Hi Dirk and all,
I've had concerns similar to Joerg's for quite a while. Indeed, I'm not
even certain how to determine who *is* a committer on any given subproject:
I'm no CVS guru, and even if I were I'm not sure run-of-the-mill committers
have the requisite privileges to look in the right place.
Two other related questions: In the past, when one was granted committer
status to one subproject, seemingly, access to other subprojects was also
granted. So, for instance, it's my belief currently that a committer to
Xerces-n does in fact have latent priviliges to Xerces-*, and so is a
committer on those subprojects. In fact, I can checkout xml-xalan using my
apache ID; so I suppose that I have commit privs there too. :) How does
this relate to votes requiring the assent of 2/3 committers on a
subproject? How does xml-commons--to which all committers to any XML
subproject have committer status, but of which a good many such committers
probably aren't even aware--fit in to this framework?
The principle of auto-deactivation seems good to me; but I wonder how the
mechanics would work? Are we proposing to write a script,that takes as
input a committer's name and a subproject, and looks through mail archives
relevant to that project for the last N months to determine whether that
person's posted? Maybe this would work to find CVS commits, but if we're
also counting general mailing list participation--as I think we should--and
even Wiki posts (of which I'm personally dubious), then how do we handle
posts from other accounts?
I think these are thorny questions but, as the Foundation matures and its
set of active participants changes over time, it's one that absolutely must
be sorted out.
Cheers,
Neil
Neil Graham
XML Parser Development
IBM Toronto Lab
Phone: 905-413-3519, T/L 969-3519
E-mail: [EMAIL PROTECTED]
|---------+---------------------------->
| | Dirk-Willem van |
| | Gulik |
| | <[EMAIL PROTECTED]|
| | .org> |
| | |
| | 03/13/2003 04:32 |
| | PM |
| | Please respond to|
| | general |
| | |
|---------+---------------------------->
>---------------------------------------------------------------------------------------------------------------------------------------------|
|
|
| To: [EMAIL PROTECTED]
|
| cc:
|
| Subject: Re: Revisions to xml.apache.org charter
|
|
|
|
|
>---------------------------------------------------------------------------------------------------------------------------------------------|
On Thu, 13 Mar 2003, J.Pietschmann wrote:
> Indeed, given that there are usually quite a few "inactive" committers
> for a specific (sub-)project. It might be necessary to keep records of
> committers currently eligible for 2/3 majorities....
Hmm - that would be rather hard, and perhaps have too many 'grays'. I
would suggest that you make this harder by for example
-> auto expire committers which have not committed for 24
months -and- have not posted to any ASF mailing list or
WiKi for 12 months.
-> auto expire committers which did not vote on certain well
defined issues for 6 months.
Or some other 'objective' principle. As this makes it easier to have a
very well and objective defined 'roster' of people.
Note that in the above _example_ I made the commit period much longer than
the vote period. The reason for this is to keep valuable 'knowledge' of
experienced committers, which are dormant at some point, in the pool.
But those are just examples - there may be betther methods for this
community.
Dw
---------------------------------------------------------------------
In case of troubles, e-mail: [EMAIL PROTECTED]
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
In case of troubles, e-mail: [EMAIL PROTECTED]
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
