Request: please consider clarifying the project's position on Spins
Jaroslav Reznik
jreznik at redhat.com
Thu Dec 2 09:15:55 UTC 2010
On Thursday, December 02, 2010 03:58:06 am Adam Williamson wrote:
> On Wed, 2010-12-01 at 21:32 -0500, Bill Nottingham wrote:
> > (Now, if we want each spin to fork off their own subproject, with their
> > own rel-eng, their own QA, and maybe even their own SCM branches?
> > That's more likely to scale.)
>
> This is the model I *really* want to avoid, because it defeats the whole
> purpose of having a project. What I'd prefer to see is the model where
> we have project-wide general groups, but SIGs contribute actual work.
I agree - that is something I'd like to avoid too and I think it's the strength
of having SIGs and higher level shared groups of qa, relengs etc. and SIGs
helping these groups. As we do not have unlimited resources - Desktop and Plasma
Desktop spins have probably enough people to live separate but I'm not sure it
will work for other desktop spins.
I think Desktop Validation worked this time and big thanks to Adam for the
coordination - this is example how we should (in the project) work together and
confirms "we have project-wide general groups, but SIGs contribute actual work"
really works.
> As I said, this worked well for QA for F14; QA group (i.e. me) set the
> framework, by providing test cases and a test matrix and notifying when
> builds were available. The spin SIGs contributed the actual testing. If
> we have each spin group have its own QA and its own releng, it's going
> to add a lot of unnecessary overhead with each group designing its own
> releng and QA processes when there's no need for these to be
> differentiated between groups; only the *work* is different.
>
> > And frankly, one of ideas behind spins was that it was a way to showcase
> > the exciting, innovative work that can be done in Fedora. If the only
> > exciting, innovative stuff we can come with as a community is just
> > 10 different implementations of a panel, terminal, window manager, and
> > file manager... that's pretty sad.
>
> It clearly isn't, given the range of spins we have at the moment, but
> the desktop spins do appear to be the most popular. And there's the
> special-case sorta desktop spins, Sugar and Meego; these are desktop
> spins, in a way, but characterizing them as just a panel, terminal,
> window manager and file manager kinda misses the point :)
It's another problem - we have two main groups of spins:
cat #1. take some desktop, add/remove some packages, call it spin (I'm not
saying it isn't hard work!!!) based on other spins (cat #2)
cat #2. make different base set of packages, with mostly completely different user
experience aiming on power users, netbook users - mostly other desktop
environments like Gnome, Plasma, XFCE or some people here are working on Server
spin. Then MeeGo - it's completely different user experience.
I try (it's not official) not to call #2 category as spin but more like edition -
imagine Shell Desktop by Fedora Project, Plasma Desktop by Fedora Project.
So yes, it's going to be more subproject but again - try to share as much as
infrastructure (qa, releng) as we can = win!
R.
PS: On the other hand - it's going to be much more bigger mess soon - classical
desktop as we know it today, that servers everyone for every purpose is quickly
dying.
--
Jaroslav Řezník <jreznik at redhat.com>
Software Engineer - Base Operating Systems Brno
Office: +420 532 294 275
Mobile: +420 602 797 774
Red Hat, Inc. http://cz.redhat.com/
More information about the advisory-board
mailing list