[Pkg-scicomp-devel] joining debian-science and pkg-scicomp

[Ondrej Certik]

On Wed, Jul 9, 2008 at 8:23 AM, Andreas Tille <tillea at rki.de> wrote:
> On Wed, 9 Jul 2008, Ondrej Certik wrote:
>
>> Great -- if it's internal to Debian, then all is fine. When I read
>> that, I said -- hell no, I don't want another derivative. :)
>
> I really love people who think "hell no" if they hear about just another
> derivative from Debian. ;-))
> Welcome in the boat of people who try to make Debian good enough to make
> any reason to derive void.  BTW, I'm constantly thinking about a better
> name for the beast than Custom Debian Distributions. [1]

Yes. I am using Debian from probably around 2001, but it took me 7
years to realize that it is so easy to fix things in Debian itself and
simply
make it the system right for my needs.

The only good derivative imho is Ubuntu, because in this particular
case it did more good than harm. Thanks to it, many people use the
packages that we fix in Debian. That is great.

>
>> change the link:
>>
>> http://wiki.debian.net/?DebianEdu
>>
>> to:
>>
>> http://wiki.debian.org/?DebianEdu
>
> Thanks for any hints.  If anybody wants to fix this kind of problems
> directly - the text is inside the cdd source package and can be accessed
> in SVN [2].
>
>> Andreas --- so the Debian Science CCD just provides the metapackages?
>> I mean -- all you have to do is to make sure that the user just
>> installs one metapackage and he gets a fully working system ready to
>> do whatever he needs in his field? And then you also make sure that
>> all the packages work nicely together?
>
> Well, this is one major point.
>
>> Or is there something more that I don't see. :)
>
> I try to propagate more things around it but things are developing slowly.
> If you consider that the tasks pages in the web [3] I frequently advertised
> here on this list took at least three years from the idea to the
> implementation
> you might imagine that there are more ideas hanging around how we could
> provide other clever tools if spare time would permit.  David Paleino had
> done a really good job with the Debian Med alioth pages [4] - the tasks
> pages
> are only one part of it.  My goal is to be the missing link between upstream
> developers and users of field specific software.  This is the cruxial idea
> in a CDD.  The rationale behind this idea is that Free Software does work
> perfectly in standard desktop applications (browser, mail, office suite),
> but it does not in special applications for a minority of users.  There
> are for sure exceptions - but try to look at the blockers in specific
> Windows - Linux migration scenarios.   The point is basically that the
> critical mass of Free Software developers is to small for specific
> applications.
>
> My dream is that CDDs might help to bring people together to reach the
> critical mass which is needed to develop high quality Free Software.
> (And trust me - it is a long winding road to make people aware of this idea.
> ;-) )
> I just learned that the way to wider acceptance is to provide techniques
> people *want* to use and the tasks pages are one of these.  Once you
> are able to provide a technical base people try to understand the idea
> behind.

It seems to me that all I need are quality packages and also make them
work together. But maybe some other people need more also.

I like the pages, for example this one:

http://cdd.alioth.debian.org/science/tasks/physics.html

unfortunately clicking the  Debian Science Project  at the top brings me to:

http://debian-science.alioth.debian.org/

which is an ugly page with no information whatsoever, so then I would
evade away thinking that Debian Science project is dead. You should
pick up a homepage and stick to it. Be it the wiki:

http://wiki.debian.org/DebianScience

or some other place and this homepage should be the page that new
users like me should go to in the first place and read what it's all
about in 5 minutes.

Ondrej