How can we gain more flexibility in debian-edu?
Hi,
being here in New York at DebConf10, I have the pleasure to be
exposed to many suggestions and ideas concerning Debian and free
software.
In a talk titled 'Managing Debian Installations using GOsa and FAI'
<URL:http://penta.debconf.org/dc10_schedule/events/556.en.html>,
Michael Banck illustrated his ideas and plans about setting up and
maintaining a system for academic work groups and the like.
After thinking a bit about the talk, I came to the conclusion that his
goal is not far away from what skolelinux does. The only difference
that comes to mind is a different set of user applications and no (or
another) artwork package.
I talked to him about working together with the debian-edu community,
as I think both sides could profit a great deal from each other. (And
I btw. consider a system for academic work groups as domain of
debian-edu too).
What is needed to make a collaboration possible?
We need to make our system more modularized and flexible. How to do
that, well, that has still to be figured out. But I think skolelinux
will profit a lot from a more modularized setup.
To me, the current setup appears to be too complicated and
specialized, resulting in many 'hacks' which make it unstructured and
hard to maintain. Perhaps that's the reason why there are only so few
people around participating in development; it seems like it is only
possible to either work full time on it or stop at all. Cruft has been
cumulated over the years, not making it easier to start contributing.
What's the alternative? The alternative is a core system that includes
all components common to all of these systems: users, groups, machines
and software management, terminals, kerberos, webserver, printers, etc.
Based on this core system, specializations are possible: Depending on
the area of use, the packages installed by default will differ. A
scientific group (debian-science, debian-med) will add their packages
and tools, skolelinux uses the task lists we have already in place and
might add school-specific administration tools. Remember the
discussions on the list about which age we want to support: We could
provide different lists supporting different ages broadening the user
base from the kindergarden up to young adults and adult evening
classes.
Hopefully, by offering more flexibility, the 'forks' and re-inventions
of skolelinux could be unified again and developers motivated to
participate also to the core project. By enlarging the field of users,
the burden of developing would be shared on more shoulders and
synergies could be used to improve the system and not only 'fix'
issues.
There might be replies that this will complicate the setup of
skolelinux, making it impossible for teachers to run it. First, I
don't know if this will necessarily be the case. Second, I think that
problems concerning the setup of skolelinux (understood as installing
from the DVD) are not our problem. A few more questions, perhaps even
some more actions after installation- this is not what makes schools
and teachers not to choose skolelinux. I guess most of the people
having installed the system in the last half year at school started to
change and modify the installation massively (more or less
successfully) right after setting it up.
Discussion opened.
Best regards,
Andi
Reply to: