29.
>
> The Maintainer then replied to me on 12/27 telling me he's waiting for a
> his new key to be added to the keyring. As i didn't heard anything from
> him again, i mailed him again on 01/24 asking about his plans about his
> package. So it seems he's not MIA completely. I hope he'll incorporate
> some of the changes i made into his package and upload a new
> version 'soon'.



He promised to work on it soon, was even faster than you.

He is alive and kicking but just a bit busy now. I CCed him so he gets
the URL below again (in case he forgot or lost it).
quote:

> For anyone interested, an updated package is available from:
> http://home.arcor.de/sfr/debian/
>
> This package fixes most of the bugs filed in the bts afaict.
> But as i'm not a developer there are probably still packaging issues
> to be resolved.
>
> ciao, Stefan



MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 11:34 am

Petter Reinholdtsen <pere@hungry.com> writes:
quote:

> [Goswin von Brederlow]
>
> The scripts that update the package list in testing ignore udebs. So
> a package which only produces udebs, do not propagate into testing.



I ment: Whats the reason for not pushing it in manually so its there
when d-i needs it?

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 1:34 pm

Wouter Verhelst <wouter@grep.be> writes:
quote:

> Op di 27-01-2004, om 04:08 schreef Goswin von Brederlow:
>
> No, we couldn't. The only requirement for a package to go into non-free
> is "Debian must be allowed to distribute it". That does not include
> automated compilation (or even compilation, for that matter) or porting
> the thing to a different architecture. Apart from that, many autobuilder
> admins (including me) will simply not want to run an autobuilder
> building non-free packages.



How about contrib?

There are reasons for packages being in contrib or non-free which
don't make me "hate" them, like the deb will Depend on something
non-free. The list should only contain packages we can legally build
and don't violate our personel or Debian philosophy (too much).

I don't see a problem with building GPL software that depends on
something outside of debian for example. Or software with a software
patent on it (which is void where I am, thank god.).

If I stick a non-free package in build/REDO (e.g. if someone asks me
to build it for example) will the buildd reject it or does that work
fine?
quote:

> (No, that doesn't mean I won't build a non-free package if the
> maintainer asks for it, which I have done in the past, but that's a
> different question)
>
>
> That's OK, but then don't start pointing fingers. Thanks.



I will get them when I run through unclaimed packages in details. For
the list i just assumed that anything kept out of sarge for a reason
like being a unstable release or cvs snapshot would get updated more
frequently than 100 days. Whats the point of a cvs snapshot otherwise?
I know its not perfect, already removed 2 packages I previously didn't
spot.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 1:34 pm

Thanks for your quick reply.

Javier =?iso-8859-15?Q?Fern=E1ndez-Sanguino_Pe=F1a?= <jfs@computer.org> writes:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
> (...)
>
> Not having a caring maintainer is not equal to not being autobuilt. Given
> the fact that for those non-free packages I maintain, some have been
> autobuilt at _some_ point



Buildd admins and DDs have build them manually from time to time. "You
see that nice package and it not build for your arch? Lets build it,
try it out and upload it." And if they loose interest and dont build
future versions your stuck in sid.
quote:

> Don't touch this unless you are willing to work on it, thanks



# 414 days old (needed 10 days)
# out of date on alpha: bass (from 1.0.7-2)
# out of date on arm: bass (from 1.0.7-2)
# out of date on ia64: bass (from 1.0.7-2)
# out of date on powerpc: bass (from 1.0.7-2)
# out of date on s390: bass (from 1.0.7-2)
quote:

> Ditto.



# The current maintainer is looking for someone who can take over
maintenance of this package. If you have some interests in this
package, please consider taking it over. Alternatively you may want
to be co-maintainer in order to help the actual maintainer. Please
see bug number #216883 for more information.

Is that still true?

# 579 days old (needed 2 days)
# out of date on alpha: lmbench (from 2.0-patch2-2)
# out of date on mips: lmbench (from 2.0-patch2-2)
# out of date on mipsel: lmbench (from 2.0-patch2-2)
# out of date on s390: lmbench (from 2.0-patch2-3)
# out of date on sparc: lmbench (from 2.0-patch2-2)
quote:

> Ditto.



# 373 days old (needed 10 days)
# out of date on alpha: satan (from 1.1.1a-3)
# out of date on arm: satan (from 1.1.1a-3)
# out of date on ia64: satan (from 1.1.1a-2)
# out of date on s390: satan (from 1.1.1a-3)
# out of date on sparc: satan (from 1.1.1-18)
quote:

> Now, I really don't remember seeing a mail on d-announce saying that
> autobuilders would not build non-free packages, but maybe I've missed it.



It was mentioned in Bug number #216883 of yours so you must have
forgotten about it since last October. But a number of DDs are
surprised to hear that so your not alone not knowing.
quote:

> Since I don't see in the Policy or Devel's reference [1] why/when this
> changed I don't intend to play the game and start begging autobuilders to
> do what they have previously done before. Now, this was discussed in -devel
> [2], with no consensus AFAIK.
>
> If you think that QA would do a better job at mantaining these packages
> (which I don't) please do explain yourself. I find it funny that some think
> that we should not waste inexpensive (autobuilders=machine) resources in
> non-free packages but we are willing to waste expensive (QA=people)
> resources in them.



Some packages can't be autobuild for legal reasons. All of those are
in non-free. There is no flag in the source packages saying what can
be legally build and what not and no list is provided otherwise. Also
some people don't want to support non-free software.

I think what would work best currently is if you get a comaintainer
for every arch you can't build the package on or at least a steady
group of sponsors. If noone is willing to build the package you can
file a bug against ftp.debian.org to get obsolete versions removed.
quote:

> Regards
>
> Javi
>
>
> [1]
> http://www.debian.org/doc/developer...rter-automation
> [2]
> http://lists.debian.org/debian-deve...1/msg00270.html
> And please don't tell me it shouldn't be documented because both the Policy
> and Reference talk about non-free.



MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Arnaud Vandyck <avdyk@debian.org> writes:
quote:

> Sorry if it's already been replied, I'm not up-to-date...



No, your still on the "Thanks for the quick reply" list. Tomorrow I
might drop the 'quick'.
quote:

> Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> writes:
>
>
> Hi,
>
>
> About every Java packages, they are not going in contrib because most
> depends on j2sdk1.(3|4) so, they are in contrib. THEY ARE NOT BROKEN!
> They just need a non-free JDK to build or to run.



Which means they need manual intervention to get pushed into sarge. It
just needs to be done where appropriate. A bunch of the Java related
packages maintainer replied that they are waiting on sablevm or
kaffe. Maybe that will be ready for sarge.
quote:

>
> If you can make the one I'll point you, come one, go ahead!



?
quote:

>
> At the moment, argouml is 0.14. It's the latest stable release of the
> software and stuck in unstable because it needs j2sdk1.(3|4). If you
> wanna write some of the Swing classes to fill the 'classpath' project
> gaps, feel free to do it.



Then it should just be hinted into sarge.
quote:

>
> Isn't it marked REMOVE? I think we already have some discussions about
> it and IBM does not provide this jdk.



Noted, will check back on it.
quote:

>
> should be resolved



Needs to be hinted too, already noted.
quote:

>
> I think for some, it's a j2sdk problem and if not, they depends on java
> packages that are stuck by the non-free jdk.
>
>
> I'll be working on next week, I'm talking with Mark.
>
> But please. We do read the lists!.. Do not mail everyone! Also,



I know one DD not reading this list. The one living nearest to me is
all I'm gonna say.
quote:

> personally, I have a lot of work at the moment, but you can notice that
> I'm moving every packages to my debian email address. Only
> libgnujaxp-java (which I'm also working on upstream) and argouml do use
> my work address.



Thats fine. No hurry. You said you are around and aware of the
packages and thats enough.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Thanks for your quick reply.

Emanuele Rocca <ema@debian.org> writes:
quote:

> * On 27-01-04 - 00:58, Goswin von Brederlow <brederlo@informatik.uni-tuebingen.de> wrote:
> [...]
>
> [...]
>
> I am working on a QA upload for qmail-qfilter.
> Actually qmail-sv, qmail-smtp-sv and qmail-pop3-sv are not orphaned.
>
> The problem here is that all these packages depends by qmail which
> isn't present because it's not allowed to distribute it in compiled
> form.
>
> What about making these packages Recommend qmail-src, perhaps giving
> additional informations in README.Debian?
>
> This is what I thought to do with qmail-qfilter.
>
> Cheers,
> ema



Suggesting/Recommending qmail-src is a good idea.

Apart from that you have to manually request that the packages are
moved to sarge after the trial period. The testing script doesn't know
how to handle the qmail Depends.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Colin Watson <cjwatson@debian.org> writes:
quote:

> On Tue, Jan 27, 2004 at 12:14:39PM +0100, Petter Reinholdtsen wrote:
>
> The responsibility is also on people orphaning packages to actually
> check why they aren't propagating into testing, rather than operating
> scattergun.
>
> Cheers,
>
> --
> Colin Watson [cjwatson@flatline.org.uk]



As I will do before actually starting on any package. This was just a
first warning shot as an attempt to get as many packages out of the
list with a minimum of investigative work for everyone. 33% are
already cleared up.

Of cause it would help to know what packages I should leave alone. If
I forget about it and NMU one of your packages or start to orphan them
by accident or if you get a reminder note that you haven't yet replied
in a while please don't whine.

Herbert: I checked the kernel-* packages against you, anything else I
should better stay away from?

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Thanks for your quick reply.

"H. S. Teoh" <hsteoh@quickfur.ath.cx> writes:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
> [snip]
>
> Please leave axe alone. I am trying to get it built for the architectures
> it is out-of-date on.



Arm, IA64 and sparc buildd admins don't have the email addresses set
on www.buildd.net.

Gerhard Tonn <gt@debian.org> is running the s390 buildd. If you ask
him or Bastian Banks (waldi on irc) nicely they can help you build for
s390. There is also a s390 emulator you can run on your pc. It might
not be the fastest but still faster than looking for someone to build
the package for every upload.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Thanks for your quick reply.

Bastian Blank <waldi@debian.org> writes:
quote:

> On Tue, Jan 27, 2004 at 12:58:33AM +0100, Goswin von Brederlow wrote:
> - they only builds udebs for non-releasable archs
>
>
> non-free but neccesary for any people which work on s/390.



# out of date on s390: 3270-common, c3270, pr3287, s3270, tcl3270,
x3270 (from 3.2.17-2)

Strange that the arch needing it is one thats keeping it out of
sarge. If you ask mellum he can compile for alpha. Don't know
anyone with hppa myself.
quote:

>
> udeb only



Need manual pushing.
quote:

> bastian



MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Dan Shearer

2004-01-27, 2:34 pm

On Wed, Jan 28, 2004 at 04:33:18AM +0100, Goswin von Brederlow wrote:
quote:

> There is also a s390 emulator you can run on your pc. It might
> not be the fastest but still faster than looking for someone to build
> the package for every upload.



You are referring to Hercules (apt-get install hercules, see
http://www.conmicro.cx/hercules/) On a decent 32-bit Intel machine it is
as fast as any of us built packages a few years ago. It's not a compile
workload, but to give an example of how real-world it is you can run an
interpretive sql-backed website on the real Internet.

If you have problems installing Debian/s390 on Hercules and the Debian
s390 list doesn't respond then I'm happy to try to help.

--
Dan Shearer
dan@shearer.org


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 2:34 pm

Nano Nano <40119.nospam@comcast.net> writes:
quote:

> On Tue, Jan 27, 2004 at 04:24:23AM +0100, Goswin von Brederlow wrote:
>
> How much faster could Debian releases occur if fewer archs. were
> supported? Has a cost/benefit ratio been established? (Bug counts v.
> installed base).



Dropping some arch would probably speed up the Debian release by 1 day
or so, if it doesn't slow the release down because less people would
work on it.

If you look back what the porters have done for debian you find that
its more than the main arch. The amount of FTBFS bugs all the
different architectures have caught alone is huge. All the different
archs make for quality control on its own.

As for the cost/benefit: 0 cost, many happy users => infinit ratio.
Could that be any better?


There are many steps that can be done _if_ one arch starts to slow a
Debian release down for more than a intermittend time. But its never
realy been a general problem so far. Check out
http://buildd.debian.org/stats/graph2-big.png for an overview. At the
current rate source are uploaded archs are at most 2-4 days behind the
uploaded sources (the number of packages missing ~ the number of
sources uploaded in 2.4 days).

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-27, 3:33 pm

Andreas Tille <tillea@rki.de> writes:
quote:

> On Tue, 27 Jan 2004, Goswin von Brederlow wrote:
>
> This reasonable approach was discussed / suggested several times
> (just search the archive) and ended always in: It's free. Would you
> volunteer to do it? I would love if someone would grab this hut.
>
> This would also work but this is kind of politics which concerns non-free.
> So if we try to go with the positive list approach above some people might
> be more happy ...



Add everything in non-free/contrib to the list and remove on request.

But I guess people would then complain how long it takes to check the
list. A list in wanna-build or quinn-diff would be better. Do it once
for all archs on a fast machine (auric).
quote:

> Kind regards
>
> Andreas.



MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Wouter Verhelst

2004-01-27, 6:33 pm

On Wed, Jan 28, 2004 at 03:11:31AM +0100, Goswin von Brederlow wrote:
quote:

> Wouter Verhelst <wouter@grep.be> writes:
>
> How about contrib?



same problem.
quote:

> There are reasons for packages being in contrib or non-free which
> don't make me "hate" them, like the deb will Depend on something
> non-free. The list should only contain packages we can legally build
> and don't violate our personel or Debian philosophy (too much).
>
> I don't see a problem with building GPL software that depends on
> something outside of debian for example.



What if the license of that particular dependency doesn't allow me to
use it? For instance, there could be a clause in the license which would
forbid people in countries X, Y and Z to use the software.

You can make a "positive" list of software that would not be of any
problem to anyone, but there might be problems in that gray area.
quote:

> Or software with a software patent on it (which is void where I am,
> thank god.).
>
> If I stick a non-free package in build/REDO (e.g. if someone asks me
> to build it for example) will the buildd reject it or does that work
> fine?



If it can reach the source... but it'd be better if you just set up your
own non-free w-b database.

That said, you'd better not configure a4000 or a1200 for this, or you'll
have to find someone else to sign their logs -- at least as long as
you're not a DD. I don't care manually building a non-free package from
time to time, but I will not run an autobuilder which does them.

[...]
quote:

>
> I will get them when I run through unclaimed packages in details. For
> the list i just assumed that anything kept out of sarge for a reason
> like being a unstable release or cvs snapshot would get updated more
> frequently than 100 days. Whats the point of a cvs snapshot otherwise?
> I know its not perfect, already removed 2 packages I previously didn't
> spot.



You shouldn't have included them in the first place.

If you send a mail to a developer claiming he's not doing his job well
(which is exactly what you did), you had better be damn sure you're
right. If you don't do that, you'll be pointing fingers to, and
accusing, people who don't deserve it. That's the best way I could think
of to demotivate a volunteer, and could, worst case, result in those
leaving the project. Is that what you want?

If you start pointing fingers at people for not maintaining their
packages right -read: for being lazy- then make sure that's not
something people can say you are.

--
Wouter Verhelst
Debian GNU/Linux -- http://www.debian.org
Nederlandstalige Linux-documentatie -- http://nl.linux.org
"Stop breathing down my neck." "My breathing is merely a simulation."
"So is my neck, stop it anyway!"
-- Voyager's EMH versus the Prometheus' EMH, stardate 51462.

Javier =?iso-8859-15?Q?Fern=E1ndez-Sanguino_Pe

2004-01-27, 7:34 pm

(Don't CC, I'm in -devel)

On Wed, Jan 28, 2004 at 03:28:50AM +0100, Goswin von Brederlow wrote:
quote:

>
> # The current maintainer is looking for someone who can take over


(...)
quote:

> to be co-maintainer in order to help the actual maintainer. Please
> see bug number #216883 for more information.
>
> Is that still true?



Yes, it is. A RFA: is disctintly different from an O: however.
quote:

>
> It was mentioned in Bug number #216883 of yours so you must have
> forgotten about it since last October. But a number of DDs are
> surprised to hear that so your not alone not knowing.



Notice that I did not said I didn't know about this, I said I have not seen
it mentioned "officially" (either in d-announce or in our documentation).
quote:

> Some packages can't be autobuild for legal reasons. All of those are
> in non-free. There is no flag in the source packages saying what can
> be legally build and what not and no list is provided otherwise. Also
> some people don't want to support non-free software.



The funny thing is that most of the packages you mentioned from me (save
for lmbench) have been autobuilt at some point and _no_one_ has ever asked
me if there were legally reasons not to do this. I can understand that
there might be _some_ packages that could not be autobuilt, but I fail to
see why this would be applied to all packages, even those that have been
already autobuild. Notice that the builds of my packages (and probably of
many other DDs) have not been removed, so those "legal" reasons don't seem
to be of much concern, but more of an excuse.
quote:

> I think what would work best currently is if you get a comaintainer
> for every arch you can't build the package on or at least a steady
> group of sponsors. If noone is willing to build the package you can
> file a bug against ftp.debian.org to get obsolete versions removed.



No, I can still keep it the current way and only provide support for i386
users running unstable, which is really fine by me too, BTW. And if someone
comes over and asks why is 'X' not available/updated in his favorite
architecture 'Y' I will forward him to the mailing list for that arch.

So, you see, there's really some in-betweens in this situation. The fact
that those packages are non-free mean that they are low priority for me,
I will probably not fix bugs in them if there are serious bugs in other
packages I maintain or co-maintain. That is not to say that I won't do it,
it's just that there are other 62 packages that are higher in the listk.
Given infinite time, they will get their share too.

Since probably many others maintainers have not rejected NMUs or help from
interested parties I don't see how this situation is better than handing
them over to our (overloaded?) QA group. Matter of fact, I would not like
to see QA people dedicating time to non-free packages when there are still
RC bugs open.

Regards

Javi

Dalibor Topic

2004-01-27, 9:34 pm

Goswin von Brederlow wrote:
quote:

> Arnaud Vandyck <avdyk@debian.org> writes:
>


quote:

>
>
> Which means they need manual intervention to get pushed into sarge. It
> just needs to be done where appropriate. A bunch of the Java related
> packages maintainer replied that they are waiting on sablevm or
> kaffe. Maybe that will be ready for sarge.



There is quite a bit of work happening on this here:

http://java.debian.net/index.php/MovingJavaToMain

feel free to jump right in.

Oh, and kaffe could need some interest from DDs on the platforms on
which it doesn't build atm (alpha, arm, m68k, mips, mipsel, sparc, ia64,
parisc). There are some ways to work around that, but in general, it'd
be great (for getting more Java packages into main) if some interested
developers on those architectures picked up the ball. The kaffe.org
upstream simply lacks experienced developers on all those platforms at
the time being, to be able to fix them all.

cheers,
dalibor topic


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Arnaud Vandyck

2004-01-27, 9:34 pm

Jochen Voss

2004-01-28, 12:34 am

On Tue, Jan 27, 2004 at 09:35:39AM +0100, Bastian Blank wrote:
quote:

> non-free but neccesary for any people which work on s/390.


Huh? I remember some

..., but we will never make the system depend
on an item of non-free software.

clause.

Jochen
--
http://seehuhn.de/

Bastian Blank

2004-01-28, 1:34 am

On Wed, Jan 28, 2004 at 04:38:59AM +0100, Goswin von Brederlow wrote:
quote:

> Need manual pushing.



No

Bastian

Please fix your mua:
Mail-Followup-To: debian-devel@lists.debian.org

--
The man on tops walks a lonely street; the "chain" of command is often a noose.

Goswin von Brederlow

2004-01-28, 1:34 am

Wouter Verhelst <wouter@grep.be> writes:
quote:

> On Wed, Jan 28, 2004 at 03:11:31AM +0100, Goswin von Brederlow wrote:
....[QUOTE][color=darkred]
> You can make a "positive" list of software that would not be of any
> problem to anyone, but there might be problems in that gray area.



I think such a positive list that a bunch of people and debian-legal
have check would be nice. Even if its just as a guideline to people
that get asked to compile something.
quote:

> [...]
>
> You shouldn't have included them in the first place.
>
> If you send a mail to a developer claiming he's not doing his job well
> (which is exactly what you did), you had better be damn sure you're
> right. If you don't do that, you'll be pointing fingers to, and
> accusing, people who don't deserve it. That's the best way I could think
> of to demotivate a volunteer, and could, worst case, result in those
> leaving the project. Is that what you want?



I didn't want to imply that any one developer from that group is not
doing their job well. All I said is that the packages are not actively
pushed into sarge by the maintainer or users. I ratteled the cage a
bit saying they look orphaned to me and I'm sorry if any maintainers
pride is hurt and he feels personally attacked. That was just cage
rattling and no attack on any of you that care for their packages.

Some packages in the list have perfectly good reasons for not
being in sarge, some don't. I'm sorry for the few that got cought up
in there. If the patch for debbugs for a machine readable interface
had been added to debbugs I could have quickly checked every package
for "keep me out of sarge" bugs.

Some maintainers didn't know non-free/contrib isn't autobuild and
never noticed or investigated whats blocking their package. Those have
been reminded.

Some packages should have been purged, like w3-el was now or where
orphaned but not correctly. 18 out of 71 confirmed packages will get
orphaned, removed, adopted or replaced. Thats 25% of the packages. And
thats just packages with an active maintainer.

There are even packages in there with a NM waiting to take over but
needing a sponsor, which he seem to have found through this action.
quote:

> If you start pointing fingers at people for not maintaining their
> packages right -read: for being lazy- then make sure that's not
> something people can say you are.



I don't consider listing such a large group as pointing at any one in
particular. A group will allways have some black sheep, some white
sheep and a lot of grey sheep. Again, I'm sorry if any one feels
personally attacked. And I thank those that found the mail helpfull
already and hope the rest can forgive me for rattleing the cage a bit.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-28, 2:36 am

Javier =?iso-8859-15?Q?Fern=E1ndez-Sanguino_Pe=F1a?= <jfs@computer.org> writes:
quote:

> (Don't CC, I'm in -devel)
>
> On Wed, Jan 28, 2004 at 03:28:50AM +0100, Goswin von Brederlow wrote:
> (...)
>
> Yes, it is. A RFA: is disctintly different from an O: however.
>
>
> Notice that I did not said I didn't know about this, I said I have not seen
> it mentioned "officially" (either in d-announce or in our documentation).



I will see if that can be rectified.
quote:


Some as in some of non-free/contrib, not just from the list.
[QUOTE][color=darkred]
>
> The funny thing is that most of the packages you mentioned from me (save
> for lmbench) have been autobuilt at some point and _no_one_ has ever asked
> me if there were legally reasons not to do this. I can understand that
> there might be _some_ packages that could not be autobuilt, but I fail to
> see why this would be applied to all packages, even those that have been
> already autobuild. Notice that the builds of my packages (and probably of
> many other DDs) have not been removed, so those "legal" reasons don't seem
> to be of much concern, but more of an excuse.



At some point someone decided that it would be best not to build
contrib / non-free by the buildd. That someone convinced the
wanna-build maintainers (if he wasn't one of them) not to autobuild
them. After that no buildd ever got asked to build the packages.
Or wanna-build didn't include them from the start and you just see
build that pople have manually added to the buildds own build list at
some point in time.
quote:

>
> No, I can still keep it the current way and only provide support for i386
> users running unstable, which is really fine by me too, BTW. And if someone
> comes over and asks why is 'X' not available/updated in his favorite
> architecture 'Y' I will forward him to the mailing list for that arch.



You are forgetting the intrests of Debians users in general here.
There are a lot of users that don't use sid for various reasons and
you deny them your package for archs you support. Can you live with
all those poor users out there being deprived of your good work?

Your lmbench is the top package that just needs to be build for the
missing archs. But since its up for adoption there is no blame
there. So please think about this in the general case of packages and
not just your three packges mentioned. Forget about your packages for
a moment, relax and imagine you were a stable user. Now ...

The maintainer should now best if a non-free/contrib package can be
build for debian by others or not and if it needs to be build on some
archs to keep current. If he sees someone that previously helped
building the package lost interest wouldn't it be good to ask for
someone new on debian-<port>? It takes maybe 5 minutes to check a
package and to write such a mail. 5 minutes per package every 3 month
or at least once a year. Is that so much to ask. Wouldn't you (as
user) ask about whats up when you see that hapening?
quote:

> So, you see, there's really some in-betweens in this situation. The fact
> that those packages are non-free mean that they are low priority for me,
> I will probably not fix bugs in them if there are serious bugs in other
> packages I maintain or co-maintain. That is not to say that I won't do it,
> it's just that there are other 62 packages that are higher in the listk.
> Given infinite time, they will get their share too.
>
> Since probably many others maintainers have not rejected NMUs or help from
> interested parties I don't see how this situation is better than handing
> them over to our (overloaded?) QA group. Matter of fact, I would not like
> to see QA people dedicating time to non-free packages when there are still
> RC bugs open.
>
> Regards
>
> Javi



Never said I want to just dump them all into the lap of the QA
group. Most non-free/contrib packages in fact just need a little nudge,
need to have some obsolete debs from unmaintained archs removed or
need someone to compile it for their arch to get into sarge. Now I
know your position on your packages and can act accordingly.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-28, 2:36 am

Jochen Voss <voss@seehuhn.de> writes:
quote:

> On Tue, Jan 27, 2004 at 09:35:39AM +0100, Bastian Blank wrote:
> Huh? I remember some
>
> ..., but we will never make the system depend
> on an item of non-free software.
>
> clause.
>
> Jochen
> --
> http://seehuhn.de/



Ever looked at the various bootloaders for other archs in detail?
..oO( Opening a can of worms )

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-01-28, 6:38 am

Thanks for the reply.

Neil Roeth <neil@debian.org> writes:
quote:

> On Jan 27, Goswin von Brederlow (brederlo@informatik.uni-tuebingen.de) wrote:
....[QUOTE][color=darkred]
> ^^^^^^^^^
>
> Don't orphan aplus-fsf. It hasn't gone in because it depends on xemacs21,
> which has an m68k bug that is preventing it from going in. I provided the
> patch to fix a different RC bug (201657).



Did you follow xemacs21 on m68k? Is there a change to get it to
compile with some patches stuck in the BTS or newer/older toolchain?

I use xemacs a lot and wish I had it for m68k too. Last three time I
tried to build it on m68k failed with various errors. Buts its just so
big and such a pain to find bugs if you are unfamiliar with the source.

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Andreas Tille

2004-01-28, 8:35 am

On Wed, 28 Jan 2004, Goswin von Brederlow wrote:
quote:

> I think such a positive list that a bunch of people and debian-legal
> have check would be nice. Even if its just as a guideline to people
> that get asked to compile something.


Why talking about such list? Here is the start:

molphy
phylip
treetool

have no visible licensing constraints which might prevent autobuilding.

Kind regards

Andreas.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Andreas Barth

2004-01-28, 9:35 am

* Andreas Tille (tillea@rki.de) [040128 22:40]:
quote:

> On Wed, 28 Jan 2004, Goswin von Brederlow wrote:
> Why talking about such list? Here is the start:



netpbm-nonfree can also be autobuild. However, this package is
(hopefully ;) due to be removed in August.


Cheers,
Andi
--
http://home.arcor.de/andreas-barth/
PGP 1024/89FB5CE5 DC F1 85 6D A6 45 9C 0F 3B BE F1 D0 C5 D1 D9 0C


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
era+debian@iki.fi

2004-02-03, 7:34 pm

On 27 Jan 2004 00:58:33 +0100, Goswin von Brederlow
<brederlo@informatik.uni-tuebingen.de> posted to debian-devel:
quote:

> Noone has cared enough about these packages to get them compiled,
> fixed or pushed into sarge so I am assuming the packages don't have a
> caring maintainer or fan community. Ergo they should be orphaned.



Speaking of which, could somebody do an NMU of dpkg-dev with all the
trivial patches it has accumulated in the BTS over the years?

Perhaps it should be split out of the dpkg source package -- AFAICT it
doesn't really contain anything which would force it to be the same
source package (except possibly the deb(5) manual page and some stuff
like that).

/* era */

--
formail -s procmail <http://www.iki.fi/era/spam/ >http://www.euro.cauce.org/
cat | more | cat<http://www.iki.fi/era/unix/award.html>http://www.debian.org/


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Jan Kesten

2004-07-30, 2:48 am

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Goswin von Brederlow wrote:

Hi Goswin!

| Noone has cared enough about these packages to get them compiled,
| fixed or pushed into sarge so I am assuming the packages don't have a
| caring maintainer or fan community. Ergo they should be orphaned.

| If you maintain one of thses packages then tell me (including the
| names of packages you maintain) during the next week. If you are using
| one of these packages and could maintain (or NMU some fixes) you
| should contact the maintainer and me to work things out. If I hear
| nothing about a package soon I will start with the oldest and do a few
| packages every day.

I use these (under debian and solaris) and if there is some need for
some work (i.e. as maintainer) I would try to spend some work (in
fact I get paid during my studying for getting at least 3270 an
x3270 to work :-)

| 3270
| ntop

So if there is anything I can do for help let me know!

Cheers,
Jan
- --
GPG-KeyID: 82201FC4
Available at my public sks keyserver sks.nerdcamp.net
Please report any problems using sks.nerdcamp.net!
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFBCb7GvvmCkIIgH8QRAoAMAJ93C6Tg7drL
v9jb3THh0qKig5fCMACggavP
CfBjuQX2ZD23FqJ7iGLheUo=
=zgRK
-----END PGP SIGNATURE-----


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Richard A Nelson

2004-07-30, 2:48 am

On Fri, 30 Jul 2004, Jan Kesten wrote:

I'm not sure if the following attributions are correct, but that is what
was in the note...

> Goswin von Brederlow wrote:
>
> Hi Goswin!
>
> | Noone has cared enough about these packages to get them compiled,
> | fixed or pushed into sarge so I am assuming the packages don't have a
> | caring maintainer or fan community. Ergo they should be orphaned.
>
> | If you maintain one of thses packages then tell me (including the
> | names of packages you maintain) during the next week. If you are using
> | one of these packages and could maintain (or NMU some fixes) you
> | should contact the maintainer and me to work things out. If I hear
> | nothing about a package soon I will start with the oldest and do a few
> | packages every day.


hrm, I've not seen the original message yet

> I use these (under debian and solaris) and if there is some need for
> some work (i.e. as maintainer) I would try to spend some work (in
> fact I get paid during my studying for getting at least 3270 an


I do, however use daily (and maintain) the 3270 series of packages.
I count 3 open bugs (yeah, one is serious - I've to remail the upstream)
one should be closed because its been almost a year without any response
from the submitter - and I can't test it locally... and nothing in the way
FTBFS so I'm kinda stumped that it shows up on whatever this list is...

What am I missing ?
--
Rick Nelson
<``Erik> 18,446,744,073,709,551,616 is a big number


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-07-30, 7:53 am

Jan Kesten <debian-lists@the-hidden-realm.de> writes:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Goswin von Brederlow wrote:
>
> Hi Goswin!
>
> | Noone has cared enough about these packages to get them compiled,
> | fixed or pushed into sarge so I am assuming the packages don't have a
> | caring maintainer or fan community. Ergo they should be orphaned.
>
> | If you maintain one of thses packages then tell me (including the
> | names of packages you maintain) during the next week. If you are using
> | one of these packages and could maintain (or NMU some fixes) you
> | should contact the maintainer and me to work things out. If I hear
> | nothing about a package soon I will start with the oldest and do a few
> | packages every day.
>
> I use these (under debian and solaris) and if there is some need for
> some work (i.e. as maintainer) I would try to spend some work (in
> fact I get paid during my studying for getting at least 3270 an
> x3270 to work :-)
>
> | 3270
> | ntop
>
> So if there is anything I can do for help let me know!
>
> Cheers,
> Jan


I have 3270 marked as 'needs orphaning', someone (presumably the
maintainer) earlier replied to my mail saing so.

I suggest you contact the maintainer proposing comaintainership or
adopting the package and, in the unlikley case he fails to reply,
hijack the package.


ntop on the other hand seems to be resolved. No RC bugs open and the
sarge and sid versions are both 2:3.0-1 (dated 2004-05-19).

MfG
Goswin


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Goswin von Brederlow

2004-07-30, 7:53 am

Richard A Nelson <cowboy@debian.org> writes:

> On Fri, 30 Jul 2004, Jan Kesten wrote:
>...
> hrm, I've not seen the original message yet


The original message was send out month ago.

>
> I do, however use daily (and maintain) the 3270 series of packages.
> I count 3 open bugs (yeah, one is serious - I've to remail the upstream)
> one should be closed because its been almost a year without any response
> from the submitter - and I can't test it locally... and nothing in the way
> FTBFS so I'm kinda stumped that it shows up on whatever this list is...
>
> What am I missing ?


http://packages.qa.debian.org/3/3270.html

At the time of the original mail the package had been stuck in sid
(out of sarge) for over 100 days. The current version is 97 days old
and also stuck in sid.

Apart from the one RC bug the package is missing on several archs. You
need to login to one of debians systems and build the package manually
(non-free is not autobuild) or find someone on the porters list to
build it for you.

The QA page does not list any version of 3270 for testing. If thats
right the upcoming stable will releas