You are viewing a plain text version of this content. The canonical link for it is here.
Posted to general@incubator.apache.org by Matthieu Riou <ma...@offthelip.org> on 2008/03/26 23:50:43 UTC
[DISCUSS] Community diversity (again)
Hi,
We've already discussed this a bit in the past [1][2][3] but no clear
consensus as emerged with respect to diversity and the current graduation
policy. There seems to be growing pain with the following policy paragraph
in the graduation checklist:
"The project is not highly dependent on any single contributor (there are at
least 3 legally independent committers and there is no single company or
entity that is vital to the success of the project)"
The problems that I have with this is that it doesn't seem to be really
applied so far. We've rejected podlings that were compliant to the rule when
applied strictly. I know that the policy also states that this is the
minimum requirements but it doesn't seem to be clear enough in the minds of
people reading this. And I'm at a point where, as a mentor, I'm not sure
anymore how high I should place the bar before pushing a podling toward
graduation.
Right now I can see 3 different ways to improve this:
1. Stick to the rule and accept graduation even if there are 20
committers from a single company and just two other independent, low
activity committers.
2. Make the barrier of graduation higher, with more rules (percentages
come to mind for example), as the current are apparently not enough.
3. Accept the fact that's it's a bit more fuzzy than that and modify
the policy to reflect this fuzziness, making it clear that the current
requirements are just a minimum indication (it already says so but given the
expectations, it's apparently not clear enough).
Honestly, I can see several problems with all 3 alternatives. But they're
still better than what we have currently.
Thoughts?
Thanks,
Matthieu
[1]
http://mail-archives.apache.org/mod_mbox/incubator-general/200711.mbox/%3Cf767f0600711171546h5b5aab29tf489715dda01ed7f@mail.gmail.com%3E
[2]
http://mail-archives.apache.org/mod_mbox/incubator-general/200710.mbox/%3Cf767f0600710130043v3fc6d5f8r70162a0671e20393@mail.gmail.com%3E
[3]
http://mail-archives.apache.org/mod_mbox/incubator-general/200710.mbox/%3C88CA85D3-6C57-4D30-A651-584F7EC09E87@hogstrom.org%3E
Re: [DISCUSS] Community diversity (again)
Posted by Yoav Shapira <yo...@apache.org>.
On Thu, Mar 27, 2008 at 5:55 AM, Endre Stølsvik <En...@stolsvik.com> wrote:
> Thilo Goetz wrote:
> > Make it absolutely clear that the diversity of the community
> > will be judged by the IPMC based on the overall conduct of the
>
> One obvious problem with such approaches is that the criteria will
> become very unevenly/unfairly applied. Some projects might just "slip
They already are, so we might as well be honest about it. Different
members of the IPMC have very different opinions, fears, and personal
guidelines, and they vote accordingly, which is totally fine. It's a
realistic reflection of the diversity of the IPMC and Apache itself ;)
Yoav
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: [DISCUSS] Community diversity (again)
Posted by Marnie McCormack <ma...@googlemail.com>.
Hi All,
I can definitely see the value in having rules that allow for discussion and
some subjective assessment of podlings by the IPMC.
I think some additional detail (best practice style ?) around what a diverse
community looks like from various angles would be really helpful for podling
projects. Some of the angles discussed recently came as a surprise to me and
it'd be good to see what other people's suggestions would be.
For example:
- Committer activity over a given period i.e. what should this look
like, how diverse should this be, what would a problematic position be, on
trunk/branches etc ?
- PMC make up. We've taken a slightly different approach to PMC
composition (no bar to entry beyond the committer bar, but on a committer
requested basis), thinking it applicable. Recent discussions have
highlighted that the PMC make up is perhaps more important than general
committer composition on a project and thus we should strive to make it
diverse by encouraging (nominating) committers on to the PMC for diversity
reasons ?
- Code vs documentation weighting. Some projects have contributers
whose focus is specialised in a particular area like documentation. What do
we include when considering diversity from this perspective i.e.
document changes, svn commits, JIRAs created, release management tasks
- List contribution. I think this one is tricky to measure, as we all
have different ways of working. Some of us (me :-) speak a lot, others less
so but perhaps in a more precise fashion and are effective. We include list
contribution as one of the factors assessed before committer-ship. I'd have
hoped that would be enough, but maybe not all projects assess this in the
same way ?
In conclusion, I can see that no objective bar applies in some of these
areas. I do think that the IPMC (and others) views on what a good example
looks like or what some differing, but diverse, projects look like would be
helpful.
I've left out 'legally independent' (argh) but I thought that the idea of
salaried % was a good one.
Maybe there's some scope for a set of measures and a guide around the
must-haves & the nice-to-haves and some idea that you must not fail the
diversity check on more than x nice-to-haves ?
Hth,
Regards,
Marnie
On 3/27/08, Thilo Goetz <tw...@gmx.de> wrote:
>
> Noel J. Bergman wrote:
> > Endre Stølsvik wrote:
> >> Thilo Goetz wrote:
> >>> Make it absolutely clear that the diversity of the community
> >>> will be judged by the IPMC based on the overall conduct of
> >>> the project, mailing list, commit activity etc.
> >
> > I'm not sure that "diversity" and "conduct" are really mapping that
> way. A non-diverse community can conduct itself properly and well, but
> still not be diverse. And although conduct is important, we do place value
> on demonstrated diversity, as well. I'm afraid that we can give you
> examples of where one or the other was present, but arguably not both, and
> we've had to deal with the consequences later. So we all try to learn from
> prior experiences, and apply that in our current and future judgments.
>
> My quote is a bit out of context like that. What I meant to
> convey was that it is not sufficient that there *be* diversity
> (in the sense that 3 independent committers exist). The business
> of the project must be conducted by a diverse set of people. At
> least that's what I got out of the recent discussions.
>
> --Thilo
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
Re: [DISCUSS] Community diversity (again)
Posted by Thilo Goetz <tw...@gmx.de>.
Noel J. Bergman wrote:
> Endre Stølsvik wrote:
>> Thilo Goetz wrote:
>>> Make it absolutely clear that the diversity of the community
>>> will be judged by the IPMC based on the overall conduct of
>>> the project, mailing list, commit activity etc.
>
> I'm not sure that "diversity" and "conduct" are really mapping that way. A non-diverse community can conduct itself properly and well, but still not be diverse. And although conduct is important, we do place value on demonstrated diversity, as well. I'm afraid that we can give you examples of where one or the other was present, but arguably not both, and we've had to deal with the consequences later. So we all try to learn from prior experiences, and apply that in our current and future judgments.
My quote is a bit out of context like that. What I meant to
convey was that it is not sufficient that there *be* diversity
(in the sense that 3 independent committers exist). The business
of the project must be conducted by a diverse set of people. At
least that's what I got out of the recent discussions.
--Thilo
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: [DISCUSS] Community diversity (again)
Posted by "Noel J. Bergman" <no...@devtech.com>.
Endre Stølsvik wrote:
> Thilo Goetz wrote:
> > Make it absolutely clear that the diversity of the community
> > will be judged by the IPMC based on the overall conduct of
> > the project, mailing list, commit activity etc.
I'm not sure that "diversity" and "conduct" are really mapping that way. A non-diverse community can conduct itself properly and well, but still not be diverse. And although conduct is important, we do place value on demonstrated diversity, as well. I'm afraid that we can give you examples of where one or the other was present, but arguably not both, and we've had to deal with the consequences later. So we all try to learn from prior experiences, and apply that in our current and future judgments.
> One obvious problem with such approaches is that the criteria will
> become very unevenly/unfairly applied. Some projects might just "slip
> past" with very little scrutiny, while others suddenly rack up a lot
> of discussion
That's life. In the end, as much as we may attempt to quantify some notions, votes really come down to the HUMAN and SUBJECTIVE judgment of the Incubator PMC Members. Hopefully, we all agree that these are people who have earned the right to make those judgments, and do so in all sincerity, even if a mistake is made from time to time. The Incubator PMC is looking after the best and long term interests of the ASF and each of our Incubating projects. We want each and every one to succeed. If we feel that QPid, or any other project, is best served by building diveristy, that iss our collective best judgment of what is important and helpful.
The rules that I want to see are those that support human judgment, not rules that make the judgments for us.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: [DISCUSS] Community diversity (again)
Posted by Endre Stølsvik <En...@stolsvik.com>.
Thilo Goetz wrote:
> It's just that to an outsider, it is totally unclear what that
> means. And it may be impossible to really convey the meaning
> of it in a few sentences. So why don't we just say so.
>
> Make it absolutely clear that the diversity of the community
> will be judged by the IPMC based on the overall conduct of the
> project, mailing list, commit activity etc. I know it's there
> already, but it could be reinforced. Perhaps at the end of the
> "Creating an Open and Diverse community" (community should be
> capitalized, btw) paragraph: "The IPMC will judge diversity of
> the project based on many criteria. These include mailing list
> activity, commit activity and the affiliations of the committers.
> There is no single sufficient criterion, it is the overall conduct
> of the development community that counts." Or something like that.
One obvious problem with such approaches is that the criteria will
become very unevenly/unfairly applied. Some projects might just "slip
past" with very little scrutiny, while others suddenly rack up a lot of
discussion, where everybody has something to say and the project gets
lots of (somewhat unwanted!) scrutiny. The amount of scrutiny and
subjective evaluation will be affected by time of year, holidays and
daily moods of the PMC members etc etc..!
An idea would be to have a pretty clear set of rules, with X committers,
where all must actively have participated in email and subversion
activity, blah blah - SOMETHING that can be objectively evaluated and
the project may strive towards - and then some final line about "this
will however be up to a final subjective evaluation where obvious
attempts at circumventing/bypassing/"minimal-efforting" the system will
(hopefully) be caught and stopped."
Endre.
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: [DISCUSS] Community diversity (again)
Posted by Thilo Goetz <tw...@gmx.de>.
Matthieu Riou wrote:
> On Wed, Mar 26, 2008 at 5:03 PM, Robert Greig <ro...@gmail.com>
> wrote:
>
>> I think this is an important topic for future incubator project groups
>> to have clarified.
>>
>>> "The project is not highly dependent on any single contributor (there
>> are at
>>> least 3 legally independent committers and there is no single company or
>>> entity that is vital to the success of the project)"
>> Perhaps the word "legally" needs to be removed since from the Qpid
>> discussion it would appear that several people do not think a strict
>> legal interpretation should apply?
>>
>> However, given that the intent (as I understand it) is to avoid the
>> case where a project dies because one entity withdraws funding,
>> perhaps some definition along the lines of "for people who are paid to
>> contribute to the project, no single entity remunerates more than 50%
>> of the committers"?
>>
>
> The thing is, for both for Qpid and Tuscany, there *were* 3 independent
> committers. And in the case of Qpid, I believe the "no single entity
> remunerates more than 50% committers" would have worked as well.
The thing that I've learned over the past 18 months as a podling
committer is that there are no hard and fast rules. The community
(in this case, the IPMC) decides on a case-by-case basis, based
on more than just the letter of the law. That's a good thing, but
it is a difficult concept to grasp when you're new to the Apache
way.
So when you're new to this, as I am, you look for the well-defined
rules and want to cling to them. You see the "3 independent
committer" passage and you aim for that, because that's something
you can understand right off the bat. However, when I go back
to the graduation guide now and read the whole passage, it says:
"Graduation tests whether (in the opinion of the IPMC) a podling
has learned enough and is responsible enough to sustain itself
as such a community."
It's just that to an outsider, it is totally unclear what that
means. And it may be impossible to really convey the meaning
of it in a few sentences. So why don't we just say so.
Make it absolutely clear that the diversity of the community
will be judged by the IPMC based on the overall conduct of the
project, mailing list, commit activity etc. I know it's there
already, but it could be reinforced. Perhaps at the end of the
"Creating an Open and Diverse community" (community should be
capitalized, btw) paragraph: "The IPMC will judge diversity of
the project based on many criteria. These include mailing list
activity, commit activity and the affiliations of the committers.
There is no single sufficient criterion, it is the overall conduct
of the development community that counts." Or something like that.
Another thing that might be helpful is to advise podlings to
engage in the incubator community. The graduation guide sounds
a bit like general@incubator is a place where you just go for
help or questions. I think that's actually misleading. We should
strongly advise podling committers to subscribe to and follow
general@incubator. It's the best way (the only way?) to understand
what it takes to graduate.
So I would change "Please post any questions about graduation to
the general incubator list" to "Subscribe to and closely follow
the general incubator list. It is *the* place to learn about
graduation votes and current policies and their interpretation.
It is also where you can post any questions about graduation."
If this goes in the right direction, I'll propose a patch.
--Thilo
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: [DISCUSS] Community diversity (again)
Posted by Matthieu Riou <ma...@offthelip.org>.
On Wed, Mar 26, 2008 at 5:03 PM, Robert Greig <ro...@gmail.com>
wrote:
> I think this is an important topic for future incubator project groups
> to have clarified.
>
> > "The project is not highly dependent on any single contributor (there
> are at
> > least 3 legally independent committers and there is no single company or
> > entity that is vital to the success of the project)"
>
> Perhaps the word "legally" needs to be removed since from the Qpid
> discussion it would appear that several people do not think a strict
> legal interpretation should apply?
>
> However, given that the intent (as I understand it) is to avoid the
> case where a project dies because one entity withdraws funding,
> perhaps some definition along the lines of "for people who are paid to
> contribute to the project, no single entity remunerates more than 50%
> of the committers"?
>
The thing is, for both for Qpid and Tuscany, there *were* 3 independent
committers. And in the case of Qpid, I believe the "no single entity
remunerates more than 50% committers" would have worked as well.
Matthieu
> Robert
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
Re: [DISCUSS] Community diversity (again)
Posted by Luciano Resende <lu...@gmail.com>.
Another point that needs discussion and more clarification is related
to who should be counted towards diversity of a project. From [1], it
states that a podling need at least 3 legally independent committers,
but from recent comments on the QPid graduation thread [2], it seems
that only PMC members are being considered.
[1] http://incubator.apache.org/incubation/Incubation_Policy.html#Graduating+from+the+Incubator
[2] http://www.mail-archive.com/general%40incubator.apache.org/msg17111.html
On Wed, Mar 26, 2008 at 5:03 PM, Robert Greig <ro...@gmail.com> wrote:
> I think this is an important topic for future incubator project groups
> to have clarified.
>
>
> > "The project is not highly dependent on any single contributor (there are at
> > least 3 legally independent committers and there is no single company or
> > entity that is vital to the success of the project)"
>
> Perhaps the word "legally" needs to be removed since from the Qpid
> discussion it would appear that several people do not think a strict
> legal interpretation should apply?
>
> However, given that the intent (as I understand it) is to avoid the
> case where a project dies because one entity withdraws funding,
> perhaps some definition along the lines of "for people who are paid to
> contribute to the project, no single entity remunerates more than 50%
> of the committers"?
>
> Robert
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
--
Luciano Resende
Apache Tuscany Committer
http://people.apache.org/~lresende
http://lresende.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: [DISCUSS] Community diversity (again)
Posted by Robert Greig <ro...@gmail.com>.
I think this is an important topic for future incubator project groups
to have clarified.
> "The project is not highly dependent on any single contributor (there are at
> least 3 legally independent committers and there is no single company or
> entity that is vital to the success of the project)"
Perhaps the word "legally" needs to be removed since from the Qpid
discussion it would appear that several people do not think a strict
legal interpretation should apply?
However, given that the intent (as I understand it) is to avoid the
case where a project dies because one entity withdraws funding,
perhaps some definition along the lines of "for people who are paid to
contribute to the project, no single entity remunerates more than 50%
of the committers"?
Robert
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org