incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Benson Margulies <>
Subject An idea you might call pre-incubation
Date Fri, 11 Feb 2011 16:13:36 GMT
Let's say that I have an idea for some new open source initiative. How
would I proceed?

Well, I *won't* proceed by asking for an ASF lab. Why? Because an ASF
lab precludes me from collaboration with anyone except ASF committers,
and that's a completely arbitrary distinction.

So, these days, my likely trajectory would be to go create a repo on
github. If, over time, the thing grows some legs, I might then bring
it to the incubator.

In the mean time, I have to come up with Java package IDs and maven
group IDs, and I have to assert the copyright myself, yada yada, and
if it makes it into the incubator all that gets changed around.

Seems pretty wasteful. Leaving aside the git versus svn issue, where I
have bright hopes that something good is coming, why can't we have a
sort of light-duty incubation that would cover this case? I'm
imagining a procedure roughly as follows:

There has to be an iPMC member as part of the starting team. Thus,
this is open to any member, or any committer who can persuade the iPMC
that they have enough experience on other Apache projects to be

The seed person would have to talk two other iPMC members into serving
as supervisors.

New committers would be invited by that group.

Code would be granted 'from birth' to the foundation.

There would be very minimal reporting requirements.

There would be no formal ASF releases.

There would be a time limit of 3 months to achieve, say, a group of at
least 5 committers.

If the group achieves critical mass including some specified threshold
number of existing ASF committers, it would be a candidate for direct
graduation as a TLP. Otherwise, it would progress from this 'lab'
status to that of a normal incubator podling.

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

View raw message