Re: Ubuntu branches on salsa

[Michael Biebl]

Hi Iain,

Am 12.06.19 um 11:30 schrieb Iain Lane:
> Hi there,
> 
> This is something we've talked about in the Ubuntu team, and I've been
> meaning to bring it up with pkg-gnome for a while but kept delaying. I'm
> wearing both my Debian and Ubuntu hats here.
> 
> When we switched to git in pkg-gnome, Ubuntu also followed suit. On the
> Ubuntu side we've been pushing to Launchpad. By and large (IMO) it's
> been positive - our packages are derived from Debian's and this makes it
> much easier to maintain the delta. But there are some warts that can
> make it difficult to work with, particularly for new and drive-by
> contributors and people who don't yet have a good mental model of how
> git and gbp work. I'll not go into all the details, but the main one is
> that the 'upstream' and 'pristine-tar' branches get really difficult to
> handle if they are out of sync downstream - when we want to take a
> release ahead of Debian or if we're occasionally on a different series,
> for example.
> 
> My proposal is to allow Ubuntu (no reason that other downstreams
> couldn't do this if they were in a similar position too) branches to be
> pushed to the repositories on salsa. This would require that we add
> members of the Ubuntu team to the group. On the Ubuntu side we'll take
> responsibility for educating our developers to avoid disrupting Debian's
> work. I think it'd be possible to arrange the permissions so that they
> can't push to 'debian/*' (make those maintainer-only, and only grant
> Ubuntu people the developer permission), but I hope we could try without
> doing that initially.
> 
> I believe that the MySQL / mariadb team does something like this:
> 
>   https://salsa.debian.org/mariadb-team/mysql (look at the branches)
> 
> Thoughts / objections / non-objections?
> 

I'm not really active anymore in pkg-gnome so take my feedback with a
grain of salt, but what I would do is to pick a couple of packages (some
trival ones (like games) and some more complicated, like glib), and just
try for a while and then re-evaluate later.
If there is a benefit for Debian to add those Ubuntu branches, then I
don't see why it shouldn't be added.
My point is that we probably can't tell without trying and we shouldn't
be afraid to doing so.

Michael





-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Bug#927667: gnome: please confirm or revert choice of Wayland for default desktop

[Michael Biebl]

Am 20.06.19 um 11:25 schrieb Jonathan Carter:
> I just have a small proposal:
> 
> Selecting "Gnome on Xorg" is really easy from GDM for anyone who has
> trouble on Wayland. It might be worth while adding that to the release
> notes so that users who are not quite ready for Wayland yet know that
> there's an easy way to get the old behavior back without having to
> re-install stretch or some other distro.

That seems like a very good idea to document this prominently in the
release notes. After all, we do install both Xorg and Wayland support,
so switching the desktop session is indeed trivial.

I was about to file a bug report against release-notes to add such a
section, but then it probably makes sense to wait for a final decision.

Related to that, we already have
https://salsa.debian.org/ddp-team/release-notes/commit/5496e24

Assuming it is decided, that the default is switched back to Xorg, this
existing paragraph in the release notes should be adapted accordingly.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?

Re: Bug#927667: gnome: please confirm or revert choice of Wayland for default desktop

[Michael Biebl]

Hi everyone

Am 20.06.19 um 11:12 schrieb Iain Lane:
>> I've left some comments on
>> https://salsa.debian.org/gnome-team/gdm/merge_requests/8 regarding the
>> technical side of the proposed change.
> Someone could probably look in Ubuntu's gdm3 package to see what we're
> doing. We provide "GNOME" (Xorg, the default) and "GNOME on Wayland"
> sessions.

Afair, this required changing gnome-session. I left a comment in the gdm MR.
If the point is, to not switch the desktop session automatically on
upgrades, then the session files would have to be renamed (back again)
to gnome.desktop (Xorg) and gnome-wayland.desktop from gnome.desktop
(Wayland) and gnome-xorg.desktop.
At least this is how I remember the details from back then in 2016.
I haven't checked if the situation is still the same today.

Regards,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Which team is best for team-maintaining libepoxy package ?

[Michael Biebl]

Am 16.10.18 um 21:53 schrieb Jeremy Bicha:
> On Tue, Oct 16, 2018 at 2:37 PM Jérémy Lal  wrote:
>> libepoxy is needed by both KDE and GNOME packages.
>> What's the best team to maintain it ?
> 
> I guess we've been using https://salsa.debian.org/utopia-team for
> cross-desktop stuff.
> 
> It might make sense to merge that team with
> https://salsa.debian.org/freedesktop-team even if much of the Utopia
> stuff isn't actually hosted on freedesktop.org .
> 
> I only recently realized that the Utopia team's name derives from
> Project Utopia, an initiative which hasn't been around for several
> years.

This is correct. The utopia name mainly has nostalgic value (it was
relevant when hal still was a thing) but does not actually describe any
current efforts.
That said, it's also very unspecific, which makes it useful as an
umbrella if a thing is not strictly GNOME, fdo, ... specific.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Salsa sub-project for Gnome shell extensions

[Michael Biebl]

Hi Jonathan,

sorry it took so long to get back to you.

On 6/25/18 14:18, Jonathan Carter (highvoltage) wrote:
> I don't have access to create sub-teams under salsa/gnome-team, could
> someone who does please create a subgroup called
> salsa/gnome-team/shell-extensions?

I've created such a subgroup and set you as owner.
If something is still missing, please holler.

Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?

Re: Salsa sub-project for Gnome shell extensions

[Michael Biebl]

Am 10.05.2018 um 21:41 schrieb Jonathan Carter:
> Hi!
> 
> Most of the gnome shell extensions in Debian are packaged by around a
> handful of people, and I've been wondering for a while whether we should
> rather form a team for those.
> 
> Since then, salsa came along and it supports subgroups, so it seems that
> it would make sense to have a subgroup in the gnome team for these
> extensions.
> 
> Any thoughts?

Have you contacted those maintainers, if they are interested in joining
such a group?
What would you use as maintainer address?

Regards,
Mihcael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Action requested: migration of alioth list pkg-gnome-maintainers

[Michael Biebl]

Am 28.01.2018 um 19:26 schrieb alioth lists migration team:
> Dear list owner,
> 
> As per the announcement on debian-devel-announce[1] the migration of
> lists.alioth.debian.org mailing lists is now underway. If you would
> like pkg-gnome-maintain...@lists.alioth.debian.org
> to be included in this process, to ensure that the list still works
> after the migration in late March/April, please let us know by replying
> to this email. Otherwise, the list will stop working at migration time
> and the archives will no longer be accessible.

Thanks for getting in touch and providing this service.

The Debian GNOME team would like to keep the
pkg-gnome-maintain...@lists.alioth.debian.org address so please transfer
that mailing list to the new setup.

pkg-gnome-commits@l.a.d.o is not longer needed and can be binned.

Regards,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Fwd: Action requested: migration of alioth list pkg-gnome-maintainers

[Michael Biebl]

Am 30.01.2018 um 14:05 schrieb Jeremy Bicha:
> On Tue, Jan 30, 2018 at 4:55 AM, Jordi Mallach  wrote:
>> I think polluting -gtk-gnome with BTS, archive and all that automated
>> email will make it as hard to follow as -maintainers is, so I'd vote
>> that given the list can continue to exist, we keep it.
> 
> One other point is that tracker.debian.org may offer an email service
> for team packages. In that case, we may want to change the Maintainer
> one more time. That may take several months before it is available
> though.

Ok. So it seems like we want to keep pkg-gnome-maintainers@a.l.d.o for
now then? And in a few months we re-evaluate if the functionality in
tracker.d.o is sufficient for our use cases by then.
The alioth replacement won't be available forever, but it should give us
enough time to see how things turn out.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?





signature.asc
Description: OpenPGP digital signature

Fwd: Action requested: migration of alioth list pkg-gnome-maintainers

[Michael Biebl]

Hi

Are we going to keep using pkg-gnome-maintain...@a.li.d.o as Maintainer?
Or do we switch to debian-gtk-gnome@l.d.o (i.e. touching every
debian/control.in again)?
Or do something entirely else?

In any case, I guess
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-gnome-commits
can be retired or is anyone actively using that one?

Regards,
Michael

 Weitergeleitete Nachricht 
Betreff: Action requested: migration of alioth list pkg-gnome-maintainers
Datum: Sun, 28 Jan 2018 18:26:14 +
Von: alioth lists migration team 
An: pkg-gnome-maintainers-ow...@lists.alioth.debian.org

Dear list owner,

As per the announcement on debian-devel-announce[1] the migration of
lists.alioth.debian.org mailing lists is now underway. If you would
like pkg-gnome-maintain...@lists.alioth.debian.org
to be included in this process, to ensure that the list still works
after the migration in late March/April, please let us know by replying
to this email. Otherwise, the list will stop working at migration time
and the archives will no longer be accessible.

Feel free to also let us know if the list is no longer needed, so we can
note that down too. Replies must be received by 15th March 2018.

More information about the new service can be found here:


Thanks,
the alioth-lists migration team.

[1] 





signature.asc
Description: OpenPGP digital signature

Re: bad experience with Gnome 3.8 on kFreeBSD

[Michael Biebl]

Am 18.10.2013 03:08, schrieb Bruno Melo:
 Hi,
 I have kFreeBSD jessie here and I have installed the last gnome 3.8
 packages for transition from sid and I came here to report my bad
 experience:
 1- Alt+Tab doesn't work.

(media) key handling, specifically the key grabbing, has been moved into
GNOME shell (previously it was a mix of gnome-settings-daemon and
gnome-shell, which was a huge mess).
This does affect the flashback (formerly known as fallback) mode, also
on Linux.

 2-GDM unusable (using LightDM now)

GDM relies on GNOME Shell now, so you need working 3D support. Not
necessarily hardware accelerated, llvmpipe works too, but I don't know
the state of llvmpipe on kfreebsd.
For i386/amd64 at least llvmpipe works okayish (tested in a VM).

 3- Wallpaper doesn't appear, just a black screen.

Also a flashback mode issue (which also affects Linux). Desktop
background drawing has been removed from nautilus and moved into GNOME
shell.

 4- Empathy and Epiphany don't show anything. Empathy doesn't show the
 conversation and Epiphany the websites.

Not sure about this one.

 5- The multimedia apps also have problems.

The multimedia apps require clutter, so this might be same problem as 2/

 6- Gnome Shell obviously can't run because kFreeBSD have not 3D
 acceleration.

I don't know the state of 3D (hardware accelerated or llvmpipe) on
kfreebsd. Would be great if someone more knowledgeable would provide
more insight.

 I recommend to you install kFreeBSD and test for themselves to see what I'm
 talking about. It's almost unusable :/

The bad experience you encouter looks to a big part due to the bad state
the flashback/fallback mode is currently in. And unfortunately I only
see it get worse over time unless someone steps up to help
maintain/develop it.
As for GNOME Shell / GDM not running on kfreebsd: We'd love to see
kfreebsd users/porters/developer investigate that.

None of us Debian GNOME maintainers use kfreebsd, so we rely on
fixes/patches from others.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: bad experience with Gnome 3.8 on kFreeBSD

[Michael Biebl]

Am 18.10.2013 10:43, schrieb Robert Millan:
 On 18/10/2013 03:08, Bruno Melo wrote:
 6- Gnome Shell obviously can't run because kFreeBSD have not 3D
 acceleration.
 
 That's not correct. kFreeBSD provides 3D acceleration and we even have
 confirmation of developers using it (e.g. Christoph).

If Christoph has working 3D support on his kfreebsd system, could he
install gnome-core (from sid) and test which bits work and which don't?

So far, the only feedback the Debian GNOME team has from kfreebsd users
(admittedly only 1 or 2 so far) is that GNOME Shell does *not* run on
kfreebsd.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME introduction help screen

[Michael Biebl]

Am 16.10.2013 15:52, schrieb Osamu Aoki:
 Hi,
 
 When I installed FEDORA19 and started GNOME 3.8, I got a nice animation
 helper telling me how GNOME 3.8 UI works.  Things like SUPER-M and
 SUPER-SPACE.  
 
 Does Debian GNOME 3.8 has trhe same feature?  How can we activate such
 helper? Package name ?

I think that would be gnome-getting-started-docs.

Not uploaded yet.


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3.8 in unstable?

[Michael Biebl]

Am 11.09.2013 13:18, schrieb Johannes Rohr:
 Emilio Pozuelo Monfort wrote:
 
 On 11/09/13 12:41, Johannes Rohr wrote:
 And I really don't get what is still holding up GNOME 3.8.

 Right now it's the evolution-data-server transition, which is being worked
 on. After that happens, it'll probably be time to get mutter, gnome-shell,
 g-s-d, g-c-c and so into unstable, bringing the last few bits of G3.8
 there. And then, we can start getting 3.10 in, hopefully much faster this
 time
 
 Emilij, thanks so much for the update. Just wondering, is that a regular 
 transition? It is not listed at http://release.debian.org/transitions/

It is. I've filed relevant bugs for it [1] and after analysing what the
evo/eds transtion entails, will file a transition bug now.
(We will need to update some other libs, like tp-logger, tp-farstream,
libgweather, etc along with evo/eds 3.8)
Some other things that blocked progress on this, is webkitgtk (= 1.10)
not being available everywhere. Hopefully that is fixed now [2]. We
needed a few tries to get it build on all architectures. That it takes
several days on slower architectures to build, doesn't help to speed up
the process.
If you actually want to help instead of just complain, you can pick up a
bug from [1] and work on it.

Michael

[1]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=pkg-evolution-maintain...@lists.alioth.debian.org
[2] https://buildd.debian.org/status/package.php?p=webkitgtk
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3.8 in unstable?

[Michael Biebl]

Am 11.09.2013 13:37, schrieb David Roguin:
 According to the evoluton-data-server package (
 http://packages.qa.debian.org/e/evolution-data-server.html) it will be part
 of db6.0 incoming transition:
 http://release.debian.org/transitions/html/db6.0.html

That's very likely not going to be an issue. db6.0 switched to a
different, incompatible license which makes it basically impossible to
start this transition. See the linked bug report.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3.8 in unstable?

[Michael Biebl]

Am 11.09.2013 12:57, schrieb Emilio Pozuelo Monfort:
 On 11/09/13 12:41, Johannes Rohr wrote:
 And I really don't get what is still holding up GNOME 3.8.
 
 Right now it's the evolution-data-server transition, which is being worked on.
 After that happens, it'll probably be time to get mutter, gnome-shell, g-s-d,
 g-c-c and so into unstable, bringing the last few bits of G3.8 there. And 
 then,
 we can start getting 3.10 in, hopefully much faster this time.

Yeah, after evo/eds, we probably should do the rest together in one go.
There is still some unsorted stuff though, like most of the
NetworkManager integration patches (to support user session connections
or make NM optional on !Linux) have been dropped in experimental.
Someone needs to work on those patches before we can move gnome-shell
and g-c-c to unstable.
Also, the situation around systemd is not fully sorted yet.
First, we need a newer systemd version, which is currently in
experimental only (also waiting for a transition slot [1]), and then
there is the problem, that working power management requires systemd to
be PID 1. That's going to cause problems which we haven't addressed yet.

So, you see, it's not like we are sitting on our lazy asses, it's just a
lot of work and coordination efforts.

Michael

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=717418
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3.8 in unstable?

[Michael Biebl]

Am 01.08.2013 03:19, schrieb Ben Finney:
 Michael Biebl bi...@debian.org writes:
 
 Am 28.07.2013 13:46, schrieb Ben Finney:
 What will constitute a [cooled-down] unstable, enough for GNOME 3.8
 to come in?

 We've filed the transitions bugs for the various library transitions
 about 2 months ago. Unfortunately, we didn't receive any replies so far
 just yet.
 
 Thanks. Is two months enough time that you could ping them again to see
 how long we'd need to wait?

We got the ball rolling a bit the last couple of days, so we should see
some updates in the near future.
Thanks for being patient with us.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3.8 in unstable?

[Michael Biebl]

Am 28.07.2013 13:46, schrieb Ben Finney:
 Emilio Pozuelo Monfort po...@debian.org writes:
 
 On 17/05/13 08:13, Johannes Rohr wrote:
 sorry for my impatience! I noticed that GNOME 3.8 is still in
 experimental, even though the Wheezy release process is over. Are
 there any indications as to when it is going to make it into
 unstable? Is it waiting for some transition to complete first?

 Yes, we're basically waiting for unstable to cool down (too many
 transitions atm).
 
 Ping.
 
 What will constitute a colled-down unstable, enough for GNOME 3.8 to
 come in?

We've filed the transitions bugs for the various library transitions
about 2 months ago. Unfortunately, we didn't receive any replies so far
just yet.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Affected by gnome-shell lockups on wheezy?

[Michael Biebl]

Am 05.07.2013 23:30, schrieb David Weinehall:

 
 I'm using gnome-shell 3.8.3 from experimental I'm still seeing lockups
 every now and then.  Not sure whether it's the same issue, but it's
 annoying, it's rather frequent, and now with the workspace grid and
 weather notification extensions rectifying the shortcomings of the
 default behaviour I'd be very disappointed to have to abandon
 gnome-shell.
 
 Is there anything I can do to help debug this issue (unless it's
 completely unrelated and should be filed separately).
 

I think we need a newer mozjs for that to be solved properly. I vaguely
remember that this requires js17 or later and gjs built with the patches
from the wip/js17 branch.

Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Affected by gnome-shell lockups on wheezy?

[Michael Biebl]

Am 06.07.2013 02:07, schrieb Michael Biebl:

 I think we need a newer mozjs for that to be solved properly. I vaguely
 remember that this requires js17 or later and gjs built with the patches
 from the wip/js17 branch.

See also:
https://bugs.launchpad.net/ubuntu/+source/mozjs/+bug/1113166
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=709434#15

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Meta-gnome3 dependency on xul-ext-adblock-plus (Bug #689858)

[Michael Biebl]

Am 13.06.2013 16:45, schrieb Bill West:
 Johannes Rohr jorohr at gmail.com writes:
 

 Dear GNOME maintainers,

 while I do use adblock plus with Iceweasel, it seems a bit arbitrary to 
 me that the gnome meta package declares a dependency on it. Adblock plus 
 is not part of the GNOME desktop. In addition, it pulls in  iceweasel 
 which also isn't. What is the rationale behind this decision?

 Until now, I had assumed, that the policy regarding Debian's GNOME 
 desktop was to closely reflect the official GNOME desktop as defined by 
 upstream. I see that the gnome metapackage is not totally strict on 
 that, but still, most packages which are not part of the official GNOME 
 release are suggests or recommends, not depends. Why this exception, for 
 a package which isn't even based on the GNOME platform?

 BTW: I do keep on using adblock plus myself, however, there are some who 
 dislike it since the developer has begun to generate revenues through 
 paid-for exclusion from the filter lists for some types of ads. 
 Therefore, as I understand, some people are now using different adblockers.

 So I would suggest to at least lower the dependency to suggests, if not 
 to drop it from the dependencies altogether.

 Thanks,

 Johannes


 
 
 I agree completely with Johannes. I use Debian/Linux because it is not
 Windows and I want to control what is installed on my computer. My opinion
 is that anything that is not strictly required in order to make a package
 work should not be a dependency, only a recommendation. I don't want ads to
 be blocked, I don't want extra code bulking up my tools, and I don't want to
 have to even think about having to disable some functionality that I never
 wanted installed in the first place.

Then don't use the meta packges (they are empty and don't provide any
functionality). Then pick and choose to your liking.

Michael




-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Suspend problem in Gnome 3.8?

[Michael Biebl]

Am 06.04.2013 12:38, schrieb Arief M Utama:
 Hi all,
 
 Been trying gnome-3.8 packages that is starting to available in
 experimental.

 Aside from the missing of network configuration from the control-center, my
 biggest annoyance with the new version is the missing of suspend ability.


 Aside from using gnome-3.8 packages, I also used systemd, network-manager
 and latest dbus and libc from experimental.
 
 Any ideas ? Please cc me on replies, not subscribed to the list.

The systemd version currently available in wheezy/unstable is too old
and doesn't provide the necessary suspend functionality.
We are working on a updated systemd package. Sorry for the inconvenience
in the mean time.


Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: The difficulty of contributing / volunteering

[Michael Biebl]

On 31.01.2013 22:22, Boruch Baum wrote:
 Hello everyone,
 
 I'm getting a bit discouraged (see below). I thought it would be
 simple to offer to volunteer helping in debian, specifically packaging
 'meld' and other packages with only older versions debianized. It
 really shouldn't be this difficult.

The best way is contact the maintainer directly. A quick look at [1] and
the changelog reveals that Balint has been  been taking care of the
package mostly. I'm sure he would welcome your support.

Michael

[1] http://packages.qa.debian.org/m/meld.html
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Updating GTK+2 in Lenny

[Michael Biebl]

On 25.11.2012 19:24, Camaleón wrote:
 (sorry if it comes duplicated)
 
 Hello,
 
 A quick question.
 
 Since Firefox/Thunderbird 17.x releases now it's required GTK+2 2.18¹ so 
 for those still running the olstable flavour (Lenny) this is a problem 
 as it comes with GTK+2 2.12.

You are strongly advised to upgrade to squeeze given that lenny no
longer has any security support.

 I've found an updated GTK+2 from the backports archive² with these two 
 packages (required to update Mozilla programs):
 
 libgtk2.0-0 2.18.6-1~bpo-50+1
 gtk2-engines-pixbuf 2.18.6-1~bpo-50+1
 
 Would there be any problem in getting just these two and keep the rest of 
 the libraries at 2.12 (that is, mixing GTK+2 versions)? I just did it 
 under a VM and all seemed to be fine but I prefer to ask the masters.

Why don't you just try?

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

GNOME 3.4: uploading remaining bits to unstable (minus evo/eds)

[Michael Biebl]

Hi,

as soon as gcr/gnome-keyring/seahorse has migrated to testing (should
happen in about 4 days), I plan to move the remaining bits of GNOME 3.4
from experimental to unstable. They basically all need to go in together.

Affected packages are:

gnome-control-center
gnome-settings-daemon
gnome-bluetooth
gnome-shell
gnome-shell-extensions
gnome-tweak-tool
mutter
metacity


gnome-bluetooth
===

Updating gnome-bluetooth involves a small library transition.
libgnome-bluetooth8 → libgnome-bluetooth10 which affects 3 packages:

gnome-phone-manager(can be binNMUed)
network-manager-applet (can be binNMUed)
gnome-user-share   (needs sourceful upload)

gnome-user-share has a package ready in experimental and I'll ping the
maintainer once gnome-bluetooth is uploaded. The other two have been
successfully build-tested against gnome-bluetooth 3.4 on amd64.
Given the size of this transition, and given that 2 of those 3 packages
are under the control of the Debian GNOME team, I don't think we need to
setup a tracker for this.

I currently don't see any of the above packages involved in any ongoing
transition, so this should be safe to start.

In case I've missed something, let us know so we can defer the upload.

Cheers,
Michael

on behalf of the Debian GNOME team


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Environment variables when using the email launcher

[Michael Biebl]

On 19.05.2012 13:25, Jean-Christophe Dubacq wrote:
 Hello,
 
 I am in the process of testing seahorse integration with my usual
 environment (wheezy/testing gnome3, without any special tweaks;
 gnupg-agent installed but disabled at session start editing
 /etc/X11/Xsession.d/90gpg-agent).
 
 I find that if I launch icedove either from the terminal
 (gnome-terminal, $GPG_AGENT_INFO is set) or from the launcher (left-hand
 side of the window), I get gpg working ok (using seahorse as a
 gpg-agent, I suppose).
 
 However, if I launch it from the shortcut key set in the Keyboard
 parameters, gpg is not working correctly (cannot contact gpg-agent: no
 gpg-agent available in this session). My guess is that GPG_AGENT_INFO is
 not set when using this special session.
 
 I would like to investigate this further, but I do not have many ideas
 about where to start. Do you have any ideas ?

This is probably
https://bugzilla.gnome.org/show_bug.cgi?id=662528

An upload of gnome-settings-daemon-3.4.2 containing this patch will be
coming soon.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Environment variables when using the email launcher

[Michael Biebl]

On 19.05.2012 23:00, Jean-Christophe Dubacq wrote:

I wouldn't have thought about searching in bugs for
 gnome-settings-daemon for such a bug. I will test the upload as soon as
 I see it.

Should be available in experimental now.

Let us know if that solves your problem.

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: getting rid of the gnome-shell network-manager dependency

[Michael Biebl]

On 04.04.2012 10:40, Miles Bader wrote:
 Any idea when gnome-shell will be fixed to not depend on
 network-manager being installed?

yesterday

 [gnome-shell dies at startup unless network-manager is installed, but
 network-manager interacts very badly with NFS, so installing NM is at
 best a painful workaround...]

Instead of whining, you could have checked the BTS


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: getting rid of the gnome-shell network-manager dependency

[Michael Biebl]

On 04.04.2012 16:07, Miles Bader wrote:
 Michael Biebl bi...@debian.org writes:
 Any idea when gnome-shell will be fixed to not depend on
 network-manager being installed?

 yesterday
 
 Hmm, it still wasn't fixed in my update today ... maybe it hasn't
 percolated through to the mirror I use?

network-manager (0.9.4.0-3) unstable; urgency=low

  * debian/patches/12-initialize-gerror.patch: Initialize GError, else invalid
free() crash can occur. Fixes the crash in gnome-shell if NetworkManager
is not running. Patch cherry-picked from upstream. (Closes: #665791)
Thank a lot to Antti-Juhani Kaijanaho for tracking down this problem!
  * Remove explicit Build-Depends on gir1.2-glib-2.0 and gir1.2-freedesktop,
let libgirepository1.0-dev pull those dependencies.

 -- Michael Biebl bi...@debian.org  Tue, 03 Apr 2012 20:20:13 +0200

You still might have the old version of the library loaded.
So after upgrading to 0.9.4.0-3, you need to restart
gnome-shell.

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: getting rid of the gnome-shell network-manager dependency

[Michael Biebl]

On 05.04.2012 05:40, Miles Bader wrote:

 I'm a little confused by this ... since this fix seems to be in NM,
 does this mean that network-manager still needs to be installed, but

This fix was in libnm-glib4, and libnm-glib4 needs to be installed as
gnome-shell links against it.


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: automount under gnome 3

[Michael Biebl]

On 07.01.2012 10:55, Takis Diakoumis wrote:
 
 i'm still tweaking a little and can't seem to get auto-mount to work when
 inserting or booting with usb drives attached.
 

Do you mean, the USB stick is already attached before you login?

Does it work correctly, when you plugin the USB stick while you are
logged in?

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

binNMUs for incorrect shlibs in libvte9

[Michael Biebl]

Hi,

libvte9 had incorrect shlibs information which lead to dependencies not
being strict enough [1].
I've uploaded a fixed package in the mean time and checked which rdeps
of libvte use the new API.
I'd appreciate if you could schedule binNMUs fro the following packages,
so they pick up the correct dependency:

 cairo-dock-plug-ins_2.4.0~2-1
 cdebconf-terminal_0.15
 evilvte_0.5.0-1
 mssh_1.2-1.1
 roxterm_1.22.2-1
 ruby-gnome2_1.0.3-1
 sakura_2.4.2-1
 xfce4-terminal_0.4.8-1

gnome-desktop-sharp2_2.26.0-7

As the ia64 build has not yet been upload, please add a dep-wait against
libvte-dev 1:0.28.2-4

Thanks,
Michael

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=633798

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Distinct names for GNOME 3 packages

[Michael Biebl]

Am 03.11.2011 23:32, schrieb Robert Ross:
 Is it too late to request distinct names for GNOME 3 packages?

This boat has sailed, I'd say.
The only case where this was used was gdm and gdm3.
We also had this case for source package names, where we needed
co-installability, like libwnck and libwnck3.

 show time servers in version 2.30.2 but not 3.0.0 count as a
 missing feature, a feature revision (perhaps the list was
 intentionally moved elsewhere), or a regression bug?

Some of them will be bugs, some design decisions. I guess we need to
discuss this on a case by case basis.

For gnome-system-tools, even if it got a GNOME 3 port, this package has
reached EOL.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: RFS: gedit-r-plugin (updated package, needed for gedit 3.0 transition, NMU)

[Michael Biebl]

Hi Tobias,

Am 04.10.2011 03:01, schrieb Tobias Hansen:
 Dear mentors,
 
 I am looking for a sponsor for my package gedit-r-plugin.
 
   * Package name: gedit-r-plugin
 Version : 0.7.1.0-Gtk3-1.0
 Upstream Author : Dan Dediu
   * URL : http://rgedit.sourceforge.net/
   * License : GTK-3+
 Section : gnome
 
 It builds those binary packages:
 
 gedit-r-plugin - Gedit plugin for R statistical computing language
 
 I contacted the maintainer, but he didn't answer and seems to be inactive.
 A copy of this mail is in the bug report of the gedit transition bug:
 

You changed the debian revision to 1.0, to suppress a lintian warning.

The correct debian revision number in this case (NMU) would be 0.1



Cheers,
Michael


[1]
http://anonscm.debian.org/gitweb/?p=collab-maint/gedit-r-plugin.git;a=commitdiff;h=f1fbfa1d39388afa38db205a1fbfa65f618127f5
[2] 
http://www.debian.org/doc/manuals/developers-reference/pkgs.html#nmu-changelog

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Ephiphany backport?

[Michael Biebl]

Am 26.09.2011 15:27, schrieb Paul van der Vlis:
 Hello,
 
 Epiphany in Squeeze is not very good.
 
 Do you think it would be possible to make a backport from Epiphany 3.04
 for Squeeze without big upgrades like GTK and Gnome?

Not possible without also backporting a lot of GNOME 3 dependencies, like GTK 3.
So this is most likely not going to happen.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Bug#640082: Please start xfce4-notifyd via the session manager

[Michael Biebl]

Package: xfce4-notifyd
Version: 0.2.2-1
Severity: normal

Hi,

there have been some recent changes in notification-daemon and the way
it is started. It previously was started via dbus activation on-demand.
This had the unpleasant side effect, that shipping different
notification systems either had to conflict with each other or divert
the dbus service file (or ship it under a different dbus service file
name where it no longer was clear which service would be started then).

In GNOME 3, the notification system is integrated into gnome-shell itself,
which is komparable with how kdebase-workspace does it.

As a consequence, the dbus service file was removed from
notification-daemon, as this could otherwise lead to a race when
gnome-shell is started [1]. In the GNOME 3 fallback or GNOME 2 case, we
start notification-daemon via the session manager now.

For GNOME 2, we ship a autostart file in gnome-session 2 in
/usr/share/gnome/autostart. gnome-session 3 will start
notification-daemon in the fallback case and we also ship a XDG
autostart file, which should handle non-GNOME environments.

It currently lists:
OnlyShowIn=LXDE;OPENBOX;GNOME;
AutostartCondition=GNOME3 unless-session gnome

I.e. XFCE is explicitly excluded, as I assumed that for XFCE the
preferred notification system is xfce4-notifyd.

We have been discussing that on IRC for a while, and we came to the
conclusion, that it should be up to the individual desktop environment
to depend on its notification system of choice.
The notification daemons should no longer be started via dbus activation
and instead by the session manager, so we have a defined and reliabe
behaviour.

If you can agree with that, please update xfce4-notifyd accordingly and
make the xfce4 metapackage depend/recommend xfce4-notifyd.

I also think, that individual applications should not directly depend on
notification-daemon directly. But this should be discussed / fixed
independently from this bug.

Cheers,
Michael


[1]
http://git.gnome.org/browse/notification-daemon/commit/?id=1ad20d22098bc7718614a8a87744a2c22d5438d0


-- System Information:
Debian Release: wheezy/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (200, 'experimental')
Architecture: i386 (i686)

Kernel: Linux 3.0.0-1-486
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages xfce4-notifyd depends on:
ii  libc6   2.13-18   
ii  libcairo2   1.10.2-6.1
ii  libdbus-1-3 1.4.14-1  
ii  libdbus-glib-1-20.94-4
ii  libgdk-pixbuf2.0-0  2.23.5-3  
ii  libglib2.0-02.28.6-1  
ii  libgtk2.0-0 2.24.5-4  
ii  libpango1.0-0   1.28.4-3  
ii  libx11-62:1.4.4-1 
ii  libxfce4ui-1-0  4.8.0-3   
ii  libxfce4util4   4.8.1-3   
ii  libxfconf-0-2   4.8.0-3   

Versions of packages xfce4-notifyd recommends:
ii  libnotify-bin  0.7.3-2

xfce4-notifyd suggests no packages.



-- 
To UNSUBSCRIBE, email to debian-gtk-gnome-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/20110902025729.5649.62309.report...@pluto.milchstrasse.xx

Re: Bug#637331: transition: nautilus

[Michael Biebl]

Am 15.08.2011 22:26, schrieb Raphael Hertzog:
 On Fri, 12 Aug 2011, Josselin Mouette wrote:
 Le jeudi 11 août 2011 à 20:31 +0200, Michael Biebl a écrit : 
 The alternative is to backport the necessary changes in g-s-d (which
 sounds insane) or to re-add the necessary support in nautilus (which is
 far from trivial).

 I'd rather just have automount support broken temporarily, tbh.

 I don’t think there is any point in testing if we break such basic
 functionality for a long period of time. But that’s for the RT to decide
 anyway.
 
 +1 That's not the kind of brokenness that we want in testing.

I would consider non-functioning automount for a (short) period of time as
acceptable compromise, but ymmv.

 It's nice to be able to split transitions, but when it's not possible our
 infrastructure should cope with large transitions.

So do you want to wait until this magically happens? When would that be?
Right now, transitions are quite a bit of manual work and the more transitions
are entangled the harder it gets.

 With everything more or less working in experimental, what exactly do you
 fear?

Not by far. See the comments about e.g. libpanel-applet. There is an awful lot
of work to be done to get everything into shape.

The nautilus transitions is already complex enough imho (involving a
mini-transition for evince, brasero, gnome-media, libgdata and possibly
gnome-bluetooth and nautilus-sendto).

Personally I'd be fine with dumping everything into unstable, but imho we could
have a transition which would last over several weeks or months and which
certainly would block other transitions.

That said, it's not really up to me to decide and I'll just wait now what the
release team wants us to do.

Cheers,
Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Bug#637331: transition: nautilus

[Michael Biebl]

Am 11.08.2011 16:42, schrieb Julien Cristau:
 On Wed, Aug 10, 2011 at 15:36:15 +0200, Michael Biebl wrote:
 
 - Updating evince from 2.32 to 3.0 involves a soname bump of libevince3 →
 libevince3-3. The only rdep of libevince3 is python-evince
 (src:gnome-python-desktop).  python-evince has been deprecated, so there is 
 no
 update for GNOME 3 and we most likely need to drop it if we are going to 
 update
 evince.
 That means rdeps on python-evince need to be updated too. Affected packages:
 sugar-read-activity-0.84
 sugar-read-activity-0.86

 Have bugs been filed against those sugar packages?

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=637823
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=637825

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Bug#637331: transition: nautilus

[Michael Biebl]

Package: release.debian.org
Severity: normal
User: release.debian@packages.debian.org
Usertags: transition

Hi,

as next step of getting GNOME 3 into unstable, we'd like to request a
slot to start the nautilus 3 transition
The API of libnautilus-extension hasn't changed with version 3.0, but it
uses GTK 3 now which changes its ABI and a process can't be using GTK 2
and GTK 3 at the same time, as this will lead to an application abort [1].
As a result, we decided to change the binary package name to
libnautilus-extension1a.

Packages which are updated to libnautilus-extension 3.0 need to also be
updated to GTK 3 (GNOME 3).

I've filed corresponding bug reports for the affected packages:

brasero
deja-dup
diff-ext
eiciel
evince
file-roller
gksu
gnome-disk-utility
gnome-user-share
gtkhash
nautilus-actions
nautilus-dropbox
nautilus-filename-repairer
nautilus-image-converter
nautilus-open-terminal
nautilus-python
nautilus-sendto
nautilus-share
seahorse-plugins
totem
tracker

Those bugs are also user-tagged [2]. Some of those packages are already
updated in experimental. I will mark those bug reports accordingly.

- gnome-user-share uses libgnome-bluetooth 3.0, which would mean
  entangling those two transition, so I'd suggest temporarily disabling
  bluetooth support in gnome-user-share.
- totem 3.0 (from exp) requires a newer libgdata (involving a soname
  bump). We probably need to do this transition first. I'll file a
  separate bug for that.
- Updating nautilus-sendto to 3.0 (available in exp) will break the
  plugins shipped by empathy and gnome-bluetooth:
  gnome-bluetooth: /usr/lib/nautilus-sendto/plugins/libnstbluetooth.so
  nautilus-sendto-empathy: /usr/lib/nautilus-sendto/plugins/libnstempathy.so
  I'd prefer to temporarily break those plugins in nautilus-sendto to
  not entangle this transition with an empathy 3.0 or gnome-bluetooth
  3.0 update, but some of the pkg-gnome team would like to at least
  update empathy. So input from the release team on this matter would be
  appreciated.

I've also started collecting notes about this transition at [3].

Cheers,
Michael



[1] http://developer.gnome.org/gtk3/3.0/ch25s02.html#id1493823
[2]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=pkg-gnome-maintain...@lists.alioth.debian.org;tag=nautilus3-transition
[3] http://whiteboard.debian.net/nautilus3.wb

-- System Information:
Debian Release: wheezy/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (200, 'experimental')
Architecture: i386 (i686)

Kernel: Linux 3.0.0-1-486
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash



-- 
To UNSUBSCRIBE, email to debian-gtk-gnome-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/20110810131157.15212.68216.report...@pluto.milchstrasse.xx

Re: Bug#637331: transition: nautilus

[Michael Biebl]

Am 10.08.2011 15:11, schrieb Michael Biebl:
 I've filed corresponding bug reports for the affected packages:
 
 brasero
 deja-dup
 diff-ext
 eiciel
 evince
 file-roller
 gksu
 gnome-disk-utility
 gnome-user-share
 gtkhash
 nautilus-actions
 nautilus-dropbox
 nautilus-filename-repairer
 nautilus-image-converter
 nautilus-open-terminal
 nautilus-python
 nautilus-sendto
 nautilus-share
 seahorse-plugins
 totem
 tracker

I forgot the following two issues:
- Updating evince from 2.32 to 3.0 involves a soname bump of libevince3 →
libevince3-3. The only rdep of libevince3 is python-evince
(src:gnome-python-desktop).  python-evince has been deprecated, so there is no
update for GNOME 3 and we most likely need to drop it if we are going to update
evince.
That means rdeps on python-evince need to be updated too. Affected packages:
sugar-read-activity-0.84
sugar-read-activity-0.86

- Updating brasero from 2.30 to 3.0 involves a soname bump of libbrasero-media0
→ libbrasero-media3-1.
Affected packages:
goobox
gthumb
python-brasero (src:gnome-python-desktop)
rhythmbox-plugin-cdrecorder (src:rhythmbox)
sound-juicer

I currently don't see an alternative from doing these transitions in one go.

Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3 in Sid?

[Michael Biebl]

Am 01.08.2011 11:02, schrieb Tshepang Lekhonkhobe:
 Hi,
 
 What is holding back GNOME 3 from entering Sid?

Getting those various transitions done:

http://lists.debian.org/debian-release/2011/04/msg00111.html







signature.asc
Description: OpenPGP digital signature

Re: remove resize handle in gnome-terminal 3?

[Michael Biebl]

Am 14.07.2011 11:31, schrieb Miles Bader:
 Michael Biebl bi...@debian.org writes:
 If you care, you can forward this issue upstream and tag the debian bug
 accordingly.
 
 How does one tag a bug accordingly...?

http://www.debian.org/Bugs/server-control#forwarded


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Analysis of the gnome3 transitions

[Michael Biebl]

Am 13.04.2011 15:03, schrieb Mehdi Dogguy:
 Sorry for the late reply, and thanks for this detailed analysis. It's very
 helpful!
 
 On  0, Josselin Mouette j...@debian.org wrote:
 Hi again,

 not all modules are ready, but I’ve had a closer look at what the GNOME
 3 library transitions imply and how we can deal with their testing
 migrations in a sane way.


..

 3. devhelp
 It is almost a self-contained transition (devhelp + anjuta).
 AFAICT it can be done at any time.

 
 and, anjuta-extras I guess. If everything needed is already ready in
 sid/wheezy, you can go ahead with this.

This took a bit longer, since we needed webkit 1.4 for this.
As soon as webkit has been succefully built on all archs[1] I'm planning to
start this transition.

This would involve:
devhelp, anjuta, anjuta-extras, gtkpod (libanjuta rdep)

We already got a go for this transition, but it's been some time since then,
so I'd like a re-affirmation.

Please shout if we should hold back with starting this mini transition now.

Cheers,
Michael

[1] https://buildd.debian.org/status/package.php?p=webkitgtk%2Bsuite=sid
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Analysis of the gnome3 transitions

[Michael Biebl]

Am 12.07.2011 09:01, schrieb Michael Biebl:
 Am 13.04.2011 15:03, schrieb Mehdi Dogguy:

 3. devhelp
 It is almost a self-contained transition (devhelp + anjuta).
 AFAICT it can be done at any time.


 and, anjuta-extras I guess. If everything needed is already ready in
 sid/wheezy, you can go ahead with this.
 
 This took a bit longer, since we needed webkit 1.4 for this.
 As soon as webkit has been succefully built on all archs[1] I'm planning to
 start this transition.
 
 This would involve:
 devhelp, anjuta, anjuta-extras, gtkpod (libanjuta rdep)
 

I forgot to add: All those package are ready in experimental and just need to be
moved to sid.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Please transition gdk-pixbuf for multiarch

[Michael Biebl]

Hi Steve,

i tried to test the patch but it not only contains some cruft (.pc) but also
does not apply cleanly on top of 2.23.5-1. Would you mind sending an updated 
patch?

Does this affect packages installing pixbuf loaders, like libwmf0.2-7 and
librsvg2-common? Do you have to take special care about upgrading those 3
packages in lock step?


That said, I have a more general question:
gtk2.0 and gtk3.0 ship a variant of this ugly 041_ia32-libs.patch found in
gdk-pixbuf:
- debian/patches/041_ia32-libs.patch (gtk+2.0) [1]
- debian/patches/041_ia32-libs.patch (gtk+3.0) [2]

I don't see any multiarch related bug reports for gtk2/gtk3 yet and wondered if
you have patches for them too and how this would affect other packages
installing gtk engines or (im)modules?
Do we need to be careful in which order we transition all those packages?


Thanks again for working on this,
Michael


[1]
http://patch-tracker.debian.org/patch/series/view/gtk+2.0/2.24.5-1/041_ia32-libs.patch
[2]
http://patch-tracker.debian.org/patch/series/view/gtk+3.0/3.0.10-1/041_ia32-libs.patch
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: gnome-terminal 3.0 already in Unstable?

[Michael Biebl]

Am 26.06.2011 11:57, schrieb Tshepang Lekhonkhobe:
 Hi,
 
 Was it a mistake that GNOME 3's gnome-terminal got uploaded to Sid?

No, it wasn't a mistake. I've been gradually uploading GNOME3 packages to
unstable for which all dependencies are there yet.

 say it because a lot of the stuff is still in Experimental, and the it
 looks out-of-place (the theme) with the rest of Sid GNOME desktop.

Choose the Adwaita Theme, which should work reasonable well in GNOME2 session
with GTK3 apps.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Analysis of the gnome3 transitions

[Michael Biebl]

Am 07.04.2011 00:50, schrieb Josselin Mouette:

 If you know of other libraries that will be involved, please speak up
 now. If you have read the whole mail, congratulations.

* network-manager 0.9

network-manager 0.8.4 is currently in unstable, 0.9 rcS in experimental.

GNOME 3 requires NM 0.9 (gnome-control-center 3.0 at least).

Fortunately the KDE plasma widget for NM 0.9 is getting into a usable shape.
What's still missing in plasma-widget-networkmanagement is support for migrating
0.8 settings to 0.9 [1]

Aside from that, we have a soname bump for libnm-glib and libnm-util, so rdeps
at least need a rebuild. I haven't made a test build of affected packages, but
can do so.

There is also a small change in the D-Bus API regarding offline/online network
status detection [2]


* tracker 0.10
libtracker-client-0.8 in unstable is used by
  - gtk+2.0 (via dlopen), nautilus (via dlopen)
  - bognor-regis, brasero, hornsey, totem-plugins

tracker-0.10 uses the library name libtracker-client-0.10, and the pkg-config
file was renamed too. So in most cases a sourceful upload is required.

I started filing bugs for that [3].

totem 3.0 (exp) is already updated to use tracker 0.10 (using libtracker-sparql)
nautilus 3.0 (exp) already supports tracker 0.10 (using libtracker-sparql)
gtk+2.0 can be patched to support both tracker-0.8 and tracker-0.10


* poppler 0.16

To update evince, we need a newer poppler (= 0.15).
I've been working on that with Pino and I expect him to send a separate email to
debian-release about this.
The current progress is tracked via [4] and we managed to make all but 3
packages binNMUable (xpdf, python-poppler, poppler-sharp).
I think we should start the poppler transition as soon as possible, as other
transitions will depend on it (e.g. evince, tracker-0.10)



Michael



[1] http://live.gnome.org/NetworkManager/ApiSimplify
[2] https://fedorahosted.org/fesco/ticket/572#comment:50
[3]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=bi...@debian.org;tag=tracker-0.10
[4]
http://bugs.debian.org/cgi-bin/pkgreport.cgi?users=toscano.p...@tiscali.it;tag=poppler-0.16
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

How to omit --dbg-package for certain binary packages

[Michael Biebl]

Hi Jonas,

in the pkg-gnome team we use cdbs and we have several packages which build
different flavors (eg. deb/udeb builds or gtk/gnome builds).

As an example

Binary: evince
Binary: evince-gtk
Binary: evince-dbg

envince-gtk is a flavour of the evince build with different configure options
(basically all gnome deps disabled).

Now, as debian/control contains a evince-dbg package, cdbs will call
dh_strip --dbg-package=evince-dbg for *all* binary packages.
In case of evince-gtk, we don't want that, as this would overwrite the detached
debugging symbols of the evince binary package.

For this we used a hack:

DEB_DH_STRIP_ARGS_evince-gtk = --dbg-package=

This no longer works with debhelper compat level 8 and one gets
Option dbg-package requires an argument
when calling dh_strip -pevince-gtk --dbg-package=

Do you have a suggestion, how to omit the --dbg-package option for evince-gtk in
a clean way using cdbs?

Thanks,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

libimobiledevice : unstable upload

[Michael Biebl]

Hi Julien,

I was wondering, when you plan to upload libimobiledevice 1.1.1-1 to unstable.

Currently, gvfs 1.8 and g-d-u 3.0 are blocked because of this, so it would be
nice to have the newer libimobiledevice soon.

Have you checked the rdeps of libimobiledevice if they are binNMUable?
I guess we should also inform the release team, if if this looks like a smaller
transition.

Cheers,
Michael



signature.asc
Description: OpenPGP digital signature

Bug#627735: dh_installgsettings should run after dh_install

[Michael Biebl]

Package: cdbs
Version: 0.4.93
Severity: important

Hi,

the current dh_installgsettings integration is broken, as
dh_installgsettings is run *before* dh_install.

See the code in dh_installgsettings:

# Get a list of the schemas
my $schemas = `find $gsettings_schemas_dir -type f \\( -name \\*.xml -o 
-name \\*.override \\) -printf '%P '`;
if ($schemas ne '') {
addsubstvar($package, misc:Depends, dconf-gsettings-backend 
| gsettings-backend);
}

If dh_installgsettings is run before dh_install, $schemas will be empty and
the dependency on on gsettings-backend via misc:Depends is not
generated.

I'd suggest running dh_installgsettings at the same time when dh_gconf
is run, as dh_gconf has basically the same logic.

Cheers,
Michael

-- System Information:
Debian Release: wheezy/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (200, 'experimental')
Architecture: i386 (i686)

Kernel: Linux 2.6.39-1-486
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages cdbs depends on:
ii  debhelper 8.1.6  helper programs for debian/rules

Versions of packages cdbs recommends:
ii  autotools-dev 20110511.1 Update infrastructure for config.{

Versions of packages cdbs suggests:
ii  devscripts2.10.73scripts to make the life of a Debi

-- no debconf information



-- 
To UNSUBSCRIBE, email to debian-gtk-gnome-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/20110524032154.30273.48227.report...@pluto.milchstrasse.xx

vala / test suite failures on kfreebsd / ongoing transitions

[Michael Biebl]

Hi Sebastian,

as you are probably aware, with the upload of vala 0.10 to unstable, the test
suite failure [1] on kfreebsd has become a major issue.
E.g. tracker requires valac to build, and as a result blocks the ongoing evo/eds
transition, as tracker also builds an evo plugin.
I suggested making the test suite failures non-fatal, but pochu was against that
and would rather see the test suite failures fixed.

Unfortunately my vala-fu is not strong enough.
Could you please look after this or suggest how you want this to be fixed?

Thanks,
Michael

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=589460
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: GNOME 3 and panel applets

[Michael Biebl]

Am 15.02.2011 09:59, schrieb Michael Biebl:
 Am 14.02.2011 18:17, schrieb Josselin Mouette:
 

 Michael Biebl bi...@debian.org
tracker
 
 The search bar applet can be disabled via a configure switch.
 I still do hope that search (and therefore) tracker is integrated more deeply
 into the new gnome shell so such a applet would be kinda obsolete anyway.

That said, the upstream unstable branch of tracker apparently has a port of
tracker-search-bar using libpanelapplet-3.0, which I assume is the new D-Bus
based panel in GNOME 3.0.

I can't seem to find such a package in Debian just yet.
Joss, when do you plan that this will be available in unstable or experimental?

Cheers,
Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Permission to upload gnome-power-manager 2.32.0 to unstable (Was: Re: Dropping libdevkit-power-gobject for squeeze.)

[Michael Biebl]

Hi Adam!

On 23.10.2010 19:50, Adam D. Barratt wrote:
 On Wed, 2010-10-20 at 00:38 +0200, Michael Biebl wrote:
 Done, 2.32.0-2 has been uploaded to unstable.


 With all rdeps updated, attached is the debdiff for upower_0.9.5-5 which 
 removes
 the deprecated libdevkit-power-gobject library, closing #595083.

 I've also included a fix for #586751, the patch is cherry-picked from 
 upstream
 Git and rather simple.
 
 Please go ahead.

Done, package has been uploaded as 0.9.5-5

unblock upower/0.9.5-5

Thanks,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Permission to upload gnome-power-manager 2.32.0 to unstable (Was: Re: Dropping libdevkit-power-gobject for squeeze.)

[Michael Biebl]

On 19.10.2010 21:52, Adam D. Barratt wrote:
 On Mon, 2010-10-18 at 23:12 +0200, Michael Biebl wrote:
 On 18.10.2010 23:05, Michael Biebl wrote:

  71 files changed, 3670 insertions(+), 4923 deletions(-)

 Looks a bit scary at first. The high number of deleted lines is due to
 libhal-glib being removed (as it is no longer used), and if you don't count 
 the
 very verbose Changelog, the diffstat looks a bit more sane.

 And the vast majority of the changes come from the switch from
 libdevkit-gobject-power to libupower-glib being a trivial search and replace.
 As you can see from the debdiff, it's mostly a s/dkp_/up_/ and s/Dkp/Up/
 
 Please go ahead.

Done, 2.32.0-2 has been uploaded to unstable.


With all rdeps updated, attached is the debdiff for upower_0.9.5-5 which removes
the deprecated libdevkit-power-gobject library, closing #595083.

I've also included a fix for #586751, the patch is cherry-picked from upstream
Git and rather simple.

Please let me know if I can upload upower to unstable, so this task can be 
closed.


Thanks for your hard work,

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?
reverted:
--- upower-0.9.5/debian/libdevkit-power-gobject-dev.install
+++ upower-0.9.5.orig/debian/libdevkit-power-gobject-dev.install
@@ -1,3 +0,0 @@
-usr/lib/libdevkit-power-gobject.{so,a}
-usr/lib/pkgconfig/devkit-power-gobject.pc
-usr/include/DeviceKit-power/devkit-power-gobject/
reverted:
--- upower-0.9.5/debian/libdevkit-power-gobject1.install
+++ upower-0.9.5.orig/debian/libdevkit-power-gobject1.install
@@ -1 +0,0 @@
-usr/lib/libdevkit-power-gobject.so.*
diff -u upower-0.9.5/debian/rules upower-0.9.5/debian/rules
--- upower-0.9.5/debian/rules
+++ upower-0.9.5/debian/rules
@@ -7,12 +7,6 @@
 
 DEB_CONFIGURE_EXTRA_FLAGS = --enable-gtk-doc
 
-# upower's version number is smaller than devicekit-power's, so we need to
-# epoch the transitional packages
-DEB_DH_GENCONTROL_ARGS_libdevkit-power-gobject1 = -- -v1:$(DEB_VERSION)
-DEB_DH_MAKESHLIBS_ARGS_libdevkit-power-gobject1 = -- -v1:$(DEB_VERSION)
-DEB_DH_GENCONTROL_ARGS_libdevkit-power-gobject-dev = -- -v1:$(DEB_VERSION)
-
 common-post-build-arch::
 	make -C po upower.pot
 
diff -u upower-0.9.5/debian/control upower-0.9.5/debian/control
--- upower-0.9.5/debian/control
+++ upower-0.9.5/debian/control
@@ -76,29 +75,0 @@
-Package: libdevkit-power-gobject1
-Architecture: any
-Section: oldlibs
-Depends: ${shlibs:Depends}, ${misc:Depends}
-Conflicts: devicekit-power
-Description: abstraction for power management - shared library (old ABI)
- upower provides an interface to enumerate power sources on the system
- and control system-wide power management. Any application can access the
- org.freedesktop.UPower service on the system message bus. Some
- operations (such as suspending the system) are restricted using PolicyKit.
- .
- This package contains an old version of the shared library to be used by
- applications. It provides the same ABI as the old DeviceKit-power library, but
- works with upower.
-
-Package: libdevkit-power-gobject-dev
-Architecture: any
-Section: oldlibs
-Depends: ${shlibs:Depends}, ${misc:Depends}, libglib2.0-dev, libdevkit-power-gobject1 (= ${binary:Version})
-Description: abstraction for power management - development files (old API)
- upower provides an interface to enumerate power sources on the system
- and control system-wide power management. Any application can access the
- org.freedesktop.UPower service on the system message bus. Some
- operations (such as suspending the system) are restricted using PolicyKit.
- .
- This package contains development files for the  old version of the shared
- library. It provides the same API as the old DeviceKit-power library, but
- works with upower.
-
diff -u upower-0.9.5/debian/changelog upower-0.9.5/debian/changelog
--- upower-0.9.5/debian/changelog
+++ upower-0.9.5/debian/changelog
@@ -1,3 +1,14 @@
+upower (0.9.5-5) UNRELEASED; urgency=low
+
+  * debian/patches/00git_dont_poll_serial_port_if_no_wup_adaptor.patch
+- Do not continue to poll the serial port if there is no Watts Up Pro
+  adaptor. Patch cherry-picked from upstream Git. (Closes: #586751)
+  * Drop libdevkit-power-gobject library packages as all reverse dependencies
+have been updated to use the new API provided by libupower-glib.
+(Closes: #595083)
+
+ -- Michael Biebl bi...@debian.org  Wed, 20 Oct 2010 00:15:28 +0200
+
 upower (0.9.5-4) unstable; urgency=low
 
   * Add 00git_initialize_polkit_gerror.patch: Ensure we've initialized errors
reverted:
--- upower-0.9.5/debian/libdevkit-power-gobject1.symbols
+++ upower-0.9.5.orig/debian/libdevkit-power-gobject1.symbols
@@ -1,58 +0,0 @@
-libdevkit-power-gobject.so.1 libdevkit-power-gobject1 #MINVER#
- dkp_client_can_hibern...@base 1:0.9.1
- dkp_client_can_susp...@base 1:0.9.1
- dkp_client_enumerate_devi...@base 1:0.9.1
- dkp_client_get_daemon_vers...@base 1:0.9.1
- dkp_client_get_t...@base 1:0.9.1

Re: Permission to upload gnome-power-manager 2.32.0 to unstable (Was: Re: Dropping libdevkit-power-gobject for squeeze.)

[Michael Biebl]

On 18.10.2010 23:05, Michael Biebl wrote:

  71 files changed, 3670 insertions(+), 4923 deletions(-)
 
 Looks a bit scary at first. The high number of deleted lines is due to
 libhal-glib being removed (as it is no longer used), and if you don't count 
 the
 very verbose Changelog, the diffstat looks a bit more sane.

And the vast majority of the changes come from the switch from
libdevkit-gobject-power to libupower-glib being a trivial search and replace.
As you can see from the debdiff, it's mostly a s/dkp_/up_/ and s/Dkp/Up/

Cheers,
Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Permission to upload gnome-power-manager 2.32.0 to unstable (Was: Re: Dropping libdevkit-power-gobject for squeeze.)

[Michael Biebl]

Hi Adam,

thanks for unblocking tracker 0.8.17-1

On 10.10.2010 16:14, Adam D. Barratt wrote:
 On Wed, 2010-09-15 at 08:56 +0200, Michael Biebl wrote:
 
 Currently there are 4 packages remaining (in squeeze) that still depend on
 libdevkit-power-gobject:
 - gdm3 2.30.2-4
 - tracker 0.8.15-1
 - gnome-session 2.30.2-1
 - gnome-power-manager 2.30.1-1
 
 gnome-session 2.30.2-2 is now in testing, and no longer depends on
 libdk-p-g; gdm3 is unblocked and waiting to age.  As discussed on IRC, a
 g-p-m 2.32 upload may be possible, depending on what the diff looks
 like.


So, the last remaining piece is gnome-power-manager.

I've uploaded gnome-power-manager 2.32.0-1 to experimental in the mean time.
buildd status looks fine so far and I've been testing it for quite a few
days
now without encountering any problems.

I tried to keep the packaging related changes in debian/ minimal, the
changelog
reads:


gnome-power-manager (2.32.0-1) experimental; urgency=low

  [ Josselin Mouette ]
  * Move autostart file to /usr/share/gnome/autostart since Xfce now has»
its own power manager. Closes: #591776.

  [ Michael Biebl ]
  * New upstream release.
- Port to libupower-glib. Closes: #595086
- Provide a pkexec helper for systems that do not have XRandR backlight
  capability, i.e. we no longer require hal.
  * debian/control.in
- Drop Build-Depends on libhal-dev.
- Change Build-Depends on libdevkit-power-gobject-dev to
  libupower-glib-dev (= 0.9.1).
- Drop Suggests: hal.
- Add Suggests: policykit-1 for pkexec.
- Bump Standards-Version to 3.9.1. No further changes.
  * debian/rules
- Drop obsolete --enable-hal from DEB_CONFIGURE_EXTRA_FLAGS.

 -- Michael Biebl bi...@debian.org  Sun, 17 Oct 2010 19:39:48 +0200

The diffstat (without translation and auto* crap):

 configure.ac |   50
 gnome-power-manager-2.32.0/ChangeLog |  988
+-
 gnome-power-manager-2.32.0/Makefile.am   |   14
 gnome-power-manager-2.32.0/NEWS  |
 116 +
 gnome-power-manager-2.32.0/README|8
 gnome-power-manager-2.32.0/applets/brightness/brightness-applet.c|4
 gnome-power-manager-2.32.0/applets/brightness/gpm-common.h   |5
 gnome-power-manager-2.32.0/applets/inhibit/gpm-common.h  |5
 gnome-power-manager-2.32.0/applets/inhibit/inhibit-applet.c  |4
 gnome-power-manager-2.32.0/data/gnome-power-manager.desktop.in   |2
 gnome-power-manager-2.32.0/data/gnome-power-manager.schemas.in   |4
 gnome-power-manager-2.32.0/data/gnome-power-preferences.desktop.in   |2
 gnome-power-manager-2.32.0/data/gnome-power-statistics.desktop.in|3
 gnome-power-manager-2.32.0/data/gnome-power-statistics.desktop.in.in |1
 gnome-power-manager-2.32.0/debian/changelog  |   23
 gnome-power-manager-2.32.0/debian/control|7
 gnome-power-manager-2.32.0/debian/control.in |7
 gnome-power-manager-2.32.0/debian/postinst   |   13
 gnome-power-manager-2.32.0/debian/rules  |7
 gnome-power-manager-2.32.0/gnome-doc-utils.make  |
 157 -
 gnome-power-manager-2.32.0/help/sv/gnome-power-manager.xml   |   10
 gnome-power-manager-2.32.0/help/zh_CN/gnome-power-manager.xml|   94
 gnome-power-manager-2.32.0/libhal-glib/Makefile.am   |   52
 gnome-power-manager-2.32.0/libhal-glib/egg-dbus-monitor.c|
 251 --
 gnome-power-manager-2.32.0/libhal-glib/egg-dbus-monitor.h|   65
 gnome-power-manager-2.32.0/libhal-glib/egg-dbus-proxy.c  |
 301 ---
 gnome-power-manager-2.32.0/libhal-glib/egg-dbus-proxy.h  |   66
 gnome-power-manager-2.32.0/libhal-glib/egg-debug.c   |
 308 ---
 gnome-power-manager-2.32.0/libhal-glib/egg-debug.h   |   83
 gnome-power-manager-2.32.0/libhal-glib/hal-device.c  |
 260 --
 gnome-power-manager-2.32.0/libhal-glib/hal-device.h  |   71
 gnome-power-manager-2.32.0/libhal-glib/hal-manager.c |
 357 ---
 gnome-power-manager-2.32.0/libhal-glib/hal-manager.h |   86
 gnome-power-manager-2.32.0/libhal-glib/hal-marshal.c |
 325 ---
 gnome-power-manager-2.32.0/libhal-glib/hal-marshal.h |   83
 gnome-power-manager-2.32.0/libhal-glib/hal-marshal.list  |9
 gnome-power-manager-2.32.0/ltmain.sh |8
 gnome-power-manager-2.32.0/man/Makefile.am   |6
 gnome-power-manager-2.32.0/po/Makefile.in.in |4
 gnome-power-manager-2.32.0/po/POTFILES.in|4
 gnome-power-manager-2.32.0/policy/Makefile.am

Re: on tracker-gui

[Michael Biebl]

On 04.10.2010 11:02, Tshepang Lekhonkhobe wrote:
 Hi,
 
 Josselin once mentioned that it's preferable to have a meta-gnome2
 alternatives if the new entrant is a drop-in replacement. In the case
 of meta-gnome2 2.30+2, tracker-gui is set to be an alternative of
 gnome-search-tool, even though it's not even close to
 gnome-search-tool offers

Well, gnome-search-tool does offer more options to *refine* the search
But full-text search is dog slow, whereas t-s-t can offer you a search as you
type interface even for full-text search.

It should be easily doable to add the filters, that g-s-t offers, to t-s-t.

But as you already said, development of t-s-t is a bit stalled.

Just my 2¢: having a dedicated search application is nice, but much more
important is integrated search (and tagging) everywhere and this is the focus of
tracker. The integration of tracker search in nautilus and the gtk file selector
is just a start.


Cheers,
Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: A first look at GNOME 2.32 - squeeze status

[Michael Biebl]

On 14.09.2010 19:25, Josselin Mouette wrote:

 * gnome-disk-utility : OK

upstream dropped the complete API documentation for libgdu, so pitti and I are
currently discussing, if the updated translations and the few bug fixes are
worth it.
Most likely not.

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

tracker 0.8 transition

[Michael Biebl]

Hi release team,

tracker 0.8.0 was released two weeks ago. It's a major new release with a lot of
features I'd like to see in squeeze. Another point is, that upstream considers
the old stable branch 0.6 (current version in sid is 0.6.96) dead.
Since the initial 0.8.0 release there was a new bug fix release 0.8.1 which I
uploaded to experimental. It's not yet built on all archs yet, but so far it
looks quite good [1].


The list of rdepends (that I know) are:

nautilus (2.28 links against libtrackerclient, 2.30 uses dlopen)
gtk+2.0 (uses dlopen)
totem-plugins (links agains libtrackerclient)
rygel-tracker (uses the D-Bus interface)
catfish (search frontend, uses the command line interface of tracker-search)

Both, the D-Bus interface and the library ABI/API, have changed.

I started tracking the progress at [2],
I've prepared and tested patches for gtk, nautilus 2.28 and totem-plugins 2.30.

The rygel-tracker (co-)maintainer, Andreas Henriksson, told me, that upstream
Git contains support for tracker 0.8 and he was either consider packaging a Git
snapshot or dropping the rygel-tracker plugin, to not block the tracker-0.8
transition.

I haven't heard back from the catfish maintainer yet, but as catfish is a search
frontend for multiple search engines, where tracker is only one of them, I
wouldn't consider it a blocker. Yet, as the tracker-search output only changed
slightly it should be simple to provide a patch.

So, I'd like the release teams and the GNOME maintainers input, when to best
schedule this transition.

For gtk+2.0 I would consider an upload as soon as 2.20.0-2 has transitioned to
testing. With the patch from [3], it will support both 0.6 and 0.8 by loading
the correct one at runtime.

For nautilus, there is a patch for 2.28 [4], which will link against the new
libtracker-client-0.8 library. An alternative would be, to directly upload
nautilus 2.30, which similar to gtk, dlopens the correct version of the tracker
library and so would decouple it from tracker's testing migration.

For totem, slomo will be preparing 2.30 packages in experimental, linking
against libtracker-client-0.8.

The tracker evolution plugin builds against evolution/evolution-data-server-dev,
so this needs to be kept in mind when planning the transition.


I'd like input from the release team and the GNOME maintainers, when to best
schedule this transition and how to move forward from here.

Cheers,
Michael


[1] https://buildd.debian.org/status/package.php?p=trackersuite=experimental
[2] http://wiki.debian.org//Tracker08
[3] 
http://svn.debian.org/wsvn/pkg-gnome/desktop/unstable/gtk%2B2.0/?rev=23689sc=1
[4] 
http://svn.debian.org/wsvn/pkg-gnome/desktop/unstable/nautilus/?rev=23634sc=1
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: tracker 0.8 transition

[Michael Biebl]

Am 12.04.2010 22:09, schrieb Josselin Mouette:
 Le lundi 12 avril 2010 à 18:35 +0200, Michael Biebl a écrit : 
 nautilus (2.28 links against libtrackerclient, 2.30 uses dlopen)
 
 Be it only for that, I’d prefer to do it after the gnome-desktop
 transition (which implies uploading nautilus, control-center and
 gnome-settings-daemon 2.30).

I'd be ok with uploading tracker 0.8 after the nautilus/gnome-desktop
transition.
pochu said on irc that the gnome-desktop transition should be a smaller one.

 The question remains, whether to do it before or after evolution, since
 it would be entangled with it.

Given David's input, I'd prefer to do the tracker transition before
evolution

 Question: have you checked whether nautilus 2.30 tracker support has the
 bug GTK+ originally had? It used to assume that when the library is
 available, the daemon is too, and didn’t try to make a connection to it.

No, I didn't. Will further investigate.

As an additional point:
Uploading totem 2.30 from experimental, will mean a new library
transition involving libtotem-pl-parser, which will entangle the two
transitions (tracker-extract depends on libtotem-pl-parser,
totem-plugins 2.30 would depend on libtracker-client-0.8).

An alternative would be, to upload totem 2.28 now and disable the
tracker-plugin. When tracker 0.8 has migrated to testing the
totem/totem-pl-parser 2.30 transition could be scheduled and the
tracker-plugin reenabled.
The benefit of this would be, to decouple the tracker from the
totem/totem-pl-parser transition.

slomo, should we do it that way? What's the release teams preference?

Cheers,
Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Sudo mode and policykit

[Michael Biebl]

Am 25.03.2010 23:34, schrieb Josselin Mouette:

 Actually there are two very simple solutions, so it is really a matter
 of what design you prefer. 
  1. (Stolen from Ubuntu) Create a new “admin” group, modify
 policykit to accept self-authentication for all members of the
 admin group. Let d-i simply add the user to group admin if in
 sudo mode. Bonus points for using the admin group in sudo too
 instead of hardcoding the username. 
  2. Let d-i create a file somewhere in /etc/polkit-1 that will add
 the created user to the list of users authorized to
 self-authenticate.
 

Speaking as policykit maintainer, I have to say that I generally like
the idea of such an admin group and thus would prefer 1.).
One simple reason is, that later on, it's easier to grant (additional)
users admin rights: simply add them to the admin group and you're done.
For option 2.) you'd have to create user specific policykit .pkla file
which is more tedious and not many users/admins know how to configure
policykit.
groups as a matter of privilege granting is a wiedly used and understood
concept though.

Not sure what other DDs think about such an admin group. I heard
rumours that this was discussed in the past and not very welcomed.
If so, I'd be interested in further references.

Cheers,
Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Sudo mode and policykit

[Michael Biebl]

On 26.03.2010 01:21, Margarita Manterola wrote:

 The amount of groups that a users needs to belong to in order to get
 the best experience from their computer is growing and growing.  This

Actually, I personally tried to work against this trend.

The groups plugdev/powerdev/netdev, which primarily were introduced to grant
access to certain D-Bus services, are more or less optional/deprecated in 
squeeze.
The recommended way nowadays is to use consolekit+policykit.

But I think this is getting slightly off-topic now (at least for -boot).
If we follow up on this, we should at least take -boot from the CC.

Cheers,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Please test experimental gvfs packages

[Michael Biebl]

Josselin Mouette wrote:
 Hi,

Hi

 since Michael Biebl fixed IDE CD support in devicekit-disks, the
 GDU-enabled version of gvfs (currently in experimental) should be usable
 again.

The packages and the versions are
gvfs_1.4.1-5+gdu
gvfs-backends_1.4.1-5+gdu
gvfs-bin_1.4.1-5+gdu (optional)
gvfs-fuse_1.4.1-5+gdu (optional)

logout/login from your current GNOME session (or reboot your machine).

 Before uploading it to unstable, I’d appreciate if some adventurous
 people could test it and report any regressions.

I'm especially interested if automounting/media detctions of cd-roms is working
properly with ide-cd. You will easily know if you are using ide-cd, if your
cd-rom device is something like /dev/hd* (whereas libata used /dev/sr* resp.
/dev/scd*).

If you have any problems, please CC me or contact me directly.
Succesfull test report are of course also welcome.

Cheers,
Michael




signature.asc
Description: OpenPGP digital signature

Re: Let’s remove scrollkeeper

[Michael Biebl]

Josselin Mouette wrote:
 Hi,
 
 is anyone aware of some leftovers that scrollkeeper can handle and
 rarian-compat cannot? If there are still some, we can handle them just
 like this was done for the DTDs since I’d like to ask for the removal of
 scrollkeeper.
 
 Once this is done, I propose to add a trigger to rarian-compat and to
 deprecate dh_scrollkeeper entirely.
 
 I don’t know what we should do to remove scrollkeeper from the users’
 desktops. Currently gnome-core forces the dependency on rarian-compat,
 maybe we should do the same with some other packages.

Hi Joss,

I have a few packages which have an (unversioned) Build-Depends on scrollkeeper.

What would you suggest: Should this be changed to rarian-compat, or will
rarian-compat keep the scrollkeeper Provides.

Cheers,
Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: CD's not automounted in GNOME desktop

[Michael Biebl]

José Alburquerque wrote:
 Would anyone happen to know why a few days now my cd's are not
 automounted in my gnome destkop?  I'm running unstable and I've been
 upgrading regularly so I think it's happened with a recent update.
 Sorry if it's a simple question.
 

Hi Jose,

this bug is related to a consolekit update [1]. I've uploaded a fixed g-v-m
version to experimental, which hopefully will be in unstable soon (I'm waiting
for the new nautilus 2.24).
As I only have limited time, I didn't invest the time to backport the fix when
it will be obsolete in a few days anyways.

Sorry for the inconvenience in the meantime.

Michael



[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=519550
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Configuration files location for GNOME applet

[Michael Biebl]

Josselin Mouette wrote:
 Le samedi 22 novembre 2008 à 00:27 +0100, Andrew M. a écrit :
 Hi.
 I recently developed a small GNOME applet and I wanted to share it among a
 few friends. It uses a couple of files to save settings and I was
 wondering what's the best location to store them.
 Is ~/.myapplet/ acceptable or should I use another folder?
 
 If it’s an applet using libpanel-applet, you should really use GConf to
 store the values so that the configuration can be instantiated once for
 each instance of the applet.
 
 If it’s merely an application using the notification area, you can use
 ~/.myapplet/ or ~/.gnome2/myapplet/, or course - but if you’re looking
 for a way to store user settings, GConf is the recommended way as well.

If you are not using GConf, I'd recommend to use the XDG basedir spec [1],
rather then ~/.myapplet/

Glib already has support for this, see the g_get_user_config_dir() function (I'd
recommend to install devhelp and libglib2.0-doc for easier access of the glib
API doc)

HTH,
Michael

[1] http://standards.freedesktop.org/basedir-spec/basedir-spec-0.6.html

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: PolicyKit and new HAL?

[Michael Biebl]

Christian Weeks schrieb:
 Hi
 I've just noticed that recent HAL changes have broken the click to mount
 of nautilus.
 
 When I click any hard disk type device, I now get a fun and useful popup
 and the disk doesn't mount:
 
 Cannot mount volume.
 Error org.freedesktop.Hal.Device.PermissionDeniedByPolicy.
 Details: hal-storage-fixed-mount refused uid 1000 (1000 is me)
 
From the hal (0.5.9-1) changelog:
 - debian/patches/55_nonpolkit-mount-policy.patch
 (containing this):
 +/* root can do everything; only allow handling removable
 devices
 + * without uid change to non-root users */
 +if (!invoked_by_uid || strcmp(invoked_by_uid, 0))
 +if (!privilege || strcmp (privilege,
 hal-storage-removable-mount))
 +permission_denied_privilege (privilege,
 invoked_by_uid);
 
 So, in summary, it appears a whole new privileges system has arrived.
 Except that I cannot control it. A bit of research indicates that the
 PolicyKit is supposed to manage these privileges. Is there any
 information on when policykit is going to be debianised? I looked at the
 pkg-utopia svn repository and there is nothing there that appears to be
 ready- or even recent (several months have passed since the last code
 commit).
 
 If PolicyKit is going to be some time in development, could we at least
 fall back to where we were before this patch: i.e. anyone can mount
 anything? Or do I have to install etch's hal (and all that pain) to
 bring back what I want?
 
 PS. Command line solutions are all well and good, but what the point of
 the UI then?
 

It would be a non negligible security risk, if every user could mount
all non-removable (fixed) devices. If you want to allow that, configure
your fixed mount points via /etc/fstab (with the user option).

For removable devices (external usb drives, cd-rom etc), everything
should work out of the box.

If and when PolicyKit will land in Debian, I can't tell yet. First it
has to mature. It changed quite a lot during the last months and simply
was/is to much in flux to be packaged.

Michael




-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Is it a bug?

[Michael Biebl]

Magnus Therning wrote:
 I thought I'd check before raising this as a bug:
 http://lists.debian.org/debian-user/2007/04/msg03070.html
 
 Is it a bug?  If so, what package is the culprit?

Mounting luks-encrypted partitions works just fine for me.
Make sure to close and teardown all existing mappings (unmount the
partition and then run cryptsetup luksClose /dev/mapper/...).
Then run gnome-mount -t -d /dev/sdb1 (given that sdb1 is the luks
encrypted partition).
This should prompt you for the password, unlock the key, and then mount
/dev/dm-0 at /media/$label_of_your_partition.
If you haven't set a label for your partition, it will be mounted at
/media/disk.

HTH,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

Re: Tracker/Beagle integration in Nautilus

[Michael Biebl]

Jose Carlos Garcia Sogo schrieb:
 El jue, 22-02-2007 a las 00:09 +0100, Xavier Bestel escribió:
 On mar, 2007-02-20 at 17:16 -0800, Alex Malinovich wrote:
 Is the extra runtime dep really so heavy ?
 Strictly from an end-user perspective, I think it's much better to have
 both alternatives easily available. 
 As long as it doesn't pull the mono interpreter, it's fine.
 
Depends if it will only use libbeagle0 or not. With that nautilus
 will be able to access beagle data, but it has to be gathered somehow.
 And beagled does need Mono.

Enabling tracker resp. beagle support in nautilus will lead to a
dependency on libtrackerclient0 (111k) [1] resp. libbeagle0 (319k) [2],
generated by dh_shlibdeps. So the dependencies are rather leight weight,
and I think we could enable both.

A Depends/Recommends/Suggests (whatever Joss seems fit best) on tracker
or beagle would have to be added to nautilus manually.
Without the tracker or beagle package installed, nautilus will fall back
to plain search.
Installing the beagle package will pull in the whole mono machinery,
which (I haven't counted) is I'd guess something around 30-50 Mb.

Cheers,
Michael


[1] which depends on libdbus-1-3, libdbus-glib-1-2 and libglib2.0-0
[2] which depends on libglib2.0-0 and libxml2
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

gtk-update-icon-cache

[Michael Biebl]

Hi all,

as maintainer of network-manager-gnome, I received this [1] bug report
today. The problem is, that n-m-g installs icons into
/usr/share/icons/hicolor. Apparently it needs to call
gtk-update-icon-cache /usr/share/icons/hicolor. (Is this a new
behaviour of Gnome 2.14?)
Shouldn't other packages that install icons into
/usr/share/icons/hicolor have the same problem? Looking at the Ubuntu
package, the Ubuntu devs wrote a debhelper utility called dh_iconcache,
which creates the necessary postinst/postrm scripts.
But reading [2] dh_iconcache does not seem to be the way to go.
So, what's the best way to deal with this bug for now? Should I create
the postinst/postrm rules manually?

Cheers,
Michael

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=372466
[2] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=369755
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature

gnome-panel 2.14 [was .la files]

[Michael Biebl]

Hi all,

I have to agree with Ross, that as soon we get rid of the *.la files the
better. I hope we will get saner/shorter dependency lists then (though I
wouldn't count on that too much as we still have a lot of broken *.pc
files).

Back to the subject. I wanted to upload a new release of NetworkManager
which fixes two important RC bugs. Unfortunately NM build depends on
libpanel-applet2-dev, so it fails because the *.la files of
libpanel-applet2-dev still reference the removed
libXcursor.la/libXrender.la files. I guess you won't be releasing a new
revision of gnome-panel 2.12 but go to 2.14 directly.
So my question is: When do you intend to upload gnome-panel 2.14 now
that e-d-s 2.6 has finally hit unstable? What's your proposed timeframe
for that.

Cheers,
Michael
-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?



signature.asc
Description: OpenPGP digital signature