Re: Debian -- the best
On Fri, Dec 19, 2008 at 11:43 AM, Michael Hanke <michael.hanke@gmail.com> wrote:
>
> Dear Debian developers,
>
>
> for some reason I am subscribed to debian-devel and even try to read
> most of the posts. I guess I do that to stay in touch with the most recent
> developments, but it is also I fairly good indicator of the projects
> climate ... which seems to be getting colder ...
>
> But I cannot understand _why_ this is happening. Posts in the thread
> started by the resignation of our secretary (but, in fact, also
> countless times before) have speculated that it might be due to an
> unfortunate (self-)selection of people generating most traffic on the
> major mailing lists, preferably about supposed-to-be-negative aspects of
> this project. What can be done?
>
> Let me express my appreciation and gratitude for Debian.
>
> I believe that the Debian project (not just the OS it produces) is an
> outstanding and unique example of what can be jointly achieved by people
> from a huge range of cultural backgrounds, access to monetary ressources
> and types (or sources) of motivation. Given the reality on this planet,
> the sheer existance of the project after so many years is so unlikely
> that Hollywood should think about a movie. I am really proud to be able
> to contribute my bits to Debian.
>
> Debian is about freedom and Debian is setting the standards. The project
> is percieved as the mothership of free-software. Software that is not in
> Debian is hardly distributed somewhere else. If you want to have
> something in Debian, you have to do it _right_. Not just on the
> software-enginering side, but also wrt the legal situation. A lot of
> people only start thinking about what a license really is about when
> forced to obey it by some Debian packager. IMHO this is very important
> as it propagates the idea in an effective and productive way -- much
> more than a disfunctional wireless device due to a missing firmware.
>
> Sorry, for the long intro -- here is my 'success story'.
>
> I work in the neurosciences. Fortunately, over the last few years the
> idea of open-source (sadly not necessarily of free _and_ open source)
> got established in this science community. More and more great pieces of
> software become available.
>
> But even better: more and more software also becomes part of Debian (see
> http://debian-med.alioth.debian.org/tasks/imaging.html for the ones
> relevant for my research).
>
> Debian can be considered the optimal environment for brain imaging
> research (compared to all other possible operating systems). It allows
> neuroscientists to setup a functional analysis environment within a few
> hours ... and keep it that way for years with minimal effort.
>
> This is only possible due to the _joint_ effort of the whole Debian
> project. I can only fail to list and thank all the subprojects and
> developers who contribute to that success, therefore I will only pick a
> few examples:
>
> You cannot make people try the universal OS if it doesn't run on their
> hardware. Thanks to the amazing Debian installer it runs on almost
> anything. In a number of neuroscience labs I know it is often the case
> that people are forced to work in some predefined environment, set up to
> fulfil the needs of the sysadmin, not the researcher. Quickly installing
> Debian in a VM is actually helping a lot of people to be more
> productive. But for sure it serves as a proper desktop, the powerful
> workstation and the computing cluster equally well. Thanks for that.
>
> I am also part of the upstream developer team of a machine-learning
> framework geared towards neurocientific datasets (http://www.pymvpa.org and
> of course http://packages.debian.org/unstable/python/python-mvpa).
> This framework is intended to glue together lots of generic packages and
> make them available for neuroscience research through a uniform
> interface. Again, Debian is the optimal environment to do that, as it
> provides almost any software package that is useful for our purpose.
>
> I went through the process of providing binary packages for this tool
> and its major dependencies on other operating systems. For some it is
> almost impossible (win), for some painful (mac). The OpenSuse build
> service is a great tool to compile stuff for a wide range of RPM-based
> distros, but still you have to do it yourself, as there is not a strong
> neuroscience-related community. In Debian however, you have a great
> Python team and the Debian-med blend, that make it a nice and pleasant
> job. Thanks for that as well.
>
> But the best is that people get used to things being to easy and just
> work that they start to demand more. With a (admittedly still low), but
> increasing frequency you hear people: 'I have this Debian setup, will
> your new tool work in it?' ... setting standards.
>
> I hope Debian will continue to provide this rich environment (even for
> the very-special-interest software) and propagate the idea of freedom.
> I could go on for a while listing examples of what makes me happy about
> Debian, but I guess this message is already long enough.
>
> I'd love if the feeling while reading -devel would become a bit more
> similar to the one I get when using the OS.
I was just about to write the same thing. Debian is great and it is
doing very well. I am very busy with my own studies (theoretical
physics) and research (electronic structure calculations), so I
haven't finish my NM yet, but I will. Because Debian is just the best
platform for scientific computing I know of.
And as to community, I was to Merida QA meeting the last year, Ubuntu
Developer Summit in Prague and Google Mentors Summit (where I met our
DPL and other develoepers), so I think I met some part of the
community and the people I had a chance to meet are great.
Ondrej
Reply to: