Fedora Board Strategic Working Group

Tue Jan 12 20:03:13 UTC 2010

On Tue, Jan 12, 2010 at 11:29:06AM -0600, Mike McGrath wrote:
> On Tue, 12 Jan 2010, Toshio Kuratomi wrote:
> 
> > On Tue, Jan 12, 2010 at 12:39:49AM -0500, Jon Stanley wrote:
> > > On Tue, Jan 12, 2010 at 12:24 AM, Toshio Kuratomi <a.badger at gmail.com> wrote:
> > > > There are potential issues.  For instance, when the Desktop spin wanted to
> > > > move to PolicyKit-1.0 and said that they'd veto a PoicyKit-compat package to
> > > > allow KDE applications to work.  Or when zope was dropped from the
> > > > repositories because it wasn't ported to python2.5 and we didn't want to
> > > > have a python-2.4 compat package.  There have also been times when certain
> > > > compilation options on one package were needed to allow other packages to
> > > > function but, because of the dependencies that the compilation options
> > > > brought in, that package didn't want to do so.
> > >
> > > Good points - but where and how do we draw the line between something
> > > that's acceptable to be called Fedora and something that is not?
> > >
> > I like the line of "in the Fedora repositories" but I'm warning that
> > sometimes we keep things (that are free software and legal to ship in the
> > US) outside of the Fedora repositories and that limits what a spin can do.
> > When we discuss a target audience we have to be careful that we continue to
> > make it possible for other audiences to be addressed by having ways to
> > mediate these differences.  Deciding that the default spin is more important
> > than any other spin and that the default target audience is more important
> > than any other audience is dangerous as we start making decisions based on
> > the importance to the target audience instead of on how the decision enables
> > more contributors to do the work that's important to them.
> >
> > I see things a bit differently from poelcat in that I think that having the
> > Board define a target audience for Fedora is not beneficial.  In fact, it is
> > detrimental to Fedora.
> 
> Pretend you're on the desktop team.  You've got lots of people working
> with you and are going to spend the next 3 years on something.  Is it
> going to be features or usability?  Sure it's a mix of both, but when an
> issue comes up that that will either sacrifice usability for a feature,
> or will sacrifice a feature for usability what do you do?  You can't
> answer that question without a target audience or what is Fedora (the OS).
> 
> At the moment though we seem to be sacrificing usability for features,
> that may be right or it may be wrong for Fedora.  I have no idea and
> neither can anyone else without some focus.
> 
I absolutely agree that there should be a target audience for a SIG to
effectively decide what to work on.  However, that definition should be
decided on by the individual SIGs, not by the Board.  The Board's job should
be to balance the needs of the various SIGs since they share the package
collection and therefore need to figure out how to work together in that
space.

The Board should be an enabler, a servant leader, for those who are doing
the work and driving their pieces of the overall Fedora Project forward.
Defining a target audience for the whole Fedora Project limits the people
who feel they can provide vision to the Project.

>   *Individual spins* (including the default spin)
> > would definitely benefit from targeting specific audiences but the Board and
> > FESCo's responsibility is to help all the sub-communities that make up
> > Fedora be able to derive usable products from the Fedora Package Collection.
> > This means mediating disputes, drumming up support for switching base parts
> > of our architecture (like moving from SysVinit to upstart), and defining the
> > absolute limitations that the Fedora Project will follow (free software,
> > legality within the US).
> >
> 
> The funny thing here is I think spins are a detriment to Fedora.  We
> pretend they're useful and interesting but they're really not.  If we
> describe them as "a subset of what is in the Everything/ directory" which
> is what they are, they're not at all compelling.
> 
Spins aren't best described as a subset of the packages in the Everything
repo.  A better description is that they are a particular group of
contributors' vision of what an operating system should look like.  This
seems like a compelling way of filling that niche.

OTOH, in terms of getting the current version of Fedora installed on
a system, I think network installs for high bandwidth situations and install
DVDs for low/no bandwidth is more compelling.  But I think that the only
people who care about doing this are people who are willing to customize
their distro. People who are willing to be a part of creating distributions.
People who just want to install and go should be the targets of spins.
Fedora the Project can do better than that because it can target the people
who want to create those visions.

> > If there's any audience that should be targeted for Fedora, the Project, it
> > is the people who want to create a free software operating system.  Those
> > people can then define audiences that their individual spins and SIGs will
> > target.  The Board can mediate disputes because they're tasked with
> > providing an environment in which people can build the free software
> > operating systems that are right for them.
> >
> 
> I agree with this, I don't think there's any doubt the project is healthy
> and thriving.  It's our OS I'm worried about.
> 
Sure.  But how do you define OS?  Is it the set of packages in the Everything
repo?  Is it the default spin?  And once you define the OS, how do you take
all the people scratching their itches via the Fedora Package Collection and
define a target audience that addreses all of them?

I think that Fedora the OS and Fedora the Project are currently too close.
The Fedora Project should be about growing the comunity that produces Fedora
OSes.  The Fedora OSes should be allowed to both grow and die as the demand
to create them arises.  Defining target audiences for the individual OSes is
fine but it should not be applied to the Project as a whole.  The Project
should be the foundation that allows the OSes to grow and meet the target
audiences that they want to address.

-Toshio
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
Url : http://lists.fedoraproject.org/pipermail/advisory-board/attachments/20100112/2af8db08/attachment.bin 