path: root/xml/htdocs/proj/en/elections/council/2009/manifesto/calchan.txt

:Author: Denis Dupeyron (Calchan)
:Contact: calchan@gentoo.org
:Date: 02 Jun 2009

(Note that this document may be updated over time, so pay attention to the
above date)

The following is presented as a fake interview. Because we all know it's a lot
easier to answer trivial questions you wrote yourself.


=============
Who are you?
=============

I've been using Gentoo almost exclusively since 2002, more exactly just before
it turned 1.0. I came as a developer 3 years ago for the sci-electronics herd,
and joined dev-embedded soon after, mostly because Daniel (Black) asked me
nicely. Since then I've touched and accepted the maintenance of various
packages. I don't maintain that many packages since I like to do things
properly and I have other tasks in Gentoo.


=========================
No, who are you *really*?
=========================

Oh, that! My real name is Denis Dupeyron, but Calchan is how my wife has always
called me. I usually sign "Denis" but feel free to call me Calchan as even my
mother-in-law calls me this way. I'm French, my wife is Japanese, my son was
born in Belgium and my daughter was born in France. I've lived in the US
previously and I'm currently back there, but also lived in France obviously, in
Belgium and in the Netherlands. This whole international thing defines me a
lot. I have been lucky to meet quite a few of you physically on both sides of
the Atlantic, and it's been a great experience every single time.


===================
And professionally?
===================

I have worked in a lot of places over the past 20 years. Mostly as an engineer
or project leader in the IC design industry, but there were a few other very
different and fun jobs. More recently I've been in the hearing implant business
for 7 years. I have a PhD EE and a couple of master's degrees.

=======================================
You said you had other tasks in Gentoo?
=======================================

First I'm a recruiter. I haven't done any recruiting since I went away about 8
months ago due to living in airports. I have resumed working on Gentoo two
months ago and I'm gradually restarting all my activities one by one, so
recruiting will come back at some point. It will basically happen as soon as
I'm done mentoring Rafael Martins who will join us as a developer for
sci-electronics. He will be the first one since plasmaroo abandoned me more
than two years ago. I consider recruiting more than just interviewing recruits.
I believe almost anybody can become a Gentoo developer and that it's the
mentor's and recruiter's task to get the candidate up to speed. And since we
can't ask mentors to always have all of the necessary skills for that, I
consider it is the recruiter's responsibility to fill in the gaps. I favor
social skills over technical skills as the formers are hard to get by and the
latters can be acquired over time. I also insist a lot on how to behave
properly as a developer during recruiting reviews, and I believe that it's what
it will take in the long run to get that cultural change we needed so badly not
so long ago. I like to think that some of it has already happened.

I'm also a devrel member. Most of what we do in there occurs in the background
(on purpose) so it isn't very visible. My goal is simply to make sure that our
devs are developing happily and that they can focus on that. Things I do are
for example resolving disputes, defusing situations before they can even be
called disputes, making sure elections of a Lead of a major project are being
held, trying to tame some of our alpha males, etc... It doesn't always work,
but it's always worth a try.


=======================================
So what are your plans for the council?
=======================================

None of the above. The council is about technical and strategic issues so I
will not confuse it with devrel. You can however expect the same kind of honest
and pragmatic management and problem-solving attitude from me. Here's what I'm
hoping the new council will work on:

The daily grind
---------------

Because we all know there's a lot to do. And in case we get bored, well, see
below.

git migration
-------------

Figure out with Robin what he needs to make this happen in a reasonable amount
of time. If that's a monthly subscription to the local strip joint I'm OK with
it. He says he needs more time (as in more than 24 hours everyday), but I'm
thinking that he could need help in planning the thing, dividing tasks,
getting more people working on it, etc... Robin is more than smart enough to do
this himself but he seems to have too much to do at the moment. The council
should follow-up on this with him and coordinate with other groups (e.g. devrel
for additional manpower, or PR to publicize his needs, etc...) in order to help
him make progress on this.

GLEP process
------------

Better follow-up on the GLEP process and have people actually respect the
spirit of it as written in GLEP 1. Things like making sure the GLEP champion is
actively looking to build consensus or documenting the thinking process that
occurs on lists come to mind. Build small teams around each GLEP and give them
clear goals and deadlines that they agree with, rearrange the teams when the
goals and/or deadlines are not respected (because life happens). Maybe review
some of the aspects of GLEP 1. The intent is to have the GLEP process work
better, clear-up the GLEP pipeline and have people work more together.

Proactively secure our critical activities
------------------------------------------

List all the critical activities in Gentoo and identify what their needs are.
For example: security, portage, infra (these are only examples). Make sure
their needs are covered. Involve e.g. devrel (for internal or external
recruitment) or trustees (for finances) if they're not. Setup a way to
continuously monitor these needs in the future, as well as a way to update the
list of critical activities as needed.

New organization
----------------

Decide if we need a new organization and work on it if needed. You have
probably seen drafts of my organizational GLEP. This has since then generated
ideas and I'm currently merging my ideas with mainly Tiziano's but also others'
into something that I find way better than my original proposal. More
discussion will certainly ensue, and more ideas will come up, but the
implementation may take enough time that a whole term may not be too much to
prepare and actually do it. Some of the changes are simple enough that they can
certainly be experimented with right now. Others can be implemented
progressively along the year. The target is to get an organization which will
be at the same time more flexible, simpler, and more importantly more efficient
than what we currently have. As soon as it will be presentable I'll send it to
the -dev@ mailing-list. Some of the proposed changes are actually already in
this manifesto.

Merge devrel and userrel
------------------------

These two groups have been wanting to merge for a long time but it just does
not happen. Maybe the help of the council and its more strategic view is
required here to make things easier and more official.  In case the switch to a
new organization does not happen it may also be a good opportunity to reshuffle
responsibilities. Some of the global issues which are more or less implicitly
incumbent on the council as per GLEP 39 could be explicitly transferred to this
new entity. This whole thing could be done in two steps, first the merge
without much change (only a consolidation of the manpower) and then the
rethinking of how the council will work with this new entity.

Experts
-------

Formation of a group of "experts" (for lack of a better name) to discuss future
EAPIs and other technical issues. The members need not necessarily be all
Gentoo developers, and may change over time on an as-needed basis. The council
will ask them to discuss specific topics and they will document their
discussions and explain their opinions. They do not need to all agree on
everything, the council will then make decisions based on their work. The idea
is to get rid of the bickering during council meetings while still allowing
discussions. This will also allow the council to have more time to work on
other topics like those listed above.

Share with other distros
------------------------

Try and see whether we can share some of our workload with akin distros. I had
a very nice discussion with Bryan Østergaard at FOSDEM 2009, and I'm sure Fabio
Erculani wouldn't mind discussing that either. The idea is the more we share,
the less we do but the farther we go. Not counting the impact on pacifying our
relationships. And before you say those three do not have enough in common to
share workload I suggest you think outside the box. This won't certainly be
easy and won't happen overnight, and it may not lead anywhere, but it's worth
exploring.


===============
*Post scriptum*
===============

I understand that all of the above is very ambitious and some of it will
probably not be achieved in one term, if ever. I listed them anyway because
that's what I believe we should work on. If you want to adopt these ideas or
even claim them as yours, that's OK with me as long as you make them happen.