Re: debian.org e-mail address and SPF/SRS

[Gustavo Franco]

On Fri, 5 Nov 2004 16:38:20 +1100, Matthew Palmer <[EMAIL PROTECTED]> wrote: 
> That's a question you'll have to ask of Yahoo and the SPF people.  My guess
> is that the pushers of these schemes want their thing adopted for whatever
> reason (corporate greed, personal gratification, whatever), but they know
> that random people don't care enough about e-mail forgery to really take it
> up.  However, most everyone online seems to be pretty pissed off about spam,
> so saying "this stops spam" will get people interested in the scheme, and
> they're hoping that people kinda forget that the system was supposed to stop
> spam when people work out, definitively, that it doesn't actually do squat
> to stop spam.
> 

"this stops spam" ? It isn't what they're saying, please read:

http://spf.pobox.com/faq.html#churn  
http://spf.pobox.com/faq.html#howworks

FYI, there are some different approaches that check different headers,
like Sender-ID but the 'publish the (spf) records' step is always
there.

Hope that helps,
Gustavo Franco -- <[EMAIL PROTECTED]>

Re: Happy new year 2003

[Gustavo Franco]

On Fri, 31 Dec 2004 23:50:52 +0100, Jean-Luc Picard <[EMAIL PROTECTED]> wrote:
> yes, debian is still 2 years late to any other distro.

If you're sleeping since Woody release, we're! Good morning Jean-Luc,
for your information:

- There' s a new Debian installer[0] that will be used in the upcoming
release, called Sarge;
- Some awards[1] since then;
- It seems that our infrastructure and QA matters[2];

[0] = http://www.debian.org/devel/debian-installer
[1] = http://www.debian.org/misc/awards
[2] = http://www.debian.org/misc/children-distros

There' s much more but i hope it' s enough to anyone sleeping for two
years see that Debian isn't `too late`. It' s far away from state of
art but the majority of work is being done by volunteers and without
it you wouldn't be booting knoppix, ubuntu and others!

--
Gustavo Franco -- <[EMAIL PROTECTED]>

Re: Canonical's business model

[Gustavo Franco]

Canonical's business model doesn't belong in -devel. If Canonical as a
company is being fair, cool, whatever with Debian project i think we
can discuss it in -project, but why not do the same exercise about
Linspire? Do they sponsor conferences? Oh, i think Canonical does it
too.

It's up to Canonical how they will contribute back to the community,
IMHO. I don't the same rant over others Debian related companies so
i'm assuming that we're wasting time shooting Canonical, (mainly)
because Ubuntu is sucessful.

I did a different opinion a month ago, but the fact is that i tried to
start a collaborative dicussion two times with Canonical employees and
it's going well. I recommend you do the same, and discuss in -project
what we (as a project) need from Canonical: free tools, better
formatted patches, whatever, ... I don't think they will "waste" time
and money thinking about it for us.

It's going to far, after all some people here and there are just
criticizing old time friends before asking them if they can share
resources and workload for the better of both projects.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/10/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Tue, Jan 10, 2006 at 12:22:03AM -0200, Gustavo Franco wrote:
> > I don't[sic] the same rant over others Debian related companies
>
> Have you ever actually subscribed to any Debian mailing lists?
>

Hi Andrew,

Don't be fooled by From mail headers.

--
Gustavo Franco - <[EMAIL PROTECTED]>

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Jan Nieuwenhuizen <[EMAIL PROTECTED]> wrote:
> Thomas Bushnell writes:
>
> > No, I think it's because Ubuntu doesn't cooperate well with Debian,
> > while pretending to cooperate.
>
> Does Debian want to cooperate with Ubuntu, and how well does Debian
> do?  What steps could Ubuntu and Debian reasonably take to improve
> cooperation?
>

Debian is cooperating with Ubuntu as it's cooperating with each Custom
Debian Distribution (no, Ubuntu isn't one). I don't think that Ubuntu
is unsatisfied with Debian volunteers, e.g: d-i merging base-config
package stuff[0], Ubuntu will benefit of this in April (Dapper
release), we will just see the impact late this year after releasing
Etch and the CDDs that are building based on our stable release, right
after.

Ubuntu runs a tool to merge Debian changes into its "unstable"
archive, and it's (almost) smart enough to not override Ubuntu changes
when necessary. Is that tool available to Debian (utnubu[1] to be
specific) do the same in a not automatic way but generate a report ?
No and i don't think that we asked for that tool, utnubu is running
with its own resources.

Please note that i'm not saying that there's one side right and other
wrong, i just reported the current status.

[0] = http://lists.debian.org/debian-boot/2006/01/msg00118.html
[1] = http://alioth.debian.org/projects/utnubu/

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Tue, Jan 10, 2006 at 11:07:43AM -0200, Gustavo Franco wrote:
> > On 1/10/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> > > On Tue, Jan 10, 2006 at 12:22:03AM -0200, Gustavo Franco wrote:
> > > > I don't[sic] the same rant over others Debian related companies
> > >
> > > Have you ever actually subscribed to any Debian mailing lists?
> > >
> >
> > Don't be fooled by From mail headers.
>
> Well, I've sure seen similar things being said about nearly every
> Debian-related company I've ever heard of (Progeny, Linspire, Nexenta,
> etc). I find it hard to see how else you could have missed them.

I agree with "similar things being said" but i'm yet to hear about the
lack of collaboration and give Debian something back. For example: I
don't remember too much people caring about PGI (Progeny) and after
that anaconda "port" to say that they weren't contributing the
installer efforts to us, even when d-i was already there. I remember
talking with Ian Murdock about "anaconda x d-i" during a dinner
(debconf 4), it was clear there that Progeny wasn't doing nothing
evil. It seems that some people is considering Canonical as the evil
itself.

I don't want to say that Canonical is a perfect company and that all
its employees and contributors are commited to free software and care
about the Debian project "health". The point is that they aren't white
or black, they're like others companies helping us. We're in a
position that some volunteers helping them (and they're less than us)
started asking themselves if it was ethical contribute to Ubuntu and
ignore Debian. It won't take too much time to they realize that
contributing to Debian is contribute to Ubuntu but the reverse isn't
always true. We just need to clarify it and brought mentors, utnubu
and similar projects in a new level of visibility and we're doing it
right now, IMHO.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 11, 2006 at 02:56:35PM -0200, Gustavo Franco wrote:
> > On 1/11/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> > > On Tue, Jan 10, 2006 at 11:07:43AM -0200, Gustavo Franco wrote:
> > > > On 1/10/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> > > > > On Tue, Jan 10, 2006 at 12:22:03AM -0200, Gustavo Franco wrote:
> > > > > > I don't[sic] the same rant over others Debian related companies
> > > > >
> > > > > Have you ever actually subscribed to any Debian mailing lists?
> > > > >
> > > >
> > > > Don't be fooled by From mail headers.
> > >
> > > Well, I've sure seen similar things being said about nearly every
> > > Debian-related company I've ever heard of (Progeny, Linspire, Nexenta,
> > > etc). I find it hard to see how else you could have missed them.
> >
> > I agree with "similar things being said" but i'm yet to hear about the
> > lack of collaboration and give Debian something back.
>
> None of the other companies ran around pronouncing how great they were
> at 'giving things back' and how 'committed' they were to free
> software, etcetera. That appears to be the relevant point.
>
> I don't think anybody seriously objects to the existence of companies
> who *don't* do these things.

Are you saying that they're spending more money with PR than really
contributing back ? I don't have the same feel, but if it's the
reality it's grave. We just need to avoid measure these things by
emotion and point out facts, and i'm not talking about you and your
opinions but being more general.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Marco d'Itri <[EMAIL PROTECTED]> wrote:
> On Jan 11, martin f krafft <[EMAIL PROTECTED]> wrote:
>
> > How do you think Canonical could *better* work with Debian, ignoring
> > whether they meet up to their promises at the moment or not.
> E.g. when I repeatedly say "I'd like to receive any change you make to
> my packages, in any form you find convenient" they could actually do
> it... I'm tired of begging for patches.
>

http://utnubu.alioth.debian.org/scottish/by_maint/[EMAIL PROTECTED]/

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Daniel Ruoso <[EMAIL PROTECTED]> wrote:
> Em Qua, 2006-01-11 às 16:48 +0100, martin f krafft escreveu:
> > What would you like to see?
>
> I think submitting bugs and patches to the BTS would already be enough.
>

It was already discussed[0], and there's no consensus on this idea of
"every Ubuntu changeset, a patch in Debian BTS" between DDs. I don't
remember Linspire, Progeny, ... employees doing the same thing so it
makes no sense rant against Canonical only. There's scott's patches
list[1] that sucks IMHO,  and utnubu one[2]. AFAIK, some PTS work was
already done too so we (probably) are listing if there's a ubuntu
patch in every Debian package from qa.d.o. After all, do you still
want annoying automatic bug reports?

We've a lot more volunteers than Canonical, if you want to change the
scenario (and i'm not writing to Daniel only) you should join
utnubu[3] and help, but i think that the "patch" argument is over.
FYI, there's a MOTU effort into Ubuntu to package universe stuff back
to Debian too[4].

[0] = http://lists.debian.org/debian-project/2005/12/msg00120.html
[1] = http://people.ubuntu.com/~scott/patches/
[2] = http://utnubu.alioth.debian.org/scottish/
[3] = http://utnubu.alioth.debian.org/
[4] = https://wiki.ubuntu.com/ContributingToDebian

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Marco d'Itri <[EMAIL PROTECTED]> wrote:
> On Jan 11, Gustavo Franco <[EMAIL PROTECTED]> wrote:
>
> > > E.g. when I repeatedly say "I'd like to receive any change you make to
> > > my packages, in any form you find convenient" they could actually do
> > > it... I'm tired of begging for patches.
> > http://utnubu.alioth.debian.org/scottish/by_maint/[EMAIL PROTECTED]/
> You are totally missing my point.
>

You said *in any form you find convenient* but which one do you
prefer: bug reports through Debian BTS, just email, ... ? Please, read
my reply to Daniel's message.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Marco d'Itri <[EMAIL PROTECTED]> wrote:
> On Jan 11, Gustavo Franco <[EMAIL PROTECTED]> wrote:
>
> > You said *in any form you find convenient* but which one do you
> > prefer: bug reports through Debian BTS, just email, ... ? Please, read
> > my reply to Daniel's message.
> Uploading the diffs on a web server is nice, but it's not much more
> different from just downloading the .diff.gz files from their archive.
> I still need to do it from time to time and then hunt for changes since
> the last time.

I see, i would like to see the utnubu patch list[0] integrated in PTS
(scott's already is[1]), with that everyone subscribed to the package
could receive a mail after a patch shows up there. I'm sure that
Rafael or someone else said something about it, but not the entire
idea. Is it possible, Rafael?

[0] = http://utnubu.alioth.debian.org/scottish/
[1] = look at Patches - http://packages.qa.debian.org/u/udev.html

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Daniel Ruoso <[EMAIL PROTECTED]> wrote:
> Em Qua, 2006-01-11 às 14:36 -0800, Thomas Bushnell BSG escreveu:
> > Gustavo Franco <[EMAIL PROTECTED]> writes:
> > > It was already discussed[0], and there's no consensus on this idea of
> > > "every Ubuntu changeset, a patch in Debian BTS" between DDs.
> > Right.  I want Ubuntu to exercise judgment, and not just give a big
> > pile of patches, some of which are Debian-relevant and some of which
> > are not.  Think, for example, of the normal way a Debian developer
> > should interact with upstream.
>
> This is exactly the point, what can I do with a patch if I don't know
> why it's there? Which problem is it trying to address (I know, I can
> read the patch and guess, but WTF), and why such solution was adopted...
> Everytime I submit a patch, I also submit this reasoning...

I disagree with a pile of patches and as i said it would be better a
revision control system and good log (and debian/changelog) entries.
We can use PTS (and we're doing already in a way) to be warned about
new patches. I don't think Canonical will put money on judging by DDs,
in the end it's up to us include or not the change.

If they want to promote that they give us something back their
reponsability is keep things more organized (for DDs and NMs) and
publish somewhere what exactly they're giving back (for the
community), IMHO. Does it mean that they're not contributing? No,
they're believe in me. We just need to do our homework, and inform
them that we can do the things in a better way.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Thomas Bushnell BSG <[EMAIL PROTECTED]> wrote:
> Gustavo Franco <[EMAIL PROTECTED]> writes:
>
> (...)
>
> > I don't
> > remember Linspire, Progeny, ... employees doing the same thing so it
> > makes no sense rant against Canonical only.
>
> On the other hand, Linspire and Progeny do not pretend to be
> cooperating with Debian.

Yes, they do. Progeny[0] even has a Ubuntu' similar page[1]. Linspire
doesn't do much noise about that but, sponsor events. I'm not saying
that both companies don't give something back to Debian, but we never
discussed if it was being done in the right way. If we started with
Canonical great, but we need to consider others too, IMHO.

[0] = http://www.progeny.com/partners/debian.htm
[1] = http://www.ubuntu.com/ubuntu/relationship

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 11, 2006 at 03:25:01PM -0200, Gustavo Franco wrote:
> (...)
> > Are you saying that they're spending more money with PR than really
> > contributing back ?
>
> I don't know about money, but I'm pretty sure their claims exceed
> their actions. I think that a sufficient response is to point this out
> whenever people start worshipping Canonical in public.

If it's a reality, it's up to us inform them that they're spending
more words than work giving Debian something back. In my view, they're
just organizing and publishing the contributions in a way that aren't
satisfiying the Debian majority.

We are complaining about different aspects and some of them were
already addressed, others aren't. Let me try list some of them:

We don't like:
- Canonical is saying more than doing (is it a consensus?);

- Scott's pile of patches and utnubu by_maint list isn't enough;

What we want:
- DDs wants to know when there's a patch for their packages into
Ubuntu that applies to Debian. PTS already lists them in "Patches"
section but it links to Scott's patches and we don't like them, right?
I think we can ask for a revision control system from them and do some
work on PTS (if possible) to mail us or just list the logs in the
website.

Solved stuff (IMHO):
- If you talk with a Ubuntu developer or contributor probably he won't
ignore you. Many of us tried and some are working together with us on
alioth projects or something else;

- The Ubuntu 'universe' has packages that Debian hasn't. There's a
initiative there to contribute back to Debian;

- We won't accept automatic bug reports with each changeset as a patch
for every changed package in Ubuntu;

I'm sure that you can help me with this list, and i think that we can
keep the discussion around "what we have", "what we need", 'what is
solved", things like that. We need to define it, and not only point
the (past and current) problems.

Closing, i hope to see the same effort spread over others Debian
"related" companies and their business model.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Gustavo Noronha Silva <[EMAIL PROTECTED]> wrote:
> Em Qua, 2006-01-11 às 19:54 -0300, Daniel Ruoso escreveu:
> > Em Qua, 2006-01-11 às 14:36 -0800, Thomas Bushnell BSG escreveu:
> > > Gustavo Franco <[EMAIL PROTECTED]> writes:
> > > > It was already discussed[0], and there's no consensus on this idea of
> > > > "every Ubuntu changeset, a patch in Debian BTS" between DDs.
> > > Right.  I want Ubuntu to exercise judgment, and not just give a big
> > > pile of patches, some of which are Debian-relevant and some of which
> > > are not.  Think, for example, of the normal way a Debian developer
> > > should interact with upstream.
> >
> > This is exactly the point, what can I do with a patch if I don't know
> > why it's there? Which problem is it trying to address (I know, I can
> > read the patch and guess, but WTF), and why such solution was adopted...
> > Everytime I submit a patch, I also submit this reasoning...
>
> That's sometimes documented in the changelog. I benefited quite a lot
> from the ubuntu patches for gksu, and I've worked quite nicely with
> seb128 and mvo on issues like this one and update-manager.

I agree, things like that are happenning all the time but many DDs
just don't know and Canonical is failing to inform the community how
and where they're helping. I think they're just saying that they're
helping a lot.

> (...)

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/11/06, Steve Langasek <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 11, 2006 at 05:48:22PM -0200, Gustavo Franco wrote:
> > On 1/11/06, Daniel Ruoso <[EMAIL PROTECTED]> wrote:
> > > Em Qua, 2006-01-11 às 16:48 +0100, martin f krafft escreveu:
> > > > What would you like to see?
>
> > > I think submitting bugs and patches to the BTS would already be enough.
>
> > It was already discussed[0], and there's no consensus on this idea of
> > "every Ubuntu changeset, a patch in Debian BTS" between DDs. I don't
> > remember Linspire, Progeny, ... employees doing the same thing so it
> > makes no sense rant against Canonical only. There's scott's patches
> > list[1] that sucks IMHO,  and utnubu one[2]. AFAIK, some PTS work was
> > already done too so we (probably) are listing if there's a ubuntu
> > patch in every Debian package from qa.d.o. After all, do you still
> > want annoying automatic bug reports?
>
> > We've a lot more volunteers than Canonical, if you want to change the
> > scenario (and i'm not writing to Daniel only) you should join
> > utnubu[3] and help,
>
> Of course people can do this, but this is so very much not the point.  The
> point is that publishing source packages on a website that people have to
> poll is not "giving back to Debian", and AFAICT the majority of changes
> Ubuntu makes to packages are only made available to Debian in this format.
> This includes many changes in Ubuntu's universe section[1] which I think it's
> bad strategy to be making externally to Debian in the first place if
> they're serious about limiting divergence from Debian.

I agree with the poll thing, but the 'giving back to Debian' applies
when you think about things like xorg (David even wrote it), gksu
(kov), pkg-ltsp, some other transitions and i'm sure that someone can
came up with a better list than me.

> I've also seen Canonical employees make comments in the past to the effect
> that Debian has an obligation to meet Ubuntu part-way (read: monitor
> Ubuntu's changes) on the question of integrating their changes back into
> sid.  This is either a wholly unrealistic assessment of the scalability
> issues with coordinating between the many CDDs and Debian derivatives in
> existence, or simply hubris regarding Ubuntu's privileged position within
> the Debian cosmos; but in either case, it does not support the thesis that
> Canonical systematically "gives back to Debian" or that they have
> succeeded in structuring Ubuntu's culture in a way that promotes such giving
> back.

They give something back to Debian (see above), the current problems
are around the way they're informing us about their patches not if
they contribute or not. They contribute, the people are just mixing up
the patch handling issue with others contributions.

> All of which is fine, and the right of anyone working off of Debian (hurray
> Free Software!), up until the point where one starts claiming to be giving
> back to Debian when by and large they are not; and I'm afraid this does seem
> to be the case with Ubuntu today.

They're saying that really, but they're not saying that every patch is
classified and informed to us. With that in mind, they're not lying
and contributing something back. If it's enough or not and how it
could be better, i'm trying to discuss.

--
Gustavo Franco

Re: Canonical's business model

[Gustavo Franco]

On 1/12/06, Joey Hess <[EMAIL PROTECTED]> wrote:
> Gustavo Franco wrote:
> > I agree with "similar things being said" but i'm yet to hear about the
> > lack of collaboration and give Debian something back. For example: I
> > don't remember too much people caring about PGI (Progeny) and after
> > that anaconda "port" to say that they weren't contributing the
> > installer efforts to us, even when d-i was already there.
>
> FWIW, progeny uploaded pgi to Debian (I forget if it ever made it out of
> incoming) and have contributed back other tools like pickaxe too (pity
> we haven't tried to use it and are still stuck with the Evil that is
> debian-cd). I think it was pretty clear by the time their anaonda port
> came around that Debian was not very interested it it except possibly as
> a fallback if d-i failed to materialize.

AFAIK, PGI reached our repositories but my point was that nobody
complained that if they're doing the things in the right way. Were all
the Progeny patches (i'm not talking about new packages) listed,
informed, considered, or even reviewed to Debian ? We're doing
considerations about Canonical way that should be involve at least
Progeny too, since we're listed at partners.

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/12/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 11, 2006 at 03:41:16PM -0800, Matt Zimmerman wrote:
> > On Wed, Jan 11, 2006 at 11:09:12PM +, Andrew Suffield wrote:
> > > Let's take this one apart and see what it is that pisses people off so
> > > much.
> >
> > I don't intend to participate in this type of email argument with you; I've
> > yet to see it pay off for anyone involved.  However, I will be in London
> > later this month and would be willing to use that opportunity to civilly
> > discuss your concerns face-to-face.
>
> The intent here being "stop people from scrutinising Ubuntu in public;
> get it off the lists so that it's less visible". Not likely.

Do you want visibility or solve current problems ? I'm glad that Matt
is open for suggestions, and even a talk in person with you. You
missed a chance to provide (yet another) sane feedback from Debian
perspective to him, and give us something back.

Thanks,
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/12/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Thu, Jan 12, 2006 at 05:31:40PM -0200, Gustavo Franco wrote:
> > On 1/12/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> > > On Wed, Jan 11, 2006 at 03:41:16PM -0800, Matt Zimmerman wrote:
> > > > On Wed, Jan 11, 2006 at 11:09:12PM +, Andrew Suffield wrote:
> > > > > Let's take this one apart and see what it is that pisses people off so
> > > > > much.
> > > >
> > > > I don't intend to participate in this type of email argument with you; 
> > > > I've
> > > > yet to see it pay off for anyone involved.  However, I will be in London
> > > > later this month and would be willing to use that opportunity to civilly
> > > > discuss your concerns face-to-face.
> > >
> > > The intent here being "stop people from scrutinising Ubuntu in public;
> > > get it off the lists so that it's less visible". Not likely.
> >
> > Do you want visibility or solve current problems ?
>
> If I were interested in solving Ubuntu's problems then I would be
> working on Ubuntu. As people keep pointing out, Ubuntu's failure to
> cooperate effectively is *not* our problem - the only 'problem' that
> *we* have is that Ubuntu-worshippers keep showing up and
> proselytizing. An effective solution to this problem is to raise
> awareness to the point where people stop believing and start
> thinking. It appears to be working.
>

We can't decide how they need to "give us something MORE back" and
it's their problem? An effective solution is discuss between us and
keep our needs clear, since there are some Canonical employees that
are DDs too, they're in a position that they can participate in these
discussions (they're doing it right now).

You started (in a way) two threads about it and there's other about
launchpad, in my view we've the following status (with comments
below):

- No, Debian isn't going to embrace launchpad and ask for "rw" status
(Debian is listed there as read-only).[0]

- Yes, Ubuntu is contributing back but some DDs judge that it isn't
Canonical but some Canonical employees that are dedicated to Debian
too, so they help because they want;[1]

- Scott's url with patches isn't part of the "give something back"
approach that we want. We need to be well informed about patches, but
we don't know exactly how;

I removed more non-Canonical related initiatives like Ubuntu MOTUs
page about contributing to Debian.

[0] = I agree, DDs and contributors are free to use any tools they
want, so we're not going to force any DD or contributor go through
launchpad. I think it's consensus and valid for other things, like
alioth and related stuff;

[1] = If they're doing it in their workhours, it isn't only because they want.

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/12/06, Daniel Ruoso <[EMAIL PROTECTED]> wrote:
> Em Qui, 2006-01-12 às 18:08 -0200, Gustavo Franco escreveu:
> > - Scott's url with patches isn't part of the "give something back"
> > approach that we want. We need to be well informed about patches, but
> > we don't know exactly how;
>
> Don't we?
>
> Debian is Ubuntu's upstream, right?

In a way, yes.

> When you modify something in the upstream code, you normally send it to
> upstream, right?

The normal "upstream" can't be applied here. There are some scenarios
where Ubuntu can patch Debian packages and it isn't a simple "debian
(upstream) - ubuntu (dd)" relationship, see:

non-native: debian/ patches; debian/patches or whatever - normally
related to upstream (do they need to report it to us or the real
upstream? both?);

native: I think everything not in debian/ could be reported back, but
each debian/ changeset should be verified first.

> Do you send it as a link to a file with a patch only? Or do you send a
> comment explaining the problem, the proposed solution and why that
> decision was made?

That's why i mentioned revision control system (on Ubuntu side not us)
in the other thread, but Manoj missed the point replying and i don't
think it's going to happen anyway.

There are cases when debian/changelog is enough (native as described
above) but others aren't.

> That's it, just it, nothing more... That's what distinguishes
> cooperating from forking...

Yes, but you're yet to fill the "how" gap since i believe you agreed
that a new bug in our BTS related with every Ubuntu changeset wouldn't
be a good idea, right? Don't came with "they need to review and
judge", it's up to us decide if we will include patches or not. They
won't  do it every time, but it's clear that some of Canonical
employees are doing it as Matt cited.  If it's not happening to your
packages, i recommend you ask who modified them to file bug reports.
It will be your policy, you're free to do that.

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/13/06, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Thu, Jan 12, 2006 at 06:08:52PM -0200, Gustavo Franco wrote:
> > We can't decide how they need to "give us something MORE back" and
> > it's their problem?
>
> Whoever said they need to do that? They just need to stop bragging
> about shit they don't do. There's at least two ways to accomplish this.
>
> If they fail to contribute in a meaningful way, it just means more
> work for them (in trying to maintain a diverging fork). Hence, that's
> their problem. It's not really a problem for us.
>

We can't say that Canonical/Ubuntu isn't contributing back. They're,
as pointed out by some of us. e.g.: David said that Daniel helped him,
but if he did that in his workhours it's under Canonical bless.

It seems that the main problem is how they're handling the list of
patches. If they want to spread the word that they're contributing, it
seems that many of us want to be informed about the patches as we
inform upstreams and not as it's today.

I can't affirm if they're saying more than they're doing, but we are
for sure. I was the only trying to prepare a list of things that we
can ask them to change, and mdz (Ubuntu/Debian) tried to collect
feedback too. We want cooperation, it seems that they want too but
Debian by nature is a complicated project and Ubuntu will never
satisfy all our needs, even for just handling and reporting back some
patches.

With that in mind, it would be good to hear about some internal
discussion in Ubuntu camp too, maybe in the next online meeting or in
London. It will proof that they want to be something different than a
simple fork, as described by mako[0].

[0] = http://mako.cc/writing/to_fork_or_not_to_fork.html  (long)

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/13/06, Matthew Garrett <[EMAIL PROTECTED]> wrote:
> David Nusinow <[EMAIL PROTECTED]> wrote:
>
> > Please stop trying to twist my words around. Canonical didn't contribute
> > back. An individual who happened to work for Canonical did. If someone
> > employed by the US government contributes to Debian of his own volition do
> > we say that the US government gives back to Debian? Do we say that your
> > employer gives back to Debian?
>
> If it's an authorised use of company time, sure. Whether or not it is in
> this case, I don't know.
>

Exactly my point Matthew, and calm down David, i wrote: "e.g.: David
said that Daniel helped him, but if he did that in his workhours it's
under Canonical bless.". Do you see ? I just pointed out that there's
a possibility that he was helping you in his workhours, but i won't
cite you as a reference anymore.

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/13/06, Matthew Palmer <[EMAIL PROTECTED]> wrote:
> On Fri, Jan 13, 2006 at 03:03:14PM -0200, Gustavo Franco wrote:
> > On 1/13/06, Matthew Garrett <[EMAIL PROTECTED]> wrote:
> > > David Nusinow <[EMAIL PROTECTED]> wrote:
> > >
> > > > Please stop trying to twist my words around. Canonical didn't contribute
> > > > back. An individual who happened to work for Canonical did. If someone
> > > > employed by the US government contributes to Debian of his own volition 
> > > > do
> > > > we say that the US government gives back to Debian? Do we say that your
> > > > employer gives back to Debian?
> > >
> > > If it's an authorised use of company time, sure. Whether or not it is in
> > > this case, I don't know.
> > >
> >
> > Exactly my point Matthew, and calm down David, i wrote: "e.g.: David
> > said that Daniel helped him, but if he did that in his workhours it's
> > under Canonical bless.". Do you see ? I just pointed out that there's
> > a possibility that he was helping you in his workhours,
>
> You've never done anything at work that wasn't officially sanctioned by your
> boss?

No, because i'm the technology coordinator in a NGO and i'm free to
contribute to the Debian project during my workhours since we develop
a CDD for telecentres.

I see your point, but you're mixing different stuff. AFAIK the
'contribute back to Debian' is endorsed by Canonical, so it's
officially sanctioned there using your own words.

--
Gustavo Franco

Re: Dissection of an Ubuntu PR message

[Gustavo Franco]

On 1/14/06, Kevin Mark <[EMAIL PROTECTED]> wrote:
> On Fri, Jan 13, 2006 at 03:03:14PM -0200, Gustavo Franco wrote:
> (...)
> > Exactly my point Matthew, and calm down David, i wrote: "e.g.: David
> > said that Daniel helped him, but if he did that in his workhours it's
> > under Canonical bless.". Do you see ? I just pointed out that there's
> > a possibility that he was helping you in his workhours, but i won't
> > cite you as a reference anymore.
> >
> > --
> > Gustavo Franco
> Hi Gustavo,
> Is it within the scope of Canonical employees to contribute code to
> Debian that is under the his copyright and not Canonical's? And
> especially since it is in the exact same area that he was employed by
> Canonical to do?  Would this apply to Progeny and Debian, Progeny and
> Canonical, Linspire and ...

Hi Kevin,

I think that Matt Zimmerman (mdz) knows the answer.

--
Gustavo Franco

Re: For those who care about lesbians

[Gustavo Franco]

On 1/14/06, Amaya <[EMAIL PROTECTED]> wrote:
> Bartosz Fenski aka fEnIo wrote:
> > I got you sarcasm, but I still think that messages posted to
> > debian-devel-announce should be more official.
>
> Come on, it is just Mr. Suffield slowly turning into the Overfiend :)
>

slowly? Are you sure? I disagree with "into the Overfiend" sentence,
it should be replaced with "into *the* Debian troll". Are we going to
list this position in our Organizational Structure[0] ?

[0] =  http://www.debian.org/intro/organization

--
Gustavo Franco

Re: Andrew Suffield

[Gustavo Franco]

On 1/18/06, Dallam Wych <[EMAIL PROTECTED]> wrote:
> On Sun, Jan 15, 2006 at 05:09:03PM -0600, Joe Wreschnig wrote:
> > On Sun, 2006-01-15 at 06:28 -0500, sean finney wrote:
> > > On Sun, Jan 15, 2006 at 11:58:51AM +0100, Adrian von Bidder wrote:
> > > > Do you think your constant bitching is funny?  Do you think it achieves
> > > > anything?
> > > >
> > > > There are other DDs who are also involved in intense debates and 
> > > > flamewars
> > > > very often, but you're the only one  where I constantly get the 
> > > > impression
> > > > that you're just being childish, insulting and annoying for the sake of 
> > > > it.
> > >
> > > ...says someone who just publicly ostracized a fellow dd
> > > on a public mailing list.  personal opinions of the matter aside,
> > > debian-devel is not the place for ridiculing other developers, no
> > > matter how justified you feel you may be.
> > >
> > > please post follow-ups to -private.
> >
> > I said this on -private, and I'll say it here -- why is it appropriate
> > to be an ass on -private, but not on -devel? It's not appropriate
> > anywhere. That goes for Adrian, and Andrew, and everyone. It also leads
> > to situations like the present, where it looks like we're doing nothing
> > to reprimand offensive behavior, because most conversation is happening
> > on -private, while the original, offensive message is sitting on d-d-a.
> >
> > If you are upset by how Andrew acted, talk to him rationally, regardless
> > of whether it's public or private.
> >
> > If you are *very* upset by how Andrew acted, there is an appropriate and
> > agreed-to policy for expelling developers. Roger Leigh has mentioned his
> > interest in seeing this through; contact him.
> > --
> > Joe Wreschnig <[EMAIL PROTECTED]>
>
> I imagine that the ubuntu people, which include those on canonicals
> payroll that are posting to this list, are really finding this kind of discord
> within the Debian community quite comical and amusing.
>

You ignore that a lot of them are part of the Debian community. This
project would be better if people like you applied part of the
imagination to contribute (at least) with useful comments.

--
Gustavo Franco -- <[EMAIL PROTECTED]>

Re: Andrew Suffield

[Gustavo Franco]

On 1/18/06, Dallam Wych <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 18, 2006 at 06:57:13PM -0200, Gustavo Franco wrote:
> > You ignore that a lot of them are part of the Debian community. This
> > project would be better if people like you applied part of the
> > imagination to contribute (at least) with useful comments.
>
> Rather, I think *you* missed my point...I tend to be a bit dimissive
> of people who change sides for a dollar/pound/whatever.  As for applying
> my imagination, I contribute to Debian by way of local install fests,
> helping steer new users in my community towards Debian, etc. Note
> all us aren't whiz-bang programmers hence one *can* help Debian
> without being a DD. I have been a Debian user for about four years
> now, a linux user since '95...so don't do the glib disrespectful bit
> on me.
>

I'm glad that you contribute to Debian, you're part of the Debian
community as some people that you're pointing that changed sides for a
dollar. I'm sure that you don't know none of them, to say for sure.
Please, stop the troll here. If you really believe in what you're
saying go in the right ubuntu list and complain.

See, i haven't excluded you in my previous message, i don't even knew
if you were a DD, before replying. I don't care, i used the term
'community' and not 'whiz-bang programmers' or whatever.

--
Gustavo Franco - <[EMAIL PROTECTED]>

Re: A great weekend for Debian

[Gustavo Franco]

On 1/23/06, Amaya <[EMAIL PROTECTED]> wrote:
> Steinar H. Gunderson wrote:
> > I think you're overly optimistic :-) Most of the simple RC bugs
> > (related to the xlibs-dev transition) have been fixed; there aren't 90
> > more like those.
>
> I got home from work and have second thoughts about the email I
> previously sent. I think I am a bit pissed off by this "simple RC bugs"
> statement. I don't personally care about my own credit, but I think the
> people who worked in this transition deserved the email Joseph Smidt
> sent and an acknowledgement for their work. Some of us have been barely
> doing anything else in our lives than this transition for 14 days.
>
> I am talking about Marc 'HE' Brockschmidt, Thomas Viehmann, Nico Golde,
> Steve Langasek, Victor Seva Lopez and Justin Pryzby, others I might not
> be aware of, and Moritz Muehlenhoff, who provided the script that helps
> find out the new Build-Depends.
>
> (...)

The point is that there are a lot of people in this project that loves
(or not) duelling banjos, cares (or not) about lesbians, uses Debian
and Ubuntu (or not) and keep considering that flamewars are a waste of
time and we still are about a *universal* operating system, that
involves "shut up and hack" sometimes and "be cool" in others.

Thanks Justin, Amaya and all the others involved in any transition or
just for you that keep your packages without RC bugs as much as you
can, or for you that look into the RC bug list and feel that you
should help.

Sometimes, it's really good feel that i'm less than 1/1000 of all this
and while i'm busy helping with something that i think is important
there are others working in others not less interesting tasks.

Etch is coming. :)

Thanks,
Gustavo Franco

ITP: gtimelog -- minimal timelogging system

[Gustavo Franco]

Package: wnpp
Severity: wishlist

* Package name: gtimelog
  Version : 0.0+svn65-1
  Upstream Author : Marius Gedminas <[EMAIL PROTECTED]>
* URL : http://mg.pov.lt/gtimelog/
* License : GPL
  Description : minimal timelogging system

 gtimelog provides a time tracking application to allow the user
 to track what they work on during the day and how long they
 spend doing it.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Automatic testing of .deb's

[Gustavo Franco]

On 2/2/06, Ian Jackson <[EMAIL PROTECTED]> wrote:
> (...)
>
> I would like to have some idea what people think I should do with the
> tests that we're hopefully going to have, eventually for lots of
> packages.  Would Debian like those tests as patches in wishlist bug
> reports, in general ?  That would seem to be best to me but before I
> go down this route I'd like to be clear that that's what Debian
> developers want.

I think you should put out a summary of how many and what packages will
be changed in Ubuntu. After that we can agree or not with automatic filling
whislist bugs (containing the patches) against Debian packages.

Since Ubuntu Dapper is actually on freeze, what's your timeline to these
patches? Will you include this stuff just on Dapper+1, in the end of
the year?

Thanks,
Gustavo Franco

Re: Automatic testing of .deb's

[Gustavo Franco]

On 2/6/06, Ian Jackson <[EMAIL PROTECTED]> wrote:
> Gustavo Franco writes ("Re: Automatic testing of .deb's"):
> > On 2/2/06, Ian Jackson <[EMAIL PROTECTED]> wrote:
> > > I would like to have some idea what people think I should do with the
> > > tests that we're hopefully going to have, eventually for lots of
> > > packages.  Would Debian like those tests as patches in wishlist bug
> > > reports, in general ?  That would seem to be best to me but before I
> > > go down this route I'd like to be clear that that's what Debian
> > > developers want.
> >
> > I think you should put out a summary of how many and what packages will
> > be changed in Ubuntu. After that we can agree or not with automatic filling
> > whislist bugs (containing the patches) against Debian packages.
>
> I think you've misunderstood the situation and what we (Ubuntu) are
> going to do.  There's no question of automatic bug filing, here.

Ok.

> At the moment there are no patches and so there is no list of
> packages.  As time goes on, Ubuntu will have more and more tests and
> each time an Ubuntu maintainer writes a test that will be an
> additional diff between Ubuntu and Debian; the question then is
> whether in general these tests are considered valueable by Debian.  If
> they are then every time Ubuntu acquires or improves a test the diff
> should be filed in the Debian BTS in the normal way (just as Ubuntu
> maintainers should, IMO, for any package).
>
> On the other hand, if the tests are not seen as valuable by Debian
> then Ubuntu should not bother the Debian maintainers with them.  But I
> think the responses so far show that Debian probably do value the
> tests[1].  That's good - as a Debian maintainer I think this automatic
> testing will be valuable for Debian too, despite any differences
> between exactly how Debian and Ubuntu will use the tests.

The Ubuntu maintainer should always open bugs with the test related stuff
and see if the Debian maintainer judge it's valuable or not.

I think it's clear and there's no "Debian decision", each maintainer or group
will judge, that's it. We don't need other flamewar to discuss the
obvious, IMHO.

> (...)
> > Since Ubuntu Dapper is actually on freeze, what's your timeline to these
> > patches? Will you include this stuff just on Dapper+1, in the end of
> > the year?
>
> This particular change to mawk will make it into Dapper.  I'm not sure
> how much effort we'll have at this stage to add tests, but since
> adding tests doesn't break anything (at least until anyone starts
> running the tests) I expect test changes to be accepted quite late.
> But this is really the kind of question for ubuntu-devel, not
> debian-devel.

Sure, but i asked because you were wearing your Ubuntu hat and in my previous
view it was relevant to the subject.

Thanks,
Gustavo Franco

Re: Problems found by piuparts

[Gustavo Franco]

On 2/20/06, Lars Wirzenius <[EMAIL PROTECTED]> wrote:
> In the past six months, I've filed about 260 bug reports based on what
> piuparts has found. About 40% of those have been fixed so far. Below is
> a summary of the common problems, hopefully the list will help everyone
> to find and especially avoid problems in their own packages.
>  (...)

Hi Lars,

I think some of these problems can be detected by lintian, adding some
more checks there. It could bring more visibility to so common errors.
Comments ?

-- stratus

Re: Problems found by piuparts

[Gustavo Franco]

On 2/24/06, Peter Samuelson <[EMAIL PROTECTED]> wrote:
>
> [Gustavo Franco]
> > I think some of these problems can be detected by lintian, adding
> > some more checks there. It could bring more visibility to so common
> > errors.  Comments ?
>
> A better way to phrase "Comments?" would be: "Here is a proof-of-
> concept patch to lintian to demonstrate which of these issues can be
> detected.  Comments?"

I won't waste my time writing a patch without hear Lars' and lintian
maintainers opinions first.  I still act as a member of the Debian
community that we used to be, and last time i see you wasn't a lintian
maintainer. You can get away with your rudeness and came up with a
patch before me, because it seems that you just don't care about
*comments*.

Thanks,
-- stratus

Re: Problems found by piuparts

[Gustavo Franco]

On 2/24/06, Russ Allbery <[EMAIL PROTECTED]> wrote:
> Gustavo Franco <[EMAIL PROTECTED]> writes:
> > On 2/20/06, Lars Wirzenius <[EMAIL PROTECTED]> wrote:
>
> >> In the past six months, I've filed about 260 bug reports based on what
> >> piuparts has found. About 40% of those have been fixed so far. Below is
> >> a summary of the common problems, hopefully the list will help everyone
> >> to find and especially avoid problems in their own packages.
> >>  (...)
>
> > Hi Lars,
>
> > I think some of these problems can be detected by lintian, adding some
> > more checks there. It could bring more visibility to so common errors.
> > Comments ?
>
> Most of those look semi-difficult to do in lintian because most of them
> require fairly deep parsing of maintainer scripts.  lintian can get a long
> way by cheating with simple heuristics, but that approach can be somewhat
> unreliable and prone to false positives.
>
> Certainly it's possible, and I expect the lintian maintainers would
> welcome patches, but I don't think it's that straightforward.
>

Hi Russ,

What i thought in a first look to the Lars' list. I think that the
best thing would include piuparts as a infrastructural test (oficially
as a part of our archive code), or due to restrict admin time to do
that, opt for something like piuparts.debian.org as we have
lintian.d.o.

I would be glad to help with a web interface to show the piuparts html
results in a organized way, but i don't have enough resources to test
every package, but i'm sure that somebody will step in at least for
i386 architecture. Since Lars already did the check (for i386 only, i
think), maybe we just need to bring more organization/visibility to
the results and keep them updated, right?

Hope that helps,
-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

What's wrong with us ? I just read some messages with a "no Martin,
can we revert it?", it seems that the default reply is "ok Martin, see
you, thanks.".

It's volunteer work, he's free to do whatever he wants and spend his
time with more pleasant tasks, but when will we try to solve some of
the real problems we have instead going through the easy way that is
"ok, who's going to take that task?". It's clear that the new stable
maintainer or group will have at least some of the current problems.
Don't you care ?

It's just something more of the same as always, we're wasting
motivated human resources waiting someone come in and take over the
task. If you think that it's always easy to replace other person and
his work and that really motivated human resource is infinite, i'm
sorry but you're wrong. It's a wake up call and isn't the first, IMHO.

Closing, thanks for your valuable contribution to this project Martin.
Hopefully, i'll be around to see the day when we'll be cool with each
other and do more to reach our goal. What about a universal operating
system?

Thanks,
-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/9/06, Amaya <[EMAIL PROTECTED]> wrote:
> Gustavo Franco wrote:
> (...)
> > Don't you care ?
>
> We do. But asking him to go back to doing something that was such a
> source of pain? No way.
>
> (...)

Hi Amaya,

I agree with you, i wasn't simple asking him to go back. My point was
that if we just let somebody or a team take over that task the problem
will still be there, maybe not all the problems but some of them.

It's harder, but we need to solve problems and stop changing people by
teams thinking it's a drop-in solution for everything. I'm in some
teams, i prefer to work this way and the 'stable updates' probably
will be in good hands in the near future but why not with Martin too?
Someone blocked him and that's my problem.

Closing, it isn't a cheap rant against ftpmasters, i would like to
thank the ftp assistants
Joerg Jaspert and Jeroen van Wolffelaar for their work too.

-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/9/06, Amaya <[EMAIL PROTECTED]> wrote:
> Gustavo Franco wrote:
> > I agree with you, i wasn't simple asking him to go back. My point was
> > that if we just let somebody or a team take over that task the problem
> > will still be there, maybe not all the problems but some of them.
>
> I am also worried that these issues continue to exist throughout
> releases/DPLs/seasons... I honestly think it is all a lack of
> transparency and communication.
>

Sure Amaya, "what we've got here is failure to communicate...". Btw, i hope we
don't end in a civil war. :-)

-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/9/06, Amaya <[EMAIL PROTECTED]> wrote:
> Gustavo Franco wrote:
> > Sure Amaya, "what we've got here is failure to communicate...". Btw, i
> > hope we don't end in a civil war. :-)
>
> 
> Civil? Not as long as we focus on attacking Ubuntu
> 
>

hehe, i think you missed my joke first.

-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/9/06, Amaya <[EMAIL PROTECTED]> wrote:
> Gustavo Franco wrote:
> > hehe, i think you missed my joke first.
>
> Maybe, I just saw a GnR quote.
>

That's what i mean with the civil war thing, not a real war against
anyone or a project. :-)

-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/9/06, Lars Wirzenius <[EMAIL PROTECTED]> wrote:
> to, 2006-03-09 kello 19:21 +0100, Amaya kirjoitti:
> > 1 - lobby (all of them)
> > 2 - get promises in exchange of votes
>
> That reminds me of something I meant to propose some time ago: someone
> with a bit of time on their hands could make a wiki page,
> DplPromises2006 say, and list all the promises of all the candidates.
> Then, during the next year, we can keep coming back to that page and
> check how well they keep their promises.
>


Liw, go ahead and do that but not in the Debian wiki. They're going to
remove^W edit the content.


-- stratus

Re: For those who care about stable updates

[Gustavo Franco]

On 3/10/06, Steve Langasek <[EMAIL PROTECTED]> wrote:
> On Thu, Mar 09, 2006 at 10:38:51AM -0300, Gustavo Franco wrote:
> > What's wrong with us ? I just read some messages with a "no Martin,
> > can we revert it?", it seems that the default reply is "ok Martin, see
> > you, thanks.".
>
> > It's volunteer work, he's free to do whatever he wants and spend his
> > time with more pleasant tasks, but when will we try to solve some of
> > the real problems we have instead going through the easy way that is
> > "ok, who's going to take that task?". It's clear that the new stable
> > maintainer or group will have at least some of the current problems.
> > Don't you care ?
>
> Which problems?  The problem where they ignore the ftpmaster trying to
> coordinate certain archive changes with the point release to make future
> point releases easier, the problem where they ignore the critical bugs that
> have been filed on the security NMU of sudo that was done with a patch that
> the maintainer disagrees with, the problem where they repeatedly make a very
> public stink whenever they were unable to resolve a conflict with other
> ftp/DSA members...?

Maybe the problem where the ftpmasters failed to communicate their
reasons and even when they're right it seems that they're wrong?!

Please Steve, i don't want to take your time and discuss with you, my
opinion is clear and it would be better if a ftpmaster or ftp
assistant spend some minutes replying Joey here. If it was already
done, i think i missed the message.

> Well, no, those seem to all be problems specific to Joey and his approach to
> the situation.  I don't see any reason at all why we should assume that a
> new SRM will fall victim to the same destructive cycle.  I guess maybe if
> you buy the idea that this is an ftpmaster conspiracy, then you might think
> that any new SRM will have the exact same problems being able to get a point
> release out; but the truth is always more complicated than that.  Joey does
> a lot of great work for Debian, but that doesn't make him a saint; and when
> a group of people are having problems working together as a team, sometimes
> the best answer really is for one of them to move on.

Don't you see that the people that are moving on, are always from the
same side of the discussion (the side that speaks) ? By the way if
Joey isn't a saint (i'm almost sure he isn't), the ftpmasters aren't
too. There's no conspiracy, it's just the way the things and decisions
are handled that isn't always clear, i'm against a yet another formal
procedure but when a Debian developer in a position listed in our
organizational structure page resigns sending a announce criticising
the work of a group listed there too, that group should came with a
response. The "we don't care" approach can't be the default response
IMHO. Isn't it strange that a ftp master or assistant never (or at
least for quite some time) resigned upset by the critics or being
blocked by someone else?

-- stratus

Re: NEW queue backing up again -- ftpmasters, any explanation or comment?

[Gustavo Franco]

On 3/13/06, Lars Wirzenius <[EMAIL PROTECTED]> wrote:
> ma, 2006-03-13 kello 08:57 +0100, Thijs Kinkhorst kirjoitti:
> > I don't think it's useful to second-guess what they're doing, so my
> > question to Nathanael: when did you post this question to them directly
> > and what was their answer?
>
> Is there a reason why the question should be made in private?
>
> I do think N.N. formulated the question in a needlessly accusatory tone,
> but I don't think -devel was the wrong place. Transparency and openness
> are good for the project.
>

I agree with you, but i don't think it's just a lack of transparency
and openness, it's just the same old way to not communicate decisions
in a proper way.

If the ftpmasters are going to stop NEW processing for a while with or
without a special criteria, they should inform us through d-d-a or the
DPL if they think it will generate too much noise, like these threads.
If they did that i'm yet to hear about.

I want the archive split and i want etch, don't get me wrong but  at
the same time i see that we're discussing communication and conduct
problems every week, it's clear that we're delaying things here.
Unfortunately, we're going nowhere once again, there's only one side
writing, people in the middle acting like lawyers and the others just
don't talk.

I'm not going to the debconf, but i would like to suggest that the
folks that are listed in the debian's organizational structure page[0]
meet up and organize a bof with others developers interested. I see
that Steve will talk about release etch in time and Enrico about the
debian community guidelines, maybe they've better ideas for this than
me.

You know, the basic idea is ftpmasters, listmasters and others that
are going to Mexico, spend some minutes sitting there, being cool with
each other and hear  the feedback from others DDs and say what they
think about some past and the current issues. Someone take notes and
post the results somewhere.

[0] = http://www.debian.org/intro/organization

-- stratus

Re: NEW queue backing up again -- ftpmasters, any explanation or comment?

[Gustavo Franco]

On 3/13/06, Jeroen van Wolffelaar <[EMAIL PROTECTED]> wrote:
> On Mon, Mar 13, 2006 at 03:23:29PM +0100, Thomas Viehmann wrote:
> > Martin Michlmayr wrote:
> > > 20:38 < Ganneff> the archive grow too fast too big. so i should not 
> > > process
> > >   NEW so fast to not grow much more in a short term. something like that 
> > > more.
> >
> > Well, if that's the reason, are updates to existing source packages
> > still allowed? I'd really like to fix my RC bugs and sync with upstream
> > at the same time but the latter would involve so-version changes.
>
> This is not the reason for any backlog, although it does limit amount of
> NEW accepts per day, but there's still plenty of room in each day to do
> a lot of NEW. The bigger bottleneck is simply human processing time.
>

It's clear for me that the NEW packages/per month processing time is
way better than sometime ago, thanks the ftpmasters for this. The
unsolved problems at this moment (as i see them) are:
- Some packages aren't accepted or rejected and stay there for too
much time with a unknown reason. If the ftpmasters agree we should add
more information in the NEW queue page[0] than the reject-faq;
- The process stopped for some time and some people thought it was due
the archive split. It's a HR issue or a PR issue and can be solved
easily IMHO.;

[0] = http://ftp-master.debian.org/new.html

-- stratus

Re: NEW queue backing up again -- ftpmasters, any explanation or comment?

[Gustavo Franco]

On 3/13/06, Michael Banck <[EMAIL PROTECTED]> wrote:
> On Mon, Mar 13, 2006 at 02:39:11PM -0300, Gustavo Franco wrote:
> > On 3/13/06, Lars Wirzenius <[EMAIL PROTECTED]> wrote:
> > > ma, 2006-03-13 kello 08:57 +0100, Thijs Kinkhorst kirjoitti:
> > > > I don't think it's useful to second-guess what they're doing, so my
> > > > question to Nathanael: when did you post this question to them directly
> > > > and what was their answer?
> > >
> > > Is there a reason why the question should be made in private?
> > >
> > > I do think N.N. formulated the question in a needlessly accusatory tone,
> > > but I don't think -devel was the wrong place. Transparency and openness
> > > are good for the project.
> > >
> >
> > I agree with you, but i don't think it's just a lack of transparency
> > and openness, it's just the same old way to not communicate decisions
> > in a proper way.
>
> I am not sure, but did you get the mails from ftp-assistant Jeroen van
> Wolffelaar to this thread?  If you haven't read them already, I suggest
> you do so, realize that there is nothing to communicate, and move on to
> other things.

I did Michael, please read my previous reply (jvw's message).

Thanks,
-- stratus

Re: NEW queue backing up again -- ftpmasters, any explanation or comment?

[Gustavo Franco]

On 3/13/06, Joerg Jaspert <[EMAIL PROTECTED]> wrote:
> On 10592 March 1977, Gustavo Franco wrote:
>
> > If the ftpmasters are going to stop NEW processing for a while with or
> > without a special criteria, they should inform us through d-d-a or the
> > DPL if they think it will generate too much noise, like these threads.
> > If they did that i'm yet to hear about.
>
> Brr, there was no such decision, so no need to inform anything anywhere.
>
> It was more a "try to not grow that fast for now, and try to remove
> stuff when you add new things". That was after the archive growing
> superfast in a very short timeframe, due to some multiple uploads of big
> packages in a row.

Oh, i see. Thanks Joerg.

> The backlog in NEW is more from some timing problems I had.

Do you consider in backlog, packages like mozilla-firefox-adblock? If
yes, couldn't us add in the NEW queue page[0] a status (or notes)
column? I'm sure you know better than me if it will be useful to
inform the maintainer that you're investigating about licensing issues
or something like that.

[0] = http://ftp-master.debian.org/new.html

Thanks,
-- stratus

Re: removal of svenl from the project

[Gustavo Franco]

On 3/14/06, Andres Salomon <[EMAIL PROTECTED]> wrote:
> Hi,
>
> I am going through the expulsion process to have Sven Luther removed
> from the project.  The process is outlined here:
> ,
> and I have already completed step 1.
>

[ Andres, don't get me wrong but in the next time you start a thread
like this one doing
cross-posting, can you set the reply-to for -devel, for example ? ]

Please Andres, the expulsion process is the last mile. Are we there yet ?

It seems that the project is splitting in two groups basically: The
people that wants to work together and release Etch, and the people
that with a reason or not wants to see it delayed. The minute after
the release team announces that we're going to delay our next release,
we will stop with these weird threads and keep arguing that we're all
volunteers and are doing our best. oh, the humanity!

I'm asking myself what's behind all that ? Ubuntu ? Probably no.
Subconcious fear to delivery in time ? Probably yes. Stop thinking
about who you're going to ask to be expelled next and spend some time
considering not my words, but just Etch.

Thanks in advance,
-- stratus

Re: Minimizing ld dependencies with --as-needed

[Gustavo Franco]

On Apr 1, 2005 7:32 AM, Andrew Suffield <[EMAIL PROTECTED]> wrote:
> On Fri, Apr 01, 2005 at 12:24:23PM +0200, Tollef Fog Heen wrote:
> > * Andrew Suffield
> >
> > | It's a method of working around bugs. Just fix the bugs
> > | instead. Update libtool to the latest version and don't -l stuff you
> > | don't need to -l.
> >
> > pkgconfig does add a bunch of gratious -l, similar to what libtool
> > used to.  This will be fixed, but I'm not sure I want to introduce
> > that so late in the release cycle.
> 
> I'm pretty sure we don't want to go meddling with this stuff at all so
> late in the release cycle, so everything here is post-sarge.
> 

I agree, but are we tracking all these post-sarge issues that are coming on d-d
and others lists? I hope that after sarge we start working on these
issues before
etch came closer.

I'm interested in put online a web page containing some things suggested to be 
investigated post-sarge (technical stuff only), anyone too?

--
Gustavo Franco -- <[EMAIL PROTECTED]>


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Minimizing ld dependencies with --as-needed

[Gustavo Franco]

On Apr 1, 2005 1:58 PM, Tollef Fog Heen <[EMAIL PROTECTED]> wrote:
> * Gustavo Franco
> 
> | I agree, but are we tracking all these post-sarge issues that are
> | coming on d-d and others lists? I hope that after sarge we start
> | working on these issues before etch came closer.
> 
> I'm doing a rebuild of sarge with a changed pkg-config now, to see
> what breaks.
> 

It's good to hear, let us known about the results. 

I'll do something about collect what's being discussed on the lists
and is being pushed to "after sarge" this weekend. It would be good
sum up these ideas in a wiki like wiki.debian.net but i accept help
and suggestions.

--
Gustavo Franco -- <[EMAIL PROTECTED]>


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Marco d'Itri <[EMAIL PROTECTED]> wrote:

On Jul 28, Amaya <[EMAIL PROTECTED]> wrote:

> Huge Troll Warning
It's sad that many people replied this way when it's obvious that Debian
has stopped innovating long ago.
But I suppose that this is part of the problem.



Debian stopped innovating? What's next? Are we going to die like Microsoft or
the BSD's? We've a bunch of really successful subprojects and the
subproject organization probably is a 'innovation' as you see what a
innovation really is.

We jumped from 'boot floppies' to 'd-i', we are working on a better
Python infrastructure, the desktop and other tasks in the installer
and after installer are a reality and hopefully will gain more
attention of our users during the Etch support cycle and more.

We aren't Google or Microsoft. We don't need cheap marketing tricks.
We need better QA (not that the team isn't doing a great job) and
polish more our work (it includes artwork). It requires people that
supports to be flamed and receive no money or congratulations from
others, they will work in tons of packages that aren't well maintained
and sometimes it will be really hard and that's why i strongly
disagree with a one person approach to maintain some core stuff.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Matt Zimmerman <[EMAIL PROTECTED]> wrote:

On Fri, Jul 28, 2006 at 10:46:57AM -0500, John Goerzen wrote:
> Debian is a project of volunteers.  I am a Debian volunteer.  I'm not
> going to write something just because you gripe at me about it.  I have
> no obligation to you.  I will work on things that are interesting to me.

Absolutely agreed.  One small addition though:

> Ubuntu has paid employees, which changes the equation entirely.

There are more than a few Debian developers who are employed to work on
Debian; they are simply less visible (and possibly more fragmented, working
for different organizations) than the developers who work for Canonical on
Ubuntu.



I agree, but while some Ubuntu employees work on the distribution
agenda, Debian has no clear agenda and these Debian developers working
on random companies follow their own way or the company roadmap, that
a lot of times involves custom Debian distributions development or
weird non integrated hacks fragmenting the efforts.

Ubuntu does a interesting job keeping the 'ecosystem' around while
Debian ignore or don't see it as a problem or something that needs
attention really.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, martin f krafft <[EMAIL PROTECTED]> wrote:

Please reply to -project only!

also sprach Matthew Garrett <[EMAIL PROTECTED]> [2006.07.28.1737 +0100]:
> If Debian had slightly less of a culture of "Keep your hands off
> my package", I'd do it here instead.

I've been thinking about this a lot for the past week.

Is there any way this could be changed?

Does Debian *want* it changed?


I would like to see it, but with some comments:

Before Etch:

* Promote NMU LowThreshold wiki list giving it some official status.
The developer needs to be logged and mark if all his packages (where
he's listed as uploader) can be NMU'ed or not. He could add comments
like "I'm listed as uploader in foo but group x is the backup
uploader";

After Etch:

For NEW packages:

* Add in the new ITPs reports what group (alioth) is responsible for
your package too (eg: vte, pkg-gnome) or could act as "backup
uploader". The package doesn't need to be in the group's svn
repository, to give it some more flexibility. I think it should be the
development/upload case for a lot of packages;

* If the ITP is without the "backup uploader" means that any DD is
free to upload it. Almost no coordination to change or upload new
packages (upstream release, RC bugs, wishlist bugs) is needed. It
would be good if  "Joe" ping the maintainer before upload, just to
avoid version conflicts, duplicated work and all that;

Note: Those that are upstream in project bar and uploads the package
for Debian just for its own use, might open RC bugs against the
package. The rationale is if you don't want that nobody uses your
stuff, touch your sacred upstream code, use it at home alone! :-)

For existing packages:

* The package that contains only the Maintainer field with the name of
a person and not a group can be uploaded by any DD. ping the current
maintainer is good but not required;

* If the package contains a group in the Maintainer field and/or a
group of people in the Maintainer field or Uploaders. It's required
that the uploader ping the group and coordinate his upload.

I think with something similar to what i wrote above we will end up
with almost all the packages maintained by groups and some packages
maintained by Debian as a whole and not individuals.  The next step
would be groups allowing other groups to upload some of "their"
packages.

The core stuff will be more flexible and well maintained, if we don't
have groups where just one person do all the work and others are there
just to look cool.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, John Goerzen <[EMAIL PROTECTED]> wrote:

On Fri, Jul 28, 2006 at 07:22:11PM +0200, Marco d'Itri wrote:
> On Jul 28, Gustavo Franco <[EMAIL PROTECTED]> wrote:
>
> > Debian stopped innovating?
> Yes. This should be obvious to people who joined the project before 2000.

I'm one, and it's not obvious to me.

Some things that have been done since I joined debian:

 * apt
 * unified menu system
 * debconf
 * debian-installer
 * buildd.debian.net
 * pbuilder
 * srcinst
 * xen integration
 * selinux integration
 * cron-apt
 * x.org integration
 * {svn,arch,darcs,bzr}-buildpackage
 * signed packages
 * packages.qa.debian.org
 * package pools
 * lintian
 * testing

Some of these are quite recent.



Thanks for burning his 'i was here first, so i know' comment. Btw, i'm
here for ages
too, i just haven't the DD red swirl in my cap.

regards,
-- stratus
ps: You forgot tasks! :D


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, martin f krafft <[EMAIL PROTECTED]> wrote:

also sprach Gustavo Franco <[EMAIL PROTECTED]> [2006.07.28.1838 +0100]:
> * Promote NMU LowThreshold wiki list giving it some official status.
> The developer needs to be logged and mark if all his packages (where
> he's listed as uploader) can be NMU'ed or not. He could add comments
> like "I'm listed as uploader in foo but group x is the backup
> uploader";

How about a flag in the package's control file and an appropriate
column in the PTS?



I'm not sure about the details. Think about this use case: You're in
the middle of the transition, so this is ok NMU all  "your" packages
related with that transition. You won't upload a revision to point out
that NMUs are welcome. We're using wiki articles and stuff like that
today. There's wotomae (DWTT) coming too..

Well, i think the general idea is ok, we need to think about the
details and side effects. I would prefer something using PTS, but not
requiring package changes. It could be a signed mail to PTS
subscribing/unsubscribing packages from our own "Low Threshold" list.
The uploader would be able to query this information by email or using
the web interface.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, Clint Adams <[EMAIL PROTECTED]> wrote:

> Yes, and we could start by really enforcing co-maintainership.  Make it 100%
> mandatory for all essential, required and base packages at first.

Are there packages which are particularly well co-maintained right now?



What about debian-installer, xorg, gnome, kde, ... ?

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, Fabio Tranchitella <[EMAIL PROTECTED]> wrote:

Hi Pierre,
  please don't Cc me, I read this list. :)

Il giorno ven, 28/07/2006 alle 19.28 +0200, Pierre Habouzit ha scritto:
> and that won't happen because I'm not very keen on leraning yet another
> VCS, and that other's think the same, and that you will find poeple
> that never used svn or just can't use it, and poeple that never used
> bzr or don't like it , or ...

I was talking about repositories, not a single monolithic
repository: you are free to use cvs, svn, monotone, bzr, darcs or
whatever else you prefer. If every developer would use a common server
for his repositories, it would be easier for the others to find and use
them.
(...)


I think we've two important points here:

* Are you working on the package foo ?

In a scenario when anybody or tons of people can upload the package foo, it's
necessary to tag somewhere that you're working on the package foo. Groups do it
using IRC or wiki articles today. We could do it using the NEWS in the
PTS. It won't
break the current groups approach since these groups can point the wiki articles
or irc channels for coordination there.

* Where are you working on the package foo ?

You could send the vcs information about the package to the PTS that would list
it in the web or answer it through a mail query. No central repository
of any vcs
is required.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Roger Leigh <[EMAIL PROTECTED]> wrote:

John Goerzen <[EMAIL PROTECTED]> writes:

> On Fri, Jul 28, 2006 at 07:22:11PM +0200, Marco d'Itri wrote:
>> On Jul 28, Gustavo Franco <[EMAIL PROTECTED]> wrote:
>>
>> > Debian stopped innovating?
>> Yes. This should be obvious to people who joined the project before 2000.
>
> I'm one, and it's not obvious to me.
>
> Some things that have been done since I joined debian:
>
[...]
>
> Some of these are quite recent.

Agreed.  However, I do feel that some package maintainers are not
exercising enough thought about how their packages can better
integrate with the Debian system as a whole, as well as other related
packages.  As an example, there are many packages not registering
their documentation with doc-base, and doc-base itself could do with
some work on better supporting documentation formats other than HTML.



This is the QA work i talked about. List these packages, open bugs (take care
about mass bug filing) and coordinate the effort to patch them and/or do NMUs.

This kind of task takes time and isn't well documented on how to do from
scratch, but hopefully we will have another web tool (wotomae) coming to
help with this and other stuff.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Matthew Garrett <[EMAIL PROTECTED]> wrote:

John Goerzen <[EMAIL PROTECTED]> wrote:

> Debian is all about not making assumptions like this.  If you want this
> feature, you are free to install it.

And, while this makes Debian a wonderful choice for all sorts of things,
it means that Debian itself isn't a compelling desktop distribution.
Instead, that's going to be left to derived distributions of one sort or
another.


No, no. Please download the latest testing image[0] and after base
install select "desktop environment", give it a try and you will see that if a
better artwork we're quite good. Btw, there's a chance that the installer will
suggest that task for you. The users are free to unselect it, of course. I think
with the tasksel revival in place, Ubuntu is going in a similar path.


Personally, I have no problem with this. But if Debian is unwilling to
fill these (not terribly niche) requirements itself, it's not reasonable
to complain when people build on Debian in order to provide a more
complete solution for a more narrow use case. It's not possible to
simultaneously believe that Debian's flexibility is what makes it
worthwhile, and that the fact that other projects treat Debian as a
supermarket is a bad thing. One or the other.


Agreed, but please review our current status as pointed out above.

[0] = http://www.debian.org/devel/debian-installer

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Amaya <[EMAIL PROTECTED]> wrote:

martin f krafft wrote:
> I am finding lots of interesting points in other people's replies.

Yes, I am also surprised about the civil tone (on most of the replies).


I've the same feeling, best thread in Debian mailing lists for me
after years and years.


> I think the discussion so far has been much constructive. Whether we
> are going to let it deteriorate to a flame fest or not will largely
> serve to show whether Debian has learnt something over the past
> years or not.

100% agreed. I wonder if the "Huge Troll Warning" had any effect.


You know, you need at least two trolls to make a flame fest. If you do
constructive discussion based on a troll, he lost arguments and die.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: wotomae? (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, martin f krafft <[EMAIL PROTECTED]> wrote:

also sprach Gustavo Franco <[EMAIL PROTECTED]> [2006.07.28.1901 +0100]:
> This kind of task takes time and isn't well documented on how to
> do from scratch, but hopefully we will have another web tool
> (wotomae) coming to help with this and other stuff.

What's this? Any links?



I'm sorry, writing too many replies. There you go[0].

[0] = http://wiki.debian.org/DWTT

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Tyler MacDonald <[EMAIL PROTECTED]> wrote:

Matthew Garrett <[EMAIL PROTECTED]> wrote:
> Personally, I have no problem with this. But if Debian is unwilling to
> fill these (not terribly niche) requirements itself, it's not reasonable
> to complain when people build on Debian in order to provide a more
> complete solution for a more narrow use case.

Isn't the "Task:" package header (and maybe the "Tag:" header as
well) and the corresponding menu options in the debian-installer supposed to
make debian magically morph into whatever you want it to be?


In a way, right.


A few messages ago there was a message about update-notifier, and
how it wouldn't be installed and/or enabled by default due to administrative
access issues, etc... if someone specifically says that they are installing
a debian desktop system, would it then make more sense to auto-install,
config, and enable update-notifier? If not for the entire "users" group,
then at least for the user that's created on system install, that gets added
to cdrom, floppy, audio, and video groups already?


We do, see my reply for Matthew and test if you want. You can install
the 'desktop' and 'gnome-desktop' task in a sid or testing system
using aptitude too.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Joey Hess <[EMAIL PROTECTED]> wrote:

John Goerzen wrote:
> Think about it -- if you manage dozens, hundreds, or thousands of Debian
> machines -- few of which even have a monitor -- how useful is this?
>
> Debian is all about not making assumptions like this.  If you want this
> feature, you are free to install it.

Debian is capable of figuring out whether it's being installed with a
monitor[1] at install time; if it is then update-notifier will be installed
(in etch, as part of the gnome-desktop task, thanks to the work of
Stratus).


our work, of course. Talking about our work, i'm unable to receive a response
from you for some time. :(


It's a fallacy to think that just because Debian wants to support
everyone that we can't target common sets of users, like desktop users,
and support them well, with polish.


This is a urban legend about Debian, a myth that hopefully we will burn
with Etch, nothing against you John, really. Unfortunately, this is a
wrong common
sense between a lot of developers.

In a way, we will "innovate" again (i hate this word, but to stay on
topic) with Etch,
and for those who don't know and care, it seems that Ubuntu partially
agreed with our
approach and probably will use tasksel in their next release (Edgy,
getting rid or not of their current metapackage approach), something
that they have disabled since their first release (Warty). Do you see
Katrina?

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Katrina Jackson <[EMAIL PROTECTED]> wrote:

I keep seeing your arguments about how some people want on;y a 100 MB system
by default.  But you do give an option to have a Desktop, even Laptop
install.  What not make those installs have better support for Desktop and
Laptop Users.  For example, why not have the package updater installed for
those who specifically want to install the desktop environment?  Those who
only want servers or 100MB installs will not be hurt by this move.  Then,
Desktop and Laptop users are happy too.


No you don't. It was just John Goerzen that commented about this and I
and Joey Hess pointed him at some progress on this area.

I and others are working very hard to polish the desktop environment
and laptop related stuff, actually the desktop environment installs
the update-manager and update-notifier, as i wrote some messages ago.


 In short, it would be nice if the Desktop and Laptop environments catered
to Desktop  users.

 I see how Debian is innovative with apt, lintian, etc..  But why not
innovative with the Desktop experience?


Suggestions after your new tests are welcome (hint: a better and
consistent artwork is pending).


 (...)
 PS.  Hardware, Hardware, Hardware, I have to confess, if there was better
hardware support I think most people would be happy.  Hardware supported by
Ubuntu 6 months ago, should be supported by Debian by now.


Do you have a list for comparison? If not, i can point out a some
hardware setup that works on Debian Sarge, our last release and not in
Ubuntu Dapper. Is it Ubuntu's fault? In a lot of cases, no. These days
even Microsoft Windows users are worried about hardware selection and
operating system compatibility.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, Mario Iseli <[EMAIL PROTECTED]> wrote:

Hi there,

(...)
I imagine if we would have a big CVS tree like Gentoo or some BSD's, i
wouldn't know where to begin with my work or what I sould do. The forest
is so large and you don't see the tree!



I don't think we need a central approach, every maintainer or group of
maintainers is free to use svn, bzr, whatever with Debian
infrastructure or not. It would be good if the maintainers pointed out
the vcs they're using, and its url through PTS.

*BSD are a different and smaller beasts in this matter and what about
when their cvs repositories are offline? :-)

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, Adeodato Simó <[EMAIL PROTECTED]> wrote:

* Gustavo Franco [Fri, 28 Jul 2006 14:38:52 -0300]:

> * Promote NMU LowThreshold wiki list giving it some official status.

And remember that (well done) NMUs are not only for bugs of RC severity.
For example, I'm going to upload to 7-delayed a fix for #368917, sending
the patch to the BTS at the same time.



Yes, i agree. My point was that we can do the "promote NMU
LowThreshold" thing before Etch, and after that do more intrusive
changes in the way we maintain packages, that involves ban the NMU
concept, except for groups or groups of groups maintaining packages.

regards,
-- stratus

Re: package ownership in Debian (was: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/28/06, Katrina Jackson <[EMAIL PROTECTED]> wrote:

  Okay here is another honest question:   Do you really honestly think not
having co-maintainers for base packages is ever a good idea?  What if
someone is busy?  You don't really feel safe noticing your base packages
aren't being co-maintained since people are busy.


No, *I* don't. But the big point in Debian is that my and your opinion
won't change anything. Debian is a meritocracy with all its pros and
cons, not that *i* (again) am not working to change something in this
maintainers, co-maintainers area.


 Also back to the innovation issue:  Don't your think having more
comaitainers there would be more innovation merging a package into Debian.
Is your goal just to make a package compile and not have RC bugs, or is it
to take a package and innovate in a way to make that package experience must
better for a Debian User?  With co-maintainers this would be easier to pull
off.


co-maintainers means more hands, but not always responsibility and
work (innovative
or not). What i've in mind is that if there's a group where only one
person works, the others well organized and balanced groups can put
some more pressure there than when it's a one man sacred thing. Read
my message about what could be done before and efter Etch on this on a
subthread closer to you. :)


 Not to cause a fight, I just wanted to point out some say all Debian is
worried about is RC bugs.  "Debian worships RC bugs".  With co-maintainers
you could do a better job "worshiping RC Bugs" While doing additional work
to innovate how that package will work in Debian.  Plus more QA support when
more then one work on the package.


You're obsessed with innovation while i'm obsessed with QA and
features. I can tell you for sure that a lot of developers cares about
their own packages and just that, QA, shiny and new stuff (well
integrated with others packages or not), their own agenda, their
companies agenda, and yadda, yadda, yadda. At the same time we fail to
merge everything possible and cook the best distribution out there,
we've some success in a lot of areas.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/28/06, Daniel Baumann <[EMAIL PROTECTED]> wrote:

Gustavo Franco wrote:
> For existing packages:
>
> * The package that contains only the Maintainer field with the name of
> a person and not a group can be uploaded by any DD. ping the current
> maintainer is good but not required;

then I will have to found a 'these-are-daniels-packages'-group
consisting only of me? *scnr* :)


I meant with group of maintainers,  number of uploaders > 1. Joerg
Jaspert said that he wouldn't like to be forced to team maintenance
and suggested 0day NMUs for >= normal bugs with current rules (patch
to the bts), so if you add this rule to my suggestion, i think it's
better than 'ping not required', would be better than now too.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/28/06, Simon Richter <[EMAIL PROTECTED]> wrote:

Hello,

Gustavo Franco wrote:

> * The package that contains only the Maintainer field with the name of
> a person and not a group can be uploaded by any DD. ping the current
> maintainer is good but not required;

I propose that under that policy, if someone NMUs a package without
clearing the patch with the maintainer first, that person is responsible
for the package until the maintainer acknowledges or reverts the NMU.

The rule of sending a patch to the BTS and giving a bit of time to reply
serves quality assurance more than it hurts, because you get a second
opinion, from a person who is familiar with the package.


I reverted my opinion, agreeing with Joerg's idea.


> The core stuff will be more flexible and well maintained, if we don't
> have groups where just one person do all the work and others are there
> just to look cool.

But that is exactly what happens if you force group maintenance on
everyone. Only that within a group, you cannot even point at somebody
and say they are responsible for anything.


Every group has at least one admin. I was talking about avoid one-working-man
groups. You always can blame the group and the group admins. Don't we blame
ftpmasters and their members ? XSF ? same deal.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Josselin Mouette <[EMAIL PROTECTED]> wrote:

Le vendredi 28 juillet 2006 à 23:10 +0200, Marco d'Itri a écrit :
> On Jul 28, Josselin Mouette <[EMAIL PROTECTED]> wrote:
>
> > Happily you are here to innovate. For example, I love that kernel/udev
> > breakage in sarge to etch upgrades. This is indeed a great innovation
> > (i.e. something we didn't have before) that was inspired from Ubuntu.
> Nice personal attack, but you are wrong.

The last time I tried a sarge -> etch upgrade, the udev upgrade stopped
until I created /etc/udev/kernel-upgrade manually.

If this has changed, I would be the first to be pleased.


Let us calm down and think twice before replying, please.

Btw, it happened with a co-worker yesterday, i asked him to open a
bug. I think you could check for the report or open the bug Joss.
Hopefully Marco or somebody else is working on a better solution.

regards,
-- stratus

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/28/06, Josselin Mouette <[EMAIL PROTECTED]> wrote:

Le vendredi 28 juillet 2006 à 18:53 -0300, Gustavo Franco a écrit :
> Let us calm down and think twice before replying, please.
>
> Btw, it happened with a co-worker yesterday, i asked him to open a
> bug. I think you could check for the report or open the bug Joss.
> Hopefully Marco or somebody else is working on a better solution.

See the bug #35, which has been closed without being actually fixed.

The better solution is known: it implies being able to install several
udev versions together. Now *that* would be a great innovation. However
it requires a lot of work and no one seems to be interested enough to
spend the necessary time - including the maintainer himself, criticizing
the lack of innovation in Debian.


Joss, please let us focus on solve the problem than keep hurt the
project with this. I disagree with Marco lack of innovation message,
but it makes no sense review every bug on "his" packages. That's what
is wrong with us. Marco has no packages, he has work, contributions
and some more. We will be responsible for Etch, at least i will feel
this way.

We don't have a sane way to "put the file there" (trigger the upgrade
process) if the user is upgrading? You know, through some dpkg compare
versions magic and in the worst scenario asking for user input through
debconf.

regards,
-- stratus

Re: package ownership in Debian

[Gustavo Franco]

On 7/29/06, Adeodato Simó <[EMAIL PROTECTED]> wrote:

* Thomas Viehmann [Fri, 28 Jul 2006 23:40:19 +0200]:

> If that is wanted, I'd consider it important enough information to have
> it in debian/control.

A couple packages of mine ship already with an X-VCS-Bzr header in the
source. Example:

  (Format: X-Vcs-${VCS}: ${URL})
  X-Vcs-Bzr: http://people.debian.org/~adeodato/code/packages/taglib

Another, perhaps more parseable format, would be:

  X-VCS-Url: ${VCS}:${URL}
  X-Vcs-Url: bzr:http://people.debian.org/~adeodato/code/packages/taglib

Though you'd had to wonder what you'd do with a svn:// url.



Looks good, but i think we could write a proposal for new headers to
address 'vcs' and 'homepage' needs, or add it somewhere else and sync
(like we do with Tags). Thoughts?

regards,
-- stratus

Re: package ownership in Debian

[Gustavo Franco]

On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:

On Fri, 28 Jul 2006 14:38:52 -0300, Gustavo Franco
<[EMAIL PROTECTED]> said:

> * Promote NMU LowThreshold wiki list giving it some official status.

What does this mean?


That you're out of date on what's going on and trying to make jokes of
my opinions before reading for a second time.

FYI, http://wiki.debian.org/LowThresholdNmu


> For existing packages:

> * The package that contains only the Maintainer field with the name
>   of a person and not a group can be uploaded by any DD. ping the
>   current maintainer is good but not required;

> * If the package contains a group in the Maintainer field and/or a
>   group of people in the Maintainer field or Uploaders. It's
>   required that the uploader ping the group and coordinate his
>   upload.

Why the disrtinction? If the maintainer is active, and is
 currently working on the package, you have just made matters worse by
 not pinging.  If the team is negligent, why should it be treated
 differently?


I've changed my opinion on this based on what Joerg wrote, already
stated this some messages ago two times.


> I think with something similar to what i wrote above we will end up
> with almost all the packages maintained by groups and some packages
> maintained by Debian as a whole and not individuals.  The next step
> would be groups allowing other groups to upload some of "their"
> packages.

I am mostly unconvinced that this would improve the quality of
  distribution.


Yes, so let us keep the lack of communication and "my packages, don't
touch team" approach or do you have suggestions for the problems on
the table? Don't you see any problem?


> The core stuff will be more flexible and well maintained, if we
> don't have groups where just one person do all the work and others
> are there just to look cool.

Why do you think that is not likely to happen?  My experience
 leads to to think that your view point is akin to the island called
 Utopia.



My experience with Debian Python Modules Team, pkg-ltsp and pkg-gnome,
show me that groups, communication in these groups, cooperation from
non DDs are better than "mail me and i'll reply in two weeks...".
Again, my point isn't that every package should be under group (as in
alioth) maintenance, some packages would be under group maintenance as
in Debian under some less restrict rules for NMU along the lines
Joerg's wrote.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:

On Fri, 28 Jul 2006 12:23:33 -0500, John Goerzen <[EMAIL PROTECTED]> said:

> On Fri, Jul 28, 2006 at 05:44:38PM +0100, Steve Kemp wrote:
>> > If Debian had slightly less of a culture of "Keep your hands off
>> > my package", I'd do it here instead.
>>
>> That seems understandable.  I'm keen on teams, but even more keen
>> on a less "ownery" stance by package owners.

> I agree.  We should do it like the BSDs: a tree that any developer
> can commit to, for any package.

How would you handle dilution of responsibility?  Or too many
 people making broad commits to to many packages without considering
 the details of the particular package being touched?

When everyone is responsible for something, no one is
 responsible.



Group admins. A group can have more than one admin, but at least one
is reachable and/or is leading the group (eg: python-modules, d-i), if
not the group organization is broken, but it happens more often with
the one man approach by default, IMHO.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:

On Sat, 29 Jul 2006 16:27:26 +, Gustavo Franco <[EMAIL PROTECTED]> said:

> On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:
>> On Fri, 28 Jul 2006 14:38:52 -0300, Gustavo Franco
>> <[EMAIL PROTECTED]> said:
>>
>> > * Promote NMU LowThreshold wiki list giving it some official
>> >   status.
>>
>> What does this mean?

> That you're out of date on what's going on and trying to make jokes
> of my opinions before reading for a second time.

Actually, it just reinforces my view that you jump to wild
 conclusions with inadequate data.

> FYI, http://wiki.debian.org/LowThresholdNmu

*Sigh*. I see I'll have to use non polysyllabic words here.

I know about the LowThresholdNmu page. What exactly do you
 mean about giving it "official status"?


I wrote that it could be integrated with PTS, somebody else suggested
a new header in control. I think subscribe/unsubscribe a package
(using signed messages) to "LowThresholdNMU" with notes that could be
queried by mail and included in PTS web interface, would do.


>> > I think with something similar to what i wrote above we will end
>> > up with almost all the packages maintained by groups and some
>> > packages maintained by Debian as a whole and not individuals.
>> > The next step would be groups allowing other groups to upload
>> > some of "their" packages.
>>
>> I am mostly unconvinced that this would improve the quality of
>> distribution.

> Yes, so let us keep the lack of communication and "my packages,
> don't touch team" approach or do you have suggestions for the
> problems on the table? Don't you see any problem?

We have a policy on NMU's. I think the release team has
 authorized 0-day NMU's, while following the rest of the NMU
 guidelines (nmudiff to the BTS, etc), in order to correct lacunnae in
 packages.


0day NMU's for RC bugs more than a week old - send the patch to the
BTS before upload apply. This is different than Joerg's idea.


There is nothing wrong with offering to help out with packages
 either -- and nothing wrong with people forming teams. Rammning it
 down people's throats won't work, though.


Don't you see that the team thing is to avoid a random developer that
have no idea what's going on with the history of that package, do the
upload ? The packages that aren't under group maintenance and will
never be, needs more not so strict NMU rules.


> My experience with Debian Python Modules Team, pkg-ltsp and
> pkg-gnome, show me that groups, communication in these groups,
> cooperation from non DDs are better than "mail me and i'll reply in
> two weeks...".  Again, my point isn't that every package should be
> under group (as in alioth) maintenance, some packages would be under
> group maintenance as in Debian under some less restrict rules for
> NMU along the lines Joerg's wrote.

You have examples pro teams.  There are also anecdotes where
 teams do not work.  Teams are akin to marriages: somethimes they work
 wonderfully, other times they result in the analogue to a nasty
 divorce. Even worse are teams that function like bad marriages: there
 is tension in the air, people distrust other members on the team,
 commits are reverted with no discussion, changes are made to SVN
 trees without any discussion, and the whole project suffers.


Please don't attack the team model, without pointing where it could be
better if it was a one-man approach. "The team foo is broken!" but it
would better with you or me maintaining the package(s) alone? Who
knows?


So, if teams form naturally, and work well, that great.

Mandating it from up on high is not.


I think the discussion is around how to put the teams to work well and
some kind of better relationship between the teams and less strict NMU
rules to non team maintaned packages.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:

On Sat, 29 Jul 2006 20:00:21 +, Gustavo Franco
<[EMAIL PROTECTED]> said:

> On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:
>> On Sat, 29 Jul 2006 16:27:26 +, Gustavo Franco
>> <[EMAIL PROTECTED]> said:
>>
> I wrote that it could be integrated with PTS, somebody else
> suggested a new header in control. I think subscribe/unsubscribe a
> package (using signed messages) to "LowThresholdNMU" with notes that
> could be queried by mail and included in PTS web interface, would
> do.

Ah. Sounds like a decent idea.

>> There is nothing wrong with offering to help out with packages
>> either -- and nothing wrong with people forming teams. Rammning it
>> down people's throats won't work, though.

> Don't you see that the team thing is to avoid a random developer
> that have no idea what's going on with the history of that package,
> do the upload ? The packages that aren't under group maintenance and
> will never be, needs more not so strict NMU rules.

Seems to me you need rules exactly as strict, since the people
 doing the NMU are less familiar with the package, and thus  need to
 exercise more care, and need to bring in the input of the person most
 experienced with the package. Lowering the upload threshold from
 people unfamiliar with the package probably would lead to a drop in
 quality, simply because it is harder to package something one is
 unfamiliar with.


I'm not here for push "new upstream releases" into your packages, for
example. We're talking about bug fixing and better integration, eg:
better hardware support, as Anthony pointed out.


(...)


regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/29/06, Manoj Srivastava <[EMAIL PROTECTED]> wrote:

On Sat, 29 Jul 2006 21:20:35 +, Gustavo Franco
<[EMAIL PROTECTED]> said:

> I'm not here for push "new upstream releases" into your packages,
> for example. We're talking about bug fixing and better integration,
> eg: better hardware support, as Anthony pointed out.

Even then, there might be different ways of implementing a bug
 fix -- and thus my input should still be sought. After all, I'll be
 the one who has to live with the code.

What about fixes for bugs that would have been labelled
 wontfix were I had been given time to  ofer an input? Not all
 reported RC bugs are indeed that, or warrant a fix.

I see no real reason for cutting the maintainer out of the NMU
 loop, even if they are not a team -- which is why provisions for
 keeping the maintainer in the loop are a Good Thing™.



I agree. Please read again[0], [1] and [2]. You replied [0] after i
wrote [0], [1] and [2].

[0] = http://lists.debian.org/debian-devel/2006/07/msg01170.html
[1] = http://lists.debian.org/debian-devel/2006/07/msg01219.html
[2] = http://lists.debian.org/debian-devel/2006/07/msg01230.html

thanks in advance,
-- stratus

Re: Etch artwork (was: Re: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

On 7/31/06, Michael Banck <[EMAIL PROTECTED]> wrote:

On Mon, Jul 31, 2006 at 10:37:54AM +0100, Jon Dowland wrote:
> At 1154196833 past the epoch, Michael Banck wrote:
> > It makes no sense to complain about this until we have a
> > good default artwork.  So let's fix that first, then
> > convince the gdm maintainer that we should use this.
>
> Well put. On that note, where is consistent art work being
> coordinated?  I'm interested in helping out here, if
> possible.

We have started trying to coordinate this at

http://wiki.debian.org/DebianDesktopArtwork

What we need first is somebody to make good screenshots of the default
GNOME, KDE (and maybe XFCE) desktops and their default toolkit/window
manager themes (and check back with the packaging teams whether any
changes until etch release are planned to this end), so we can figure
out whether one set of artwork is possible, or (due to big
color/whatever differences) we need several sets.


We already have some good artwork. The problem is where it is and
visual consistency between the different desktop environments. Btw,
some wallpapers will work better in a desktop environment and won't in
others (eg: text or image in the bottom is evil, we usually have
panels there).

I want to make a call for a online meeting to discuss some points with
gdm, kdm, pkg-gnome,
pkg-kde and xfce maintainers. I just need to talk (again) with Joss
about desktop-base
plans and see what he think about my ideas (see below).


Then, decide on some basic things like color (Debian purple? (ugh!)
Blueish? (yay!) Greenish?) and then announce a Call for Artwork through
appropriate channels (to be determined, d-d-a might not cut it)


I agree.


The package integration stuff seems to have happened at least for GNOME
thanks to Gustavo, the KDE team should check what needs to be done on
their side.


We've desktop-base package, that i would like to:
- Move desktop-base from pkg-gnome to a "neutral" alioth project where
at least two members of each group has commit access. I've asked for
the second time debian-desktop group members (2, non-DDs) input about
this yesterday;
- Use to build gdm and kdm themes;
- GNOME, KDE and XFCE splash screens;
- sid wallpaper (common for all environments);
- etch wallpaper (to be easily changed during freeze).

That is what i want to discuss in a online meeting with the parties
involved, this is just what i've in mind, nothing was decided yet.
Feedback about it, specially from the maintainers involved in the
changes is welcome.

The sid->etch wallpaper is a release issue and should be discussed
with the release team too, but it won't hurt change to Etch wallpaper
right before the freeze if required by them. The kdm is maintained by
the pkg-kde group, but the gdm isn't. Actually gdm is configured to
pick a random theme in each login. I think Joey Hess talked with Ryan
Murray about change this in the near future.

FYI, i'm including GNOME and KDE because there are tasks gnome-desktop
and kde-desktop respectively. I've prepared a xfce-desktop task, but
it won't be in d-i beta3 due to tasksel freeze. Hopefully it will be a
possibility preseeding the Etch installer.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: package ownership in Debian

[Gustavo Franco]

On 7/31/06, Hubert Chan <[EMAIL PROTECTED]> wrote:

On Sat, 29 Jul 2006 20:00:21 +0000, "Gustavo Franco" <[EMAIL PROTECTED]> said:

[...]

> The packages that aren't under group maintenance and will never be,
> needs more not so strict NMU rules.

Why?



Due to the "my stuff, don't touch that!" current approach, but (again)
this is just IMHO.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Etch artwork (was: Re: Why does Ubuntu have all the ideas?)

[Gustavo Franco]

Message sent to the related projects mailing lists and maintainers.
Time to deal with the goodies, pressure, critics and work...

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Debian should have a weekly debate

[Gustavo Franco]

On 7/31/06, martin f krafft <[EMAIL PROTECTED]> wrote:

also sprach alfredo diega <[EMAIL PROTECTED]> [2006.07.31.1616 +0100]:
> I really believe Debian would benefit if they had a weekly debate.
> I understand you debate on things you are voting on, but I think
> it could be better if you debated on general concerns.  I was
> reading  the "Why Ubuntu has all the ideas" thread and think two
> things were apparent.

I don't think a debate is in any way a useful format for F/OSS, but
there's nothing preventing you from inviting to one and moderating
it.


I thought #debian-tech @ irc.oftc.net was started with the "technical
debates" goal in mind, no?


> Every day when I watch CNN they have a panel of people debating
> key issues which need to be addressed.

I generally end up wanting to throw up seeing those people discuss.
Words and no action. And I more than often wonder why some people
are even given the chance chance to speak about a topic. Like the
other day I read a statement by Britney "hussy" Spears about the
Lebanon situation. WTF?

I'd prefer if Debian got work done rather than talk (too) much about
issues. We know we're going to disagree on many things, and mailing
list discussions at least give us links with some of the useful
comments to link to from DWN or other discussions.


I agree, but i think that the online meetings about specific projects
work. (eg: d-i, ltsp, ...)

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Why does Ubuntu have all the ideas?

[Gustavo Franco]

On 8/1/06, Josselin Mouette <[EMAIL PROTECTED]> wrote:

Le dimanche 30 juillet 2006 à 08:36 +0200, Christian Perrier a écrit :
> To be fair with Ryan here, I seem to remember that he mentioned (maybe
> not in the bug report) that he would consider making a Debian theme
> the default...if one gets enough acceptance.

Great!


Sure.


> So, someone has to come with a nice Debian theme for gdm..:-)

The ayo, debian, debian-dawn and debian-greeter themes all received very
good feedback. Maybe we can organize an informal vote to choose which of
them will become the default.


I think we have a parallel discussion going on with the pkg-xfce,
pkg-kde and pkg-gnome teams that hopefully we will end in a online
meeting really soon.

Christian, i haven't mailed you but if you're interested let me know
and i'll forward the messages for you.

regards,
-- stratus

Re: Thanks for the work

[Gustavo Franco]

On 9/3/06, David Barker <[EMAIL PROTECTED]> wrote:

I'm writing this mail to you the developers to say thanks for the work
you put into Debian. I am really impressed at how polished etch as a
desktop OS is. Apart from the few odd things (mostly because I'm on
64bit and its still testing) its really good. I thought the installer
did a great job at figuring out all my hardware (sound, graphics,
network, CD drives etc etc all perfectly detected). I like the gnome
install it does and the selection of packages it installs for you, I
have no problems browsing samba shares, copying songs onto my ipod,
playing different videos which are all things that I struggled to do
under Linux before. In fact to be honest I was just testing the new
version to see what it was like but its so good I have no problems at
all sticking with it and ditching windows completely on my PC! Thanks
again, David


Hi David,

Thanks for taking your time to test our next release. If you've
usability suggestions for the default desktop environment
installation, drop us a message at [EMAIL PROTECTED]

Keep in mind that (hopefully) we will do a better job with the artwork
really soon. We will be adding some more stuff (packages), and general
testing and feedback is welcome.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Desktop task(sel) in Etch? (Bug #389092)

[Gustavo Franco]

On 9/26/06, Petter Reinholdtsen <[EMAIL PROTECTED]> wrote:


[Christoph Haas]
> Currently the way to install KDE is to provide the option
> "tasksel/first=kde-desktop" when booting the Etch installer CD. No
> user will guess that.

Ah, is that how it is done.  Thank you for the clue.  It should be
mentioned in the documentation available on F2-10 when the CD boot.


I agree with the documentation bit, but as i said it will be possible
use 'tasks=kde-desktop' in d-i etch rc1.


In Debian Edu, we use KDE.  The choice was mostly because of the
availability of translation for the languages we care about, but also
because the KDE tools were (and are) very good and well integrated.

Perhaps someone should wrap up a separate set of CDs with KDE as the
default desktop?  It should not be too hard.  Debian/Gnome and
Debian/KDE could exist side by side.


FYI, the three current desktop tasks (desktop, gnome-desktop,
kde-desktop) won't fit entirely in the first CD. They fit in DVD1
though. My proposal is generate:

Debian GNOME Desktop CD (installs by default desktop and gnome-desktop tasks)
Debian KDE Desktop CD (installs by default desktop and kde-desktop tasks)
Debian XFCE Desktop CD (xfce-desktop wasn't added yet but is in my
tasksel' branch)
Debian regular CDs
Debian regular DVDs

Do we need more than just work on debian-cd for this?


I can perfectly understand the d-i and debian-cd teams unwillingness
to dive into the desktop selection discussion.  But on the other hand
one is already taking side in the discussion by keeping gnome as the
default.


I can't understand how people ignored the discussion and the work that
has being done in the desktop underground of the project ! You're
still free to read or ignore my blog posts and follow debian-desktop
mailing list.


Anyway, I believe the users of KDE need to work on improving the
default KDE desktop and make it easily available (for example by
making a KDE version of the Debian CD), and then let the users choose
which desktop they prefer.


They're free to do that. Your point is that it will be harder to
install KDE desktop than GNOME desktop, stop arguing about freedom
here please.


> What I would expect at least: Rename the task from "Desktop" to
> "Desktop (Gnome)" so more experienced users know what's coming up.

Good idea.


I disagree as i explained in my previous message.


> What I would prefer: Tasks for Gnome, KDE and optionally Xfce (for
> the less resource-fortunate).

I believe it is a bad idea to increase the cognitive strain on new
users by adding more options to the task lists.  Most new users do not
have any opinion, and should thus not be forced to choose between
Gnome and KDE.  They rightfully expect us to choose for them, and we
should.  Though I am unhappy with the Gnome choice selected by the
tasksel maintainer, I agree that such choice must be made for the
default installation system, to avoid confusing new users
unnecessarily.


Well said.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Desktop task(sel) in Etch? (Bug #389092)

[Gustavo Franco]

On 9/26/06, Christoph Haas <[EMAIL PROTECTED]> wrote:

Hi, devs...

I have tried the new Etch installer last weekend and I'm impressed. The
installation works perfectly well and the graphical installer - while it
doesn't add much value IMHO to the dialog based installer - looks nice and
will surely be liked by many users. Kudos to the installer team.

I chose to install the "Desktop" task when tasksel asked me. And I ended up
with a Gnome desktop. Since Gnome is not really my favorite desktop
manager I deinstalled Gnome (probably leaving a lot of Gnome cruft on the
system) and installed KDE through apt-get. I wondered why Debian installed
Gnome as a default and does not even offer the choice between - say -
Xfce, Gnome and KDE during the installation. I just didn't expect Gnome.
Sarge at least installed Gnome and KDE which appears to be a waste of
resources but at least gives users the choice later when logging in.


There's desktop, gnome-desktop and kde-desktop tasks, xfce-desktop
task is coming.


Some people on #debian-boot argued that it might be a problem with
preseeding.


No it wasn't.


Joey Hess argued that users may not know what KDE and Gnome are and just
expect a working desktop. Opinions on
http://kitenet.net/~joey/code/tasksel/faq.html


I agree with  Joey.


I know that tasksel tries to keep things simple. But I want to question the
way the desktop installation is handled. So I'd like to hear a few
opinions on how happy other people are with that choice.

Currently the way to install KDE is to provide the
option "tasksel/first=kde-desktop" when booting the Etch installer CD. No
user will guess that.


It'll be possible use 'tasks=kde-desktop' in d-i etch rc1, as i've
discussed previously with joeyh.


What I would expect at least:
Rename the task from "Desktop" to "Desktop (Gnome)" so more experienced
users know what's coming up.


I think it could be done in 'expert' mode, but not in the normal
installation mode. The next step will be that somebody else will
suggest "Desktop (KDE)" listed too (in normal mode).


What I would prefer:
Tasks for Gnome, KDE and optionally Xfce (for the less resource-fortunate).


See above :-)


Ubuntu uses Gnome because they package just one part of Debian to create a
specialized preconfigured desktop. But the Debian user should get the
freedom to choose. And that begins with the installer. Doesn't it?


FYI, Ubuntu now uses tasks too.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Is something wrong to XGL, Compiz, Cgwd be packaged?

[Gustavo Franco]

On 10/28/06, David Nusinow <[EMAIL PROTECTED]> wrote:

On Fri, Oct 27, 2006 at 11:23:27PM +1000, Paul TBBle Hampson wrote:

> Given you're talking about cgwd, you prolly mean beryl (nee
> compiz-quinnstorm) rather than the original compiz which is
> in Debian.
>
> The beryl packaging is being undertaken by the X Strike Force, who're
> also maintianing the compiz packages. So no problem yet, you're better
> of reading the debian-x list archives to discover why it's not shipped
> yet. (I believe they're doing an internal code-freeze for Etch, but I
> dunno if Beryl's in that freeze or not)

You're right about the internal code freeze, although I think beryl should
probably go in to unstable so we can start getting the packages out there.
According to Shawn, who's maintaining it for us, it's not stable enough to
release with Etch though, so seeing it there is pretty well out of the
question. The current set of beryl packages are in good shape, and are
sitting in our svn repo, although shawn hasn't finished packaging emerald
yet, which appears to be the missing piece.


Hi David,

If we're going to ship xorg with aiglx and composite enabled by
default (actually i dunno really), beryl in etch and in default
desktop environment (just listed not enabled by default) would be a
huge win, maybe it's too late now. I would like to point out this
anyway.

My opinion is based as a beryl user in an ibook with 'heavily'
customized Debian and not so customized Ubuntu (just needed to push
beryl from 3rd party repository).


> As for cgwd, I presume that also became something in the beryl
> packages (beryl-manager?) and so is being packaged as part of the
> beryl packaging.

I don't know what cgwd is either, but I don't follow beryl at all at this
point, so I rely on Shawn to teach me those new acronyms. :-)



I guess this is somewhere in emerald (part of beryl) source. Btw,
we're talking about window decorators here.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Is something wrong to XGL, Compiz, Cgwd be packaged?

[Gustavo Franco]

On 10/28/06, David Nusinow <[EMAIL PROTECTED]> wrote:

On Sat, Oct 28, 2006 at 02:30:19PM -0200, Gustavo Franco wrote:
> If we're going to ship xorg with aiglx and composite enabled by
> default (actually i dunno really), beryl in etch and in default
> desktop environment (just listed not enabled by default) would be a
> huge win, maybe it's too late now. I would like to point out this
> anyway.

AIGLX is enabled by default in etch already. Composite is not, although we
could enable it. I don't know what the impact would be on kde users though,
as I understand kwin has a compositing manager, although it might not be
able to take advantage of the acceleration hooks like compiz and beryl do.
Compiz should ship with etch, and turning on composite in xorg.conf is
trivial, so I'm not too worried about having it off by default in this
release. I don't really  understand your definition of "default desktop
environment", so I can't comment on that...


In other words, with the d-i in Etch our users will be able to easily
install KDE and Xfce with the most important related packages and
GNOME will be easier than that. Just the result of the tasksel team,
d-i team, related desktop environment teams, debian-desktop and some
others work. :-)

If there will be no regressions, please add the composite bit in the
xorg.conf by default.


> My opinion is based as a beryl user in an ibook with 'heavily'
> customized Debian and not so customized Ubuntu (just needed to push
> beryl from 3rd party repository).

How heavily customized does your debian install have to be? I just
installed the packages from the XSF svn repo and beryl worked out of the
box, once I enabled composite. This is pretty minimal.


I bet that  the XSF svn repo already contains that endianess fix,
right? as i told you, i'm using ppc (ibook). :-)

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Is something wrong to XGL, Compiz, Cgwd be packaged?

[Gustavo Franco]

On 10/28/06, Frans Pop <[EMAIL PROTECTED]> wrote:

On Saturday 28 October 2006 23:11, Gustavo Franco wrote:
> If there will be no regressions, please add the composite bit in the
> xorg.conf by default.

Is that really a good idea for something that is so young and untested, so
shortly before the release?
Is it wanted for all architectures, for all systems, irrespective of their
speed?


Calm down Frans, what about aiglx then? I wrote 'if there will be no
regressions', that's up to XSF and the users using unstable and even
testing tell us. I still trust our release process (as in
unstable->testing).

Btw, i like the debconf suggestion too.


(...)


regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Is something wrong to XGL, Compiz, Cgwd be packaged?

[Gustavo Franco]

On 10/29/06, David Nusinow <[EMAIL PROTECTED]> wrote:

On Sat, Oct 28, 2006 at 08:43:07PM +0200, Hendrik Sattler wrote:
> Am Samstag 28 Oktober 2006 20:30 schrieb David Nusinow:
> > For etch+1, I'm planning on making it enabled by default and doing away
> > with most of the debconf stuff anyway though.
>
> AFAIK this can be very bad when looking at performance and CPU usage, doesn't
> it? In this case, it should be only be enabled by default, if the underlying
> hardware and it's driver supports hardware acceleration for that. For myself,
> I would not appreciate it on my i815.

Only if a compositing manager is enabled, from what I understand. If
there's no compositing manager turned on, the server doesn't redirect
drawing and you get normal rendering. So you could simply not run a
compositing manager and be fine.



Since it won't hurt metacity users, if kde people tell us that it
won't be enabled by default in kwin i think we could enable composite
by default in xorg, no?

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Bits from the debian-cd team; more CD/DVDs being built regularly

[Gustavo Franco]

On 12/20/06, Steve McIntyre <[EMAIL PROTECTED]> wrote:

(...)

Gnome vs. KDE vs. XFCE
==

The KDE and XFCE variants of CD#1 are now being produced to give more
choice to people for initial installation. By default, CD#1 has always
meant to be enough to install a fully-functioning system, including a
desktop. On sarge, we (just about) managed to make all that fit,
including a choice of KDE or Gnome. However, since the sarge release
the size of the set of packages needed to cover each desktop *and* the
rest of the base system has grown substantially. We're now at the
point where that's just not possible any more. The default CD#1 will
now install a Gnome desktop, and there are replacement CDs that will
install KDE or XFCE instead. Download your own choice. Thanks to Joey
Hess for the work to make these happen.


Let me thanks the debian-desktop, debian-installer and tasksel team
too. Without them, that stuff would be useless.

I would like to clarify that debian-desktop (alioth, svn, mailing
list) is now a unit composed by pkg-gnome, pkg-kde and pkg-xfce and
tasksel members. Joey Hess (d-i and debian-cd) give us all the support
to make stuff happen.

Unfortunately, even the GNOME desktop environment doesn't fit on the
CD#1. You will need broadband connection to fetch some more packages
or use the full DVD#1 for your architecture. The bonus is that KDE
will be there too (full DVD #1). I'm not sure about the latest
statistics on KDE and XFCE alternative desktop CD#1. Joey, could you
give us that information? Btw, i would like to suggest add it
somewhere in the release notes "where's what" in terms of medias. I
can submit patches, after you feed me with info.


Multi-arch
==

Quick warning to avoid possible confusion: this is nothing to do with
the multi-arch binary support that people are working on. The new
multi-arch discs are designed to be convenient for people who
need/want to be able to install Debian onto multiple different types
of machine from a single disc.

To that end, there is now a combined amd64/i386/powerpc netinst that
is basically an amalgamation of the contents of the individual
netinsts for those arches. This is based on the "etch" d-i version as
described above.

There is also a combined amd64/i386/powerpc/source DVD which should
contain most of what people will commonly want to install on each of
those 3 arches, roughly equivalent to the contents of the first 3-4
CDs or so each. Binary-all package overlaps mean that there is space
on this disc for more packages than might be expected. Plus, sources
for all the binaries on the DVD will be included too. This DVD is
therefore probably the ideal choice of disc to sell or give away at
Expos.


Does it contains desktop, kde-desktop, gnome-desktop and xfce-desktop?
I guess not, but who knows. ;)


In case it's not obvious, the 3 arches picked for these discs are
simply the most common end-user systems out there. There is scope for
producing more multi-arch discs, but the possible combinations are
*massive* :-).


I would like to suggest 'split' the 9 archs left in 3, something like:
- arm, mips, mipsel
- hppa, ia64, s390
- sparc, m68k, alpha

Is it possible? Do you think it will fit ? If yes, i can prepare
patches for debian-cd once i figure out all that pile of changes.


To do
=

I'm still expecting to add Live CDs to the list here before we
release. Otherwise, I think we're about set. If there's anything else
you'd like us to do or anything you'd like to ask about, please reply
to this mail - note the Reply-To to the debian-cd list.


Do you plan to use the live-package to build the live cds or some new
code into debian-cd ? If you're going to pick live-package i submitted
a patch to Daniel that would permit use a tasksel based approach to
select packages, and not the fixed list as they actually do. I have no
feedback about that until now. It seems they were going to do a major
change in the code.


(...)


regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Bits from the debian-cd team; more CD/DVDs being built regularly

[Gustavo Franco]

On 12/20/06, Joey Hess <[EMAIL PROTECTED]> wrote:

Gustavo Franco wrote:
> Unfortunately, even the GNOME desktop environment doesn't fit on the
> CD#1.

True, but enough does fit to have a basically useable gnome desktop.


Right, but we need to make it clear for the users and the first line
of them are subscribed to -devel. ;)


> You will need broadband connection to fetch some more packages
> or use the full DVD#1 for your architecture. The bonus is that KDE
> will be there too (full DVD #1). I'm not sure about the latest
> statistics on KDE and XFCE alternative desktop CD#1. Joey, could you
> give us that information?

All of kde doesn't fit on the kde CD, the key packages for the task do.
I expect that all of xfce will fit on its CD, but as its CD is currently
apparently broken and doesn't include much of anything, I'm not yet
sure.


Do you know what's broken in Xfce cd? Is it tasksel related? I hope not.


> >those 3 arches, roughly equivalent to the contents of the first 3-4
> >CDs or so each. Binary-all package overlaps mean that there is space
> >on this disc for more packages than might be expected. Plus, sources
> >for all the binaries on the DVD will be included too. This DVD is
> >therefore probably the ideal choice of disc to sell or give away at
> >Expos.
>
> Does it contains desktop, kde-desktop, gnome-desktop and xfce-desktop?
> I guess not, but who knows. ;)

I don't know yet, it might since IIRC kde starts around CD 3 or 4 of the
regular CD set.


Sounds great, could you give me a link with the log or check it in the
next build please?


> I would like to suggest 'split' the 9 archs left in 3, something like:
> - arm, mips, mipsel

Not enough machines for these arches with CD drives to be useful, I think.


You've a point.


> - hppa, ia64, s390

AFAIK the use cases for s390 CDs are very limited and don't include
booting, so it's not needed.


I'm sure they're, so can't we go for a new multiarch containing hppa
and ia64 only or even i386, hppa and ia64 ?


regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Bits from the debian-cd team; more CD/DVDs being built regularly

[Gustavo Franco]

On 12/20/06, Steve McIntyre <[EMAIL PROTECTED]> wrote:

On Wed, Dec 20, 2006 at 05:04:28PM -0200, Gustavo Franco wrote:
(...)
>>In case it's not obvious, the 3 arches picked for these discs are
>>simply the most common end-user systems out there. There is scope for
>>producing more multi-arch discs, but the possible combinations are
>>*massive* :-).
>
>I would like to suggest 'split' the 9 archs left in 3, something like:
>- arm, mips, mipsel
>- hppa, ia64, s390
>- sparc, m68k, alpha
>
>Is it possible? Do you think it will fit ? If yes, i can prepare
>patches for debian-cd once i figure out all that pile of changes.

There is a problem that I didn't mention here specifically yet: many
of the arches clash in terms of how to make a CD bootable. Many expect
to find their own special metadata in sector 0 of the CD, and most of
them are completely incompatible. :-( We're actually quite lucky that
the common 3 arches are compatible.


I thought about that but since you've not cited the technical limitation
i asked anyway. ;)


The other thing is that the number of CD/DVD-bootable machines in the
other arches is really quite small, probably small enough not to be
worth bothering with the extra effort here for the multi-arch discs.


No problem.


>> (...)
>>I'm still expecting to add Live CDs to the list here before we
>>release. Otherwise, I think we're about set. If there's anything else
>>you'd like us to do or anything you'd like to ask about, please reply
>>to this mail - note the Reply-To to the debian-cd list.
>
>Do you plan to use the live-package to build the live cds or some new
>code into debian-cd ? If you're going to pick live-package i submitted
>a patch to Daniel that would permit use a tasksel based approach to
>select packages, and not the fixed list as they actually do. I have no
>feedback about that until now. It seems they were going to do a major
>change in the code.

Yep, I'm expecting to use live-package. I need to get playing with
that soon...


Sounds good. Please, let us be sure that the during a GNOME, KDE or
XFCE 'live session' the user will have the same or similar as possible
set of packages as installed by d-i (tasksel), as i told you, over the
current codebase there's a patch i submitted that is a step on this
direction. We just need to tweak it a bit and build at least a set
with 3 live desktop cd's (one for each desktop environment and i386),
IMHO. Thoughts?

The possibility of give to the users during a event the DVD#1
containing the 3 desktop environments and 1 live-cd containing the one
he prefers to use (if any) to play with the stuff and test the
hardware compatibility before install is exciting.

regards,
-- stratus


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

for those who care about GNOME^w glib applications.

[Gustavo Franco]

Well, if you're running your favourite glib application and it seems
to be broken with the output like:
(...)
*** glibc detected *** free(): invalid pointer: 0x08ced388 ***

It's probably due to the new glib (GSlice) memory allocator[0], but
it's an application bug. Please first check if it's really glib
related[1].

There's a workaround that is run the application as follows:
$ G_SLICE=always-malloc application

FYI, evolution seems to be affected and it was already fixed upstream.
Sebastian Bacher has a patch to galculator too. If you hit this bug in
other random application, please open a bug against the application
and feel free to cc: me. I'll try keep track of these bugs with
usertags.

[0] = http://www.gnome.org/start/2.14/notes/en/rndevelopers.html
 http://developer.gnome.org/doc/API/2.0/glib/glib-Memory-Slices.html

[1] = ldd `which application` | grep libglib

-- stratus

Re: for those who care about GNOME^w glib applications.

[Gustavo Franco]

On 3/21/06, David Pashley <[EMAIL PROTECTED]> wrote:
> On Mar 21, 2006 at 14:42, Gustavo Franco praised the llamas by saying:
> > Well, if you're running your favourite glib application and it seems
> > to be broken with the output like:
> > (...)
> > *** glibc detected *** free(): invalid pointer: 0x08ced388 ***
> >
> > It's probably due to the new glib (GSlice) memory allocator[0], but
> > it's an application bug. Please first check if it's really glib
> > related[1].
> >
> > There's a workaround that is run the application as follows:
> > $ G_SLICE=always-malloc application
> >
> > FYI, evolution seems to be affected and it was already fixed upstream.
> > Sebastian Bacher has a patch to galculator too. If you hit this bug in
> > other random application, please open a bug against the application
> > and feel free to cc: me. I'll try keep track of these bugs with
> > usertags.
>
> I appear to have a bug filed today against irssi with a similar error.
> Bug number is 358172.
> >
> > [0] = http://www.gnome.org/start/2.14/notes/en/rndevelopers.html
> >  http://developer.gnome.org/doc/API/2.0/glib/glib-Memory-Slices.html
> >
> > [1] = ldd `which application` | grep libglib
> >

Right, i just commented in that bug and usertagged the others that i
found[0]. If you see any other related bug, please tell me. I
commented on the evo' related bug and AFAIK seb128 already has a patch
that fixes the galculator one.

[0] = http://bugs.debian.org/cgi-bin/[EMAIL PROTECTED];tag=gslice

Thanks,
-- stratus

Re: for those who care about GNOME^w glib applications.

[Gustavo Franco]

On 3/22/06, Steve Langasek <[EMAIL PROTECTED]> wrote:
> Hi Gustavo,
>
> On Tue, Mar 21, 2006 at 11:42:49AM -0300, Gustavo Franco wrote:
> > Well, if you're running your favourite glib application and it seems
> > to be broken with the output like:
> > (...)
> > *** glibc detected *** free(): invalid pointer: 0x08ced388 ***
>
> > It's probably due to the new glib (GSlice) memory allocator[0], but
> > it's an application bug. Please first check if it's really glib
> > related[1].
>
> > There's a workaround that is run the application as follows:
> > $ G_SLICE=always-malloc application
>
> Thanks for the information.  Can you explain (or provide a pointer to) why
> these bugs are application bugs rather than bugs in glib, and does glib need
> to be adding << conflicts with those applications in order to avoid broken
> partial upgrades?
>

Hi Steve,

Yes, the new GSlice allocator is triggering those[0] bugs[1]. It happens due to
different (but similar) reasons, e.g: In nettool and galculator[2] the
code should
call g_object_unref() - that decrease the ref count and frees the memory when
the ref count is zero, instead call g_free() directly.

FYI, GMemChunk (old and deprecated by the upstream) was reimplemented to
use GSlice, so no need to change or rebuild code to be affected due to buggy
code. I don't know exactly why asking GSlice to force allocate and free memory
through "standard malloc" solves the problem, maybe someone else can
clarify this point.

Talking about the << conflicts thing. The fact that we don't have a list of
affected applications yet, i'm working on a partial list[3] based on
the bugs already
opened. Hopefully  it's less grave than you think.

[0] = http://bugzilla.gnome.org/show_bug.cgi?id=329344
[1] = http://bugzilla.gnome.org/show_bug.cgi?id=163639
[2] = http://bugs.debian.org/358064
[3] = http://bugs.debian.org/cgi-bin/[EMAIL PROTECTED];tag=gslice

Cheers,
-- stratus

Re: for those who care about GNOME^w glib applications.

[Gustavo Franco]

On 3/22/06, Gabor Gombas <[EMAIL PROTECTED]> wrote:
> On Wed, Mar 22, 2006 at 11:29:01AM -0300, Gustavo Franco wrote:
>
> > FYI, GMemChunk (old and deprecated by the upstream) was reimplemented to
> > use GSlice, so no need to change or rebuild code to be affected due to buggy
> > code. I don't know exactly why asking GSlice to force allocate and free 
> > memory
> > through "standard malloc" solves the problem, maybe someone else can
> > clarify this point.
>
> Because those applications are abusing the API and make assumptions
> about how some pieces of memory were allocated. More specifically, they
> think that they can call g_free() (or equivalent) directly instead of
> the appropriate API function. With the introduction of GSlice that is no
> longer the case.
>

Sure Gabor, i was saying that i don't know what these assumptions are, and
the exact difference between use GSlice and force GSlice use malloc through
'G_SLICE=always-malloc', that is a (slow) workaround.

Thanks,
-- stratus

Re: for those who care about GNOME^w glib applications.

[Gustavo Franco]

On 3/22/06, Steve Langasek <[EMAIL PROTECTED]> wrote:
> On Wed, Mar 22, 2006 at 11:29:01AM -0300, Gustavo Franco wrote:
> > > > There's a workaround that is run the application as follows:
> > > > $ G_SLICE=always-malloc application
>
> > > Thanks for the information.  Can you explain (or provide a pointer to) why
> > > these bugs are application bugs rather than bugs in glib, and does glib 
> > > need
> > > to be adding << conflicts with those applications in order to avoid broken
> > > partial upgrades?
>
> > Yes, the new GSlice allocator is triggering those[0] bugs[1]. It happens 
> > due to
> > different (but similar) reasons, e.g: In nettool and galculator[2] the
> > code should
> > call g_object_unref() - that decrease the ref count and frees the memory 
> > when
> > the ref count is zero, instead call g_free() directly.
>
> Right, I saw the patch posted to the evolution bug, which made perfect sense
> -- objects are being allocated via one API and then being freed by another,
> unrelated API that happened to work before because they were both
> implemented on top of malloc()/free().

Yes, evo was NMUed.

> So, unquestionably an application bug.
>
> > FYI, GMemChunk (old and deprecated by the upstream) was reimplemented to
> > use GSlice, so no need to change or rebuild code to be affected due to buggy
> > code. I don't know exactly why asking GSlice to force allocate and free 
> > memory
> > through "standard malloc" solves the problem, maybe someone else can
> > clarify this point.
>
> It works because g_free() is a wrapper around free(), so if the address
> passed to g_free() doesn't point to memory allocated using malloc(), glibc
> will abort as shown.  Making sure everything is allocated with malloc()
> works around this.

Oh, i see. Thanks!

> > Talking about the << conflicts thing. The fact that we don't have a list of
> > affected applications yet, i'm working on a partial list[3] based on
> > the bugs already
> > opened. Hopefully  it's less grave than you think.
>
> Ok, seems like you have this under control then. :)

Yes, just a few bugs reported and probably two without a patch
(classpath and gnome-games) atm. The others were patched through the
BTS, NMUed or Ubuntu has a fix. I'll take a look into the classpath,
but the bug was already forwarded to the upstream, so there's a
possibility that they've a patch.

Cheers,
-- stratus

Re: glibc_2.3.6-6_i386.changes REJECTED

[Gustavo Franco]

Aurelien, with the feedback provided by aj and your own. After the
small changes, i think you should upload it for experimental and ask
for tests. If you think it's ok for tests "as is", just drop the
package in experimental soon and let us see.

I think the experimental upload (or just a people.d.o/~foo thing would
do) will clarify some points and put us on track, avoiding unnecessary
noise.

Thanks,
-- stratus

Re: utnubu-desktop for the masses

[Gustavo Franco]

On 4/23/06, Joey Hess <[EMAIL PROTECTED]> wrote:
> Gustavo Franco blogged:
> > In Ubuntu there' s ubuntu-meta source package that results in
> > [ubuntu-minimal][1], [ubuntu-standard][2] and [ubuntu-desktop][3]. They're
> > metapackages and the list of packages is built with a tool called germinate
> > based on a seed in the web. I've a [branch of cjwatson' ubuntu seed][4] and
> > asked him to upload [germinate][5] in Debian, he did! To avoid confusion, 
> > i've
> > renamed "our ubuntu-meta" to utnubu-meta and it will be included in [utnubu
> > alioth group ][6]in svn soon.
>
> I've very confused by the approach you are taking here. Debian already
> has its own way to install a desktop, namely tasksel's desktop tasks.
> Any help with maintaining that would be appreciated; but introducing a
> competing thing taken from Ubuntu into Debian doesn't seem at all
> helpful from my perspective, unless I've misunderstood what you're
> doing.
>
> This seems to be a metapackage that depends on 239 packages[3]. Debian
> has already rejected using large metapackages such as that for many
> reasons, including:
>
>  * The way they clog up britney by tying a lot of otherwise unrelated
>packages together. (As a sometime member of the release team, I feel
>like I'm in the shower seeing the shadow of a figure with a knife.)

That's up to me open a RC bug to avoid this problem, and i plan to do
so based on your feedback.

>  * Their general fragility, breaking if any one semi-unimportant package
>in the metapackage is removed from testing for any reason, or is
>unavailable for any one architecture for any reason.

Yes, that's a problem but i would like to see these metapackages only
in sid, see below my purposes...

>  * Their all or nothing nature making it a pain to put them onto CDs,
>if any one semi-unimportant package doesn't fit the whole metapackage
>won't go on.

I'm trying to address part of the divergence here, not promising that
we will have the same solution ready to be released with Etch. You
known, it's a wip thing and i've more things in mind for that. The
fact is that i've talked with Otavio Salvador about a 'utnubu task' or
something like that in the installer. I think it would depend on some
infrastructure to the tasksel knows if he's into the installer itself
and which image (netinst, ...), in a system already installed, ... .
right?

>  * Their lack of a clean way to remove the metapackage (semi-addressed by
>aptitude).

Yes but pointless IMHO, since there's no plan around to remove or
check the usability of others metapackages.

> We tried it, it doesn't work for us[4]. I still have the scars. Tasksel
> avoids all of these problems. If you are interested in maintaining
> tasksel's desktop or other tasks, that could be arranged.

I'm interested in a "utnubu desktop/minimal/standard" now in sid so
you see my metapackage upload. With Etch, i would like to add a
"utnubu desktop task" yes, but as i pointed out above it seems that we
will need more than a simple task. I can help with code if -boot
agree.

> (...)

Closing, i don't think we should discard utnubu-meta in sid now. It
would be a good test and will show us how much work needs to be done
in utnubu front until Etch. Btw, i already received some 'off blog'
feedback from Ubuntu users that would like to move from Ubuntu Dapper
(that will be released in June) to Debian Sid back again. They're all
power users, of course.

Thanks,
-- stratus

Re: Intent to hijack Bacula

[Gustavo Franco]

On 5/9/06, John Goerzen <[EMAIL PROTECTED]> wrote:

Hello,

I intend to take over the Bacula package.  I would first like to say
thanks to Jose Luis Tallon for initially packaging it for Debian and
maintaining it for these years.
(...)


Hi John,

Thanks for this. I'm using backuppc at work and was considering to
move our backups to bacula after upgrading our current hardware setup.
Package updates and bug squashing in general was on the roadmap.

That would be good if you, Jose and probably others joined a 'bacula'
group in alioth to keep this in group maintenance. Hopefully i would
be able to join with real work in the next month. Thoughts?

Closing, do you think it will be possible to ship in Etch a "backup
server" task using bacula ? I think that's all up to add more stuff in
debconf and prepare the task itself. Probably a goal for the first
'group upload', if you agree.

regards,
-- stratus

Re: gcc 4.1 or not

[Gustavo Franco]

On 5/11/06, Martin Michlmayr <[EMAIL PROTECTED]> wrote:

* Andreas Barth <[EMAIL PROTECTED]> [2006-05-11 10:00]:
> > One: What's the easiest way to extract the list of gcc-4.1 related bugs
> > from the BTS?
>
> There is none I know - I asked Martin already yesterday on IRC to
> provide such a way.

I've created the following meta bug: 366820



Hi Martin,

Why you did this metabug thing, and not just usertagged the bugs ? The
results seems to be similar, but i don't think that a metabug can be
managed by email, usertags are.

regards,
-- stratus

Re: gcc 4.1 or not

[Gustavo Franco]

On 5/11/06, Martin Michlmayr <[EMAIL PROTECTED]> wrote:

* Gustavo Franco <[EMAIL PROTECTED]> [2006-05-11 11:20]:
> Why you did this metabug thing, and not just usertagged the bugs ? The
> results seems to be similar, but i don't think that a metabug can be
> managed by email, usertags are.

What can not be managed by email?


The metabug itself, AFAIK just individual bugs can be managed, no?

Thanks,
-- stratus