geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geir Magnusson Jr." <>
Subject Re: What is a subproject anyway? (was: Trifork Code)
Date Sat, 09 Jul 2005 23:16:30 GMT

On Jul 9, 2005, at 7:08 PM, Davanum Srinivas wrote:

> Please see below.
> On 7/9/05, David Blevins <> wrote:
>> Thanks!  More questions below...
>> On Sat, Jul 09, 2005 at 05:42:08PM -0400, Davanum Srinivas wrote:
>>> - is one big happy family - means one ACL (same  
>>> list of
>>> committers for CVS and SVN as some projects are in SVN and some  
>>> are in
>>> CVS)
>>> - Each entity in is a sub project. which implies each
>>> one has a subset of ws committers working on it and each sub project
>>> has its own life cycle including releases.
>> So essentially, what you're saying with these two points is that for
>> WS, subproject means (still having a hard time quite getting it)
>> something like this:
>>   [      committers      ]
>>   [code][code][code][code]
>> Rather than:
>>   [committers][committers]
>>   [         code         ]
> Yes. breakup and manage code.
>> So, in the big picture the focus is on breaking up and managing the
>> code, not breaking up and managing the people.  Is that the right way
>> to think of it?
>>> - Some projects were bootstrapped from inside (example scout,  
>>> wss4j).
>>> some came from outside (muse, pubsub).
>> How do you deal with code coming from the outside?
> You *could* maintain separate ACL's till the incubation is done. But
> it's better to get everyone involved as soon as possible.

We won't be "incubating" anything in here - if we want to incubate,  
we should go to the incubator.  I see incubating as building a  
separate, distinct community, and I don't think any of us want to do  
that here in Geronimo, but rather incorporate the Trifork people  
right into our community from day 1.

That said, separate ACLs doesn't mean we're incubating, IMO.

>>> - Folks who want to work on code outside their primary project (one
>>> for which they acquired committership via incubation or merit)  
>>> can get
>>> active on other projects if they want to.
>>> - Each project can have its own list of active folks (usually on the
>>> web site for each project)
>> So it sounds like there is some form of people grouping.  Bunch of
>> questions here, mostly overlapping (no need to answer point for
>> point):
>>  - How formal is that and how does one get grouped?
>>  - How is someone determended to be on the ative/non-active list?
>>  - How does someone "get active" on a project?
>>  - Are you always considered "active" once you make the list?
>>  - How does that affect voting?  Is it bad taste to vote on a project
>>    on which you aren't active?
> Nothing formal, its usually historical. i mean if you got in as a
> committer on axis, usually you are involved mostly in that. but if you
> want to work on other things, you can announce your intentions on the
> mailing lists, get started with patches (and get them reviewed), get
> some handle on the sub-project and the people, then once you gain
> confidence add your name to the docs. we still haven't figured out how
> to make someone emeritus. possible yard sticks are commits in the last
> year, posts to dev/user mailing lists etc. vote is a yardstick too.
> people don't usually barge in and vote, we usually have to prod people
> for example, i had to prod folks working on Axis2 to vote even for
> Axis 1.X releases. we are usually very relaxed about such stuff. yes,
> we do want 3 +1's for a release as usual.

Sounds a little like Jakarta commons, with people just noting that  
they are working on a subproject and getting involved?


Geir Magnusson Jr                                  +1-203-665-6437

View raw message