ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Greg Stein" <>
Subject Re: [Proposal] Resolution for Ant top level project
Date Tue, 05 Nov 2002 21:35:30 GMT
In article <aq91jd$fl6$>, "Costin Manolache"
<> wrote:

> Conor MacNeill wrote:
>> 1. Why do this?
>> Many of you may not be aware of recent discussions on the reorg and
>> community mailing lists. Many issues have come up about the management
>> of
> IMO this is the worse reason to do it. There is no clear legal position,
> and there are other solutions.

Actually, it is quite clear. The ASF cannot provide its protection to Ant
committers as Jakarta and Ant are currently structured. The best solution
is to create a new PMC.

> Gump ( or a similar tool ) could very well be part of ant scope ( it is
> open-source software related to apache ant build tool, and it may have
> some specific task libs ). I think it should be a case-by-case decision,
> just like in jakarta.
> I agree that sets of task libraries as ant subprojects would be the best
> fit.

Anything could be defined as "in scope", but the more that gets lumped in,
the higher the chance that we'll end up in the same situation with
oversight problems. That oversight is necessary so that the ASF can assume
full liability for the Ant code. Without oversight, then you have
individuals working on the code and releasing it. With oversight, the ASF
can say "it is our code and our responsibility; if you're going to sue,
then you can only sue the ASF." (and the corollary, an individual can say
"hey. you can't sue me for that")

>> It is important to get the scope right, without it being too narrow.
>> Another of the issues raised about the Jakarta project is how out of
>> scope many of the Jakarta subprojects are, Ant being used as a prime
>> example.
> People will raise issues doesn't matter what :-) The real question is
> what is best for the community. The scope can be changed if needed (
> like anything else ).

Best for the community and the ASF. The ASF is seeking to protect the
community, so it would be nice to help it in its job :-)

>> PMC Chair
>> We will need to choose a chair for the board resolution. The chair
>> becomes an officer of the ASF and therefore needs to be approved by the
>> board as part of the resolution. Ideas about how we can choose the
>> chair are welcome although the process might depend on how many
>> candidates there are.
> I don't know if the chair must be ASF member - but probably it would
> help if he is.

The Chair is not required to be an ASF member. The Chair is appointed by
the Board as an officer of the corporation. It is quite orthogonal to ASF

(that said, there may be more "comfort" in knowing that the Chair is an
ASF member)

>> The term of the chair and the mode of future elections would be covered
>> by the bylaws of the Ant project. The development of these would b the
>> first task of the new PMC. I would expect the chair to be elected by
>> the PMC annually as is currently done for the Jakarta PMC. I don't
>> think some sort of lifetime appointment is the way to go.

Section 6.4 of the ASF Bylaws provides the rules for officers (PMC
Chairs). Specifically, it states that officers hold their position for one
year (by default; any number of things can change that).

The notion of an "election" for the PMC Chair is somewhat of a misnomer.
In actuality, what is happening is that the "voters" are making a
recommendation to the Board on who should be appointed as the PMC Chair.
We've never veered from that recommendation :-), so the net effect is that
you're directly elected your Chair.

There has been some discussion of term limits and rotation of the Chair
and whatnot. Note that Ben Hyde has been the HTTP PMC Chair for years. I
think since the inception of the ASF. The role of Chair is quite minimal
(and it should be minimal). Nominally, they are responsible for defining
the bylaws and rules and whatnot, but the understanding is that they would
be doing that using input/feedback from the whole PMC. The Chair's only
real "work item" is preparing and delivering the quarterly report to the
Board. Of course, the Chair is responsible for the work performed within
the group, as are the rest of thePMC members.

What I'm trying to say here is that the role of Chair is probably a lot
involved than people may be perceiving it to be.

>> Bylaws
>> As I said above we will need to come up with a set of bylaws that
>> govern the operation of the Ant project. These will cover things such
>> as the terms of the Chair, how the chair is elected, how the PMC is
>> composed, how PMC decisions are made, concepts such as emeritus status,
>> adding new members, etc. Here I would expect that we would not have
>> anything peculiar to Ant, something like Jakarta or httpd would be a
>> good basis.

I believe the Incubator project has been working on documenting much of
this. In particular, there are specific needs for PMC membership. The rule
of "active committers" is totally fine.

>> I'd be happy to have a discussion now, to try and capture the essential
>> elements of the bylaws if there is interest.
> I would be very unhappy if we make the move without discussing all the
> details ( before !).
> I really don't understand how the board works - i.e. aprove projects
> without clear bylaws. I would very much like the vote on this be after
> we have the bylaws and all those details.

The last part of the Board resolution is:

         RESOLVED, that the initial Ant PMC be and hereby is tasked with
         the creation of a set of bylaws intended to encourage open
         development and increased participation in the Ant Project.

The Board is concerned with getting the Project set up. From that point,
you can figure out the specifics. There isn't any reason to block Board
action on those details.

The basic rationale here is that the Board does not have the necessary
insight, nor does it care to, nor does it need to be *that* tightly
involved. The Project knows the best way to organize itself, so the Board
says "make it so". The Board provides the structure to enable the
community, and then it requests periodic reports on that community.


-- ... ASF Chairman ...

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

View raw message