Proposal for the new Fedora Project

[Mike McGrath]

On Fri, 1 Oct 2010, Bill Nottingham wrote:

> Mike McGrath (mmcgrath at redhat.com) said:
> > > So what am I proposing?  I think Fedora should slowly transition itself to
> > > be similar to how the Apache Software Foundation is setup.  We should put
> > > more resources into fedorahosted and grow it.  (Perhaps our new
> > > infrastructure lead would agree?  ;)  but the infrastructure is only one
> > > tiny part of it.  We build these applications, get communities around them
> > > then let OTHERS actually run them.  We'd need engineering coordinators,
> > > architects, planners.  Not just from Red Hat but from other major
> > > stakeholders as well.  We create these tools for others.
> > >
> > > Our best plan for Google isn't to take them on directly but to build tools
> > > that let everyone take them on a little bit at a time.  Clearly this isn't
> > > something that will be done next month.  This is an ambitious, long term
> > > goal that would take place over the next several years.  The reason it
> > > will work is we'd be getting on the HTML5 bandwagon early.  Very early.
> > > Others are already doing work here.  Like Mozilla's skywriter:
> > >
> > > http://mozillalabs.com/skywriter/ - https://bespin.mozillalabs.com/
> > >
> > > Take a look at that thing.  That's the future of office/productivity
> > > applications, the future of communication, the future of computing.
> > > Don't just admire what skywriter does.  Imagine what it will do, what it
> > > could do.  Imagine what Google's applications will look like when they're
> > > converted.
> > >
> > > Businesses are already moving to cloud computing for their backend.  What
> > > are they going to run on the front end?  At the moment?  Not free
> > > software.  We're no where near that market right now..  But we can be.
> > >
> > > There's no reason in the world we can't spread free software via web
> > > applications / cloud computing.  Even though someone chooses to run
> > > windows or OSX, there's no reason they can't do their primary computing on
> > > free software.  Perhaps provided by their ISP, their business, other ISPs.
>
> I suspect going in this direction would have significant consequences on our day
> to day operations.
>
> By not focusing on an end-user/developer product, our contributor base is
> likely to take a pretty large hit. Our normal progression is user ->
> contributor, and if you're starting out as an app incubator, you're not
> going to have any users until the apps reach a certain stage of maturity.
>

I'm not a big fan of turning all of our users into contributors.  The
whole quality vs quantity problem.

> If we do have this drop of contributor/user base, do we still have critical
> mass to proceed? Most of our infrastructure contributors, for example, came
> from the same group - users of Fedora-the-OS who used it, as a desktop, as a
> workstation, as a preview of upcoming enterprise technology. If we're not
> producing that sort of product within our project any more, I'm not sure
> we'll be able to draw from the same set of people to help out.
>
> Of course, that discussion may be sort of moot (especially given certain
> parties interest in Fedora, and how that would change if we made a full
> switch in this direction). Even if you're targeting the internet-based apps of
> the future, those apps can't exist in a vacuum - they'd need a) a platform to
> run the server end on b) a client platform to access them. So, we could produce
> both of those as well, as a secondary focus. If we want to be really serious
> about that, we might even decide to split those two platforms.
>

This is a fair concern.  I'd argue that since we're doing this early, the
desktop will be around for some time.  FESCo's got a handle on it, let
them deal with it while the project adds focus.  I think Fedora is a great
place to start something like this because we already have the platform.
The platform is still needed, both on the server side and the end user
side.

What I'm proposing is a thinning of the end user side with added focus on
the server side.  I'm proposing this mostly because I think it's going to
happen with is or without us anyway.

	-Mike