incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Justin Erenkrantz <jerenkra...@apache.org>
Subject Role of incubator was Re: [Tapestry-contrib] Re: Tapestry?
Date Sun, 05 Jan 2003 12:03:43 GMT
--On Sunday, January 5, 2003 10:01 PM +1100 Conor MacNeill 
<conor@cortexebusiness.com.au> wrote:

> What problems do you see that Jakarta has managing what it has?
> BTW, this is a genuine question, not flamebait. I have seen this
> statement repeated often and would like to understand its basis.

*sigh*

The ability of the Jakarta PMC to maintain oversight of its 
constituent projects.  Sam has stated that the Jakarta PMC takes a 
reactionary response to issues rather than a proactive response 
because it can't keep up.  I believe that is contrary to the original 
intentions of the PMC structure, and indicates a flaw in the model 
used by Jakarta's PMC.

I believe this is partly because there isn't adequate representation 
in the PMC by all of its projects.  One of the key tenets to the ASF 
model is the meritocracy.  The Jakarta PMC, as a popularly elected 
body, isn't based on merit.

The management structure needs to be localized near the people doing 
the work - i.e. the code itself.  This allows active oversight to be 
maintained.  The people doing the work get to have a say in the 
management.  This is what makes them aware of how the ASF works. 
(Why aren't more people from Jakarta subprojects members of the ASF?)

For example, I believe that Tomcat should certainly have its own PMC. 
There is no reason to believe that the Tomcat committers themselves 
can't be legally responsible for the project and its management.  In 
fact, IIRC, according to Roy, only actions by PMC members will be 
protected by the ASF.  Actions by committers may not be protected. 
Therefore, under this interpretation, the bulk of the Jakarta 
participants aren't covered by the protection of the ASF.  How many 
people involved in Jakarta projects realize this?  (It was brought up 
on the reorg@ list and it didn't seem to matter to some.)

> What would be the compelling reason that you would see for the ASF
> to accept any project? If Tapestry does not satisfy these
> requirements, then what sort of project would, IYHO, meet them?
> IOW, can the incubator function at all?

I believe the incubator should be about nuturing new communities. 
Projects that already have a viable community have little need for 
the ASF.  About the only thing that they can leverage is either our 
infrastructure and brand name.  Those are things I do not want us to 
allow just any project to use - we can't be SourceForge - we'd 
collapse.  I'd rather us restrict our limited resources to helping 
new communities to form rather than helping already established 
communities.  I think there's a critical mass that every project 
needs to achieve to be self-sustaining.  Projects need help achieving 
that.

My point is that our infrastructure and brand name can't be the 
reason for joining the ASF.  Tapestry already seems to have a 
community and several major releases.  A new project that is just 
starting out might only have one or two interested people and perhaps 
a little bit of code.  Or, it might be a company looking to build a 
community off donated code (see Tomcat, Ant).  Those are the types of 
things I'd rather see the ASF pursue.  I'm not terribly interested in 
importing medium-or-large size communities.

Quality over quantity.  Smaller rather than larger.

> In this case I think the incubator is for the incubation of the
> Apache Way of doing things in an existing project and its
> participants. The resolution that formed the incubator states:
>
> RESOLVED, that the Apache Incubator PMC be and hereby is
>         responsible for the acceptance and oversight of new products
>         submitted or proposed to become part of the Foundation;

I'm not really sure what you mean by this.  Yeah, the incubator gets 
to decide what the ASF takes in...

> Isn't the incubator supposed to decide exactly that question? One

Well, yes, and that's why we're having this conversation on the 
incubator list not on a wiki.

> of the problems with the incubator is when the ultimate answer is
> "No", what then for a project such as Tapestry that has undergone
> such changes? I'd like to see some discussion around that, for I
> feel it may be very difficult to say No after acceptance into the
> Incubator.

Obviously, I'm favoring a much flatter organizational model.  In 
fact, it's so flat that almost every project we'd incubate would have 
its own PMC.  There isn't another PMC that would have to approve it 
when it leaves the incubator.  The incubator PMC is responsible for 
the oversight of these new PMCs.  My hunch is that the incubator PMC 
would judge when it reaches that critical mass of participation and 
then withdraw from any involvement and do the promotion.  If it 
determines that it will never reach the critical mass, it'll shut the 
PMC down or just leave it in the incubator.

That's my take on the incubator and its role.  -- justin

Mime
View raw message