incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sam Ruby <ru...@apache.org>
Subject Re: [RT] Incubator Reorg
Date Wed, 03 Dec 2003 18:56:49 GMT
Nicola Ken Barozzi wrote:

> Sam Ruby wrote:
> 
>> Nicola Ken Barozzi wrote:
> 
> ...
> 
>>> Lenya will become a sub-project of Cocoon. Or better, a sub-codebase 
>>> of the same project. It's fair to assume that once incubation is 
>>> finished, the Lenya codebase will go to the Cocoon PMC and that all 
>>> Lenya committers will go in the Cocoon PMC.
>>>
>>> Now, IIUC you suggest that in this case we don't need a PPMC.
>>>
>>> Problem: the project is under the Incubator PMC that has to keep 
>>> oversight. How can it do so effectively without being on the "Lenya" 
>>> PMC?
>>
>>
>> I think that you are missing a wonderful opportunity to delegate to, 
>> and hold accountable, the sponsoring PMC.  Ultimately, the goal is 
>> that they should be the ones held accountable, no?
>>
>> This goes back to the discussion of the incubator PMC being the vetoer 
>> of last resort as opposed to being the decision makers.
> 
> Call me stupid, but I still don't get this.

Forgive me, but I'll pass on the opportunity...  ;-)

> (not that I don't like where you are heading, as I really would like the 
> Cocoon PMC take oversight as they are eager to, but because I still 
> cannot easily see in which way it may ultimately work).
> 
> A project enters the Incubator. It's part of the Incubator. It's a 
> codebase and a set of developers in the Incubator.

Again, the semantic gap.  Is Lenya destined to become a project (i.e. to 
be listed separately in cvs:committers/board/committee-info.txt ?

If so, then the Cocoon PMC is irrelevant.

If not, then does this match what Maven is trying to do with Wagon or 
whatever they are trying to import today?  I.E.  take code which 
previously existed elsewhere and incorporate it into an pre-existing, 
active, and vibrant project?

If so, then the Cocoon PMC is extremely relevant, and the incubator's 
role should be one of IP clearance.

> But you say that the developers are on another PMC list who is to be 
> held accountable... then the project is not under the Incubator, as the 
> board holds the Incubator accountable for projects in incubation, no?

The IP clearance is still very much an an issue.  And judging by your 
recent responses to Jason, this should be a matter of days.

Beyond that, we are talking about potentially adding committers and new 
PMC members, something that Cocoon is certainly authorized to do.

> Held accountable... how, and by whom?

The exact same way that Maven is accountable in the incorporation of 
Wagon, and to the board.

> How can the Incubator PMC veto something that is not even known, as 
> decisions are taken by the Sponsoring PMC?

That seems like a curious statement, given your recent statement:

--- begin cut/paste from separate email...
> Yeah, but it seems a bit artificial to create a PPMC mailing list 
> just for this. IMHO having all Project Chairs on the Incubator PMC
> + anyone that is bringing in new codebases on the Incubator PMC 
> list should suffice IMHO. 
--- end cut/paste from separate email...

It seems to me that if each project chair is a member of the Incubator 
PMC, then it would be impossible for a PMC to take an action without the 
Incubator PMC knowing about it, right?

The following offer is meaningless as you are already an ASF member and 
an officer of the corporation, so you can already review all PMC mailing 
lists, but I would certainly support a reciprocity arrangement whereby 
you are allowed to subscribe to all PMC mailing lists.

Does this solve your concern?

> For example, committers must be voted in by a PMC. The project is in the 
> Incubator. Does it make sense/is it legal for committers that are under 
> the Incubator PMC oversight to be voted by another PMC?

Again with the project term.  If Lenya is destined to be a separate 
project, then it has no rights being "under" Cocoon, and yes, the 
Incubator PMC should be the ones doing the voting here.  If Wagon is 
destined to be incorporated into the existing Maven project, the it 
should be Maven that votes in new committers.

>>> Second problem: since the Project is not accepted in Apache, is it 
>>> good to add Lenya developers to the Cocoon PMC?
>>
>> If they are truly part of the Cocoon community, then sure.  However, 
>> if the intent is to form a second community and "hide" it under the 
>> Cocoon umbrella, then no.
> 
> I mean add them *before* the project exits incubation. Do you think that 
> adding them to that PMC before successfully incubation is ok?

If Avalon can add Noel, then I see no reason why Cocoon can't add anyone 
they feel is part of their community to their PMC.  Besides, if we limit 
the scope the incubator's role in donations to existing projects to IP 
clearance, we are talking about a matter of days, right?  It already 
takes a minimum of 72 hours to add a person to a PMC...

>>> So, it would seem that we have two solutions:
>>>
>>> 1 - add Lenya developers and Incubator members to the Cocoon PMC list,
>>>     but not the PMC
>>
>> I don't object to that, but I don't see that as necessary.
> 
> The "add Lenya committers" part or the "add Incubator members" part?
> And why?

The "add Lenya committers" part.  As already discussed above, there 
should already be a significant overlap between the incubator and cocoon 
pmcs.

As to why: as I said above, if Cocoon feels it appropriate to add mere 
"users" of their technology to their PMC, then that's OK with me.

>>> 2 - make a PPMC consisting of all the above
>>
>> If, and only if, one anticipated Lenya becoming a separate community 
>> and PMC, thus:
>>
>>> 3 - have Lenya become an indipendent project, and not stay under Cocoon
>>
>> In theory, that would be OK with me.  In practice, we need to strike a 
>> balance between the board having to manage several hundred projects 
>> (something we are far from having to concern outselfs from now), and 
>> projects with several hundred PMC members (a potential real problem 
>> for Jakarta).
>>
>> I personally don't see evidence that Cocoon is exibiting any symptoms 
>> of the latter problem, so I have no problem with Cocoon sponsoring the 
>> adoption of Lenya.  Furthermore, I have full confidence that the 
>> Cocoon PMC would be the first to recognize such a problem if it were 
>> ever to become an issue in the future, and would take actions to 
>> correct it (such as spinning off Lenya to a separate project).
> 
> I agree on this last point.
> 
> What still bugs me is how to do oversight of projects under other projects.

The ASF does not recognize the existance of projects under other 
projects.  This was unclear before.  We are actively working to correct 
this.

- Sam Ruby


---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org


Mime
View raw message