Re: Questions to candidates
* Andreas Tille <tillea@rki.de> [2004-03-02 15:20]:
> this morning I wrote in private to the DPL candidates but tbm asked
> me to foreward my questions to debian-vote which I'm doing hereby
Yes, I think it's important to share these answers with everyone who
is interested.
> 1. My concern is to propagate Custom Debian Distributions because I
> think we should set a stronger focus to the end user. I see Debian
> as a missing link between upstream developers and end users and
> Custom Debian distributions are a good way to care for end users.
>
> What are your plans according to Custom Debian Distributions?
I partly cover this in my section "External/internal - Debian based
Distributions". I see two developments: sub-projects in Debian with a
special focus, and projects outside of Debian based on our system. I
think that both developments are very good and beneficial, and show
the success of Debian. As DPL, I intend to work with Debian
sub-projects to make sure that they can achieve their goals, and that
they are not isolated in the project. Also, as described in my
platform, I think Debian can profit to a great extend from Debian
based distributions, and I intend to work with them to make sure that
their work gets integrated into Debian. Ideally, they would join
Debian and lead their project as a Debian sub-project. This is why I
mentioned this point as "external/internal" rather than just
"external" -- I think everyone will benefit from getting them closer
to Debian.
As to your statement about Debian being the "missing link between
upstream developers and end users". I fully agree with this. I think
Debian provides a great service to the community, e.g. by forwarding
bug reports to upstream, compiling and testing upstream software on
many architectures, etc. I am also excited to see a growing number of
upstream authors directly getting involved in Debian!
> 2. Recently we had some flamewars about concentration of "power" for
> some people inside Debian. [...] How do you see the role of James
> Troup in the project?
I think James is an excellent contributor to the project. I know him
personally, and I can assure you that he is not on a power trip. He
performs so many tasks in the project and holds key positions simply
because of the amount of work he puts into Debian, and because there
are often no other volunteers with the time or knowledge. Most
people are not aware of this, but after the compromise James stayed up
until 4am-5am or even longer every night, and even took off a day of
work to work on the restoration of our services. I'd like to see such
a devotion to Debian from more people! In any case, I am fully aware
that there are complaints about James ...
> While I think that he did a great job in terms of finding technical
> solutions he absolutely fails in communication with people.
... he does not _absolutely_ fail to communicate with people; there is
a large number of people who communicate with him without any
problems. For example, I'm in contact with him on an almost daily
basis. However, it's true that his communication can be improved, and
that there are some problems. However, I think the problems are much
smaller than they appear to outsiders. Usually, you don't notice when
something works as expected. You only notice when it suddenly breaks.
So if 95% of communication with James works well, we will never hear
about it. But we hear about the 5% which fails.
In any case, what can be done to improve communication? I think one
important step to take, and one I've been working on and which I
emphasize in my platform, is to assist James with his tasks. In many
cases, he is not unwilling to communicate but is simply too busy to
respond to everyone. If he would respond to everyone, he would not
get the important tasks he performs done. I think the situation can
be improved if more people assist James in his tasks. Finding people
for core teams is quite complicated (see my platform), but this is
what I will work on, and have been working on. For example, another
ftpmaster was added to help with NEW processing, and this certainly
helped. Myself, I respond to questions about the NM process. There
is also a second person responding to keyring requests. So, the first
step will be to clearly identify where help is needed (not just in the
teams James is involved, but in general), and to find people who can
provide assistance. This is a delicate task, and requires people's
skills. (When talking about evolution, there is this metaphor that
it's not possible to put the parts of an aeroplane in a box, shake it
and hope an aeroplane comes out; it's the same with people - you
cannot expect to put some random people together and hope they'll be a
good team. You have to select the right people, and the team has to
form evolutionary/naturally.)
> This starts with the fact that he is known to actively maintain a
> quite long killfile
For personal mail - not for role accounts.
> So what are your plans to enhance communication with people on
> important positions in Debian and how do you think that important
> jobs might be split onto different shoulders?
The section "Internal - Core Teams, Delegates, Communication,
Transparency" of my platform is devoted to this topic. I think our
core teams have not grown as much as the rest of Debian, and this
leads to overworked developers, bad communication, etc. We have to
find more people for core teams, and also to provide better
documentation to make the teams and its processes more transparent.
FAQs would also help answer common questions.
> 3. Do you think Debian should continue to support non-free?
No. Debian is about creating a operating system with free software,
and I don't think we should be in the business of distributing
non-free software. I think we should focus on what we do best (create
and integrate free software), and this would also get us closer to
other players in the community, such as the FSF.
Having said this, I don't think the current non-free removal vote is
being done correctly. If we decide to remove non-free, we have to
provide a good upgrade plan for our users. Thus, I think we should
*first* move non-free to something like non-free.org, encourage people
to use new APT sources list while at the same time supporting the old
APT lines (i.e. still keeping it on Debian mirrors) for a while.
I don't think that distributing non-free has benefits for our users in
the long run. By distributing it, we basically sanction its use and
give companies and other developers the message that it's "okay" to
have non DFSG-free licenses. I don't think it is, and we should
present a clear, coherent message. Moving non-free to an outside
project preserves Debian philosophy about free software, and at the
same time ensures that users will be supported.
Also, non-free does not meet the quality of the rest of Debian in many
aspects. As a QA member, I once looked through non-free and found
much more unmaintained packages than in main. This is probably
because a large number of Debian developers don't care about non-free.
non-free is also not built on all architectures we support, and often
bugs cannot be fixed (e.g. acrobat 4, which had to be removed because
we couldn't fix a security bug).
> 4. Does your normal live allow sparsing time for Debian leadership
> which seems to include much additional work (perhaps you will not be
> able to continue on working on your packages) and does your current
> employer accept your intention.
I've addressed this in my platform, but I'd like to elaborate. Last
year, I was a Master's student at the University of Melbourne and
there was a great deal of overlap between my studies and research, and
Debian. I worked full-time on Debian during this time, probably even
more than that. (I did not totally neglect my studies, however, even
if my priority was Debian. I completed my degree with excellent
results.) I partly chose to do a PhD about free software because
this would allow me to devote time to Debian; (and partly because I'm
interested in finding ways to improve the quality of free software; I
think we're doing pretty well, but we can still make it better!).
Therefore, I see myself working on Debian full-time like before, and
have done so for the first two months of my PhD. So it's working
pretty well.
> A. Meta-question: Do you know that your jpb as a Debian leader has
> the consequence to travel in several countries all over the world
> which might lead to the situation that some countries handle you
> like a criminal by taking your finger prints?
While I don't like these practices, I don't consider them off-putting
enough not to visit a country if there's a good reason to go there.
However, this has to be decided on a case by case basis. As far as I
know, EU citizens also don't have to get their finger prints recorded.
--
Martin Michlmayr
tbm@cyrius.com
Reply to: