incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Upayavira>
Subject Re: Does an incubation proposal require a codebase?
Date Mon, 10 Aug 2009 20:06:43 GMT
On Fri, 2009-08-07 at 08:04 -0700, Ralph Goers wrote:
> On Aug 7, 2009, at 2:17 AM, Niclas Hedhman wrote:
> > On Fri, Aug 7, 2009 at 5:01 PM, Bertrand
> > Delacretaz<> wrote:
> >
> >> I would oppose to a podling with no code base and no  
> >> names
> >> in the list of initial committers - the chances of morphing that into
> >> a working Apache project would seem to slim to me.
> >
> > I agree, and with one or more committers, Labs is open for business.
> > Now, it could be that the non-ASFers will be major contributors in
> > non-bulk manner (i.e. patches are too burdensome), in which case I
> > don't think we have a solution here and I would suggest Google Code,
> > GitHub or SourceForge as the bootstrapping hosting solution.
> >
> I think that is very sad. Our goal is to grow communities that develop  
> software using "the Apache way".  Sending them somewhere else so they  
> can come back later when they are a bit more mature is a great way to  
> get them to never come back.

The creation of a podling is heavyweight. Very heavy compared to
registering a project with Sourceforge, GitHub or GoogleCode. Because of
the effort involved in the creation of a podling, it is natural that the
Incubator PMC wants some reasonable chance of success.

There are two reasons why it is more heavyweight: because we haven't
spent huge amounts of time/money automating our infrastructure and
putting it behind a pretty web UI, but more importantly because we pay
more attention to the legals, ensuring we have grants, CLAs etc, and
that we know precicely what piece of code came from where.

This latter requirement can be stifling for a very young/nascent project
that is still trying to work out whether it has a life or not. For
something a little more established, perhaps the effort can more easily
be absorbed.


> A solution to this is for Noah to create 4 or 5 bogus bits of code and  
> call that the original code base to start from.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message