Re: (seemingly) declinging bug report numbers

[Christian PERRIER]

Quoting Thomas Preud'homme (robo...@debian.org):

> > This is sometimes hidden by the incredible work and investment of
> > several people in the project (yes, that's probably mean whoever is
> > reading this).
> 
> I don't think it's mean to recognize the amazing work some people do. At 
> least 
> I don't feel offended.


My sentence was misspelled (and probably bad English anyway).

I wanted to write : "yes, that probably means whoever is reading
this". 




signature.asc
Description: Digital signature

Re: Popularity of bzr-builddeb and dh-make

[Steve Langasek]

On Fri, Oct 19, 2012 at 09:13:04AM -0700, Russ Allbery wrote:
> Steve Langasek  writes:
> > On Fri, Oct 19, 2012 at 01:02:59PM +0800, Chow Loong Jin wrote:

> >> Actually most users of pristine-tar in git don't have the *first* of
> >> these branches.  They usually have an upstream branch which is
> >> synthesized solely from importing tarballs using git import-orig.  In
> >> other words, the typical practice is to avoid sharing git history with
> >> the upstream VCS, which in turn works out very well for git-dch,
> >> because you don't get unnecessary upstream changes documented in
> >> debian/changelog.

> > This seems utterly broken to me and optimized for the wrong priority.

> It makes perfect sense if you're packaging software where upstream doesn't
> use a DVCS, which I'll point out is still the most common case.  :)

Upstream doesn't have to be using a DVCS for your DVCS to be able to import
their history (git and bzr both have bridges to a number of other VCS
sources).  Evidently my experience as a maintainer is not as typical as I
thought it was.  I find that having the entire history (upstream+packaging)
at my fingertips in a single place is *the* biggest multiplier for my
productivity, and when I don't have it (generally because the tools are
getting in my way), I feel its absence profoundly.

-- 
Steve Langasek   Give me a lever long enough and a Free OS
Debian Developer   to set it on, and I can move the world.
Ubuntu Developerhttp://www.debian.org/
slanga...@ubuntu.com vor...@debian.org


signature.asc
Description: Digital signature

Re: (seemingly) declinging bug report numbers

[Игорь Пашев]

I will fill your bugreports for $1.99 per bug :-)

2012/10/19 Peter Samuelson :
>
> [Kelly Clowers]
>> But I basically never report bugs. I have used Sid for years, and in
>> fact I often don't notice bugs in my personal workflow (maybe if I
>> can think of myself as a user? I notice end-user-impacting bugs in
>> other areas). If someone comes over and sees me working the might
>> say, "wow that is an annoying bug" and I say "what bug? Oh that.  I
>> didn't notice, I just worked around it." Even with bugs I do notice,
>> I usually just ignore and work around until it is fixed.
>
> Don't feel bad about that.  Reporting a bug is a _burden_, especially
> if you care enough to produce a high-quality report.  Even if the
> actual reporting part is pretty easy, you have to gather a lot of
> information: is it reproduceable and if so, how?  How sure am I that it
> isn't user error or local configuration?  How sure am I that it hasn't
> already been fixed by a newer upload?  Is there anything strange in my
> environment that I am forgetting to mention, that would make the bug
> hard for anyone else to reproduce?  And of course that's not even
> counting the time investment of working with the maintainer after the
> initial report.
>
> I don't fault anyone for deciding that the return on investment for
> producing a high-quality bug report is higher than for just working
> around it.  I often do the same.  We of course appreciate when users
> are willing to contribute a good bug report, but we don't require or
> expect everybody to do it.  Mostly we produce Debian so you can _use_
> it, not so you can spend your time helping us make it better.
>
> Peter
>
>
> --
> To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
> Archive: http://lists.debian.org/20121019164112.gf4...@p12n.org
>


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/call-q8wn+y3kadso5dykw7yw-rjw6entrusk1+aaax0rz5j...@mail.gmail.com

Bug#690973: ITP: pgmodeler -- PostgreSQL Database Modeler

[Markus Frosch]

Package: wnpp
Severity: wishlist
Owner: Markus Frosch 

* Package name: pgmodeler
  Version : 0.3.4
  Upstream Author : Raphael Araújo e Silva 
* URL : http://pgmodeler.com.br
* License : GPLv3
  Programming Lang: C++
  Description : PostgreSQL Database Modeler

Easily create and edit database models with simple and intuitive
interface. The pgModeler have forms that indicates the fields
that must be filled to provide the correct generation of SQL code.


--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20121019182329.2BB4723A91@frosch-nb.local

Re: EFI support is not ready

[Jerome BENOIT]

Hello:

grub support for EFI is clearly a WIP:
you may specify the grub version and test with a recent one,

On 19/10/12 19:42, HacKurx wrote:

Hi,

I tested with 2 pc efi and the first gives me a kernel panic during the launch 
of the second installation when it crashes to the installation of grub.

Here the hardware tested
http://kmuto.jp/debian/hcl/Giga-byte/GA-990XA-UD3
http://kmuto.jp/debian/hcl/ASUS/1015BX+C-60

Thanks


Jerome


Best regards,

HacKurx
www.hackurx.info 




--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/508195ab.2070...@rezozer.net

Re: (seemingly) declinging bug report numbers

[Peter Samuelson]

[Kelly Clowers]
> But I basically never report bugs. I have used Sid for years, and in
> fact I often don't notice bugs in my personal workflow (maybe if I
> can think of myself as a user? I notice end-user-impacting bugs in
> other areas). If someone comes over and sees me working the might
> say, "wow that is an annoying bug" and I say "what bug? Oh that.  I
> didn't notice, I just worked around it." Even with bugs I do notice,
> I usually just ignore and work around until it is fixed.

Don't feel bad about that.  Reporting a bug is a _burden_, especially
if you care enough to produce a high-quality report.  Even if the
actual reporting part is pretty easy, you have to gather a lot of
information: is it reproduceable and if so, how?  How sure am I that it
isn't user error or local configuration?  How sure am I that it hasn't
already been fixed by a newer upload?  Is there anything strange in my
environment that I am forgetting to mention, that would make the bug
hard for anyone else to reproduce?  And of course that's not even
counting the time investment of working with the maintainer after the
initial report.

I don't fault anyone for deciding that the return on investment for
producing a high-quality bug report is higher than for just working
around it.  I often do the same.  We of course appreciate when users
are willing to contribute a good bug report, but we don't require or
expect everybody to do it.  Mostly we produce Debian so you can _use_
it, not so you can spend your time helping us make it better.

Peter


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20121019164112.gf4...@p12n.org

Re: non-developer packages depending on gettext?

[Ivan Shmakov]

> Neil Williams  writes:

[…]

 > Check if the package contains a shell script which supports
 > translated output strings — such packages should Depend: gettext-base
 > rather than drop the dependency entirely.

 > I've had a quick look at gnas and it does seem that this is a case
 > where gettext-base is required, but not gettext.

ACK, thanks for the information.

To note is that Source: gnunet has contrib/report.sh, which
calls gettext(1), but it doesn't seem to be propagated to any of
the binaries currently depending on gettext.

 > gettext should only be necessary for packages or tasks which
 > *manipulate* PO files directly, rather than use the processed .mo
 > files to generate translated output.  So, Build-Depends: yes,
 > Depends: probably a bug.  gettext-base is only really needed when the
 > package provides shell scripts with translated output because those
 > evaluate the gettext process at runtime but that only requires
 > gettext-base, not gettext.

 > Could be worth filing a wishlist bug against lintian because it
 > should be quite easy to spot.

?  I see no easy way to discern between these three cases (the
dependency is valid, depend on gettext-base instead, drop the
dependency altogether.)

-- 
FSF associate member #7257


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/86d30e8mco@gray.siamics.net

Re: Popularity of bzr-builddeb and dh-make

[Russ Allbery]

Steve Langasek  writes:
> On Fri, Oct 19, 2012 at 01:02:59PM +0800, Chow Loong Jin wrote:

>> Actually most users of pristine-tar in git don't have the *first* of
>> these branches.  They usually have an upstream branch which is
>> synthesized solely from importing tarballs using git import-orig.  In
>> other words, the typical practice is to avoid sharing git history with
>> the upstream VCS, which in turn works out very well for git-dch,
>> because you don't get unnecessary upstream changes documented in
>> debian/changelog.

> This seems utterly broken to me and optimized for the wrong priority.

It makes perfect sense if you're packaging software where upstream doesn't
use a DVCS, which I'll point out is still the most common case.  :)

> I cannot imagine why anyone would endure git's user interface and then
> not even use the DVCS functionality for collaboration with upstream.

I've used both Git and bzr for real-world projects and found Git's user
interface significantly superior to bzr's, which is why I've since
converted all of my bzr projects to Git.  YMMV.

-- 
Russ Allbery (r...@debian.org)   


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/878vb2tp7z@windlord.stanford.edu

Re: Candidates for removal from testing

[Niels Thykier]

On 2012-10-19 17:43, Wouter Verhelst wrote:
> On Thu, Oct 18, 2012 at 10:32:39AM +0200, Niels Thykier wrote:
>> We are considering removing the following packages from testing as
>> they have unfixed RC bugs filed against them. The packages can be
>> found in the attached dd-list.  The bugs that put them on this list
>> can be found in the removals file (also attached) just above the
>> package name.
> [...]
> 
> Shouldn't this have gone to debian-devel-announce instead?
> 

We have been using d-d for these mails every since we introduced
them[0][1], so the destination was intentionally d-d.
  That said, I do not (personally) have any strong feelings for or
against using d-d-a given the current rate at which we produce these
lists.  However, it is my understanding that we intended to do them
vastly more frequently than we are now[2].

If you are concerned about people not being subscribed to d-d getting
their packages removed, then I have (despite not adding it to the
original mail) Bcc'ed $pack...@packages.debian.org.  So the maintainers
should have gotten the heads up[3].

On the other hand, if you were missing something like "Bits from the RT"
then you are probably right.  I think we have been pretty much drowning
in unblocks and haven't had time to do a bits mail.

~Niels

[0] https://lists.debian.org/debian-devel-announce/2011/08/msg0.html

[1] https://lists.debian.org/debian-devel-announce/2012/01/msg9.html

and https://lists.debian.org/debian-devel/2012/01/msg00653.html

[2] The original mail[0] even go so far to suggest "daily".  I am not
certain that will happen with the current level of manual
work/filtering.  However, weekly or bi-weekly is not out of the question.

[3] Though, 4 team lists put my mail in the "needs moderator approval"
(for various reasons), so I hope their list moderators are not too
inactive.  :)


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/50817bfd.4050...@thykier.net

Re: (seemingly) declinging bug report numbers

[Kelly Clowers]

On Fri, Oct 19, 2012 at 12:40 AM, Christian PERRIER  wrote:
>
> For sure, this kind of "decline" is not that visible. We still have
> new contributors, we still manage to do releases, we still have an
> ever growing number of packages. But, we have less bug reports. We have
> partly abandoned packages, including in the "core" of the
> distribution. We have an installer that has just been "rescued" by
> nearly a "one-man" effort. And I probably forget many other examples.
>
> This is sometimes hidden by the incredible work and investment of
> several people in the project (yes, that's probably mean whoever is
> reading this).
>
> But, still, yes, I feel we are in danger in some way. That may sound alarming
> (death of Debian predicted, film at 11), but, really, getting new
> blood is important for usif we don't want to shrink into a club of
> old chaps who are doing Debian "just for their needs" but can't manage
> to do it anymore because there is too much to do..:-).
>

A thread like this really makes me feel bad personally. I am no real
programmer, but I am a power user for sure, and on my way to being a
sysadmin. But I basically never report bugs. I have used Sid for
years, and in fact I often don't notice bugs in my personal workflow
(maybe if I can think of myself as a user? I notice end-user-impacting
bugs in other areas). If someone comes over and sees me working the
might say, "wow that is an annoying bug" and I say "what bug? Oh that.
I didn't notice, I just worked around it." Even with bugs I do notice,
I usually just ignore and work around until it is fixed. I guess it is
a combo of never having gotten comfortable with the bug report
workflow and never feeling like I could go deep enough into debugging
to be very useful (I have gotten useful info out of gdb maybe twice
over the years). Now I am just in a habit of not doing it.


Kelly Clowers


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CAFoWM=-m_bpbsc4krvfdwzt3+0o3_fwx7wyeppn_9yb-_ok...@mail.gmail.com

Re: Candidates for removal from testing

[Wouter Verhelst]

On Thu, Oct 18, 2012 at 10:32:39AM +0200, Niels Thykier wrote:
> We are considering removing the following packages from testing as
> they have unfixed RC bugs filed against them. The packages can be
> found in the attached dd-list.  The bugs that put them on this list
> can be found in the removals file (also attached) just above the
> package name.
[...]

Shouldn't this have gone to debian-devel-announce instead?

-- 
Copyshops should do vouchers. So that next time some bureaucracy requires you
to mail a form in triplicate, you can mail it just once, add a voucher, and
save on postage.


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20121019154307.go2...@grep.be

Re: (seemingly) declinging bug report numbers

[Wouter Verhelst]

On Fri, Oct 19, 2012 at 03:04:45AM +0200, Christoph Anton Mitterer wrote:
> On Wed, 2012-10-17 at 17:43 +0200, Wouter Verhelst wrote:
> > Ah, well, I think you misunderstood me here. What I meant is that ubuntu
> > is gaining ground on things like Windows and MacOS. I didn't mean to
> > refer to non-free software packaged with ubuntu, nor to non-free
> > producers who support ubuntu.
> Ah I see.
> 
> Well but ultimately that could mean harm to Debian, too.

No, I disagree with that.

I see many people in my friends and family who try out using some
distribution for one reason or another. Usually, they start with Ubuntu.
Often they stay with Ubuntu for a very long time, but just as often they
do not, and move to Debian or another distribution instead. And this is
not because I'm actively advocating Debian (or advocating against
Ubuntu); in fact, when people ask me for advice about which distribution
to run, I usually reply with "I'm not the right guy to ask, I'm biased".

When people start using Linux for the first time, they often go for what
seems to be the easiest solution. This was true for myself, too: when I
first installed Linux, I used RedHat, since in the sixpack that I bought
at the time, containing RedHat, Debian, and Slackware, only RedHat could
be installed from CD-ROM immediately (at the time, Debian and Slackware
both required that you boot from floppy).

If people start using a Linux distribution at some point or another,
that means they're another Linux user, and that's a good thing. If
they're using Ubuntu, then at some level they're also a Debian user, and
that's also a good thing. If Ubuntu is the right distribution for them,
they will probably remain an Ubuntu user for a long time. If it isn't,
they will migrate to something else; maybe Debian, maybe not.

What's important for Debian is that we continue to attract enough
developers to sustain our distribution's technical level. I think we're
doing that. I do think there are things we can do to improve the inflow
of developers into Debian (as I also said during this year's DPL
campaign), but "competing against Ubuntu" isn't one of them.

(anyway, this is getting more and more off-topic for debian-devel, so
EOT for me here)

-- 
Copyshops should do vouchers. So that next time some bureaucracy requires you
to mail a form in triplicate, you can mail it just once, add a voucher, and
save on postage.


-- 
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20121019153226.gn2...@grep.be

Re: (seemingly) declinging bug report numbers

[Thomas Preud'homme]

Le vendredi 19 octobre 2012 09:40:12, Christian PERRIER a écrit :

Greetings Christian,

First, let me take advantage of this mail to thank you for your tireless work 
on localization.

> 
> Nowadays, would someone bet a coin that the same is happening? I would
> not. In my daily job, I see students coming for post-graduate or
> thesis work in scientific research departments (in optics, fluid
> mechanics, material science, etc.). I talk to them (this
> is my job to  manage the needs of scientific departments wrt IT, in
> our institution), I talk about their needs for their research work, I
> talk to their staff.
> 
> Definitely, the free software and Linux "culture" exists among people
> in our universities (or in our typically French "Grandes Écoles"). But
> Debian? Really? Not that much. I even heard (when people learn that I
> am involved in the project) questions like "oh, Debian? Does it still
> exist?". And, yes, here, Ubuntu comes up more often (I would bet that
> more than half of students laptops installed with a Linux brand
> nowadays are using it).

Couldn't it be that the proportion of Debian users among Linux users has 
decreased but not the raw number? It's probably safe to say that Ubuntu 
attracted many new Linux users which could have changed that proportion. About 
the contributors, I believe most contributors in Ubuntu have an interest in 
working in Debian as well to reduce the diff size.

> 
> For sure, this kind of "decline" is not that visible. We still have
> new contributors, we still manage to do releases, we still have an
> ever growing number of packages. But, we have less bug reports. We have
> partly abandoned packages, including in the "core" of the
> distribution. We have an installer that has just been "rescued" by
> nearly a "one-man" effort. And I probably forget many other examples.

Maybe the problem stems from the fact that contributions are focused on 
different components of the distribution than its core. Maybe it's even worse 
and potential contributors are more interested in doing android apps than 
working on Debian packages.

> 
> This is sometimes hidden by the incredible work and investment of
> several people in the project (yes, that's probably mean whoever is
> reading this).

I don't think it's mean to recognize the amazing work some people do. At least 
I don't feel offended.


Best regards,

Thomas Preud'home


signature.asc
Description: This is a digitally signed message part.

Re: Popularity of bzr-builddeb and dh-make

[Julien Cristau]

On Thu, Oct 18, 2012 at 22:29:53 -0700, Steve Langasek wrote:

> On Thu, Oct 18, 2012 at 09:54:27PM -0700, Russ Allbery wrote:
> > > AIUI, most users of pristine-tar in git don't have the second of these
> > > branches, which means the pristine-tar binary delta is done against the
> > > upstream branch - so each pristine-tar blob contains all the information
> > > about autogenerated files in the tarball, in a format that doesn't in
> > > turn compress well in the git repository.
> 
> > Oh.  No, I'm fairly certain that you're wrong, since any user of
> > git-buildpackge will have the second.  Rather, what's normally missing
> > from most Git-based packaging is the *first* branch, since the
> > git-buildpackage workflow was designed originally around importing
> > upstream tarballs to create the second branch.
> 
> Ok.  Well, bear in mind that this is all second-hand.  I was complaining on
> IRC about having to work with the designated Vcs-Git branch on a package (I
> don't remember which) that didn't use pristine-tar, and multiple developers
> rallied to the defense of this practice, claiming that pristine-tar caused
> git repositories to rapidly balloon in size.  Perhaps one of them can speak
> for themselves about what they think the issues are. :)
> 
That might well have been me.  I don't know what "most Git-based
packaging" does, as I don't have much git-based packaging experience
outside the XSF...

I've tried importing the xorg-server tarballs using pristine-tar against
the upstream git release tags once, and it made the repo unbearably
large (each delta was huge, and git couldn't compress it meaning the
repo size increased linearly with the number of imported tarballs).

Note that I don't use git-buildpackage (I tried it a couple of times a
while ago, it didn't seem to add anything I cared about to
dpkg-buildpackage, other than complexity), and I don't store or want to
store the autotools noise in the packaging or upstream branches, so in
order to import the tarballs properly I'd have to add a new branch (or a
bunch of tags) for their contents and remember to update and push it
along with the other branches at each upstream release.  And if I ever
forgot to push one of those that'd make the pristine-tar branch useless
for anyone else, since they wouldn't have the needed data.  Easier to
just use uscan to get the tarball from upstream, so bothering with
pristine-tar didn't seem worth it at the time.

Cheers,
Julien


signature.asc
Description: Digital signature

Re: (seemingly) declinging bug report numbers

[Christian PERRIER]

Quoting Russ Allbery (r...@debian.org):

> distribution that make it more popular.  But, unlike commercial
> distributions, we don't *have* to be popular to succeed.  We have a much
> broader range of successful outcomes than a business that has to make
> money.


I will take this last sentence from Russ' mail to give out my own
feeling about these issues.

I work on Debian for about the same reasons Russ gave in his mail and
I agree with the way he says things. I indeed agree with most of what
was said in this thread.

Still, I see a "threat" against the project, somewhere.

The majority of us works on Debian for these reasons, fine.  We all
mostly don't really care about Debian being "popular" or what. Fine.

However, not being "popular" also means a declining number of
contributors. If I look back to my Debian years (they start nearly to
the days where Ian created the project, though I started contributing
to it around 2000), during the late 90's and early 2000 years, Debian
*was* a kind of a "hype" in the geek community. The reference, the
clever thing people are *attracted* to contribute to.

If I look around to my fellow French developers friends, several (not
to say "many", not to say "most") of them originate from a generation
that was in their university years in these late 90's, early
2000's. Even though I was not there, I can imagine that, among the
geeky students at that time, Debian was an attractive project,
something "people" talk about, something you want to be part of.

And, imho, that perfectly explains why we were "powerful" enough to
grow up as we did. The "new blood" was in some way constantly coming
in. We even had problems dealing with that new blood if you remember
(I think Cyril Brulebois, who is currently  handling so many things in
the project, including our beloved installer, remembers how much time
it took for him to become a DD, around 2007 IIRC).

Nowadays, would someone bet a coin that the same is happening? I would
not. In my daily job, I see students coming for post-graduate or
thesis work in scientific research departments (in optics, fluid
mechanics, material science, etc.). I talk to them (this
is my job to  manage the needs of scientific departments wrt IT, in
our institution), I talk about their needs for their research work, I
talk to their staff.

Definitely, the free software and Linux "culture" exists among people
in our universities (or in our typically French "Grandes Écoles"). But
Debian? Really? Not that much. I even heard (when people learn that I
am involved in the project) questions like "oh, Debian? Does it still
exist?". And, yes, here, Ubuntu comes up more often (I would bet that
more than half of students laptops installed with a Linux brand
nowadays are using it).

For sure, this kind of "decline" is not that visible. We still have
new contributors, we still manage to do releases, we still have an
ever growing number of packages. But, we have less bug reports. We have
partly abandoned packages, including in the "core" of the
distribution. We have an installer that has just been "rescued" by
nearly a "one-man" effort. And I probably forget many other examples.

This is sometimes hidden by the incredible work and investment of
several people in the project (yes, that's probably mean whoever is
reading this).

But, still, yes, I feel we are in danger in some way. That may sound alarming
(death of Debian predicted, film at 11), but, really, getting new
blood is important for usif we don't want to shrink into a club of
old chaps who are doing Debian "just for their needs" but can't manage
to do it anymore because there is too much to do..:-).

We *will* be old chaps anyway. Several of us already are (and are even
happy with that). But we should worry about a possible start of
decline and we should avoid denying it. That was indeed the exact
purpose of my original blog post. The first reaction we can have is
probably to start facing that reality.

Hopefully such thread in one of our mailing list is kind of a way to
do it.




signature.asc
Description: Digital signature