[fab] Proposed Project: Fedora Testing

[Jeremy Katz]

I don't have a _lot_ of answers... but a few random thoughts

On Wed, 2006-07-05 at 16:53 -0700, Karsten Wade wrote:
> On Wed, 2006-07-05 at 17:16 -0400, Will Woods wrote:
> > On Sat, 2006-07-01 at 10:53 -0400, Greg DeKoenigsberg wrote:
> > > On Sat, 1 Jul 2006, Rahul Sundaram wrote:
> > > 
> > > > I would let the test suite run on a full development cycle, get the 
> > > > community involved and show the results before accepting it as a Fedora 
> > > > Project. 
> 
> Whose development cycle?  FCs?  $UNNAMED_PROJECT's?

The FC6 cycle -- given a large part of the testing, at least at this
point, will revolve around distro testing.

> > Sure, it's up to the board to decide whether to accept this as a Fedora
> > Project. We'll be running tests, releasing code, and hopefully writing
> > new stuff all through the rest of the FC6 development cycle either way.
> 
> Since this new project approval process is, well, new, I want to do a
> sanity check on what I think I see here.  Is this statement accurate?
> 
> "You want proof of the viability of the project _before_ it becomes a
> formal Fedora Project project, i.e., gets to use the name Fedora
> Something Project."

Yes.

> These points come to mind from this:
> 
> 1. If this barrier were in place previously some projects (Fedora
> Ambassadors) wouldn't have gotten very far.  "We are the Ambassadors
> Project.  Ambassadors for what?  Uh, stuff."

Actually, I think this is one of the reasons ambassadors succeeded.
Proof of viability doesn't mean you're going full steam ahead -- it
means that you have a plan that's credible.  What's involved for that
will differ depending on the type of project.

> 2. How then does a project name itself in a meaningful and memorable
> way, if it *hopes* but is not *guaranteed* acceptance as a project?
>
> For example, "Testing Project" doesn't stick that well without the
> "Fedora" at the beginning.  So, Will decides to call it "Velodrome", a
> catchy name that captures the speedy and testing (proving ground) nature
> of the project.  Then it gets accepted as a Fedora project after the
> Velodrome name has stuck.  Do we call it "Fedora Velodrome"?
> "Velodrome, the Fedora Testing Project"?  "Velodrome, a division of
> Fedora Project?"
>
> In some cases, the Velodrome brand might be worth more.  But perhaps the
> project gains more from the Fedora renaming?
> 
> We go through this in the corporate/consumer world with product
> acquisition and renaming.  Do we want to introduce this to Fedora?

While there's value in the brand, you dilute that by calling everything
"Fedora foo".  This is much like how you're starting to see GNOME
projects that aren't just named gfoo anymore and are trying to have some
identity of their own.  Sure, they take advantage of their association
with GNOME, but that's not the only thing that makes them valuable.

> 4. What does it mean to be a formal Fedora Project project, beyond the
> name?  That is, is someone restricted from using the Wiki for the
> project until then?  How about CVS?  Plone?

This is one of the recurring questions in my mind.  And I don't have an
obvious answer to it.

> 5. How many contributors does FP bring to the party?  Conversely, how
> many are held back (unaware, resistant, etc.) by withholding project
> recognition?  How much risk is a nascent project put to by trying to
> breath on its own as an unrecognized child?
> 
> The concern here is that we can miss out on valuable projects coming to
> life because we withhold vital support when it could help the most.
> Knowing when to apply the spatula and when the knife is sometimes a
> subtle thing, what happens when he hand over that decision to a process?
> 
> As much as I love a good, well-documented process, I want to keep care
> not to smother the rise of good idea.

The opposite side of your delicate knife edge is the one where you have
so many projects that just don't go anywhere that people are reluctant
to join in and help because "well, why will this be different than the
last 3 I joined that failed"?

Jeremy