incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Christian Grobmeier <>
Subject Re: should podlings have informal chairs?
Date Tue, 22 Nov 2011 14:08:33 GMT
On Tue, Nov 22, 2011 at 2:03 AM, Joe Schaefer <> wrote:
> To me a lot of the problem stems from the fact that the reports are
> misdirected- instead of informing the board about the activities of
> the IPMC, it tells them about the podling's activities, which doesn't
> scale properly.
> We should be reporting to the board about OUR work, not the work of
> the podlings.  Podlings should only be brought in for a few specific
> examplesto mention.  That's the first thing to correct.

I really don't see a huge report actually. What are we doing?
- we vote podlings in and do some discussions on them. Thats already
on the report
- from time to time we work on the website. This is/was on the report too.

So what else do we do?

Mentoring, of course. What would you like to report here, what is not
on the podlings report? Actually Mentors don't do much stuff if we
look honest at it. Maybe help with graduation or starting with
incubation, writing some emails with "how it all works" and looking at
the releases. This is not so interesting alone.

> Once we start reporting about the crap WE did, then we can start figuring
> out all the crap that's not getting done by mentors who aren't participating.

How can we see this? Mentors who don't contribute to a report?

> It doesn't matter that there are lots of well-intentioned but otherwise useless
> people mentoring projects, the fact is that they only harm the org by not prodding
> these projects along a graduation path or funneling them towards the exit door.  Part
> of how they manage to get away with that is that we pretend its important to a podling
> to create a sustainable community around itself, which is something most of them
> have no control over.  That is the reason for the long bouts of stalling on many
> levels, we need to do the sane thing and drop that bit of pretense, and yes even
> graduating projects that haven't necessarily met the silly developer diversity
> requirements- rules are not appropriate here, only very fuzzy benchmarks.

I agree on the fuzzy benchmarks.

But how can you, as mentor, help with community building (hope I got
this right). Just because a mentor says "thou shalt build up a
community" this will not cause people to do it. I have had tried that
out, but it doesn't work. It depends on the podling people, not on the

> WE are responsible for evaluating the progress of our podlings, ALL of them, and
> clutch can help us do that at a basic level as a group.  But we need to figure out,
> quickly, how to change the review process for podling reports in a scalable way
> without us all being burnt out all at once.


> I think the review needs to take place
> over a few days, on the podling's own dev lists, by 3 IPMC members actively voting
> on them.  We can still collate the podling reports on the wiki, but the report we
> hand to the board should come from us, and it should be the product of those reviews.

How does that look like? A podling says, we work to increase community
participation. How can one say this is not the case - because what
actually does "work on increasing community participation" mean. Other
thing mentioned often is: "we work towards a new release". I doubt a
mentor will find different, if a podling writes this on the report.

> We can do this wiki-style if we want to, or just have Noel poll this list for "mentor
> comments" to be included in the report.

+1 on "Mentor Comments" on the podling reports, but optional. What
should a mentor write if everything on the report is ok? Actually I
read and correct my podlings reports before I sign them. I guess this
should be enough.

> A quick scan of the podling lists wrt those
> report votes should be sufficient to determine if a podling needs more IPMC representation,
> and can be done by Noel or collectively if we'd like to start doing more cross-checking.

Reading this, I don't think we should rely on more bureaucracy. But I
think you mentioned something important, the fuzzy benchmarks. There
are some we can make more critical, like: sign copyright stuff within
3 months or leave. Something like that.

You also mentioned that the IPMC should take care on the podlings as a
whole. I like this idea. Actually I have helped on the Wave project
even when not a mentor to it. The IPMC is the only project I know at
the ASF which is not really community. People join the IPMC when they
see an interesting project and leave without a further words
sometimes. Actually we have more than 100 IPMC members, but only a few
active (see the recent threads and you know who is active).

This leads me to the impression, that not everybody familiar with the
Apache way (aka Committers, Members) must join the IPMC to mentor a
project. Yes, it is about voting. Lets assume 3 non-ipmc mentors would
vote +1 on a release. The same vote does happen on general@ usually
were IPMC members should agree too. As the votes are backuped on
general@, why is it necessary to have these 3 mentors on the IPMC?
After graduation, these 3 people are probably not interested in doing
anything else here.

This would lead to a smaller IPMC with people interested in the
incubator as a whole. These small group is the actual community here
and they speak about processes, websites, reports and so on. And they
are required finally to reread the reports of the podlings.


> ----- Original Message -----
>> From: Benson Margulies <>
>> To:
>> Cc:
>> Sent: Monday, November 21, 2011 7:42 PM
>> Subject: Re: should podlings have informal chairs?
>> On Mon, Nov 21, 2011 at 7:17 PM, Sam Ruby <> wrote:
>>>  On Mon, Nov 21, 2011 at 6:13 PM, Benson Margulies
>> <> wrote:
>>>>  Sam,
>>>>  Do you see any validity in my theory that the ipmc is so large and
>>>>  diffuse as to be directionless?
>>>  I don't see that as a necessary consequence.  The ASF is large and
>>>  diffuse, yet each month we pretty consistently get 6+ Directors to
>>>  review and sign off on each report.  The board is careful to not set
>>>  technical direction, but we do create and track action items, and work
>>>  to make sure that the individual PMCs are self-governing and get the
>>>  help that they need from the relevant board committees.
>> Compare, if you would, the board of six to the ipmc. There aren't six,
>> or sixteen, ipmc members who feel it's their job to review every PPMC
>> report before the whole business goes to the board. There's a chair,
>> who due to his volunteer status like the rest of us, shows more or
>> less engagement with the goings-on on this mailing list at different
>> times.
>> The ipmc more or less delegates to the mentors, and passes the PPMC
>> reports up to the board, with not much digestive activity in between.
>> In this sense I guess I'm trying to agree with you, but I wonder how
>> to get a giant committee of people, most of whom signed up just to
>> mentor one project, to actually step up and exercise more oversight.
>> Of course we've got a few people like Sebb who try to stay on top of
>> everything.
>> Since there are only six board members, they all know that they,
>> themselves, have to read this stuff and think about it. If there were
>> 106, I doubt that anything would get attended to unless a subset were
>> somehow tasked. So I suppose that I'm trying to float the idea that,
>> somehow, less than the full ipmc needs to focus. I suppose that the
>> committee could create a category of meta-mentor, and people who sign
>> up for that role would be signing up to read all the reports and
>> perhaps even look over the shoulders a bit of the projects.
>> Should I believe
>> that
>> there are 878 ipmc members, or is this some sort of ldap artifact?
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> For additional commands, e-mail:
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:


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

View raw message