[Server-devel] Clarifying some XSCE issues.

[David Farning]

 After a couple of eventful, and fruitful, weeks with XSCE, I would
like to to clarify a few points where our message seemed to be
unclear.

-- Emphasis on small deployments. As a community, there is currently
an emphasis towards small deployments. Due to their limited
bureaucracies, we have been able to establish very quick turnaround
times for feedback and fixes. This has enabled us to keep our rate of
development quite high.

The design and implementation of XSCE should scale from 1 school to
100,000 schools. There is currently an "Ansible rewrite"  (
http://www.ansibleworks.com/docs/ ) in process. We believes that the
use of ansible as a configure management tool has the potential to
reduce the maintenance required at large deployments. A team of
developers at AC is working on this now with the goal of landing in
the beginning of the 0.5 release cycle

-- Emphasis on ARM. XSCE runs on any hardware which runs Fedora 18.
The current releases heavily favored XOs because they were a logical
starting point from a Quality Assurance perspective. Debugging is an
order of magnitude easier when testers and developers are running the
same equipment. Starting with XOs enabled us to focus on school server
developments rather than hardware quirks.

Each quarterly release we have added additional hardware types. This
is allowing the project to scale gradually to more and more device
types without overwhelming developers and testers.

-- Core OS. The decisions to base on Fedora was not taken lightly. If
we wanted to run on XO-4s we had to run on a recent Fedora. The second
issue is the fact that most of  the interesting low powered servers
are ARM based.

Moving forward, our goal is to use the flexibility of the ansible
provisioning system to be able to configure XSCE on fedora and CentOS.
On an interesting note, we have been contacted by a team of ARM on
CentOS developers who offered to work with us to get CentOS a a couple
of ARM based servers.

-- Our goals is not to tell anyone what they should do. We have
avoided discussions on public lists such as this to avoid the
temptation for people to tell us what to do... or for us to tell them
what to do. Instead, we have focused on adding value.

We hope to add value to the ecosystem. We hope that adding value earns
us influence. In exchange we invite people to participate in the XSCE
project by adding value. That value will help them earn influence
within the project.

I hope this help clarify some of the issues that have come up over the
past couple of days.

-- 
David Farning
Activity Central: http://www.activitycentral.com