incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen McConnell <>
Subject Re: Exit Criteria
Date Sat, 27 Sep 2003 12:33:10 GMT

Rodent of Unusual Size wrote:

>Stephen McConnell wrote:
>>The incubator has a scope concerning "incubation".  I hope the incubator 
>>aims to to provide the role of gatekeeper together with a support 
>>infrasture the accelerate the sucessful exit of incubated projects.  
>so far so good.
>>Within the objective and scope, opinions have been put forward 
>>concerning the process of incubation focussed rightly on the subject of 
>>incubation.  However - the discussion concerning the publication of 
>>artifacts moves us out of the domain of incubation and into the domain 
>>of managing a incubated project.  This is in my opinion a fault - it is 
>>a subject that is the concern of the incubated project - not the incubator.
>i really don't understand why this is so difficult to understand.
>perhaps it's because you're coming from the jakarta subproject

I don't think that's the reason. 

I think the real reason is that I have a lot of scepticism about the 
successful functioning of the Incubator.  I imaging future scenarios 
where candidates are keep waiting with little or no feedback while the 
Incubator PMC debates some highly important and engaging topic, ignoring 
operational responsibilities.  I imagine a future where a project is 
left to its own devices because a Shepherd has wondered off into the 
mountains and the Sponsor is preoccupied with greater things.  I also 
imagine a future where irrespective of the failure of the system - an 
individual subject to the overhead of incubation can reach out and make 
things happen simply by holding the PMC and Chair responsible relative 
to a set of policies and procedures. 

The "holding the PMC responsible" is coming together - however, the 
distinction between Incubator responsibility, Sponsoring Entity 
responsibility, and Podling responsibility is blurred with respect to 
this notion of publication.  It is blurred because there are valid 
reasons for each party to claim a sense of jurisdiction:

* the podling has the right to claim jurisdiction because it is the 
domain expert
* the Sponsoring Entity has the right to claim jurisdiction as they are 
the entity that has implicit responsibility
* the Incubator has a claim on jurisdiction only in that the podling is 
implementing Apache practices (e.g. infrastructure release guidelines 
could be an example)

However, there is already is implicit release simply via publication of 
code under CVS as a result of acceptance of a candidate into 
incubation.  So how does one correlate these different claims to 

Here is a suggestion:

* the incubator establishes a specific incubation license template that 
includes relevant statements concerning incubation status
* every incubated project uses the incubator template
* an incubated project should be allowed to demonstrate its ability to 
form a quasi PMC and vote on a release/publication
* such a decision should be ratified by the Sponsoring Entity
* but the Incubator PMC should be able to veto a publication recommendation

This places the incubated project in a position where is has to act like 
a PMC - get it structure in place - and go though the motions of a 
release/publication cycles.  The appropriate due-diligence is assured by 
the decisions of the Sponsoring Entity while maintaining an effective 
oversite is a role maintained by the Incubator PMC.

>everything the podling does is most especially the concern of the
>incubator.  until the podling graduates, it is *not* an independent
>part of the asf or even of another tlp.  the incubator has the
>responsibility of getting the podling to the point at which is
>can be such -- but until that happens, it is in a grey area.


>and until it's an approved asf entity, it can't release anything
>with the asf stamp of 'this is 100% apache software' on it.
>releases yes, afaic, of course -- but 'this is asf software' releases,

Which is I think addressed in the division of responsibilities I have 
outlined above.



Stephen J. McConnell

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

View raw message