Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Apr 3, 2013, at 1:20 AM, Bertrand Delacretaz wrote: On Tue, Apr 2, 2013 at 11:18 PM, Benson Margulies bimargul...@gmail.com wrote: ...Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic... That's my case, and I'm not interested in arguing this much more - deconstructing the Incubator PMC does not look like a good idea to me, both as a board member and as an ASF member. I am interested in graduating podlings into TLPs when ready. I think the incubator is an excellent place that has done tremendous efforts in fulfilling the mission of the ASF - software for the public good. It pains me to continue to read Chris's continual efforts to disband. It is wearying and demotivating. Upayavira made an effort to just discuss problems, yet Chris continues to push his poison pill solution. Now I'll do my best to do the shepherd thing and look at long time podling's like VXQuery - which is slowly working towards their second release and probably needs a new Mentor. I have less time this year for the IPMC, I am not sure I have time to Mentor for awhile, but Shepherding is fun. It's time for some fledgling's to be pushed out of the nest. Regards, Dave -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Wed, Apr 3, 2013 at 3:25 PM, Ross Gardler rgard...@opendirective.comwrote: On 3 April 2013 14:41, ant elder ant.el...@gmail.com wrote: On Wed, Apr 3, 2013 at 2:12 PM, Noah Slater nsla...@apache.org wrote: Thanks for the clarification, Ant. Is the documentation ignored? Whenever I look through it, it seems like the problem is that it is incomplete and confusing. It's hardly a wonder people disagree. ;) (This is just a bit of rhetoric. I hardly mean to imply the documentation is responsible for the whole problem...) Yep I don't know that ignored is the best word, and i agree the doc can be incomplete and confusing. For another example take the minimum graduation requirements documented on the policy page: 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) - http://incubator.apache .org/incubation/Incubation_Policy.html#Graduating+from+the+Incubator Great example - it's reasonably clear but incorrect (as well as being imprecise as you illustrate). We don't require a minimum of 3 independent committers. We require a community that doesn't exclude anyone. I don't have the time to look it up but there was quite some discussion about this point some time ago. I seem to remember the IPMC agreeing the docs need to be updated. Ross That would be further evidence that the doc is often ignored right? (Would be interested in a link if you/anyone can find it, to see if a decision was clearly made about this) ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 2:26 AM, Dave Fisher dave2w...@comcast.net wrote: On Apr 3, 2013, at 1:20 AM, Bertrand Delacretaz wrote: On Tue, Apr 2, 2013 at 11:18 PM, Benson Margulies bimargul...@gmail.com wrote: ...Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic... That's my case, and I'm not interested in arguing this much more - deconstructing the Incubator PMC does not look like a good idea to me, both as a board member and as an ASF member. I am interested in graduating podlings into TLPs when ready. I think the incubator is an excellent place that has done tremendous efforts in fulfilling the mission of the ASF - software for the public good. It pains me to continue to read Chris's continual efforts to disband. It is wearying and demotivating. Chris is jumping towards the end result. Don't get upset by that. The simpler answer: *try* his new approach on a singular podling basis. That can run in parallel to the Incubator. The Board can easily absorb one special incubation project. There is an entirely separate discussion of what that really means, how much Board-offloading is performed, what kinds of mentoring/input is provided, etc. But the short answer is that we can run trials *without* dismantling the Incubator. Consider it. Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Tue, Apr 2, 2013 at 5:18 PM, Benson Margulies bimargul...@gmail.com wrote: Ant is reflecting a real dilemma here. At Apache, we try to be egalitarian, and we try to work by consensus. The natural conclusion is that the many people needed to vote on releases are also part of the decision-making body for policy that controls those releases. Please step back to the orignal roots of the problem here. It may offer alternate solutions. In essence, the *BOARD* needs to approve every release from the ASF. The Board delegates this responsibility to reach scale. Those delegates, operating in the Board's name, will then approve the releases. The problem at hand: find and designate those delegates. It doesn't matter how. The current solution is the release votes have been delegated to (I)PMC Members, and it requires three to establish ASF approval. Maybe there is another path. Talk about it, run it by the Board, and get it approved. I believe there are *many* solutions. Just fine one that works in this large, variant environment. ... Ross' other proposal :-), to move documentation (and thus some/much of the locus of policy decision) making to comdev, reduces the load of decision-making that the IPMC has to find consensus on, and thus proposes to reduce the stress. Documentation about the ASF *should* move to ComDev. It is history that has left that doc scattered around. Its logical (and eventual) home lies with ComDev. I forsee ComDev as a group who documents who we are, and how we work, and ... I forsee the Incubator as the *mechanism* of teaching what ComDev has explained. ... Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013 at 8:20 PM, Ross Gardler rgard...@opendirective.com wrote: On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. I believe this is simply a matter of training and mentor oversight. If a podling required three Members to sign off, and the report required your points... then we (the Board) might actually have better insight than some TLPs. The Board has 50+ reports to review each month. Thankfully, there hasn't been much push back on that yet. We seem to be keeping up (the shepherd/comment system helps us). Throwing in some podlings shouldn't upset us, as it actually drops the [giant] Incubator report down to (maybe?) empty. Ross says the Board pays less attention to these (by implication) than say the 137 TLPs at present. Ross is one Director. Good for him. I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm active in the IPMC and thus have more visibility. That doesn't mean they should be expected to by me or by anyone else. If we alter the incoming-project mechanism, then yes: maybe we *should* expect the Directors to read the reports with a little more attention. But if we demand that N ASF Members track the podling, and approve the report, then sure... the Board may be able to delegate/slack a little bit on those reports. Point is: the Incubator is not the only solution here. Think about other options. Maybe the Board can accept the podling, and designate some pseudo-VPs to be held responsible? I know other directors (Greg IIRC at least) didn't want the Incubator specific podling reports to go away (and to only have the summary at the top of the Incubator report). I don't think any of the Directors want them to go away. But board reports are not what the IPMC is about. That is the reporting process within the foundation and provides the level of oversight into the PPMCs that the board requires. But the IPMC does *much* more than submit a monthly board report with a verbatim copy of the podlings individual reports. Agreed! And this is a very important point that seems to be left behind a bit. I would counter that the IPMC doesn't tend to satisfy this oversight/educational role consistently well. In the end, it simply depends upon the Mentors' attention. There are very few (none?) solutions to that basic problem. ... Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
As I see it, the incubator as we have it is a mechanism for coping with the lack of mentor commitment. As Ross often writes, it's easy to say that Mentors *should* make this commitment, but mentors are volunteers, and things happen. Upayavira wonders if Mentor 'harvest glory' and then wander away. I think it's more likely that they do the best they can, given the constraints of their lives. If three members showed up and were willing to state the necessary level of commitment to a new project to be a TLP, sure, let them be a TLP instantly. But if that's the standard for starting a new project at Apache, I predict that we'll start very few new projects. This leads me to a question back to Greg: what do you want to do if a new project has troubles: a Mentor has a kid, or a new job, or whatever? Shut it down? I'd suspect that you'd hope to recruit a replacement, but that's a messy procedure for the board to be stuck with. I end up thinking that this looks like a _reductio_ argument that leads back to the IPMC. How about the following more incremental experiment: we do what Upayavira says: we set a higher bar for mentors at podling start time. We ask them to make a public statement of commitment that for some period of time (six months) they commit to thinking of themselves _as a PMC_, not just as some sort of diffuse advisors or coaches. I expect that this will make it harder to start podlings, and I think that this constraint would reflect reality. Such a group could then graduate as soon as it picked up a few more PMC members and did a release. If we stick to this, maybe the IPMC will wither away, or just shrink to a smaller group. I think that this experiment comes before Greg's, as we should see if anyone will make this claim, and whether they live up to it, before we launch such a group into the exosphere of pure board supervision. Ultimately, this might mean that we retire the word 'mentor' and replace it with, well, 'PMC member'. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Hi, On Thu, Apr 4, 2013 at 1:42 PM, Benson Margulies bimargul...@gmail.com wrote: ...How about the following more incremental experiment: we do what Upayavira says: we set a higher bar for mentors at podling start time. We ask them to make a public statement of commitment that for some period of time (six months) they commit to thinking of themselves _as a PMC_, not just as some sort of diffuse advisors or coaches... I like that - I'd say 3 months for the initial commitment, and ask mentors to indicate in the project proposal how many hours per week they think they can dedicate to the task. And also make sure each podling has a champion who agrees to fulfill the during incubation role listed at http://incubator.apache.org/incubation/Roles_and_Responsibilities.html#Champion So far we tried to make sure sufficient mentor power is available by having 2 to 3 mentors with vague requirements - with this new model a champion and a mentor can be sufficient if they are clearly engaged and available, and if the champion raises alarms if that's not the case anymore. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On 4 April 2013 09:06, Greg Stein gst...@gmail.com wrote: On Sun, Mar 31, 2013 at 8:20 PM, Ross Gardler rgard...@opendirective.com wrote: On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. I believe this is simply a matter of training and mentor oversight. That is the key issue. I can name many really good mentors. The problem is that prior to the new processes introduced by Jukka we had a great many projects that stagnated because of inattentive mentoring. The current IPMC reporting process picks those up and addresses them internally within the IPMC. This is the reason that we have seen more podlings graduate in the last year. If we remove that aspect of the IPMCs oversight then who will catch these projects that don't have mentors actively looking after them? It will be the boards responsibility to do that. I contest that this does not scale. We need a solution that will scale appropriately whilst also removing the inefficiencies introduced by a large IPMC. Ross says the Board pays less attention to these (by implication) than say the 137 TLPs at present. Ross is one Director. Good for him. I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm active in the IPMC and thus have more visibility. That doesn't mean they should be expected to by me or by anyone else. If we alter the incoming-project mechanism, then yes: maybe we *should* expect the Directors to read the reports with a little more attention. But if we demand that N ASF Members track the podling, and approve the report, then sure... the Board may be able to delegate/slack a little bit on those reports. In principle this is fine, but in principle the IPMC already demands that N ASF Members track the podling so what is changing? Point is: the Incubator is not the only solution here. Think about other options. Maybe the Board can accept the podling, and designate some pseudo-VPs to be held responsible? OK. This seems to be similar to my overlapping proposal in a different message to allow the board to sponsor podlings. Are we onto something here? I know other directors (Greg IIRC at least) didn't want the Incubator specific podling reports to go away (and to only have the summary at the top of the Incubator report). I don't think any of the Directors want them to go away. But board reports are not what the IPMC is about. That is the reporting process within the foundation and provides the level of oversight into the PPMCs that the board requires. But the IPMC does *much* more than submit a monthly board report with a verbatim copy of the podlings individual reports. Agreed! And this is a very important point that seems to be left behind a bit. I would counter that the IPMC doesn't tend to satisfy this oversight/educational role consistently well. In the end, it simply depends upon the Mentors' attention. There are very few (none?) solutions to that basic problem. I don't think I fully agree with that today (12 months ago I would agree). The shepherding process has caught a great many situations where mentors are inattentive and has addressed them directly. II do agree it can be further improved (hence my original proposal as one way of doing this). Ross
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 9:42 AM, Upayavira u...@odoko.co.uk wrote: Just a thought. Chris' solution says 'make mentors the initial PMC'. They vote in other project team members as appropriate to be peers. This creates a positive egalitarian setup which mirrors that of a PMC, which is a good thing. There was a poddling a while ago that had this approach of having the initial PPMC be just the mentors. It ended in an argument when the PPMC/mentors voted in a new committer that the other committers weren't so keen on, and it ended with the PPMC rebooted to add in all the initial committers. That approach also goes against the change we made so that poddlings could vote in their own committers without needing binding votes from Incubator PMC members. ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 2:08 PM, Ross Gardler rgard...@opendirective.comwrote: Having said that, here's an idea that builds on your proposal. There is already the opportunity to name the board as the sponsoring organisation. Why not say where the board is willing to sponsor the project it can go straight to TLP (e.g. exactly as Apache Steve did). This would allow a larger scale experiment around Chris' proposal but provides the opportunity for the board to control how much of the oversight role is pushed towards it and how much remains with the IPMC. Ross If the board are ok with that experiment then i think it sounds like an fine thing to try. ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On 4 April 2013 08:46, Greg Stein gst...@gmail.com wrote: On Thu, Apr 4, 2013 at 2:26 AM, Dave Fisher dave2w...@comcast.net wrote: On Apr 3, 2013, at 1:20 AM, Bertrand Delacretaz wrote: On Tue, Apr 2, 2013 at 11:18 PM, Benson Margulies bimargul...@gmail.com wrote: ...Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic... That's my case, and I'm not interested in arguing this much more - deconstructing the Incubator PMC does not look like a good idea to me, both as a board member and as an ASF member. I am interested in graduating podlings into TLPs when ready. I think the incubator is an excellent place that has done tremendous efforts in fulfilling the mission of the ASF - software for the public good. It pains me to continue to read Chris's continual efforts to disband. It is wearying and demotivating. Chris is jumping towards the end result. Don't get upset by that. The simpler answer: *try* his new approach on a singular podling basis. That can run in parallel to the Incubator. The Board can easily absorb one special incubation project. There is an entirely separate discussion of what that really means, how much Board-offloading is performed, what kinds of mentoring/input is provided, etc. But the short answer is that we can run trials *without* dismantling the Incubator. I'm all for such tests. If it were not for this oversight role of the IPMC then Chris' plan would be a fine one. However, it seems that almost every month the board looks at an issue in the IPMC reports and says that's an IPMC issue, they seem to be handling it well so lets move on. Without the IPMC those conversations would be different. A single test of a single project would not highlight this. Having said that, here's an idea that builds on your proposal. There is already the opportunity to name the board as the sponsoring organisation. Why not say where the board is willing to sponsor the project it can go straight to TLP (e.g. exactly as Apache Steve did). This would allow a larger scale experiment around Chris' proposal but provides the opportunity for the board to control how much of the oversight role is pushed towards it and how much remains with the IPMC. Ross Consider it. Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 7:42 AM, Benson Margulies bimargul...@gmail.com wrote: As I see it, the incubator as we have it is a mechanism for coping with the lack of mentor commitment. As Ross often writes, it's easy to say that Mentors *should* make this commitment, but mentors are volunteers, and things happen. Upayavira wonders if Mentor 'harvest glory' and then wander away. I think it's more likely that they do the best they can, given the constraints of their lives. If three members showed up and were willing to state the necessary level of commitment to a new project to be a TLP, sure, let them be a TLP instantly. But if that's the standard for starting a new project at Apache, I predict that we'll start very few new projects. The only commitment necessary is to take *care* of the project. The ASF has no specific demand about velocity of a project. As long as somebody is caring for it, then we let the community continue. I have no problem with three Members starting arbitrary projects. This leads me to a question back to Greg: what do you want to do if a new project has troubles: a Mentor has a kid, or a new job, or whatever? Shut it down? I'd suspect that you'd hope to recruit a replacement, but that's a messy procedure for the board to be stuck with. Not the Board's problem. The project is responsible for finding a replacement, in order to continue, in order to get out of probation status (or, podling status, or whatever). I end up thinking that this looks like a _reductio_ argument that leads back to the IPMC. Nope. The IPMC was created to properly handle incoming projects, in terms of community and IP. At the time, all podlings were sponsored by an existing TLP. We've jettisoned umbrellas, so now all podlings have no particular sponsor. But. The concept was sponsoring TLP provides manpower to help the podling. No sponsor? No manpower. Oops. That's where we are today. How about the following more incremental experiment: we do what Upayavira says: we set a higher bar for mentors at podling start time. We ask them to make a public statement of commitment that for some period of time (six months) they commit to thinking of themselves _as a PMC_, not just as some sort of diffuse advisors or coaches. Eh? We don't expect this kind of commitment from *anybody*. I have never said if you want to join the Subversion PMC, then you MUST make a public declaration of a six month commitment. Upayavira talked about *ability* .. not *commitment*. I expect that this will make it harder to start podlings, and I think that this constraint would reflect reality. Such a group could then graduate as soon as it picked up a few more PMC members and did a release. If we stick to this, maybe the IPMC will wither away, or just shrink to a smaller group. I think that this experiment comes before Greg's, as we should see if anyone will make this claim, and whether they live up to it, before we launch such a group into the exosphere of pure board supervision. Ultimately, this might mean that we retire the word 'mentor' and replace it with, well, 'PMC member'. before Greg's? Why? If I want to be really antagonistic, I might point out that the decision is the Board's :-) (and I might suggest to podlings out there: hey, why not... petition the Board to try a new approach; can't hurt; worst is they'll say no; but you better have lots of Members handy) Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 9:22 AM, Ross Gardler rgard...@opendirective.com wrote: On 4 April 2013 09:06, Greg Stein gst...@gmail.com wrote: On Sun, Mar 31, 2013 at 8:20 PM, Ross Gardler rgard...@opendirective.com wrote: On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. I believe this is simply a matter of training and mentor oversight. That is the key issue. I can name many really good mentors. The problem is that prior to the new processes introduced by Jukka we had a great many projects that stagnated because of inattentive mentoring. The current IPMC reporting process picks those up and addresses them internally within the IPMC. This is the reason that we have seen more podlings graduate in the last year. If we remove that aspect of the IPMCs oversight then who will catch these projects that don't have mentors actively looking after them? It will be the boards responsibility to do that. I contest that this does not scale. We need a solution that will scale appropriately whilst also removing the inefficiencies introduced by a large IPMC. The Board easily deals with this. Today, we look to the VP to give us a report. Let's say that a provisional/podling/probationary TLP requires (3) Members (mentors) to sign off on each report. If a report fails to receive those three sign-offs, then it does not get accepted. Simple as that. The project then needs to hit up (say) general@community.a.o to troll for new Members/mentors to fill the slacker slot that occurred. I see no undue burden on the Board in this approach. Ross says the Board pays less attention to these (by implication) than say the 137 TLPs at present. Ross is one Director. Good for him. I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm active in the IPMC and thus have more visibility. That doesn't mean they should be expected to by me or by anyone else. If we alter the incoming-project mechanism, then yes: maybe we *should* expect the Directors to read the reports with a little more attention. But if we demand that N ASF Members track the podling, and approve the report, then sure... the Board may be able to delegate/slack a little bit on those reports. In principle this is fine, but in principle the IPMC already demands that N ASF Members track the podling so what is changing? Because there is slack space. gee. a mentor didn't sign off on the podling report is very different from the Board saying rejected. get/find your Members to sign off. we will close you in six months if we see no valid reports by then. I also believe that the Board is more active than the IPMC. The Incubator shepherd process (modeled after our Board shepherds) has brought out the *active* IPMC Members. Those correlate to the Directors -- they are active in the PMC-level concerns. They have dedication to the Incubator aspect of our Foundation, yet I don't think they provide as much coverage (yet!) as the Directors. If that aspect of the Incubator expanded, then we'd likely be in great shape. But I'd also see that as Board replacement, and something I believe the Board is quite capable of doing anyways. etc etc. The short answer in my mind is: podlings need to learn how to report to the Board. So fine: make them do that. And make them have N Members on staff to avoid the worst of problems, which may really waste the Board's time. (but even then, the Board is quite good at marking a report as bogus via the comment section, and later reviewers learning to skip, or detail review, or whatever; a minor change in our process/time) And in the end, the Board adapts to the needs of the Foundation. Believe me, when I was Chairman and spent two hours go over reports on the conference call... something had to break. *THAT* was non-scalable. We're nowhere near that kind of hell. ... Cheers, -g - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Benson writes: We ask them to make a public statement of commitment that for some period of time (six months) they commit to thinking of themselves _as a PMC_, not just as some sort of diffuse advisors or coaches... +1 to the change of mentality. Bertrand replies: I like that - I'd say 3 months for the initial commitment, and ask mentors to indicate in the project proposal how many hours per week they think they can dedicate to the task. The task that I think it's very important for the initial group of Mentors to steward through to completion is the absorption of the code base and the approval of the first incubating release. Freelance IPMC members performing release reviews can check whether files have the necessary license headers, but can't see how those headers got there, whether all copyright relocations were performed appropriately, and so on. If we're not entirely comfortable asking candidate Mentors to make specific individual time commitments, there's an alternative: emphasize that the initial Mentors **as a group** are signing up to supervise phase 1 of incubation, which involves IP clearance and concludes with a successful vote on the first incubating release. If Mentors fall away after phase 1 ends, it's less of a problem. Replacing Mentors is less consequential once the code base has reached the known good state of having made it through the release process. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Thu, Apr 4, 2013 at 4:39 PM, Marvin Humphrey mar...@rectangular.com wrote: ...If Mentors fall away after phase 1 ends, it's less of a problem. Replacing Mentors is less consequential once the code base has reached the known good state of having made it through the release process Agreed, though it's good to still make sure a Champion and/or Shepherd is watching over the podling to make sure it still gets adequate supervision. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Sent from a mobile device, please excuse mistakes and brevity On 4 Apr 2013 15:17, Greg Stein gst...@gmail.com wrote: On Thu, Apr 4, 2013 at 9:22 AM, Ross Gardler rgard...@opendirective.com wrote: On 4 April 2013 09:06, Greg Stein gst...@gmail.com wrote: On Sun, Mar 31, 2013 at 8:20 PM, Ross Gardler rgard...@opendirective.com wrote: On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. I believe this is simply a matter of training and mentor oversight. That is the key issue. I can name many really good mentors. The problem is that prior to the new processes introduced by Jukka we had a great many projects that stagnated because of inattentive mentoring. The current IPMC reporting process picks those up and addresses them internally within the IPMC. This is the reason that we have seen more podlings graduate in the last year. If we remove that aspect of the IPMCs oversight then who will catch these projects that don't have mentors actively looking after them? It will be the boards responsibility to do that. I contest that this does not scale. We need a solution that will scale appropriately whilst also removing the inefficiencies introduced by a large IPMC. The Board easily deals with this. Today, we look to the VP to give us a report. Agreed. Let's say that a provisional/podling/probationary TLP requires (3) Members (mentors) to sign off on each report. If a report fails to receive those three sign-offs, then it does not get accepted. Simple as that. Well I've proposed we require mentor signoff in the past. It was rejected because our mentors are volunteers. I proposed we require shepherd signoff. It was rejected because our shepherds are volunteers. I, and apparently you Greg, don't think think it is unreasonable to expect the mentors to take collective responsibility like this. I would continue to support this idea. Doing so addresses my concerns about the missing oversight in Chris' proposal because the board need not visit IPMCs to verify the report is accurate (as shepherds do now). ... I also believe that the Board is more active than the IPMC. The Incubator shepherd process (modeled after our Board shepherds) has brought out the *active* IPMC Members. Those correlate to the Directors -- they are active in the PMC-level concerns. They have dedication to the Incubator aspect of our Foundation, yet I don't think they provide as much coverage (yet!) as the Directors. If that aspect of the Incubator expanded, then we'd likely be in great shape. Agreed. This was a large part of the reasoning behind my proposal to formally recognise shepherds. But I concede that your modification/clarification of Chris' proposal makes it viable and even preferable to my own proposal (assuming ComDev coverage of the non-oversight aspects). Ross
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, April 4, 2013 6:22 AM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On 4 April 2013 09:06, Greg Stein gst...@gmail.com wrote: On Sun, Mar 31, 2013 at 8:20 PM, Ross Gardler rgard...@opendirective.com wrote: On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. I believe this is simply a matter of training and mentor oversight. That is the key issue. I can name many really good mentors. The problem is that prior to the new processes introduced by Jukka we had a great many projects that stagnated because of inattentive mentoring. The current IPMC reporting process picks those up and addresses them internally within the IPMC. This is the reason that we have seen more podlings graduate in the last year. I see it a teeny bit differently (though later emails from you and Greg seem to have brought our ideas into alignment). You directly equate Jukka's processes with graduating so many podlings. While I think Jukka's processes were great, they were a means to an ends -- they along with some key IPMC members who are active (you will see later that you are in that short list of active ones ^_^) are the reasons more podlings graduated. Coupled with Joe's experiment, and coupled with the removal of the IPMC 3 +1s for releases which both came before Jukka's time. Another key was the clarification of the role of Champion and Champions really stepping up. Note that those burdens being removed are precisely the initial steps towards the removal of the meta committee that is the IPMC since both steps in effect reduced the power of the umbrella to stall and stagnate podlings. I went back, starting in March 2012 [1] when Jukka took over to cull a list of shepherds and active mentors that signed off on at least 1 report for the IPMC (before there were shepherds). Here is the list and tallies for each month that the mentors signed off on at least 1 report. Note in tallies below I count the mentor or shepherd Nx per month so if they signed N multiple reports, they still get a count of N. I went ahead and uploaded these scripts to [2] in case folks are interested in how I tallied (note I also removed some nonsense from these files by hand mainly stop words since I didn't do a ton of data cleansing): Mentors [ tallies per month since March 2012] --- 18 rgardler 18 bdelacretaz 16 mattmann 15 phunt 15 kevan 13 tomwhite 9 jukka 9 jim 7 greddin 7 cdouglas 7 adc 7 Alan 6 joes 6 bodewig 6 ate 5 wave 5 tommaso 5 omalley 5 olamy 5 elecharny 4 simonetripodi 4 lresende 4 hwright 4 gstein 4 gianugo 4 Gates 3 struberg 3 nick 3 mnour 3 jbonofre 3 coheigea 3 Struberg 3 Petracek 3 Mark 3 Gerhard 3 Cabrera 2 wavw 2 twilliams 2 rfrovarp 2 marrs 2 ddas 2 cutting 2 berndf 2 ant 2 Ralph 2 Olivier 2 Lamy 2 Goers 2 Devaraj 2 Das 2 (struberg) 2 (rgoers) 2 (gates) 1 yegor 1 wrowe 1 thorsten 1 rfeng 1 mfranklin 1 line 1 jvermillard 1 grobmeier 1 generic 1 dkulp 1 dennisl 1 dashorst 1 brett 1 bmargulies 1 asavory 1 Williams 1 Upayavira 1 Tim 1 Reddin 1 Martijn 1 Lundberg 1 Greg 1 Fisher 1 Dennis 1 Dave 1 Dashorst 1 (wave) 1 (greddin) 1 (gates@) If we cut off the above at 3 sign offs or more, we see that there are 31 mentors that fit that criteria. If we say at least 6 sign offs on reports in the last year (averaging less than 1 sign off every 2 months) then that number drops to 15. The point being that whatever number we pick any of those 15-31 (or whatever N) mentors could simply be considered candidates for these new VPs for incoming projects without an Incubator whilst the incoming projects are learning the Apache way from their 3 ASF members and others in the community. In fact, this is really what the role of the Champion is now. Sort of a provisional podling VP until the incoming project community VP (aka real VP) is elected. If we remove that aspect of the IPMCs oversight then who will catch these projects that don't have mentors actively looking after them? It will be the boards responsibility to do that. I contest that this does not scale. We need a solution that will scale appropriately whilst also removing the inefficiencies introduced by a large IPMC. Ross says the Board pays less attention to these (by implication) than
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Tue, Apr 2, 2013 at 11:18 PM, Benson Margulies bimargul...@gmail.com wrote: ...Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic... That's my case, and I'm not interested in arguing this much more - deconstructing the Incubator PMC does not look like a good idea to me, both as a board member and as an ASF member. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Tue, Apr 2, 2013 at 5:26 PM, Noah Slater nsla...@apache.org wrote: As far as I understand your comment, Ant, you mean to say that he problem is that there is too much variation in opinion and approach. (Primarily, I understand, in relation to releases.) Hi Noah, i suggested that one of the problems was the variation in opinion and in who happens to decide to be active at particular moment means it can be hard to tell what the reaction will be to any particular action, and once there is a disagreement the diversity of opinion means it can be hard to find any consensus. I didn't offer any solutions yet, just getting some agreement on what the issues are first would be good. But I'm not convinced more doc is going to help this much, and moving the doc to be under the control of comdev would IMHO just make the doc even more ignored than it is today. ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Thanks for the clarification, Ant. Is the documentation ignored? Whenever I look through it, it seems like the problem is that it is incomplete and confusing. It's hardly a wonder people disagree. ;) (This is just a bit of rhetoric. I hardly mean to imply the documentation is responsible for the whole problem...) On 3 April 2013 09:29, ant elder ant.el...@gmail.com wrote: On Tue, Apr 2, 2013 at 5:26 PM, Noah Slater nsla...@apache.org wrote: As far as I understand your comment, Ant, you mean to say that he problem is that there is too much variation in opinion and approach. (Primarily, I understand, in relation to releases.) Hi Noah, i suggested that one of the problems was the variation in opinion and in who happens to decide to be active at particular moment means it can be hard to tell what the reaction will be to any particular action, and once there is a disagreement the diversity of opinion means it can be hard to find any consensus. I didn't offer any solutions yet, just getting some agreement on what the issues are first would be good. But I'm not convinced more doc is going to help this much, and moving the doc to be under the control of comdev would IMHO just make the doc even more ignored than it is today. ...ant -- NS
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Wed, Apr 3, 2013 at 2:12 PM, Noah Slater nsla...@apache.org wrote: Thanks for the clarification, Ant. Is the documentation ignored? Whenever I look through it, it seems like the problem is that it is incomplete and confusing. It's hardly a wonder people disagree. ;) (This is just a bit of rhetoric. I hardly mean to imply the documentation is responsible for the whole problem...) Yep I don't know that ignored is the best word, and i agree the doc can be incomplete and confusing. For another example take the minimum graduation requirements documented on the policy page: 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) - http://incubator.apache .org/incubation/Incubation_Policy.html#Graduating+from+the+Incubator That seems reasonably clear. Based on that policy we have people saying a poddling can't graduate yet because they don't have three independent committers. Or maybe they do have three committers listed but some haven't been active for ages. How long is ages though? Or what is active - actually committing something or is the odd email enough? Or what about if we think they would vote in a new person if someone came along in the future, maybe thats enough? Or how about if some of the mentors agree to stick around on the new PMC to make up the numbers? All of those things get debated. Not so long ago we had a what to do with small slow poddlings debate and a couple of small poddlings were allowed to graduate anyway despite not quite meeting that minimum requirement, then just a little while later Chuwka in a very similar state was nearly retired. ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On 3 April 2013 14:41, ant elder ant.el...@gmail.com wrote: On Wed, Apr 3, 2013 at 2:12 PM, Noah Slater nsla...@apache.org wrote: Thanks for the clarification, Ant. Is the documentation ignored? Whenever I look through it, it seems like the problem is that it is incomplete and confusing. It's hardly a wonder people disagree. ;) (This is just a bit of rhetoric. I hardly mean to imply the documentation is responsible for the whole problem...) Yep I don't know that ignored is the best word, and i agree the doc can be incomplete and confusing. For another example take the minimum graduation requirements documented on the policy page: 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) - http://incubator.apache .org/incubation/Incubation_Policy.html#Graduating+from+the+Incubator Great example - it's reasonably clear but incorrect (as well as being imprecise as you illustrate). We don't require a minimum of 3 independent committers. We require a community that doesn't exclude anyone. I don't have the time to look it up but there was quite some discussion about this point some time ago. I seem to remember the IPMC agreeing the docs need to be updated. Ross -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Chris, What I was trying to do with this particular thread is to identify the problems the incubator has before deciding on solutions. If we can get a common agreement on that, specific solutions will be much easier for us all to accept. So, my question to you is are you able/willing to articulate the problems do you see the incubator as having, that need to be solved? That is, without (yet) suggesting how it should be fixed? I'd be very curious to hear how you see it. Upayavira On Tue, Apr 2, 2013, at 02:00 AM, Mattmann, Chris A (388J) wrote: Hi Niall, First off, thanks for reading my proposal! Specific comments below: -Original Message- From: Niall Pemberton niall.pember...@gmail.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Monday, April 1, 2013 7:00 AM To: general-incubator general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On Mon, Apr 1, 2013 at 5:30 AM, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Sunday, March 31, 2013 5:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. Yep this is the sticking point where we disagree -- b/c I disagree with that. 2 tasks are not a huge range. Also my table of responsibilities in the proposal [1] I believe clearly specifies where any responsibility is shifted and not one of them is the Board. There are two responsibilities you list that shift to the board - 1) spots problems with mentoring 2) fixes problems with mentoring. I agree, spots problems with mentoring could potentially fall on the board, if the incoming new project and its 3 ASF members don't spot the problem beforehand. Just like it falls to the the board related to all projects if there are issues (mentoring, the Apache Way or otherwise). Before I add that to my proposal, I would ask you to consider some of my recent emails related to mentoring issues with podlings, and ask that if you think that the IPMC is currently doing a great of spotting problems with mentoring. If you, if you can please cite examples that would be great. I have counter examples (Mesos, for one) wherein which emails requesting help have gone unanswered, but if you have others in support of that, I'd appreciate hearing them. Also in your proposal oversight of releases is discarded and therefore I would add spots problems with releases is also therefore ultimately the boards responsibility. My philosophy is that there can remain a small set of mentors, e.g., these shepherds if you will that ought to volunteer for projects as they come into the ASF and be part of their initial PMC. The initial Champion role should also be an ASF member, until the incoming project is ready to elect its own chair (should happen in 1 year). These same people are likely going to be the same people who are great at checking releases now (sebb, Marvin, others). There doesn't need to be an IPMC for them to do that. Jukka now Benson have IMO been successful in focusing podlings on what they need to do to graduate and pushing them through the process - rather than staying for years in the incubator. So I would add this to the list of what the board would need to pick up. Well sorry, I don't think it's Jukka and Benson. I've never been a shepherd once, neither has Chris Douglas, neither has a bunch of people that have successfully brought podlings through the Incubator as of late. Jukka and Benson aren't necessarily the only reasons that things shaped up around here though I wholly appreciate their efforts. Lastly I would also say that shifting voting on new projects from a public to private list is not an improvement and would exclude those proposing from answering any objections or concerns. Yes, I am +1 for that. In my proposal, I've gone and updated it to shift that responsibility to voting on general@incubator (which as I mention in my proposal will remain). One note I'd add -- in my proposal, responsibility is not directly shifted onto the board, until it's shown that the incoming project's committee is unable to handle it: (see shifted responsibility: The project's PMC. And if not, the project's VP
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Tue, Apr 2, 2013 at 10:17 AM, Upayavira u...@odoko.co.uk wrote: Chris, What I was trying to do with this particular thread is to identify the problems the incubator has before deciding on solutions. If we can get a common agreement on that, specific solutions will be much easier for us all to accept. So, my question to you is are you able/willing to articulate the problems do you see the incubator as having, that need to be solved? That is, without (yet) suggesting how it should be fixed? I'd be very curious to hear how you see it. Upayavira This is what i think is a big part of the problem: The PMC is so big and diverse, and made up of people who just join by choice not by being invited, and sometimes they don't even care about the PMC they just join to mentor their poddling, so there isn't so much sense of respect or working together. Those people all have different points of view and expectations on how things should happen, some are liberal while some are more conservative, and the set of people who are active varies over time. So what that means is it can be hard to tell what the reaction will be to any particular action, and when something unexpected happens its understandable that sometimes someone is going to get surprised or upset. Take voting on a release as an example, sometimes that will get three quick +1s with minimal review, sometimes it will take weeks of pleading for votes, sometimes a problem will be pointed out but people will still vote +1 anyway, sometimes it will be +1'd with a request to fix the issue later, other times it will be demanded that a respin is done to fix the issue. Theres no way of knowing really, it just depends who happens to be around and active at the time. And the same thing happens for just about every situation where there is some rule or policy or guideline documented. There are things I think we could do to fix some of that, but i agree with Upayavira, we would need some common understanding and agreement on what the issues are first. ...ant
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
As far as I understand your comment, Ant, you mean to say that he problem is that there is too much variation in opinion and approach. (Primarily, I understand, in relation to releases.) This doesn't seem related to the size of the PMC, to me. We're always going to need a large pool of people with the ability to cast binding votes on releases. (Just because of the effort involved.) So whether those people are on the PMC or whether we allow non-PMC mentors to exist and have binding votes — it makes no difference. We still end up with a large pool of people with wildly diverging approaches and world-views. That seems like a different sort of problem to me. Perhaps a documentation problem. Perhaps the recent suggestion to spin out the documenting of policy and guidelines to ComDev — or some proposal like it — could improve that situation. The problem seems larger than the Incubator in any case. I get the strong impression that from TLP to TLP, people are doing things very differently. (And this is where our mentors are recruited from.) On 2 April 2013 11:16, ant elder ant.el...@gmail.com wrote: On Tue, Apr 2, 2013 at 10:17 AM, Upayavira u...@odoko.co.uk wrote: Chris, What I was trying to do with this particular thread is to identify the problems the incubator has before deciding on solutions. If we can get a common agreement on that, specific solutions will be much easier for us all to accept. So, my question to you is are you able/willing to articulate the problems do you see the incubator as having, that need to be solved? That is, without (yet) suggesting how it should be fixed? I'd be very curious to hear how you see it. Upayavira This is what i think is a big part of the problem: The PMC is so big and diverse, and made up of people who just join by choice not by being invited, and sometimes they don't even care about the PMC they just join to mentor their poddling, so there isn't so much sense of respect or working together. Those people all have different points of view and expectations on how things should happen, some are liberal while some are more conservative, and the set of people who are active varies over time. So what that means is it can be hard to tell what the reaction will be to any particular action, and when something unexpected happens its understandable that sometimes someone is going to get surprised or upset. Take voting on a release as an example, sometimes that will get three quick +1s with minimal review, sometimes it will take weeks of pleading for votes, sometimes a problem will be pointed out but people will still vote +1 anyway, sometimes it will be +1'd with a request to fix the issue later, other times it will be demanded that a respin is done to fix the issue. Theres no way of knowing really, it just depends who happens to be around and active at the time. And the same thing happens for just about every situation where there is some rule or policy or guideline documented. There are things I think we could do to fix some of that, but i agree with Upayavira, we would need some common understanding and agreement on what the issues are first. ...ant -- NS
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Ant is reflecting a real dilemma here. At Apache, we try to be egalitarian, and we try to work by consensus. The natural conclusion is that the many people needed to vote on releases are also part of the decision-making body for policy that controls those releases. The dilemma is that consensus doesn't always scale so well. Neither does supervision: when everyone is responsible, no one is responsible. There are several directions to go to on this. Chris M's proposal dissolves the IPMC into many, small, egalitarian communities, and leaves overarching policy for the board and comdev, where it lives for all the other projects. Ross' proposal sacrifices some egalitarianism to achieve better scaling of both decision-making and supervision. Ross' other proposal :-), to move documentation (and thus some/much of the locus of policy decision) making to comdev, reduces the load of decision-making that the IPMC has to find consensus on, and thus proposes to reduce the stress. I sense that Chris M finds my writing on his proposal frustrating. To try to do a better job of explaining myself: Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic. It's one thing for this community to self-govern, but self-destruction strikes me as outside of the mandate. It just strikes me as sideways to seek consensus inside a community that the community is incapable of reliable reaching consensus, amongst other things. If I believed that the IPMC was unfixably nonfunctional in supervision or decision-making, I wouldn't be seeking a consensus. I'd be reporting my view to the board, making a recommendation, and asking for direction. That's how I see my duty as an officer. In other words, if there's something functional to be the chair of, my job is to be the chair of it. If there's nothing functional to be the chair of, it's my job to say so, recognizing that the board might just disagree. Now that we've cleared up some other matters, I'll try to help us all discover if we have a consensus on one of these proposals. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
A simple question for you all. If the current amount of podlings (38): [terra:~/tmp/podlings-content] mattmann% cat podlings.xml | grep current | wc -l 38 [terra:~/tmp/podlings-content] mattmann% Graduated over the next 3 months (~13 a month), or even the next 6 months (which is ~6 near the current rate +/- a few), would the Board members suddenly cease to function? And then in that time, if we gain 1-2 new projects a month (probably greater than the average the past few years), would the Board again cease to function? My proposal is to dissolve the self questioning, TL;DR, binding VOTEs and wild west that is the Incubator PMC. The rest of the situation stays the same. Keep the stinkin' documentation at http://incubation.apache.org, and folks can continue to work/crank on it there if they desire. Why is a (meta)/umbrella committee needed for this? And stop identifying the Board as the folks who shoulder the load. The committees (incoming and graduated) shoulder the load -- the Board only acts rarely, and when provoked. We also have committees for Legal and otherwise that can be leveraged here as I have stated. BTW, note the first step you all seem to agree on is also the first step in my proposal, that Greg and I proposed over a year ago. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Benson Margulies bimargul...@gmail.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Tuesday, April 2, 2013 2:18 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) Ant is reflecting a real dilemma here. At Apache, we try to be egalitarian, and we try to work by consensus. The natural conclusion is that the many people needed to vote on releases are also part of the decision-making body for policy that controls those releases. The dilemma is that consensus doesn't always scale so well. Neither does supervision: when everyone is responsible, no one is responsible. There are several directions to go to on this. Chris M's proposal dissolves the IPMC into many, small, egalitarian communities, and leaves overarching policy for the board and comdev, where it lives for all the other projects. Ross' proposal sacrifices some egalitarianism to achieve better scaling of both decision-making and supervision. Ross' other proposal :-), to move documentation (and thus some/much of the locus of policy decision) making to comdev, reduces the load of decision-making that the IPMC has to find consensus on, and thus proposes to reduce the stress. I sense that Chris M finds my writing on his proposal frustrating. To try to do a better job of explaining myself: Chris proposes that this committee recommend its own demise to the board, to be replaced, in large part, by the board itself. Every board member who has been heard from so far has been less than enthusiastic. It's one thing for this community to self-govern, but self-destruction strikes me as outside of the mandate. It just strikes me as sideways to seek consensus inside a community that the community is incapable of reliable reaching consensus, amongst other things. If I believed that the IPMC was unfixably nonfunctional in supervision or decision-making, I wouldn't be seeking a consensus. I'd be reporting my view to the board, making a recommendation, and asking for direction. That's how I see my duty as an officer. In other words, if there's something functional to be the chair of, my job is to be the chair of it. If there's nothing functional to be the chair of, it's my job to say so, recognizing that the board might just disagree. Now that we've cleared up some other matters, I'll try to help us all discover if we have a consensus on one of these proposals. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On 2 April 2013 22:18, Benson Margulies bimargul...@gmail.com wrote: Ross' proposal sacrifices some egalitarianism to achieve better scaling of both decision-making and supervision. It is not my intention to sacrifice some egalitarianism. My intention is to allow those who have signed up to mentor projects to get on with mentoring them without the well-meaning interference of a large body of ill-informed bystanders (with respect to individual projects needs). This involves both decision-making and supervision. In this regard I believe my proposal is similar in intent to Chris M's. However, unlike Chris I don't see the IPMC failing in this regard. Usually it does a great job. Where my proposal differs from Chris' is in the oversight role of the IPMC. I see oversight as the vital function of the collective IPMC, it is the ability to identify when mentors and their podlings need additional support as they progress towards graduation. When mentors are doing fine this part of the IPMC role is just a case of signing off the board report. It's when something needs adjusting that the IPMC becomes inefficient. It is this aspect that I am seeking improvement for. BUT... Maybe these situations are rare enough to not worry too much about it and rather than change the structure of the IPMC we simply look to thrash out the odd issue that arises, one at a time. For the record I am pleased that you pushed for a change in voting rules driven by a consensus issue. Maybe we just need more of that on the odd occasionally becomes necessary. Ross
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Tue, Apr 2, 2013 at 2:00 AM, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hi Niall, First off, thanks for reading my proposal! Specific comments below: -Original Message- From: Niall Pemberton niall.pember...@gmail.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Monday, April 1, 2013 7:00 AM To: general-incubator general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On Mon, Apr 1, 2013 at 5:30 AM, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Sunday, March 31, 2013 5:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. Yep this is the sticking point where we disagree -- b/c I disagree with that. 2 tasks are not a huge range. Also my table of responsibilities in the proposal [1] I believe clearly specifies where any responsibility is shifted and not one of them is the Board. There are two responsibilities you list that shift to the board - 1) spots problems with mentoring 2) fixes problems with mentoring. I agree, spots problems with mentoring could potentially fall on the board, if the incoming new project and its 3 ASF members don't spot the problem beforehand. Just like it falls to the the board related to all projects if there are issues (mentoring, the Apache Way or otherwise). Before I add that to my proposal, I would ask you to consider some of my recent emails related to mentoring issues with podlings, and ask that if you think that the IPMC is currently doing a great of spotting problems with mentoring. If you, if you can please cite examples that would be great. I have counter examples (Mesos, for one) wherein which emails requesting help have gone unanswered, but if you have others in support of that, I'd appreciate hearing them. The issue with Mesos mentors was raised by a shepherd - it did result in one new mentor for a short while. Then when that persons stepped down, you volunteered. I agree its still an issue for Mesos - so hasn't been successfully resolved. What I do see on a regular basis is shepherds reporting back on mentor status as part of the oversight role. I would be surprised if the board wanted to do this work. Also in your proposal oversight of releases is discarded and therefore I would add spots problems with releases is also therefore ultimately the boards responsibility. My philosophy is that there can remain a small set of mentors, e.g., these shepherds if you will that ought to volunteer for projects as they come into the ASF and be part of their initial PMC. The initial Champion role should also be an ASF member, until the incoming project is ready to elect its own chair (should happen in 1 year). These same people are likely going to be the same people who are great at checking releases now (sebb, Marvin, others). There doesn't need to be an IPMC for them to do that. Actually there does. PMC's are responsible for releases - they do the checks AND vote as part of the IPMC - so its an IPMC release. Take away their PMC role and the checks alone are just an opinion with no power to approve or not. I'm not necessarily opposed to that (haven't given it thought) - but if the project is given responsibility for releases - then the board will have to take ultimate responsibility for catching resolving issues. Jukka now Benson have IMO been successful in focusing podlings on what they need to do to graduate and pushing them through the process - rather than staying for years in the incubator. So I would add this to the list of what the board would need to pick up. Well sorry, I don't think it's Jukka and Benson. I've never been a shepherd once, neither has Chris Douglas, neither has a bunch of people that have successfully brought podlings through the Incubator as of late. Jukka and Benson aren't necessarily the only reasons that things shaped up around here though I wholly appreciate their efforts. Apologies - I didn't mean to diminish anyone elses efforts. Lastly I would also say that shifting voting on new projects from a public to private list is not an improvement and would exclude those proposing from answering any objections or concerns. Yes, I am +1
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Chris,
Individuals who contribute to Podlings are doing so as mentors and members
of theY IPMC. There is nothing in the Director role that says they are
required to do so.
You can add me to that list of people who have mentored projects in the
last year (I've been part of something like 7 graduating podlings this
year). In every case I did so as an IPMC member. Doing so did not increase
my responsibilities to other podlings.
I do not understand why you are listing peoples engagement in individual
podlings as evidence that Directors have not delegated oversight of
podlings to the IPMC. It simply is not true, two directors have told you
this, as have multiple IPMC members.
Nobody, that I can see, is proposing another layer in the IPMC. You are
proposing moving a layer to the board (although you don't accept that will
be an outcome). I've suggested refining the decision making process. In the
meantime Benson *has* refined the decision making process and the IPMC
seems to have agreed to it.
Let's focus on whether ComDev and the IPMC want to share some
responsibilities.
Ross
Sent from a mobile device, please excuse mistakes and brevity
On 1 Apr 2013 05:31, Mattmann, Chris A (388J)
chris.a.mattm...@jpl.nasa.gov wrote:
Hi Ross,
-Original Message-
From: Ross Gardler rgard...@opendirective.com
Reply-To: general@incubator.apache.org general@incubator.apache.org
Date: Sunday, March 31, 2013 5:20 PM
To: general@incubator.apache.org general@incubator.apache.org
Subject: Re: Incubator structure (was Re: Vote on personal matters:
majority vote vs consensus)
On 31 March 2013 17:08, Mattmann, Chris A (388J)
chris.a.mattm...@jpl.nasa.gov wrote:
Why is it so hard to see that the board is already watching those 22
nascent projects in the same manner they watch the 137 TLPs?
Because they are not watching with the same manner. They are delegating a
huge range of tasks such as IP oversight and mentoring to the IPMC.
Yep this is the sticking point where we disagree -- b/c I disagree with
that.
2 tasks are not a huge range. Also my table of responsibilities in the
proposal [1]
I believe clearly specifies where any responsibility is shifted and not
one of
them is the Board. So I've enumerated at least the concerns of myself and
many others about
a range of tasks, and addressed them (for well over a year). I've heard
zero feedback
from you about what's wrong with my table, and what I've missed, what
could be improved
and have heard nothing but it's wrong (paraphrased) or it doesn't cover
all the tasks
that of course will get dropped on the Board? I've done the work to
document
my thoughts. You don't get to then just keep telling me it's wrong without
specifying
what precisely is wrong about it.
Ross says the Board pays less attention to these (by implication) than
say the 137 TLPs at present. Ross is one Director. Good for him.
I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm
active in the IPMC and thus have more visibility. That doesn't mean they
should be expected to by me or by anyone else.
Actually it should be expected -- there is a reason that people like Jim
mentored AOO -- people like Sam joined in, and so did Greg with AOO and
Bloodhound (all 3 are directors). There is a reason that Bertrand has been
very active in the Incubator with Flex and other recent projects. Same as
Rich with Allura -- Roy helps a lot too with clarifications when needed.
I've seen more than a handful of emails from Brett Porter too, so he's
definitely around.
So, sorry Directors too pay just as much attention to PPMCs and to the
Incubator based on their
own individual Incubator and Director hats, and based on their reporting.
I know other directors (Greg IIRC at least) didn't want the Incubator
specific podling reports to go away (and to only have the summary
at the top of the Incubator report).
I don't think any of the Directors want them to go away. But board reports
are not what the IPMC is about. That is the reporting process within the
foundation and provides the level of oversight into the PPMCs that the
board requires. But the IPMC does *much* more than submit a monthly board
report with a verbatim copy of the podlings individual reports.
What i can see, and what I think even Upayavira and Ross
agree
with -- and you too Benson -- is that there is a grave problem here and
it
needs' a fixin'. My deconstruction proposal does that.
No, I do not agree there is a grave problem. I have denied that
repeatedly.
The IPMC has problems, but in the main it works extremely well.
Fine you don't think it's grave. I don't care how it's classified
('grave',
'purple', 'pink', 'yellow', whatever). There is a problem is what I
probably
should have said.
Look, I hear you that, it's probably possible that folks can come up with
even yet another layer beyond the Shepherds, etc., and that that can goad
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Mon, Apr 1, 2013 at 5:30 AM, Mattmann, Chris A (388J)
chris.a.mattm...@jpl.nasa.gov wrote:
Hi Ross,
-Original Message-
From: Ross Gardler rgard...@opendirective.com
Reply-To: general@incubator.apache.org general@incubator.apache.org
Date: Sunday, March 31, 2013 5:20 PM
To: general@incubator.apache.org general@incubator.apache.org
Subject: Re: Incubator structure (was Re: Vote on personal matters:
majority vote vs consensus)
On 31 March 2013 17:08, Mattmann, Chris A (388J)
chris.a.mattm...@jpl.nasa.gov wrote:
Why is it so hard to see that the board is already watching those 22
nascent projects in the same manner they watch the 137 TLPs?
Because they are not watching with the same manner. They are delegating a
huge range of tasks such as IP oversight and mentoring to the IPMC.
Yep this is the sticking point where we disagree -- b/c I disagree with
that.
2 tasks are not a huge range. Also my table of responsibilities in the
proposal [1]
I believe clearly specifies where any responsibility is shifted and not
one of
them is the Board.
There are two responsibilities you list that shift to the board - 1) spots
problems with mentoring 2) fixes problems with mentoring.
Also in your proposal oversight of releases is discarded and therefore I
would add spots problems with releases is also therefore ultimately the
boards responsibility.
Jukka now Benson have IMO been successful in focusing podlings on what
they need to do to graduate and pushing them through the process - rather
than staying for years in the incubator. So I would add this to the list of
what the board would need to pick up.
Lastly I would also say that shifting voting on new projects from a public
to private list is not an improvement and would exclude those proposing
from answering any objections or concerns.
Niall
So I've enumerated at least the concerns of myself and
many others about
a range of tasks, and addressed them (for well over a year). I've heard
zero feedback
from you about what's wrong with my table, and what I've missed, what
could be improved
and have heard nothing but it's wrong (paraphrased) or it doesn't cover
all the tasks
that of course will get dropped on the Board? I've done the work to
document
my thoughts. You don't get to then just keep telling me it's wrong without
specifying
what precisely is wrong about it.
Ross says the Board pays less attention to these (by implication) than
say the 137 TLPs at present. Ross is one Director. Good for him.
I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm
active in the IPMC and thus have more visibility. That doesn't mean they
should be expected to by me or by anyone else.
Actually it should be expected -- there is a reason that people like Jim
mentored AOO -- people like Sam joined in, and so did Greg with AOO and
Bloodhound (all 3 are directors). There is a reason that Bertrand has been
very active in the Incubator with Flex and other recent projects. Same as
Rich with Allura -- Roy helps a lot too with clarifications when needed.
I've seen more than a handful of emails from Brett Porter too, so he's
definitely around.
So, sorry Directors too pay just as much attention to PPMCs and to the
Incubator based on their
own individual Incubator and Director hats, and based on their reporting.
I know other directors (Greg IIRC at least) didn't want the Incubator
specific podling reports to go away (and to only have the summary
at the top of the Incubator report).
I don't think any of the Directors want them to go away. But board reports
are not what the IPMC is about. That is the reporting process within the
foundation and provides the level of oversight into the PPMCs that the
board requires. But the IPMC does *much* more than submit a monthly board
report with a verbatim copy of the podlings individual reports.
What i can see, and what I think even Upayavira and Ross
agree
with -- and you too Benson -- is that there is a grave problem here and
it
needs' a fixin'. My deconstruction proposal does that.
No, I do not agree there is a grave problem. I have denied that
repeatedly.
The IPMC has problems, but in the main it works extremely well.
Fine you don't think it's grave. I don't care how it's classified
('grave',
'purple', 'pink', 'yellow', whatever). There is a problem is what I
probably
should have said.
Look, I hear you that, it's probably possible that folks can come up with
even yet another layer beyond the Shepherds, etc., and that that can goad
people into thinking stuff is fixed around here. Jukka's work was great,
and
I applaud him for it, but as I said at the time, to me we're just adding
more
and more layers to the onion, instead of stripping it down to its roots and
core.
Also it's possible that if you guys continue to add layers, and suggest
mechanisms
for organizing those that are active around here
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Hi Niall, First off, thanks for reading my proposal! Specific comments below: -Original Message- From: Niall Pemberton niall.pember...@gmail.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Monday, April 1, 2013 7:00 AM To: general-incubator general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On Mon, Apr 1, 2013 at 5:30 AM, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Sunday, March 31, 2013 5:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus) On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. Yep this is the sticking point where we disagree -- b/c I disagree with that. 2 tasks are not a huge range. Also my table of responsibilities in the proposal [1] I believe clearly specifies where any responsibility is shifted and not one of them is the Board. There are two responsibilities you list that shift to the board - 1) spots problems with mentoring 2) fixes problems with mentoring. I agree, spots problems with mentoring could potentially fall on the board, if the incoming new project and its 3 ASF members don't spot the problem beforehand. Just like it falls to the the board related to all projects if there are issues (mentoring, the Apache Way or otherwise). Before I add that to my proposal, I would ask you to consider some of my recent emails related to mentoring issues with podlings, and ask that if you think that the IPMC is currently doing a great of spotting problems with mentoring. If you, if you can please cite examples that would be great. I have counter examples (Mesos, for one) wherein which emails requesting help have gone unanswered, but if you have others in support of that, I'd appreciate hearing them. Also in your proposal oversight of releases is discarded and therefore I would add spots problems with releases is also therefore ultimately the boards responsibility. My philosophy is that there can remain a small set of mentors, e.g., these shepherds if you will that ought to volunteer for projects as they come into the ASF and be part of their initial PMC. The initial Champion role should also be an ASF member, until the incoming project is ready to elect its own chair (should happen in 1 year). These same people are likely going to be the same people who are great at checking releases now (sebb, Marvin, others). There doesn't need to be an IPMC for them to do that. Jukka now Benson have IMO been successful in focusing podlings on what they need to do to graduate and pushing them through the process - rather than staying for years in the incubator. So I would add this to the list of what the board would need to pick up. Well sorry, I don't think it's Jukka and Benson. I've never been a shepherd once, neither has Chris Douglas, neither has a bunch of people that have successfully brought podlings through the Incubator as of late. Jukka and Benson aren't necessarily the only reasons that things shaped up around here though I wholly appreciate their efforts. Lastly I would also say that shifting voting on new projects from a public to private list is not an improvement and would exclude those proposing from answering any objections or concerns. Yes, I am +1 for that. In my proposal, I've gone and updated it to shift that responsibility to voting on general@incubator (which as I mention in my proposal will remain). One note I'd add -- in my proposal, responsibility is not directly shifted onto the board, until it's shown that the incoming project's committee is unable to handle it: (see shifted responsibility: The project's PMC. And if not, the project's VP. And if not that, the board or the membership. Just like the current way it works for existing TLPs. ) HTH. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA
Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Fri, Mar 29, 2013, at 01:56 AM, Chris Douglas wrote: On Thu, Mar 28, 2013 at 4:30 PM, Benson Margulies bimargul...@gmail.com wrote: Your position is that the IPMC fails to supervise. The consensus of the IPMC is that this is not true. Otherwise, someone would be reading the monthly report and objecting to the failure to report 'failure' to the board. If your statement were true, then someone would make the assertions you're making. If you want to change minds about this, you might need to come up with some concrete evidence of actual failure: bad commits, bad doings on mailing lists, etc. Is this a question of standing, where material harm needs to be demonstrated? The IPMC is needlessly inefficient and abusive of its podlings. Novel compliance mechanisms are literally invented and argued about on general@ during podlings' release votes.[1] The cultural clashes that Chris's proposal refers to generate huge amounts of traffic on general@ and private@, as ASF members argue the semantics of core concepts. And it's not just edge-case legal issues; some are as basic as the definition of veto. These discussions create needless confusion and deeply resented churn for podlings. The asymmetry in power teaches submissiveness to ASF members, rather than independence and self-sufficiency. There are, in truth, *many* active interpretations of the Apache Way practiced across the ASF. Reconciling them is not the mission of the incubator. Putting esoteric debates on the critical path of new projects is absurd and harmful. [1] http://s.apache.org/lFI What the IPMC now does is use the shepherd process to compensate for mentor weakness. It's not perfect. Under your plan, instead, the board would have to cope, directly, with 'starter' projects suffering from inevitable attrition -- or the Foundation would need to start many less projects, as only those who could attract very strongly committed foundation members could start. The incubator doesn't deal with this effectively, either. To take one example, Chukwa's retirement was a fiasco. there are many possible places to take the conversation if you start from the premise that no variation on the existing scheme is workable. I personally don't know how, organizationally, to reach that conclusion, especially insofar as the recent disfunction is not about supervision, it's _merely_ about sorting out who can be a member and thus a mentor. The recent discussion is about the mentor role. The ongoing dysfunction is about supervision, and the IPMC failing to discharge its purpose efficiently. It exists to put its podlings through a curriculum that transfers some cultural norms, makes podlings aware of resources, and establishes clean licensing. Even if it succeeds well enough not to be disbanded by the board, its inefficiency is worth correcting. -C The incubator has the feel of the PRC before it was disbanded (actually, it was split into trademarks/brand, fundraising and press). That split made the structure match the territory, and all three areas have been pretty quietly effective at what they do. We need to find a similar transformation for the Incubator. This is what many of the suggestions that have been floating are, in my view, trying to address, we just don't yet have a suggestion that is garnering sufficient collective support. What we need is an effective way to reduce the number of people who are 'responsible' for the day to day running of the incubator. We need one set who are 'incubator people' and another who are 'mentors'. We need a structure where the 'many' are happy to delegate responsibility to 'the few'. We have that in press@, fundraising@ and trademarks@ (perhaps too much so!!). So, the question is, what models are there that will achieve this? Chris has suggested disbanding the Incubator PMC. Ant has suggested mentors leave the Incubator PMC, and the Incubator PMC become 'the few'. Ross has suggested that the Incubator PMC votes a group of shepherds to be 'the few'. It seems to me there are two principal roles that the Incubator has had - one is to help define processes, and social mores. This, it could be argued, has been done well enough - it doesn't take as much effort anymore, and perhaps we have reached a point where it is not possible to 'define' it anymore. The second role is to provide 'oversight' or 'supervision', providing a layer above mentors to ensure that podlings are progressing and that mentors are active. It is this latter role that is still very much needed, as we would expect there to be more issues in newer podlings than in established TLPs. It seems to me that the board *wants* to delegate this responsibility to another committee. To summarise. The incubator *is* broken (but not necessarily beyond repair). We need as many mentors as we can get, and a smaller group of people who are delegated responsibility for the incubator. The board wants a
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
To summarise. The incubator *is* broken (but not necessarily beyond repair). We need as many mentors as we can get, and a smaller group of people who are delegated responsibility for the incubator. The board wants a group of folks to take responsibility for overseeing the early life of communities at the ASF. These are, to my mind, the criteria that we should be using to evaluate any suggestions as to how the incubator should be structured. If it doesn't meet these, it won't float. I think there's something missing here. The incubator structure as we know it is, very intentionally, a variation on the standard TLP model. There is a tree in svn (ok, also some git repos). There are PMC members. There are committers. Releases and committer karma are controlled by the PMC. All nice and neat, just like all other TLPs. The IPMC is, in fact, a PMC. Chris M observes, if I may parody, that it's 'just like' the discredited umbrella projects, and proposes to fix this by making podlings even more like the standard model -- each one a TLP supervised by The Board. Ross proposes to establish an interior structure to the incubator. People seem to prefer to stay close to the standard model, for good reason, as the legal structure is all worked out. I think that any alternative has to specifically address the alternate legal structure. Who votes on releases? Who votes on karma? I personally don't have a problem with a plan in which the incubator isn't really a PMC at the end of the day. One way to combine Ross and Chris is to say, 'well, the Board could decide that it can't watch 22 nascent projects, so it's delegated that to a committee. The 'podlings' are projects, and they have an initial PMC composed entirely of Foundation members, which grows over time, and the 'incubator committee' serves as the board's representatives in watching them.' Really, the big argument here is whether it's broken enough to fix. On the 'endless discussion' front, I think not. It would be easier to repair the chair to than to change the structure. On the supervision front, well, there's a big disagreement here. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Hi Benson,
-Original Message-
From: Benson Margulies bimargul...@gmail.com
Reply-To: general@incubator.apache.org general@incubator.apache.org
Date: Sunday, March 31, 2013 8:02 AM
To: general@incubator.apache.org general@incubator.apache.org
Subject: Re: Incubator structure (was Re: Vote on personal matters:
majority vote vs consensus)
[..snip..]
Chris M observes, if I may parody, that it's 'just like' the
discredited umbrella projects, and proposes to fix this by making
podlings even more like the standard model -- each one a TLP
supervised by The Board.
That's one part of it.
[..snip..]
I think that any alternative has to specifically address the alternate
legal structure. Who votes on releases? Who votes on karma? I
personally don't have a problem with a plan in which the incubator
isn't really a PMC at the end of the day. One way to combine Ross and
Chris is to say, 'well, the Board could decide that it can't watch 22
nascent projects, so it's delegated that to a committee.
Why is it so hard to see that the board is already watching those 22
nascent projects in the same manner they watch the 137 TLPs? And if they
are not then I call b. to the s. -- we ask the podlings to start operating
like TLPs on day 1 -- we ask the mentors to do the same -- and to teach
the PPMC that. Yet, the board isn't watching them in the same way?
Ross says the Board pays less attention to these (by implication) than
say the 137 TLPs at present. Ross is one Director. Good for him.
I know other directors (Greg IIRC at least) didn't want the Incubator
specific podling reports to go away (and to only have the summary
at the top of the Incubator report). That's at least one other Director
(there were probably more since there was consensus on the podling
specific reports not going away when it was discussed) that IMHO watches
the podlings the same way as the TLPs are watched.
And, why is it so hard to see that the Board may watch, but in the
end, it's on the specific committees ('podling' as we currently call
them or 'PPMC' or otherwise [TLP]) to manage their stuff? The Board
is the bazooka, the elephant gun, remember? That's why we're still
here discussing ad nauseum this topic a year later -- because to bazooka
the Incubator would be some monuments event, similar to the bazooka, of
PRC, etc. -- something that's discussed at ApacheCon over beer about the
'old ways' and 'can you believe when that happened, wow??!'. In the end,
it's not a monuments event. As Upayavira said, life went on in PRC, in
the sub committees; Apache went on.
What I've done is suggest [in the Apache vein], what is IMO, a logical,
incremental (and even potentially reversible) next step. Upayavira's
latest email on this was right -- he sees that the Incubator is broken,
and perhaps it needs to be split into smaller, separate committees. Ross
is right too -- maybe something else needs to be elected in the form of a
committee (his shepherds) to watch the incoming projects. My point is
-- great -- I can't see the forrest through the trees on the answer to
that
question yet. What i can see, and what I think even Upayavira and Ross
agree
with -- and you too Benson -- is that there is a grave problem here and it
needs' a fixin'. My deconstruction proposal does that.
I've suggested a logical next step to fixing it -- don't let the wacko
committee try and fix itself. That's like asking an insane asylum to
form a committee for how it will itself. Besides the insanity, they are
naturally in a conflict of interest state. Instead, I'm proposing, rubble
the asylum, transition delegation and authority if only temporarily until
some next step proposal can be agreed upon and discussed without the
inmates.
[..snip..]
Cheers,
Chris
-
To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org
For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013 at 6:08 PM, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: ...Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs?... It's not. Well, maybe it is, but up to a point. The good thing is that podlings currently get some level of double supervision: the Incubator PMC watches over them (including via incubator shepherds, which as a board member I like a lot) and is responsible for them. Board members can additionally have a look at the detailed podling reports, in addition to the very useful incubator summary that the IPMC chair produces. Depending on available time, as a board member I sometimes read all of the detailed reports, sometimes only the ones that seem to warrant attention, and sometimes just the chair's summary. All this works quite well IMO, as a board member I don't see a need to fix anything. -Bertrand
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013 at 3:13 AM, Upayavira u...@odoko.co.uk wrote: We need one set who are 'incubator people' and another who are 'mentors'. Disenfranchising mentors and hoarding power within a small circle of IPMC aristocrats is both unworkable and hypocritical. * It is unworkable because the people who watch over the IP clearance process and subsequently endorse incubating releases must follow the podling day-to-day and must be empowered with binding votes. Superficial review by freelance IPMC members is useful but cannot substitute for close supervision. * It is hypocritical because the IPMC needs to recognize and reward merit if we expect podlings to do likewise. The incubator *is* broken (but not necessarily beyond repair). I don't see the clashes in the IPMC's immediate past as arising from structural defects -- we experienced ordinary personnel issues which could have happened to any project. The IPMC's size wasn't even much a factor, since none of our dormant members played any part in prolonging the conflicts. Judging the success of any new structure will be easy: does it create peace and quiet (and more effective working) like the breakup of the PRC did?? The Incubator has two acute, serious problems. 1. First releases are too hard. 2. Mentor attrition. The first problem is being addressed by building consensus around clarified release approval criteria. The second problem is being addressed by making it easier to recruit outstanding podling contributors to serve on the IPMC. In my view, the various radical approaches being proposed either do not help, actively hinder, or add a lot of work and uncertainty -- so for the time being, I'd rather work to improve the current system. I'll only join the revolution if the incremental improvements are blocked. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013, at 07:12 PM, Marvin Humphrey wrote: On Sun, Mar 31, 2013 at 3:13 AM, Upayavira u...@odoko.co.uk wrote: We need one set who are 'incubator people' and another who are 'mentors'. Disenfranchising mentors and hoarding power within a small circle of IPMC aristocrats is both unworkable and hypocritical. * It is unworkable because the people who watch over the IP clearance process and subsequently endorse incubating releases must follow the podling day-to-day and must be empowered with binding votes. Superficial review by freelance IPMC members is useful but cannot substitute for close supervision. * It is hypocritical because the IPMC needs to recognize and reward merit if we expect podlings to do likewise. The incubator *is* broken (but not necessarily beyond repair). I don't see the clashes in the IPMC's immediate past as arising from structural defects -- we experienced ordinary personnel issues which could have happened to any project. The IPMC's size wasn't even much a factor, since none of our dormant members played any part in prolonging the conflicts. Judging the success of any new structure will be easy: does it create peace and quiet (and more effective working) like the breakup of the PRC did?? The Incubator has two acute, serious problems. 1. First releases are too hard. 2. Mentor attrition. The first problem is being addressed by building consensus around clarified release approval criteria. The second problem is being addressed by making it easier to recruit outstanding podling contributors to serve on the IPMC. In my view, the various radical approaches being proposed either do not help, actively hinder, or add a lot of work and uncertainty -- so for the time being, I'd rather work to improve the current system. I'll only join the revolution if the incremental improvements are blocked. Marvin, Just to clarify - I am attempting to clarify the problem, not posit a specific solution. I did not suggest that responsibility for voting should rest with a smaller group - clearly voting rests with those close to the project - we agree there. Also, I'm not suggesting we need a revolution, just that we need to recognise that there are structural issues, and to explore the nature of the problem. The better we can do that, the easier finding incremental solutions that everyone can put themselves behind will happen. I'd say your two issues are definitely worth consideration. Upayavira - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013 at 8:12 PM, Marvin Humphrey mar...@rectangular.com wrote: On Sun, Mar 31, 2013 at 3:13 AM, Upayavira u...@odoko.co.uk wrote: The Incubator has two acute, serious problems. 1. First releases are too hard. No surprise. This is incredible hard to read: http://incubator.apache.org/guides/releasemanagement.html and contains 88 TODOs. 2. Mentor attrition. The first problem is being addressed by building consensus around clarified release approval criteria. The second problem is being addressed by making it easier to recruit outstanding podling contributors to serve on the IPMC. If we make other people easier to join, we *need* to make a regular health check on projects. I already proposed the miss to sign your reports 2x times and we ask you if you are still there process. In my view, the various radical approaches being proposed either do not help, actively hinder, or add a lot of work and uncertainty -- so for the time being, I'd rather work to improve the current system. I'll only join the revolution if the incremental improvements are blocked. Being sceptic makes sense of course. Radical changes are coming with a risk. That said, I totally appreciate what Upayavira wrote and bascially support the idea of separating out mentors from the IPMC. I have heard a few people say they just want to mentor, without the rules discussion crap (see ml). Thats perfectly OK. But what do we need them on the IPMC? The argument is, we want binding votes. I believe we can give them binding votes without being on the IPMC. We now seem to have 3/4 majority vote on new people. That helps with easier recruiting. Now lets do the miss to sign process and actively work on the release guide. If that doesn't help we can become more radical. Cheers Christian - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On 31 March 2013 17:08, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Why is it so hard to see that the board is already watching those 22 nascent projects in the same manner they watch the 137 TLPs? Because they are not watching with the same manner. They are delegating a huge range of tasks such as IP oversight and mentoring to the IPMC. Ross says the Board pays less attention to these (by implication) than say the 137 TLPs at present. Ross is one Director. Good for him. I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm active in the IPMC and thus have more visibility. That doesn't mean they should be expected to by me or by anyone else. I know other directors (Greg IIRC at least) didn't want the Incubator specific podling reports to go away (and to only have the summary at the top of the Incubator report). I don't think any of the Directors want them to go away. But board reports are not what the IPMC is about. That is the reporting process within the foundation and provides the level of oversight into the PPMCs that the board requires. But the IPMC does *much* more than submit a monthly board report with a verbatim copy of the podlings individual reports. What i can see, and what I think even Upayavira and Ross agree with -- and you too Benson -- is that there is a grave problem here and it needs' a fixin'. My deconstruction proposal does that. No, I do not agree there is a grave problem. I have denied that repeatedly. The IPMC has problems, but in the main it works extremely well. Upayavira does get it right in his email when he says: To summarise. The incubator *is* broken (but not necessarily beyond repair). We need as many mentors as we can get, and a smaller group of people who are delegated responsibility for the incubator. The board wants a group of folks to take responsibility for overseeing the early life of communities at the ASF. These are, to my mind, the criteria that we should be using to evaluate any suggestions as to how the incubator should be structured. If it doesn't meet these, it won't float. The rest of his excellent mail got me thinking. When we discussed this before Jukka took over there was a push for ComDev to take on the documentation of policy, best practice and process. As VP of ComDev at the time I pushed back. I was concerned that moving the process of documentation wasn't going to make any real difference. However, I did say that if the IPMC could start to get its house in order then I agreed ComDev would be the right place for this. I think the IPMC has resolved many of its problems over the last 12 months. So it is time to revisit that suggestion. However, I would suggest we revisit in the light of Upayavira's excellent mail, specifically: It seems to me there are two principal roles that the Incubator has had - one is to help define processes, and social mores. ... The second role is to provide 'oversight' or 'supervision', providing a layer above mentors to ensure that podlings are progressing and that mentors are active. It is this latter role that is still very much needed, as we would expect there to be more issues in newer podlings than in established TLPs. It seems to me that the board *wants* to delegate this responsibility to another committee. The first of these two roles is, for the most part, where the IPMC can sometimes reach stagnation and can become extremely confusing to podlings (getting multiple answers for one question for example). Maybe it is time to move this to ComDev and take that area of conflict away from the IPMC. This would leave the IPMC to focus on providing the oversight that Jukka's new processes have started to heal and Benson is now fine-tuning. Ross -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
Hi Ross,
-Original Message-
From: Ross Gardler rgard...@opendirective.com
Reply-To: general@incubator.apache.org general@incubator.apache.org
Date: Sunday, March 31, 2013 5:20 PM
To: general@incubator.apache.org general@incubator.apache.org
Subject: Re: Incubator structure (was Re: Vote on personal matters:
majority vote vs consensus)
On 31 March 2013 17:08, Mattmann, Chris A (388J)
chris.a.mattm...@jpl.nasa.gov wrote:
Why is it so hard to see that the board is already watching those 22
nascent projects in the same manner they watch the 137 TLPs?
Because they are not watching with the same manner. They are delegating a
huge range of tasks such as IP oversight and mentoring to the IPMC.
Yep this is the sticking point where we disagree -- b/c I disagree with
that.
2 tasks are not a huge range. Also my table of responsibilities in the
proposal [1]
I believe clearly specifies where any responsibility is shifted and not
one of
them is the Board. So I've enumerated at least the concerns of myself and
many others about
a range of tasks, and addressed them (for well over a year). I've heard
zero feedback
from you about what's wrong with my table, and what I've missed, what
could be improved
and have heard nothing but it's wrong (paraphrased) or it doesn't cover
all the tasks
that of course will get dropped on the Board? I've done the work to
document
my thoughts. You don't get to then just keep telling me it's wrong without
specifying
what precisely is wrong about it.
Ross says the Board pays less attention to these (by implication) than
say the 137 TLPs at present. Ross is one Director. Good for him.
I, personally, pay as much attention to the PPMCs as I do to TLPs. I'm
active in the IPMC and thus have more visibility. That doesn't mean they
should be expected to by me or by anyone else.
Actually it should be expected -- there is a reason that people like Jim
mentored AOO -- people like Sam joined in, and so did Greg with AOO and
Bloodhound (all 3 are directors). There is a reason that Bertrand has been
very active in the Incubator with Flex and other recent projects. Same as
Rich with Allura -- Roy helps a lot too with clarifications when needed.
I've seen more than a handful of emails from Brett Porter too, so he's
definitely around.
So, sorry Directors too pay just as much attention to PPMCs and to the
Incubator based on their
own individual Incubator and Director hats, and based on their reporting.
I know other directors (Greg IIRC at least) didn't want the Incubator
specific podling reports to go away (and to only have the summary
at the top of the Incubator report).
I don't think any of the Directors want them to go away. But board reports
are not what the IPMC is about. That is the reporting process within the
foundation and provides the level of oversight into the PPMCs that the
board requires. But the IPMC does *much* more than submit a monthly board
report with a verbatim copy of the podlings individual reports.
What i can see, and what I think even Upayavira and Ross
agree
with -- and you too Benson -- is that there is a grave problem here and
it
needs' a fixin'. My deconstruction proposal does that.
No, I do not agree there is a grave problem. I have denied that
repeatedly.
The IPMC has problems, but in the main it works extremely well.
Fine you don't think it's grave. I don't care how it's classified
('grave',
'purple', 'pink', 'yellow', whatever). There is a problem is what I
probably
should have said.
Look, I hear you that, it's probably possible that folks can come up with
even yet another layer beyond the Shepherds, etc., and that that can goad
people into thinking stuff is fixed around here. Jukka's work was great,
and
I applaud him for it, but as I said at the time, to me we're just adding
more
and more layers to the onion, instead of stripping it down to its roots and
core.
Also it's possible that if you guys continue to add layers, and suggest
mechanisms
for organizing those that are active around here, I may just go back to my
merry
way of getting podlings through the Incubator, graduated, and taught in
the ASF
way.
But it's also possible that the existence of this super/meta committee and
its
super awesome badges that many of the folks here are just too blind to
give them up
will wain on individuals.
[..snip..]
The first of these two roles is, for the most part, where the IPMC can
sometimes reach stagnation and can become extremely confusing to podlings
(getting multiple answers for one question for example). Maybe it is time
to move this to ComDev and take that area of conflict away from the IPMC.
This would leave the IPMC to focus on providing the oversight that Jukka's
new processes have started to heal and Benson is now fine-tuning.
I suggested this in my proposal -- and also creating
http://incubation.apache.org/
which is home to all the documentation/processes, etc. This is step #1
in my proposal BTW (moving to ComDev).
Also
Re: Incubator structure (was Re: Vote on personal matters: majority vote vs consensus)
On Sun, Mar 31, 2013 at 12:01 PM, Christian Grobmeier grobme...@gmail.com wrote: I have heard a few people say they just want to mentor, without the rules discussion crap (see ml). Thats perfectly OK. But what do we need them on the IPMC? One of the chief responsibilities for a Mentor is performing oversight of the podling's code base on behalf of the foundation. Until a podling graduates and gets a resolution passed by the Board establishing a PMC, it is the _IPMC_ which is responsible for legal oversight of the code base. A podling needs people who teach the social aspects of the Apache Way, but we assume that the IPMC members we assign as Mentors are doing that in addition to performing their legal role. Contributions by non-IPMC members are welcome, but giving such people the title of Mentor only corrupts our accounting mechanisms and makes it harder to detect when the Incubator is failing to provide legal oversight of a code base for which it has assumed responsibility. I believe we can give them binding votes without being on the IPMC. Efforts to keep the IPMC pure and exclusive cause only harm. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 5:01 PM, Joe Schaefer joe_schae...@yahoo.com wrote: As Doug points out, votes are structured away from the status quo- we don't ever vote to continue on with previously agreed to issues just to circumvent the voting process. Ok thanks Joe and Doug. So to be absolutely clear, the wording of votes is important and votes need to be structured the right way. You can tell how to do it by looking at if the resulting action changes the status quo. Using the previous example if Joe wants to send an email to me to saying I'm now not a mentor, or remove me from a status file, etc then thats changing the status quo so now requires a 75% vote. Right? ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Fri, Mar 29, 2013 at 8:47 AM, Matthias Friedrich m...@mafr.de wrote: As someone who is relatively new to the ASF and who's first behind the scenes contact with Apache was the incubation process, I can tell that this is absolutely true. Podlings find themselves in a kafkaesque world where many rules are undocumented or can only be found in old mailing list discussions. Apache and this list especially can feel like a really hostile place for newbies. Regards, Matthias Unfortunately it can feel quite hostile to even not so newbies too. Once out of the Incubator its a much more friendly place though. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thursday, 2013-03-28, Chris Douglas wrote: [...] Is this a question of standing, where material harm needs to be demonstrated? The IPMC is needlessly inefficient and abusive of its podlings. Novel compliance mechanisms are literally invented and argued about on general@ during podlings' release votes.[1] The cultural clashes that Chris's proposal refers to generate huge amounts of traffic on general@ and private@, as ASF members argue the semantics of core concepts. And it's not just edge-case legal issues; some are as basic as the definition of veto. These discussions create needless confusion and deeply resented churn for podlings. The asymmetry in power teaches submissiveness to ASF members, rather than independence and self-sufficiency. There are, in truth, *many* active interpretations of the Apache Way practiced across the ASF. Reconciling them is not the mission of the incubator. Putting esoteric debates on the critical path of new projects is absurd and harmful. [1] http://s.apache.org/lFI As someone who is relatively new to the ASF and who's first behind the scenes contact with Apache was the incubation process, I can tell that this is absolutely true. Podlings find themselves in a kafkaesque world where many rules are undocumented or can only be found in old mailing list discussions. Apache and this list especially can feel like a really hostile place for newbies. Regards, Matthias - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
We clearly differ with our view if how much is delegated from board to IPMC. The amount of work the board does on x podlings weach month is less than the work they do on x TLPs. That is without the IPMC addressing issues that come up every now and again. We can go into detail if it becomes necessary. There are problems of efficiency, that is what I believe is the problem. But as I said we need to agree to differ at this time. Where I differ from you is that not when each podling had 3 active and engaged mentors, all would be good in those cases. That is rarely the case though. Therefore your proposal means either a reduction in the number of accepted projects (problem: how do we know which to accept), a reduction in the quality of TLPs (problem: reduction in perceived quality of all ASF brands), or a bigger oversight role for the board (problem: will the board accept this?) For me the first option is the only outcome that can be considered. If that is a desired change (it is not for me) then your proposal is great. For now though the change I want is a more efficient IPMC and this is why we need to agree to differ at least until more of the IPMC have a stomach for radical change. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 01:49, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 4:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. What specific actions would fall to the board in my proposal [1] outside of what the board already does for PMCs? I count a total of 0 in the right hand column of my table. Being specific myself: 1. Directors review the IPMC report, and are charged (at least the Director shepherd for the Incubator is; but so are other board members) with reviewing the podlings present in the Incubator report. There was discussion before about removing specific podling reports, and only leaving the summary -- this was nixed. Directors are still charged with reviewing podling individual reports, same as they are with actual project reports. Thus, if you say there are no more podlings, as I do in my proposal, please define, specifically, where the extra work is? 2. We always wax at the ASF about there being extremely little centralized authority. Oh, there's a problem? The board can't fix that -- it's a bazooka! Fix it yourself, PMC! OK, so with that said, what's the problem then by saying, no more podlings, there are simply PMCs? New projects come in to the ASF via steps 3-5 in my proposal -- through discussion on general@incubator that includes discussions of merit, community, etc, guided by the existing Incubator documentation. When a VOTE is ready, the board VOTEs on the incoming project(s). This is true today. Incubator podlings are *not officially endorsed projects of the ASF* until they are turned into TLPs by board resolution. Again, so what's changed? What's even more hilarious and illustrative of the guise towards decentralization is that there have been discussions within this very same thread that instead of telling the board there are problems with the Incubator, that we should fix them ourselves here. Hehe. Kind of a reflexive but powerful look in the mirror about the desire to move *away* from centralization. Thus, I ask, why do we have a *centralized* (fake Board) IPMC if the goal of the ASF is for the PMCs to be self governing? The Apache way is intimated through tribal knowledge of its members. Activeness of a member (and 3 of them on a PMC) is something that the board is aware of, so these things will get caught at project creation, and/or through personnel additions incrementally. That being said, I think we ought to let this drop for now. Benson has stated he wants to address the specific problem that brought all this up again. For now lets agree to differ. No problem -- I think we're closer than it seems, but yes, I'm fine with dropping it. Cheers, Chris [1] http://wiki.apache.org/incubator/IncubatorDeconstructionProposal Ross On 28 March 2013 16:19, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make
Re: Vote on personal matters: majority vote vs consensus
I don't suppose that I could convince folks to start a new thread for the topic, 'The Incubator is a FAIL', and leave this one to the patch for the decision process? On Fri, Mar 29, 2013 at 6:09 AM, Ross Gardler rgard...@opendirective.comwrote: We clearly differ with our view if how much is delegated from board to IPMC. The amount of work the board does on x podlings weach month is less than the work they do on x TLPs. That is without the IPMC addressing issues that come up every now and again. We can go into detail if it becomes necessary. There are problems of efficiency, that is what I believe is the problem. But as I said we need to agree to differ at this time. Where I differ from you is that not when each podling had 3 active and engaged mentors, all would be good in those cases. That is rarely the case though. Therefore your proposal means either a reduction in the number of accepted projects (problem: how do we know which to accept), a reduction in the quality of TLPs (problem: reduction in perceived quality of all ASF brands), or a bigger oversight role for the board (problem: will the board accept this?) For me the first option is the only outcome that can be considered. If that is a desired change (it is not for me) then your proposal is great. For now though the change I want is a more efficient IPMC and this is why we need to agree to differ at least until more of the IPMC have a stomach for radical change. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 01:49, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 4:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. What specific actions would fall to the board in my proposal [1] outside of what the board already does for PMCs? I count a total of 0 in the right hand column of my table. Being specific myself: 1. Directors review the IPMC report, and are charged (at least the Director shepherd for the Incubator is; but so are other board members) with reviewing the podlings present in the Incubator report. There was discussion before about removing specific podling reports, and only leaving the summary -- this was nixed. Directors are still charged with reviewing podling individual reports, same as they are with actual project reports. Thus, if you say there are no more podlings, as I do in my proposal, please define, specifically, where the extra work is? 2. We always wax at the ASF about there being extremely little centralized authority. Oh, there's a problem? The board can't fix that -- it's a bazooka! Fix it yourself, PMC! OK, so with that said, what's the problem then by saying, no more podlings, there are simply PMCs? New projects come in to the ASF via steps 3-5 in my proposal -- through discussion on general@incubator that includes discussions of merit, community, etc, guided by the existing Incubator documentation. When a VOTE is ready, the board VOTEs on the incoming project(s). This is true today. Incubator podlings are *not officially endorsed projects of the ASF* until they are turned into TLPs by board resolution. Again, so what's changed? What's even more hilarious and illustrative of the guise towards decentralization is that there have been discussions within this very same thread that instead of telling the board there are problems with the Incubator, that we should fix them ourselves here. Hehe. Kind of a reflexive but powerful look in the mirror about the desire to move *away* from centralization. Thus, I ask, why do we have a *centralized* (fake Board) IPMC if the goal of the ASF is for the PMCs to be self governing? The Apache way is intimated through tribal knowledge of its members. Activeness of a member (and 3 of them on a PMC) is something that the board is aware of, so these things will get caught at project creation, and/or through personnel additions incrementally. That being said, I think we ought to let this drop for now. Benson has stated he wants to address the specific problem that brought all this up again. For now lets agree to differ. No problem -- I think we're closer than it seems, but yes, I'm fine with dropping it. Cheers, Chris [1] http://wiki.apache.org/incubator/IncubatorDeconstructionProposal Ross On 28 March 2013 16:19, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote
Incubator Deconstruction (was Vote on personal matters: majority vote vs consensus)
[Note subject line change for Benson] Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, March 29, 2013 3:09 AM To: general general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus We clearly differ with our view if how much is delegated from board to IPMC. The amount of work the board does on x podlings weach month is less than the work they do on x TLPs. Yeah I guess this is the crux. I respect your opinion, but honestly feel strongly for my own too :) No worries, such is life. That is without the IPMC addressing issues that come up every now and again. We can go into detail if it becomes necessary. Well yeah that's the point. I've gone into details, ad nauseum. They are literally extrapolated on my proposal and in numerous email threads too. I've done the work to document them. There are problems of efficiency, that is what I believe is the problem. But as I said we need to agree to differ at this time. Without knowing the specifics, saying that we differ I don't think is Constructive at least on my end. IOW, I don't think it's anything that a bar camp, with some good IPA wouldn't solve ^_^ Where I differ from you is that not when each podling had 3 active and engaged mentors, all would be good in those cases. That is rarely the case though. Ross, if it's rarely the case, I wouldn't be here helping 6 podlings at the moment in the Incubator (and those podlings wouldn't have all come to me asking me directly to help mentor). I'm talking about 14 podlings over the last few years. Here you go I'll enumerate: ---graduated OODT Airavata SIS Gora Lucy Giraph cTAKES Any23 ---current HDT Mesos Tez Knox Climate Tajo Therefore your proposal means either a reduction in the number of accepted projects (problem: how do we know which to accept), a reduction in the quality of TLPs (problem: reduction in perceived quality of all ASF brands), or a bigger oversight role for the board (problem: will the board accept this?) The above is anecdotal -- the board has scaled from 90 projects a few years ago to 137 currently over that time. Based on that, I don't think any of the 3 above suggestions will happen. Cheers, Chris For me the first option is the only outcome that can be considered. If that is a desired change (it is not for me) then your proposal is great. For now though the change I want is a more efficient IPMC and this is why we need to agree to differ at least until more of the IPMC have a stomach for radical change. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 01:49, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 4:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. What specific actions would fall to the board in my proposal [1] outside of what the board already does for PMCs? I count a total of 0 in the right hand column of my table. Being specific myself: 1. Directors review the IPMC report, and are charged (at least the Director shepherd for the Incubator is; but so are other board members) with reviewing the podlings present in the Incubator report. There was discussion before about removing specific podling reports, and only leaving the summary -- this was nixed. Directors are still charged with reviewing podling individual reports, same as they are with actual project reports. Thus, if you say there are no more podlings, as I do in my proposal, please define, specifically, where the extra work is? 2. We always wax at the ASF about there being extremely little centralized authority. Oh, there's a problem? The board can't fix that -- it's a bazooka! Fix it yourself, PMC! OK, so with that said, what's the problem then by saying, no more podlings, there are simply PMCs? New projects come in to the ASF via steps 3-5 in my proposal -- through discussion on general@incubator that includes discussions of merit, community, etc, guided by the existing Incubator documentation. When a VOTE is ready, the board VOTEs on the incoming project(s). This is true today. Incubator podlings are *not officially endorsed projects of the ASF* until they are turned into TLPs by board resolution. Again, so what's changed? What's even more hilarious and illustrative of the guise towards decentralization is that there have been discussions within this very same thread that instead of telling the board
Re: Incubator Deconstruction (was Vote on personal matters: majority vote vs consensus)
Chris, The fundamental issue is that I don't agree the IPMC needs deconstructing. I believe it finds it difficult to come to a decision when unusual circumstance arises, but most of the time it does fine. I don't accept that using yourself as an example of how we can find sufficient mentors for all new entries is evidence that your proposal will scale and thus address the concerns I have expressed. You are not a typical mentor, most of us need sleep. I don't believe this topic needs debating as I don't believe the incubation process is broken. Your proposal doesn't actually solve the core problems of whether policy says this or that or whether best practice is this or that - which ultimately is the only thing the IPMC gets bogged down in. Your proposal simply moves all the hard parts to the membership and thus to the board. Moving problems does not solve them. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 16:51, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: [Note subject line change for Benson] Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, March 29, 2013 3:09 AM To: general general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus We clearly differ with our view if how much is delegated from board to IPMC. The amount of work the board does on x podlings weach month is less than the work they do on x TLPs. Yeah I guess this is the crux. I respect your opinion, but honestly feel strongly for my own too :) No worries, such is life. That is without the IPMC addressing issues that come up every now and again. We can go into detail if it becomes necessary. Well yeah that's the point. I've gone into details, ad nauseum. They are literally extrapolated on my proposal and in numerous email threads too. I've done the work to document them. There are problems of efficiency, that is what I believe is the problem. But as I said we need to agree to differ at this time. Without knowing the specifics, saying that we differ I don't think is Constructive at least on my end. IOW, I don't think it's anything that a bar camp, with some good IPA wouldn't solve ^_^ Where I differ from you is that not when each podling had 3 active and engaged mentors, all would be good in those cases. That is rarely the case though. Ross, if it's rarely the case, I wouldn't be here helping 6 podlings at the moment in the Incubator (and those podlings wouldn't have all come to me asking me directly to help mentor). I'm talking about 14 podlings over the last few years. Here you go I'll enumerate: ---graduated OODT Airavata SIS Gora Lucy Giraph cTAKES Any23 ---current HDT Mesos Tez Knox Climate Tajo Therefore your proposal means either a reduction in the number of accepted projects (problem: how do we know which to accept), a reduction in the quality of TLPs (problem: reduction in perceived quality of all ASF brands), or a bigger oversight role for the board (problem: will the board accept this?) The above is anecdotal -- the board has scaled from 90 projects a few years ago to 137 currently over that time. Based on that, I don't think any of the 3 above suggestions will happen. Cheers, Chris For me the first option is the only outcome that can be considered. If that is a desired change (it is not for me) then your proposal is great. For now though the change I want is a more efficient IPMC and this is why we need to agree to differ at least until more of the IPMC have a stomach for radical change. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 01:49, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 4:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. What specific actions would fall to the board in my proposal [1] outside of what the board already does for PMCs? I count a total of 0 in the right hand column of my table. Being specific myself: 1. Directors review the IPMC report, and are charged (at least the Director shepherd for the Incubator is; but so are other board members) with reviewing the podlings present in the Incubator report. There was discussion before about removing specific podling reports, and only leaving the summary -- this was nixed. Directors are still
Re: Incubator Deconstruction (was Vote on personal matters: majority vote vs consensus)
Hahah, we all need sleep? What?!! :) Take care duder, we'll spent some cycles doing other emails while we let this sit. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, March 29, 2013 10:11 AM To: general general@incubator.apache.org Subject: Re: Incubator Deconstruction (was Vote on personal matters: majority vote vs consensus) Chris, The fundamental issue is that I don't agree the IPMC needs deconstructing. I believe it finds it difficult to come to a decision when unusual circumstance arises, but most of the time it does fine. I don't accept that using yourself as an example of how we can find sufficient mentors for all new entries is evidence that your proposal will scale and thus address the concerns I have expressed. You are not a typical mentor, most of us need sleep. I don't believe this topic needs debating as I don't believe the incubation process is broken. Your proposal doesn't actually solve the core problems of whether policy says this or that or whether best practice is this or that - which ultimately is the only thing the IPMC gets bogged down in. Your proposal simply moves all the hard parts to the membership and thus to the board. Moving problems does not solve them. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 16:51, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: [Note subject line change for Benson] Hi Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Friday, March 29, 2013 3:09 AM To: general general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus We clearly differ with our view if how much is delegated from board to IPMC. The amount of work the board does on x podlings weach month is less than the work they do on x TLPs. Yeah I guess this is the crux. I respect your opinion, but honestly feel strongly for my own too :) No worries, such is life. That is without the IPMC addressing issues that come up every now and again. We can go into detail if it becomes necessary. Well yeah that's the point. I've gone into details, ad nauseum. They are literally extrapolated on my proposal and in numerous email threads too. I've done the work to document them. There are problems of efficiency, that is what I believe is the problem. But as I said we need to agree to differ at this time. Without knowing the specifics, saying that we differ I don't think is Constructive at least on my end. IOW, I don't think it's anything that a bar camp, with some good IPA wouldn't solve ^_^ Where I differ from you is that not when each podling had 3 active and engaged mentors, all would be good in those cases. That is rarely the case though. Ross, if it's rarely the case, I wouldn't be here helping 6 podlings at the moment in the Incubator (and those podlings wouldn't have all come to me asking me directly to help mentor). I'm talking about 14 podlings over the last few years. Here you go I'll enumerate: ---graduated OODT Airavata SIS Gora Lucy Giraph cTAKES Any23 ---current HDT Mesos Tez Knox Climate Tajo Therefore your proposal means either a reduction in the number of accepted projects (problem: how do we know which to accept), a reduction in the quality of TLPs (problem: reduction in perceived quality of all ASF brands), or a bigger oversight role for the board (problem: will the board accept this?) The above is anecdotal -- the board has scaled from 90 projects a few years ago to 137 currently over that time. Based on that, I don't think any of the 3 above suggestions will happen. Cheers, Chris For me the first option is the only outcome that can be considered. If that is a desired change (it is not for me) then your proposal is great. For now though the change I want is a more efficient IPMC and this is why we need to agree to differ at least until more of the IPMC have a stomach for radical change. Ross Sent from a mobile device, please excuse mistakes and brevity On 29 Mar 2013 01:49, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general
Re: Incubator Deconstruction (was Vote on personal matters: majority vote vs consensus)
On Fri, Mar 29, 2013 at 11:38 AM, Shane Curcuru a...@shanecurcuru.org wrote: 2) more direct leadership that seeks basic consensus on very specific and clear new changes, but doesn't let discussions get weighed down with too many options, or stalled by a relative handful of -0s. The hard work of forging consensus is wasted when the IPMC does not follow its own rules. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 11:19 PM, Doug Cutting cutt...@apache.org wrote: On Wed, Mar 27, 2013 at 3:11 PM, Niall Pemberton niall.pember...@gmail.com wrote: I think it should be 3/4 majority. I agree that supermajority would be better than simple majority here. Moving to simple majority seems too radical... +1 on requiring 3/4 majority on Incubator PMC votes on personal matters (PMC members elections etc.). And as Benson says we should require a DISCUSS thread on our private list (where those votes happen as well) before such votes, to help build consensus. I don't think we need more changes than that. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 9:21 AM, Bertrand Delacretaz bdelacre...@apache.org wrote: On Wed, Mar 27, 2013 at 11:19 PM, Doug Cutting cutt...@apache.org wrote: On Wed, Mar 27, 2013 at 3:11 PM, Niall Pemberton niall.pember...@gmail.com wrote: I think it should be 3/4 majority. I agree that supermajority would be better than simple majority here. Moving to simple majority seems too radical... +1 on requiring 3/4 majority on Incubator PMC votes on personal matters (PMC members elections etc.). +1, makes sense. And as Benson says we should require a DISCUSS thread on our private list (where those votes happen as well) before such votes, to help build consensus. +1 Christian - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
It appears to me that we have a consensus here on using a majority system with a 3/4 supermajority. I'd like to establish the existence of this consensus with a minimum of fuss, and begin to stop wasting everyone's time. Our goal here is to achieve consensus, not to hold votes. So, I'm going to treat this as a lazy consensus issues. I'm going to watch this thread for an additional 72 hours. If anyone objects to this consensus, send along a brief summary of your objection with a -1. If there are, in fact, substantive objections, I'll organize a separate vote process to resolve this. Please do not debate objections here, we'll do that later if we have to. Everyone's had a fair opportunity to state their opinion, so the only thing we're doing now is ensuring that no one is harboring a serious objection that I have somehow overlooked. Acquiescing in this process does not prejudice the discussion of changing the PMC. On Thu, Mar 28, 2013 at 4:21 AM, Bertrand Delacretaz bdelacre...@apache.org wrote: On Wed, Mar 27, 2013 at 11:19 PM, Doug Cutting cutt...@apache.org wrote: On Wed, Mar 27, 2013 at 3:11 PM, Niall Pemberton niall.pember...@gmail.com wrote: I think it should be 3/4 majority. I agree that supermajority would be better than simple majority here. Moving to simple majority seems too radical... +1 on requiring 3/4 majority on Incubator PMC votes on personal matters (PMC members elections etc.). And as Benson says we should require a DISCUSS thread on our private list (where those votes happen as well) before such votes, to help build consensus. I don't think we need more changes than that. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 12:44 PM, Benson Margulies bimargul...@gmail.com wrote: It appears to me that we have a consensus here on using a majority system with a 3/4 supermajority. I'd like to establish the existence of this consensus with a minimum of fuss, and begin to stop wasting everyone's time. Our goal here is to achieve consensus, not to hold votes. So, I'm going to treat this as a lazy consensus issues. I'm going to watch this thread for an additional 72 hours. If anyone objects to this consensus, send along a brief summary of your objection with a -1. If there are, in fact, substantive objections, I'll organize a separate vote process to resolve this. Please do not debate objections here, we'll do that later if we have to. Everyone's had a fair opportunity to state their opinion, so the only thing we're doing now is ensuring that no one is harboring a serious objection that I have somehow overlooked. Acquiescing in this process does not prejudice the discussion of changing the PMC. I'd prefer we stick with consensus but not enough to vote against this. However as no one has mentioned this I do think its worth pointing out that when using supermajority instead of consensus or simple majority then the phrasing of the vote becomes important. Eg. Say 16 people participate in a vote then: Throw the guy out needs 5 -1s to stop it happening Let the guy stay needs 12 +1s to make it happen. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Waah. Look this just DEFINES consensus as 75% instead of the old 100%. It doesn't throw consensus out the window. Please stop with all of these exaggerations and try to self-moderate- half of the volume in these debates is all you talking to yourself. On Mar 28, 2013, at 9:18 AM, ant elder ant.el...@gmail.com wrote: On Thu, Mar 28, 2013 at 12:44 PM, Benson Margulies bimargul...@gmail.com wrote: It appears to me that we have a consensus here on using a majority system with a 3/4 supermajority. I'd like to establish the existence of this consensus with a minimum of fuss, and begin to stop wasting everyone's time. Our goal here is to achieve consensus, not to hold votes. So, I'm going to treat this as a lazy consensus issues. I'm going to watch this thread for an additional 72 hours. If anyone objects to this consensus, send along a brief summary of your objection with a -1. If there are, in fact, substantive objections, I'll organize a separate vote process to resolve this. Please do not debate objections here, we'll do that later if we have to. Everyone's had a fair opportunity to state their opinion, so the only thing we're doing now is ensuring that no one is harboring a serious objection that I have somehow overlooked. Acquiescing in this process does not prejudice the discussion of changing the PMC. I'd prefer we stick with consensus but not enough to vote against this. However as no one has mentioned this I do think its worth pointing out that when using supermajority instead of consensus or simple majority then the phrasing of the vote becomes important. Eg. Say 16 people participate in a vote then: Throw the guy out needs 5 -1s to stop it happening Let the guy stay needs 12 +1s to make it happen. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
No what it means Joe is that who chooses the wording of the vote gets a lot of control the outcome. ...ant On Thu, Mar 28, 2013 at 1:25 PM, Joseph Schaefer joe_schae...@yahoo.com wrote: Waah. Look this just DEFINES consensus as 75% instead of the old 100%. It doesn't throw consensus out the window. Please stop with all of these exaggerations and try to self-moderate- half of the volume in these debates is all you talking to yourself. On Mar 28, 2013, at 9:18 AM, ant elder ant.el...@gmail.com wrote: On Thu, Mar 28, 2013 at 12:44 PM, Benson Margulies bimargul...@gmail.com wrote: It appears to me that we have a consensus here on using a majority system with a 3/4 supermajority. I'd like to establish the existence of this consensus with a minimum of fuss, and begin to stop wasting everyone's time. Our goal here is to achieve consensus, not to hold votes. So, I'm going to treat this as a lazy consensus issues. I'm going to watch this thread for an additional 72 hours. If anyone objects to this consensus, send along a brief summary of your objection with a -1. If there are, in fact, substantive objections, I'll organize a separate vote process to resolve this. Please do not debate objections here, we'll do that later if we have to. Everyone's had a fair opportunity to state their opinion, so the only thing we're doing now is ensuring that no one is harboring a serious objection that I have somehow overlooked. Acquiescing in this process does not prejudice the discussion of changing the PMC. I'd prefer we stick with consensus but not enough to vote against this. However as no one has mentioned this I do think its worth pointing out that when using supermajority instead of consensus or simple majority then the phrasing of the vote becomes important. Eg. Say 16 people participate in a vote then: Throw the guy out needs 5 -1s to stop it happening Let the guy stay needs 12 +1s to make it happen. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
No more so than they already had. Sent from my iPhone On Mar 28, 2013, at 9:56 AM, ant elder ant.el...@gmail.com wrote: No what it means Joe is that who chooses the wording of the vote gets a lot of control the outcome. ...ant On Thu, Mar 28, 2013 at 1:25 PM, Joseph Schaefer joe_schae...@yahoo.com wrote: Waah. Look this just DEFINES consensus as 75% instead of the old 100%. It doesn't throw consensus out the window. Please stop with all of these exaggerations and try to self-moderate- half of the volume in these debates is all you talking to yourself. On Mar 28, 2013, at 9:18 AM, ant elder ant.el...@gmail.com wrote: On Thu, Mar 28, 2013 at 12:44 PM, Benson Margulies bimargul...@gmail.com wrote: It appears to me that we have a consensus here on using a majority system with a 3/4 supermajority. I'd like to establish the existence of this consensus with a minimum of fuss, and begin to stop wasting everyone's time. Our goal here is to achieve consensus, not to hold votes. So, I'm going to treat this as a lazy consensus issues. I'm going to watch this thread for an additional 72 hours. If anyone objects to this consensus, send along a brief summary of your objection with a -1. If there are, in fact, substantive objections, I'll organize a separate vote process to resolve this. Please do not debate objections here, we'll do that later if we have to. Everyone's had a fair opportunity to state their opinion, so the only thing we're doing now is ensuring that no one is harboring a serious objection that I have somehow overlooked. Acquiescing in this process does not prejudice the discussion of changing the PMC. I'd prefer we stick with consensus but not enough to vote against this. However as no one has mentioned this I do think its worth pointing out that when using supermajority instead of consensus or simple majority then the phrasing of the vote becomes important. Eg. Say 16 people participate in a vote then: Throw the guy out needs 5 -1s to stop it happening Let the guy stay needs 12 +1s to make it happen. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Sent from a mobile device, please excuse mistakes and brevity On 28 Mar 2013 14:04, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hi Ross, On 3/27/13 11:33 AM, Ross Gardler rgard...@opendirective.com wrote: On 27 Mar 2013 16:43, Greg Reddin gred...@gmail.com wrote: On Wed, Mar 27, 2013 at 11:18 AM, Ross Gardler rgard...@opendirective.comwrote: Perhaps it would make sense to see how the model that has scaled well for the foundation can be applied here: ... [snip] ... Why can't the IPMC work like that? Well, to a large extent it does. Here are the same items expressed from the perspective of the IPMC and its relationship with PPMCs. Your proposal is not terribly different from proposals like what Chris floated a year or so ago. Yours adds a layer of entities. Chris' removes a layer. I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. Ross IOW, my proposal recognized that the IPMC model is broken, and that the 100s of e-mails and threads that constantly go on are better spent doing actual work. It also recognized that the work done by the Incubator has been fantastic over the years. So much so, that having a specific committee to manage/steward execution of its processes and procedures no longer is needed. In fact, it's that super committee that stands in the way of progress in current days, rather than the way that it used to enable progress. My proposal involves these steps: 1. Move the Incubator process/policy/documentation, etc., to ComDev http://wiki.apache.org/incubator/ComDev - I agree with gstein on this. I think it could be maintained by the ASF community folks there, and updated over time. But it's not vastly or rapidly changing really anymore. 2. Discharge the Incubator PMC and the role of Incubator VP -- pat everyone on the back, go have a beer, watch the big game together, whatever. Call it a success, not a failure. 3. Suggest at the board level that an Incubation process still exists at Apache, in the same way that it exists today. New projects write a proposal, the proposal is VOTEd on by the board at the board's next monthly meeting, and those that cannot be are QUEUED for the next meeting, or VOTEd on during out of board inbetween time on board@. Refer those wanting to Incubate at Apache to the existing Incubator documentation maintained by the ComDev http://wiki.apache.org/incubator/ComDev community. Tell them to ask questions there, about the process, about what to do, or if ideas make sense. But not to VOTE on whether they are accepted or not. 4. Require every podling to have at least 3 ASF members on it, similar to the current Incubator process. 5. Operate podlings exactly the same as a TLP. There is a chair. There is a committee. Committee members have binding VOTEs on releases. Pretty simple. 5 steps. Anyhoo, yeah, happy to clarify further, though I must say I have zero hope or feeling that I'm going to be able to keep up with all these e-mails. Yours was honestly the first one I flagged in days since I saw Greg R's mention of my proposal and your response to it. Take care. Cheers, Chris Mine simply proposes a way to break the occasional deadlocks in the IPMC by using an existing, but informal, layer - shepherds. Nothing else changes. The IPMC is not broken, it just has growing pains. Ross Chris' proposal is essentially that Incubating projects become PMCs instead of PPMCs. IIRC, his proposal still incorporated mentoring and oversight to ensure that incubating projects are operating according to Apache principles. Perhaps there's a model where incubating PMCs report directly to the board, as with Chris' proposal, but with a dotted-line reporting structure to a mentoring body. This mentoring body would be responsible for vetting releases and new committers as the IPMC does now. But its role would be more of a guiding role than an oversight role. The podlings I've participated in would not have suffered from such a model. Flex, for example, had a board member
Re: Vote on personal matters: majority vote vs consensus
Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings are TLPs (well the ones that graduate within a fixed set of time as Sam was trying to measure before, or simply point out that is) or you still believe that there is oversight within the IPMC. I personally don't. That's why I wrote the proposal. And I think that's at least evident to me and more than a few others that that's the problem here and that's why I don't think the Incubator should exist anymore in its current form and should be deconstructed :) Thanks for your comments and conversation and for listening. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 2:17 PM, Joseph Schaefer joe_schae...@yahoo.com wrote: No more so than they already had. It does Joe, let me give you a more clear example. Lets imagine i've done something that you deem shows i'm a terrible incubator mentor, and its not the first time. There's a big debate within the PMC, no clear consensus, so in the end you say enough is enough and call a vote to remove me so i don't do any more damage - a vote on removing ant. With this new supermajority approach you'd need 75% or more of voters to agree with you to get me gone. Alternatively, you could say enough is enough and to end the debate you're going to call a vote to demonstrate i've the PMCs support - a vote on letting ant stay on. That sounds like you're being nice, but in fact you're being clever, because now you only need 25% of voters to vote -1 and i'm gone. 25% is much easier to get than the 75% in the previous vote example. The problem here i think is that the policy is being rushed through and people are only thinking of the voting people to the PMC case whereas the policy is going to apply much more widely that - to all votes on personal matters - so lots of scope for wording bias. Isn't this the case? Apologies in advance for the noise if my logic is screwed up somewhere. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 5:29 PM, ant elder ant.el...@gmail.com wrote: ...With this new supermajority approach you'd need 75% or more of voters to agree with you to get me gone. Alternatively, you could say enough is enough and to end the debate you're going to call a vote to demonstrate i've the PMCs support - a vote on letting ant stay on. That sounds like you're being nice, but in fact you're being clever, because now you only need 25% of voters to vote -1 and i'm gone IMO the goal of a 3/4 supermajority is to make sure the result is sufficiently clear to mean rough consensus, so we probably need to say that a vote with = 75% yes passes a vote with = 25% yes does not pass a vote 25% and 75% is failed, no consensus emerges and no action is taken -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 9:29 AM, ant elder ant.el...@gmail.com wrote: Alternatively, you could say enough is enough and to end the debate you're going to call a vote to demonstrate i've the PMCs support - a vote on letting ant stay on. That sounds like you're being nice, but in fact you're being clever, because now you only need 25% of voters to vote -1 and i'm gone. This sounds like a vote to support the status quo, which isn't something we normally do. Votes are typically phrased as changes to the status quo, where a +1 indicates a vote for the change and a -1 indicates a vote to keep things as they are. So there's a natural valance to voting and the phrasing of the ballot should not alter that. Doug - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Yes your logic is flawed- what you are actually arguing for is majority voting not consensus voting, and bringing the criterion down from 100% to 75% only helps mitigate your concerns. As Doug points out, votes are structured away from the status quo- we don't ever vote to continue on with previously agreed to issues just to circumvent the voting process. From: ant elder ant.el...@gmail.com To: general@incubator.apache.org Sent: Thursday, March 28, 2013 12:29 PM Subject: Re: Vote on personal matters: majority vote vs consensus On Thu, Mar 28, 2013 at 2:17 PM, Joseph Schaefer joe_schae...@yahoo.com wrote: No more so than they already had. It does Joe, let me give you a more clear example. Lets imagine i've done something that you deem shows i'm a terrible incubator mentor, and its not the first time. There's a big debate within the PMC, no clear consensus, so in the end you say enough is enough and call a vote to remove me so i don't do any more damage - a vote on removing ant. With this new supermajority approach you'd need 75% or more of voters to agree with you to get me gone. Alternatively, you could say enough is enough and to end the debate you're going to call a vote to demonstrate i've the PMCs support - a vote on letting ant stay on. That sounds like you're being nice, but in fact you're being clever, because now you only need 25% of voters to vote -1 and i'm gone. 25% is much easier to get than the 75% in the previous vote example. The problem here i think is that the policy is being rushed through and people are only thinking of the voting people to the PMC case whereas the policy is going to apply much more widely that - to all votes on personal matters - so lots of scope for wording bias. Isn't this the case? Apologies in advance for the noise if my logic is screwed up somewhere.  ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 9:42 AM, Doug Cutting cutt...@apache.org wrote: This sounds like a vote to support the status quo, which isn't something we normally do. The original proposal was limited to VOTEs on personnel issues (misspelled as personal). Has that changed? I hope not. One of the downsides of resolving this issue through lazy consensus is that no codified proposal with clear language has emerged, setting the stage for later disputes about just what we agreed to. I anticipate that except in the rarest of circumstances, there will be no confusion about what constitutes the status quo for personnel issues. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Would anyone be willing to write up the text that we would post on the web site someplace to document a procedure for voting upon IPMC membership that reflects this discussion? Perhaps we could then lazily converge upon that? On Thu, Mar 28, 2013 at 1:16 PM, Marvin Humphrey mar...@rectangular.comwrote: On Thu, Mar 28, 2013 at 9:42 AM, Doug Cutting cutt...@apache.org wrote: This sounds like a vote to support the status quo, which isn't something we normally do. The original proposal was limited to VOTEs on personnel issues (misspelled as personal). Has that changed? I hope not. One of the downsides of resolving this issue through lazy consensus is that no codified proposal with clear language has emerged, setting the stage for later disputes about just what we agreed to. I anticipate that except in the rarest of circumstances, there will be no confusion about what constitutes the status quo for personnel issues. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Mar 28, 2013, at 9:19 AM, Mattmann, Chris A (388J) wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. My experience as a shepherd shows that you can in fact recognize podling issues and eventually get them to the point where graduation of one kind or another happens. Two examples: EasyAnt graduating to Apache Ant. Etch finally graduated with a small, but sufficient PMC. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings are TLPs (well the ones that graduate within a fixed set of time as Sam was trying to measure before, or simply point out that is) or you still believe that there is oversight within the IPMC. I don't think it is an either / or. The current amount of oversight for any podling is a function of mentors, current IPMC dynamics, and real life influences. The shepherds serve a purpose as more of a divining rod into that dynamic, many solutions are possible, and the podling needs to be pushed into making choices. Regards, Dave I personally don't. That's why I wrote the proposal. And I think that's at least evident to me and more than a few others that that's the problem here and that's why I don't think the Incubator should exist anymore in its current form and should be deconstructed :) Thanks for your comments and conversation and for listening. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. That being said, I think we ought to let this drop for now. Benson has stated he wants to address the specific problem that brought all this up again. For now lets agree to differ. Ross On 28 March 2013 16:19, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings are TLPs (well the ones that graduate within a fixed set of time as Sam was trying to measure before, or simply point out that is) or you still believe that there is oversight within the IPMC. I personally don't. That's why I wrote the proposal. And I think that's at least evident to me and more than a few others that that's the problem here and that's why I don't think the Incubator should exist anymore in its current form and should be deconstructed :) Thanks for your comments and conversation and for listening. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org -- Ross Gardler (@rgardler) Programme Leader (Open Development) OpenDirective http://opendirective.com
Re: Vote on personal matters: majority vote vs consensus
Chris, Your position is that the IPMC fails to supervise. The consensus of the IPMC is that this is not true. Otherwise, someone would be reading the monthly report and objecting to the failure to report 'failure' to the board. If you want to change minds about this, you might need to come up with some concrete evidence of actual failure: bad commits, bad doings on mailing lists, etc. What the IPMC now does is use the shepherd process to compensate for mentor weakness. It's not perfect. Under your plan, instead, the board would have to cope, directly, with 'starter' projects suffering from inevitable attrition -- or the Foundation would need to start many less projects, as only those who could attract very strongly committed foundation members could start. Strangely, I find myself thinking that the logical endpoint of your line of thought is to use the existing projects as incubators for new ones. A healthy TLP could, I think, easily ride herd on a small group of people starting something new in a related technology area, and then calve them off. I write this mostly to make the point that there are many possible places to take the conversation if you start from the premise that no variation on the existing scheme is workable. I personally don't know how, organizationally, to reach that conclusion, especially insofar as the recent disfunction is not about supervision, it's _merely_ about sorting out who can be a member and thus a mentor. On Thu, Mar 28, 2013 at 6:38 PM, Dave Fisher dave2w...@comcast.net wrote: On Mar 28, 2013, at 9:19 AM, Mattmann, Chris A (388J) wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. My experience as a shepherd shows that you can in fact recognize podling issues and eventually get them to the point where graduation of one kind or another happens. Two examples: EasyAnt graduating to Apache Ant. Etch finally graduated with a small, but sufficient PMC. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings are TLPs (well the ones that graduate within a fixed set of time as Sam was trying to measure before, or simply point out that is) or you still believe that there is oversight within the IPMC. I don't think it is an either / or. The current amount of oversight for any podling is a function of mentors, current IPMC dynamics, and real life influences. The shepherds serve a purpose as more of a divining rod into that dynamic, many solutions are possible, and the podling needs to be pushed into making choices. Regards, Dave I personally don't. That's why I wrote the proposal. And I think that's at least evident to me and more than a few others that that's the problem here and that's why I don't think the Incubator should exist anymore in its current form and should be deconstructed :) Thanks for your comments and conversation and for listening. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 4:30 PM, Benson Margulies bimargul...@gmail.com wrote: Your position is that the IPMC fails to supervise. The consensus of the IPMC is that this is not true. Otherwise, someone would be reading the monthly report and objecting to the failure to report 'failure' to the board. If your statement were true, then someone would make the assertions you're making. If you want to change minds about this, you might need to come up with some concrete evidence of actual failure: bad commits, bad doings on mailing lists, etc. Is this a question of standing, where material harm needs to be demonstrated? The IPMC is needlessly inefficient and abusive of its podlings. Novel compliance mechanisms are literally invented and argued about on general@ during podlings' release votes.[1] The cultural clashes that Chris's proposal refers to generate huge amounts of traffic on general@ and private@, as ASF members argue the semantics of core concepts. And it's not just edge-case legal issues; some are as basic as the definition of veto. These discussions create needless confusion and deeply resented churn for podlings. The asymmetry in power teaches submissiveness to ASF members, rather than independence and self-sufficiency. There are, in truth, *many* active interpretations of the Apache Way practiced across the ASF. Reconciling them is not the mission of the incubator. Putting esoteric debates on the critical path of new projects is absurd and harmful. [1] http://s.apache.org/lFI What the IPMC now does is use the shepherd process to compensate for mentor weakness. It's not perfect. Under your plan, instead, the board would have to cope, directly, with 'starter' projects suffering from inevitable attrition -- or the Foundation would need to start many less projects, as only those who could attract very strongly committed foundation members could start. The incubator doesn't deal with this effectively, either. To take one example, Chukwa's retirement was a fiasco. there are many possible places to take the conversation if you start from the premise that no variation on the existing scheme is workable. I personally don't know how, organizationally, to reach that conclusion, especially insofar as the recent disfunction is not about supervision, it's _merely_ about sorting out who can be a member and thus a mentor. The recent discussion is about the mentor role. The ongoing dysfunction is about supervision, and the IPMC failing to discharge its purpose efficiently. It exists to put its podlings through a curriculum that transfers some cultural norms, makes podlings aware of resources, and establishes clean licensing. Even if it succeeds well enough not to be disbanded by the board, its inefficiency is worth correcting. -C On Thu, Mar 28, 2013 at 6:38 PM, Dave Fisher dave2w...@comcast.net wrote: On Mar 28, 2013, at 9:19 AM, Mattmann, Chris A (388J) wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. My experience as a shepherd shows that you can in fact recognize podling issues and eventually get them to the point where graduation of one kind or another happens. Two examples: EasyAnt graduating to Apache Ant. Etch finally graduated with a small, but sufficient PMC. So, that's my whole point. You either agree with me that there is no IPMC oversight at the
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 11:23 AM, Benson Margulies bimargul...@gmail.com wrote: Would anyone be willing to write up the text that we would post on the web site someplace to document a procedure for voting upon IPMC membership that reflects this discussion? Perhaps we could then lazily converge upon that? Patch below. I will commit unless someone objects. For what it's worth, the only passage I could find on the website describing the process of joining the IPMC was in the Glossary of the Mentor guide: http://incubator.apache.org/guides/mentor.html#who-ipmc IPMC Member All activities of incubated projects are supervised by the Incubator Project Management Committee, or IPMC. Releases and new committers of incubated projects must be approved by vote of the IPMC. Incubated project mentors must be IPMC members. Any member of the Apache Software Foundation may join the IPMC by request. The committee may also grant membership to others by vote; typically, this allows a PMC member of some other project to serve as a mentor. I'm tempted to strike that passage as it will now be redundant and we have too much glossary stuff on the website IMO. Marvin Humphrey Index: content/incubation/Roles_and_Responsibilities.xml === --- content/incubation/Roles_and_Responsibilities.xml (revision 1451965) +++ content/incubation/Roles_and_Responsibilities.xml (working copy) @@ -118,6 +118,9 @@ PMC may choose to provide an experienced Mentor to assist the main Mentor in discharging their duty. /p +pIndividuals may be nominated to join the IPMC after a vote which +passes with more than 3/4 of those voting. Additionally, any Member of the +Apache Software Foundation may join the IPMC by request. pSee also: /p ul - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Hi Dave, -Original Message- From: Dave Fisher dave2w...@comcast.net Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 3:38 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus On Mar 28, 2013, at 9:19 AM, Mattmann, Chris A (388J) wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. My experience as a shepherd shows that you can in fact recognize podling issues and eventually get them to the point where graduation of one kind or another happens. Two examples: EasyAnt graduating to Apache Ant. Etch finally graduated with a small, but sufficient PMC. I wasn't stating that you can't do this :) In fact, you are precisely the example of what I'm talking about when I say, 'good' mentors. You actively volunteered for a new system created by Jukka, to care about other podlings and sign off on their reports, and monitor them. Awesome sauce. Great job. OTOH, I've *never* volunteered to be a shepherd, b/c honestly I think it's an additional name/responsibility that's fairly meaningless. I've been signing off on other podling reports that I see for years. As have other mentors, Joe S used to do that -- Ant and Alex and others have done that too. Even before there was a name, 'shepherd' for this. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings are TLPs (well the ones that graduate within a fixed set of time as Sam was trying to measure before, or simply point out that is) or you still believe that there is oversight within the IPMC. I don't think it is an either / or. The current amount of oversight for any podling is a function of mentors, current IPMC dynamics, and real life influences. The shepherds serve a purpose as more of a divining rod into that dynamic, many solutions are possible, and the podling needs to be pushed into making choices. Sure, I agree with that. And rather than generically stating there are choices, and we have many to choose amongst, I've actually written my thoughts down about a choice, a series of observations, a proposed solution and deconstruction of the Incubator. Cheers, Chris Regards, Dave I personally don't. That's why I wrote the proposal. And I think that's at least evident to me and more than a few others that that's the problem here and that's why I don't think the Incubator should exist anymore in its current form and should be deconstructed :) Thanks for your comments and conversation and for listening. Cheers, Chris ++ Chris Mattmann, Ph.D. Senior Computer Scientist NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA Office: 171-266B, Mailstop: 171-246 Email: chris.a.mattm...@nasa.gov WWW: http://sunset.usc.edu/~mattmann/ ++ Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA ++ - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h
Re: Vote on personal matters: majority vote vs consensus
Hey Ross, -Original Message- From: Ross Gardler rgard...@opendirective.com Reply-To: general@incubator.apache.org general@incubator.apache.org Date: Thursday, March 28, 2013 4:20 PM To: general@incubator.apache.org general@incubator.apache.org Subject: Re: Vote on personal matters: majority vote vs consensus I do not agree there is no IPMC oversight. The IPMC performs many actions each month which would fall to the board if the IPMC were disbanded. That is why the IPMC submits a board report. What specific actions would fall to the board in my proposal [1] outside of what the board already does for PMCs? I count a total of 0 in the right hand column of my table. Being specific myself: 1. Directors review the IPMC report, and are charged (at least the Director shepherd for the Incubator is; but so are other board members) with reviewing the podlings present in the Incubator report. There was discussion before about removing specific podling reports, and only leaving the summary -- this was nixed. Directors are still charged with reviewing podling individual reports, same as they are with actual project reports. Thus, if you say there are no more podlings, as I do in my proposal, please define, specifically, where the extra work is? 2. We always wax at the ASF about there being extremely little centralized authority. Oh, there's a problem? The board can't fix that -- it's a bazooka! Fix it yourself, PMC! OK, so with that said, what's the problem then by saying, no more podlings, there are simply PMCs? New projects come in to the ASF via steps 3-5 in my proposal -- through discussion on general@incubator that includes discussions of merit, community, etc, guided by the existing Incubator documentation. When a VOTE is ready, the board VOTEs on the incoming project(s). This is true today. Incubator podlings are *not officially endorsed projects of the ASF* until they are turned into TLPs by board resolution. Again, so what's changed? What's even more hilarious and illustrative of the guise towards decentralization is that there have been discussions within this very same thread that instead of telling the board there are problems with the Incubator, that we should fix them ourselves here. Hehe. Kind of a reflexive but powerful look in the mirror about the desire to move *away* from centralization. Thus, I ask, why do we have a *centralized* (fake Board) IPMC if the goal of the ASF is for the PMCs to be self governing? The Apache way is intimated through tribal knowledge of its members. Activeness of a member (and 3 of them on a PMC) is something that the board is aware of, so these things will get caught at project creation, and/or through personnel additions incrementally. That being said, I think we ought to let this drop for now. Benson has stated he wants to address the specific problem that brought all this up again. For now lets agree to differ. No problem -- I think we're closer than it seems, but yes, I'm fine with dropping it. Cheers, Chris [1] http://wiki.apache.org/incubator/IncubatorDeconstructionProposal Ross On 28 March 2013 16:19, Mattmann, Chris A (388J) chris.a.mattm...@jpl.nasa.gov wrote: Hey Ross, I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Please let me respectfully disagree with your interpretation of my Incubator deconstruction proposal [1]. In fact, it does not make the board legally responsible in any different way than the board is currently responsible for its plethora of TLPs -- IOW, it doesn't change a thing. It basically suggests that incoming projects can simply fast track to (t)LPs from the get go, so long as they have = 3 ASF members present to help execute and manage the Incubator process which still exists in my proposed deconstruction. My point is that all the oversight currently provided by the IPMC would have to be provided by the board. We already know that having three mentors does not guarantee adequate support for podlings. I guess I would ask what oversight? There is no global IPMC oversight. Ever since Joe's experiment, and even before, the podlings that get through the Incubator (and I've taken quite a few now, and recently, so I think I can speak from a position of experience here within the last few years), are the ones that have active mentors and *distributed*, not *centralized* oversight. IOW, I'm not seeing any IPMC oversight at the moment. I'm seeing good mentors, located in each podling, distributed, that get podlings through. Those that stall well they need help. Usually the help is debated endlessly, and not solved, or simply solved with more active/better mentors. So, that's my whole point. You either agree with me that there is no IPMC oversight at the moment (for years now), and that really podlings
Re: Vote on personal matters: majority vote vs consensus
On Fri, Mar 29, 2013 at 2:19 AM, Marvin Humphrey mar...@rectangular.com wrote: On Thu, Mar 28, 2013 at 11:23 AM, Benson Margulies bimargul...@gmail.com wrote: Would anyone be willing to write up the text that we would post on the web site someplace to document a procedure for voting upon IPMC membership that reflects this discussion? Perhaps we could then lazily converge upon that? Patch below. I will commit unless someone objects. +1 to the patch For what it's worth, the only passage I could find on the website describing the process of joining the IPMC was in the Glossary of the Mentor guide: http://incubator.apache.org/guides/mentor.html#who-ipmc IPMC Member All activities of incubated projects are supervised by the Incubator Project Management Committee, or IPMC. Releases and new committers of incubated projects must be approved by vote of the IPMC. Incubated project mentors must be IPMC members. Any member of the Apache Software Foundation may join the IPMC by request. The committee may also grant membership to others by vote; typically, this allows a PMC member of some other project to serve as a mentor. I'm tempted to strike that passage as it will now be redundant and we have too much glossary stuff on the website IMO. +1 Cheers Christian Marvin Humphrey Index: content/incubation/Roles_and_Responsibilities.xml === --- content/incubation/Roles_and_Responsibilities.xml (revision 1451965) +++ content/incubation/Roles_and_Responsibilities.xml (working copy) @@ -118,6 +118,9 @@ PMC may choose to provide an experienced Mentor to assist the main Mentor in discharging their duty. /p +pIndividuals may be nominated to join the IPMC after a vote which +passes with more than 3/4 of those voting. Additionally, any Member of the +Apache Software Foundation may join the IPMC by request. pSee also: /p ul - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org -- http://www.grobmeier.de https://www.timeandbill.de - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Mon, Mar 25, 2013 at 7:52 AM, ant elder ant.el...@gmail.com wrote: Your second suggestion sounds like the thing to do to me - separating IPMC-ship and Mentor-ship - that would solve several of the problems we've being having including this one, it would open up a much bigger pool of potential mentors, and IPMC'ers would get much more visibility of people as they work here which should make the PMC voting easier. Ok some people sound like they like this suggestion, and some others have expressed concern about the lack of a binding vote. I'd like to try this, perhaps as a sort of experiment like we've done for other changes in the past. I'll post a more articulate proposal in a new thread but doesn't anyone else have any feelings about this either way? ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 8:35 AM, ant elder ant.el...@gmail.com wrote: On Mon, Mar 25, 2013 at 7:52 AM, ant elder ant.el...@gmail.com wrote: ...Your second suggestion sounds like the thing to do to me - separating IPMC-ship and Mentor-ship... ...I'd like to try this, perhaps as a sort of experiment like we've done for other changes in the past. I'll post a more articulate proposal in a new thread but doesn't anyone else have any feelings about this either way?... As I said before I'm currently against having mentors who are not Incubator PMC members, because it creates more options that require more work when checking things and because mentors need to have binding votes. But as I said elsewhere, only idiots never change their minds...let's see your proposal. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Hi, As I said before I'm currently against having mentors who are not Incubator PMC members, As an aside it seems (and please correct me if I'm mistaken) in order to become a IPMC member you first need to be an Apache member (see bottom of [1]).This may exclude people with practical experience of being involved in a podling and making it a top level project who want to help out with other podlings. Of course they can help out in ways that doesn't involve them being a mentor but (IMO) the barrier to entry seems a little high. I also been unable to find any info on the incubator site on shepherds and what's the process to become involved there. If someone could point me in the right direction I'd be grateful. Thanks, Justin 1. http://incubator.apache.org/guides/participation.html - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 11:44 AM, Justin Mclean justinmcl...@gmail.com wrote: ...As an aside it seems (and please correct me if I'm mistaken) in order to become a IPMC member you first need to be an Apache member (see bottom of [1])... you don't - Apache members can become IPMC members just by asking, but others can also be elected as incubator PMC members. We do have some such mentors currently. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013, at 10:44 AM, Justin Mclean wrote: Hi, As I said before I'm currently against having mentors who are not Incubator PMC members, As an aside it seems (and please correct me if I'm mistaken) in order to become a IPMC member you first need to be an Apache member (see bottom of [1]).This may exclude people with practical experience of being involved in a podling and making it a top level project who want to help out with other podlings. Of course they can help out in ways that doesn't involve them being a mentor but (IMO) the barrier to entry seems a little high. I also been unable to find any info on the incubator site on shepherds and what's the process to become involved there. If someone could point me in the right direction I'd be grateful. Shepherding is a new thing, loosely based upon an approach that the board has used for some time. I'm sure it is largely undocumented. I suspect that generally speaking it is incubator PMC members that do shepherding, but if someone reads reports and provides useful analysis, I suspect this will be greatly appreciated. Upayavira - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
I suppose that as chair I ought to be heard from here. I've been off for Passover for a bit. In my view, the IPMC manifests two problems. I'd like to label them as 'operational' and 'decision-making'. This thread is about decision-making, but with some people seeing using terms like 'disfunctional', I think it's important to keep 'function' in context. Operationally, we 'started' 1.3 years ago with an acute problem of under-supervised and/or 'malingering' podlings. Under Jukka's leadership, we made a series of incremental changes that have considerably improved the situation. On the other hand, the recent influx of many new podlings worries me, because 'improved' is not the same as 'fixed'. And I'm not entirely sure that 'fixed' is possible. I'd like to see us find more incremental changes that help further, and I'd like them to scale via some mechanism other than my own personal time. I see this as a reason to put more thought into shepherds and champions. But I don't see this situation as 'disfunctional'. On the decision-making front, recent phenomena have demonstrated to me that this group is not succeeding in applying consensus process to decision making. I could write five paragraphs on what that process is and what it requires, but I'm not inclined to. I support the proposal here to apply majority rules to IPMC membership. When consensus process fails here, we have endless email threads. Many of us find these stressful, time-consuming, and disheartening. Under the proposal at hand, we'd still DISCUSS, and I'd hope that we would all try to be thoughtful and constructive and look for ways to agree. However, after a certain amount of discussion, there would be a vote, and that would be that. If this 'works' -- if people here find that it strikes a good balance between seeking consensus and limiting time and stress, we're good. It might not work. Or it might 'work', but some might feel that this large, diffuse, group, operating by majority rules is either inconsistent with Apache policy or a bad example for the podlings. In which case someone might want to dust off the proposals from 1.3 years ago that offered more or less radical alternatives. I'm personally not ready to go there yet. On Wed, Mar 27, 2013 at 7:03 AM, Bertrand Delacretaz bdelacre...@apache.org wrote: On Wed, Mar 27, 2013 at 11:44 AM, Justin Mclean justinmcl...@gmail.com wrote: ...As an aside it seems (and please correct me if I'm mistaken) in order to become a IPMC member you first need to be an Apache member (see bottom of [1])... you don't - Apache members can become IPMC members just by asking, but others can also be elected as incubator PMC members. We do have some such mentors currently. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 11:55 AM, Benson Margulies bimargul...@gmail.com wrote: Or it might 'work', but some might feel that this large, diffuse, group, operating by majority rules is either inconsistent with Apache policy or a bad example for the podlings. Thats more how i see it. Using consensus instead of majority votes is one of the main things that makes the ASF special, so we should avoid changing from that if we can, for both those reasons you suggest. And there is nothing that needs this changed presently so IMHO its not necessary to change anything. I'd like to see us find more incremental changes that help further Ok, i propose we have an experiment [1] where we try having a mentor or two who are not PMC members. Have some other experienced mentors helping to make sure nothing unfixable can go wrong, and just see how it goes for a while, reporting each month with the status. Maybe it will fail and we'll know not to try that again, but i think and hope it should work ok. If it does work ok then changing to make this more common would avoid a lot of the times were we need to make new not well known people PMC members or have these debates, which solves part of the problem here. If that works then there are some further small incremental changes we can make which will also help with reasons for people needing to join the PMC, and those will help with the PMC size and disparity issues. Wouldn't a small experiment like this be worth trying before we drop something as fundamental as using consensus? ...ant [1] We tried a similar experiment with the change back to allow poddlings to vote for their own committers again which was similarly controversial before it was tried - http://mail-archives.apache.org/mod_mbox/incubator-general/201008.mbox/%3C974721.3581.qm%40web54401.mail.re2.yahoo.com%3E - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
The incubator is currently of a scale that means it can no longer operate as a standard consensus driven PMC. It is not that much smaller than the TLPs part of the foundation. Perhaps it would make sense to see how the model that has scaled well for the foundation can be applied here: ASF Members elect a board board delegates to PMCs (no micromanagement, just delegation of broad goals) PMCs answer to the board Board answers to the Membership PMCs delegate to committers (no micromanagement just broad goals) Committers do the work and make localised decisions Committers are answerable to the PMCs Why can't the IPMC work like that? Well, to a large extent it does. Here are the same items expressed from the perspective of the IPMC and its relationship with PPMCs. No equivalent of ASF Members elect a board IPMC members delegate to the PPMCs (including Mentors) PPMCs answer to the IPMC IPMC answers to the Board PPMCs delegate to committers (no micromanagement just broad goals) Committers do the work and make localised decisions Committers are answerable to the PMCs Where this model breaks down is that the IPMC is too large to act with consistency and efficiency. The IPMC is kind of like the ASF Membership. The ASF Members list, for those unfortunate enough to be on it, occasionally blows up with problems for which there is no single, perfect solution. The IPMC list does just the same. The rest of the time everything runs smoothly. The way the foundation survives these episodes is to have an elected board that is responsible for navigating through those situations. They don't seek the one single right answer (because there isn't one), instead they listen to a set of options, each flawed in some way an they pick the one deemed to be least flawed from the foundations point of view. The board breaks deadlocks when they occur, the rest of the time the board does nothing more than provide sufficient oversight to ensure the PMCs can operate as Apache PMCs (it gets out of the way). The IPMC has no equivalent. So when it hits a problem with multiple potential outcomes - all flawed, it ends up in deadlock. Look back of the archives, this is happening increasingly often as the IPMC grows. We need a way to efficiently break deadlocks. Let me ask a question... Why shouldn't the IPMC create an equivalent to the one item in the above governance structure that is missing today. That is why shouldn't it have an equivalent of ASF Members elect a board. It would something like IPMC elect 9-15 Shepherds. These Shepherds are responsible for ensuring that the IPMC membership is heard and that decisions are made for the good of the IPMC. They approve membership of the IPMC, they approve project entry/graduation/retirement but, and this is critical, they report to the IPMC. Most of the time their role is one of delegation to the PPMCs, occasionally their role is to break a deadlock by listening to the IPMC and making the best decision it can. This need not change any other line in the existing governance. It need not change the IPMC relationship with the board. Mentors will still be IPMC members, with binding votes, etc. Since the Shepherds are accountable to the IPMC they must seek to do the right thing, or they will be replaced. Just as the Board can be replaced at any time if the Membership so desires. Of course, we could argue that the IPMC chair already has the authority to do all this. Indeed they do. However, expecting one individual to keep track of all the activity in our podlings is unreasonable. Furthermore, it is harder for one individual to make the hard decisions and suffer the mud slinging from those that don't like the outcome. Just a thought of course, my solution is as flawed as anyone elses and I look to the IPMC Chair to find the good enough solution that will allow us to move on (sorry Benson). Ross On 27 March 2013 11:55, Benson Margulies bimargul...@gmail.com wrote: I suppose that as chair I ought to be heard from here. I've been off for Passover for a bit. In my view, the IPMC manifests two problems. I'd like to label them as 'operational' and 'decision-making'. This thread is about decision-making, but with some people seeing using terms like 'disfunctional', I think it's important to keep 'function' in context. Operationally, we 'started' 1.3 years ago with an acute problem of under-supervised and/or 'malingering' podlings. Under Jukka's leadership, we made a series of incremental changes that have considerably improved the situation. On the other hand, the recent influx of many new podlings worries me, because 'improved' is not the same as 'fixed'. And I'm not entirely sure that 'fixed' is possible. I'd like to see us find more incremental changes that help further, and I'd like them to scale via some mechanism other than my own personal time. I see this as a reason to put more thought into shepherds and champions. But I don't see this situation as
Re: Vote on personal matters: majority vote vs consensus
On 27 March 2013 15:54, ant elder ant.el...@gmail.com wrote: On Wed, Mar 27, 2013 at 11:55 AM, Benson Margulies bimargul...@gmail.com wrote: Ok, i propose we have an experiment [1] where we try having a mentor or two who are not PMC members. Have some other experienced mentors helping to make sure nothing unfixable can go wrong, and just see how it goes for a while, reporting each month with the status. In general I'm all for more mentors, but if they are not IPMC members where will the binding votes for releases come from? Wouldn't a small experiment like this be worth trying before we drop something as fundamental as using consensus? There are so many issues being discussed in parallel I'm not sure which problem this experiment is designed to solve? Ross
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 4:23 PM, Ross Gardler rgard...@opendirective.com wrote: On 27 March 2013 15:54, ant elder ant.el...@gmail.com wrote: On Wed, Mar 27, 2013 at 11:55 AM, Benson Margulies bimargul...@gmail.com wrote: Ok, i propose we have an experiment [1] where we try having a mentor or two who are not PMC members. Have some other experienced mentors helping to make sure nothing unfixable can go wrong, and just see how it goes for a while, reporting each month with the status. In general I'm all for more mentors, but if they are not IPMC members where will the binding votes for releases come from? From PMC members on general@ which is what happens for lots of poddlings now anyway as most don't get three mentors voting on their releases and have to come to general@. I'm not suggesting that we start off the experiment with a poddling having all non-PMC mentors so there would be other mentors with binding votes being active in the poddling. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 11:18 AM, Ross Gardler rgard...@opendirective.comwrote: Perhaps it would make sense to see how the model that has scaled well for the foundation can be applied here: ... [snip] ... Why can't the IPMC work like that? Well, to a large extent it does. Here are the same items expressed from the perspective of the IPMC and its relationship with PPMCs. Your proposal is not terribly different from proposals like what Chris floated a year or so ago. Yours adds a layer of entities. Chris' removes a layer. Chris' proposal is essentially that Incubating projects become PMCs instead of PPMCs. IIRC, his proposal still incorporated mentoring and oversight to ensure that incubating projects are operating according to Apache principles. Perhaps there's a model where incubating PMCs report directly to the board, as with Chris' proposal, but with a dotted-line reporting structure to a mentoring body. This mentoring body would be responsible for vetting releases and new committers as the IPMC does now. But its role would be more of a guiding role than an oversight role. The podlings I've participated in would not have suffered from such a model. Flex, for example, had a board member and 2 ASF members as mentors. The IPMC did not have to provide much insight beyond what we provided. The Incubator provides the following benefits to incubating projects: * mentoring on ASF principles and procedures * vetting of releases * help growing community * a temporary community while podling community develops It seems to me that a model like what Chris has proposed or what I am proposing could still provide all those benefits without the bureaucracy of a super-IPMC. Greg
Re: Vote on personal matters: majority vote vs consensus
On 27 Mar 2013 16:43, Greg Reddin gred...@gmail.com wrote: On Wed, Mar 27, 2013 at 11:18 AM, Ross Gardler rgard...@opendirective.comwrote: Perhaps it would make sense to see how the model that has scaled well for the foundation can be applied here: ... [snip] ... Why can't the IPMC work like that? Well, to a large extent it does. Here are the same items expressed from the perspective of the IPMC and its relationship with PPMCs. Your proposal is not terribly different from proposals like what Chris floated a year or so ago. Yours adds a layer of entities. Chris' removes a layer. I disagree. Chris' proposal removes the IPMC thus making the board legally responsible for everything that committee does today. Yes it replaces it with an oversight body, but how does that scale? Mine simply proposes a way to break the occasional deadlocks in the IPMC by using an existing, but informal, layer - shepherds. Nothing else changes. The IPMC is not broken, it just has growing pains. Ross Chris' proposal is essentially that Incubating projects become PMCs instead of PPMCs. IIRC, his proposal still incorporated mentoring and oversight to ensure that incubating projects are operating according to Apache principles. Perhaps there's a model where incubating PMCs report directly to the board, as with Chris' proposal, but with a dotted-line reporting structure to a mentoring body. This mentoring body would be responsible for vetting releases and new committers as the IPMC does now. But its role would be more of a guiding role than an oversight role. The podlings I've participated in would not have suffered from such a model. Flex, for example, had a board member and 2 ASF members as mentors. The IPMC did not have to provide much insight beyond what we provided. The Incubator provides the following benefits to incubating projects: * mentoring on ASF principles and procedures * vetting of releases * help growing community * a temporary community while podling community develops It seems to me that a model like what Chris has proposed or what I am proposing could still provide all those benefits without the bureaucracy of a super-IPMC. Greg
Re: Vote on personal matters: majority vote vs consensus
Hi, this is a very interesting proposal. Let me ask a few questions. On Wed, Mar 27, 2013 at 5:18 PM, Ross Gardler rgard...@opendirective.com wrote: Why shouldn't the IPMC create an equivalent to the one item in the above governance structure that is missing today. That is why shouldn't it have an equivalent of ASF Members elect a board. It would something like IPMC elect 9-15 Shepherds. These Shepherds are responsible for ensuring that the IPMC membership is heard and that decisions are made for the good of the IPMC. They approve membership of the IPMC, they approve project entry/graduation/retirement but, and this is critical, they report to the IPMC. Most of the time their role is one of delegation to the PPMCs, occasionally their role is to break a deadlock by listening to the IPMC and making the best decision it can. it sounds a little bit as the Shepherds would be the true PMC of the IPMC. It's interesting, because you would bypass the problem with IPMC members have binding votes, therefore everybody needs to be there. The term PMC is already in use, so you create a new term and give these group actual PMC-ship. Therefore it would be logic to me if the IPMC chair would be elected out of the Shepherds. Also if we follow this, Shepherds doesn't sound so nice. Actually it is a kind of Board. This need not change any other line in the existing governance. It need not change the IPMC relationship with the board. Mentors will still be IPMC members, with binding votes, etc. Since the Shepherds are accountable to the IPMC they must seek to do the right thing, or they will be replaced. Just as the Board can be replaced at any time if the Membership so desires. Nice. Of course, we could argue that the IPMC chair already has the authority to do all this. Indeed they do. However, expecting one individual to keep track of all the activity in our podlings is unreasonable. Furthermore, it is harder for one individual to make the hard decisions and suffer the mud slinging from those that don't like the outcome. I don't think a chair should act with authority. I also have not seen any place where it is said that a chair should dictate decisions or so. You would also bypass this problem with making multiple chairs. I believe it would relax the chair (generally spoken) a lot. And you bypass the problem that a strong chair is needed, which seems to be the case with 172 members. I was having the mentor = committer model in mind, which is similar to Chris model (i think). At least I would have the problem with the binding votes. One could only bypass it to give Mentor a binding vote as exception to other projects. Your proposal has an exception (the Shepherds) too. I like it though, but I am still not convinced if we need another layer of people - or if we just minimize the IPMC and give Mentors (= Committers) that binding vote. Just a thought of course, my solution is as flawed as anyone elses and I look to the IPMC Chair to find the good enough solution that will allow us to move on (sorry Benson). I look to all IPMC members. Cheers Christian Ross On 27 March 2013 11:55, Benson Margulies bimargul...@gmail.com wrote: I suppose that as chair I ought to be heard from here. I've been off for Passover for a bit. In my view, the IPMC manifests two problems. I'd like to label them as 'operational' and 'decision-making'. This thread is about decision-making, but with some people seeing using terms like 'disfunctional', I think it's important to keep 'function' in context. Operationally, we 'started' 1.3 years ago with an acute problem of under-supervised and/or 'malingering' podlings. Under Jukka's leadership, we made a series of incremental changes that have considerably improved the situation. On the other hand, the recent influx of many new podlings worries me, because 'improved' is not the same as 'fixed'. And I'm not entirely sure that 'fixed' is possible. I'd like to see us find more incremental changes that help further, and I'd like them to scale via some mechanism other than my own personal time. I see this as a reason to put more thought into shepherds and champions. But I don't see this situation as 'disfunctional'. On the decision-making front, recent phenomena have demonstrated to me that this group is not succeeding in applying consensus process to decision making. I could write five paragraphs on what that process is and what it requires, but I'm not inclined to. I support the proposal here to apply majority rules to IPMC membership. When consensus process fails here, we have endless email threads. Many of us find these stressful, time-consuming, and disheartening. Under the proposal at hand, we'd still DISCUSS, and I'd hope that we would all try to be thoughtful and constructive and look for ways to agree. However, after a certain amount of discussion, there would be a vote, and that would be that. If this 'works' -- if people
Re: Vote on personal matters: majority vote vs consensus
Sent from a mobile device, please excuse mistakes and brevity On 27 Mar 2013 20:12, Christian Grobmeier grobme...@gmail.com wrote: Hi, this is a very interesting proposal. Let me ask a few questions. On Wed, Mar 27, 2013 at 5:18 PM, Ross Gardler rgard...@opendirective.com wrote: Why shouldn't the IPMC create an equivalent to the one item in the above governance structure that is missing today. That is why shouldn't it have an equivalent of ASF Members elect a board. It would something like IPMC elect 9-15 Shepherds. These Shepherds are responsible for ensuring that the IPMC membership is heard and that decisions are made for the good of the IPMC. They approve membership of the IPMC, they approve project entry/graduation/retirement but, and this is critical, they report to the IPMC. Most of the time their role is one of delegation to the PPMCs, occasionally their role is to break a deadlock by listening to the IPMC and making the best decision it can. it sounds a little bit as the Shepherds would be the true PMC of the IPMC. No. The IPMC would be the true IPMC, but they elect a representative body to make the IPMC function more efficiently. That body remains answerable to the IPMC. This is important because the mentors should be the ones making recommendations about graduation, report approval etc. More importantly only PMC members have binding votes on releases. The shepherds delegate all this to the PPMC (and its mentors). The shepherds only act to ensure the PPMC is capable, unhindered and healthy. Also if we follow this, Shepherds doesn't sound so nice. Actually it is a kind of Board. Yes it is. I avoided new names to prevent the false impression that this is adding new layers. The Shepherds already do almost everything I'm suggesting. The only addition is for them to be the ones who break consensus deadlocks. Why them? Because their role means they have more visibility into the breadth of the Incubator than many IPMC members. I don't think a chair should act with authority. Sometimes it is necessary, but I agree that it should be very rare. The problem with the IPMC is that it is needed too frequently. My proposal, as you observe, provides a representative group, answerable to the IPMC, to build consensus on these occasions. I am still not convinced if we need another layer of people - or if we just minimize the IPMC and give Mentors (= Committers) that binding vote. I have no doubt that my proposal is imperfect, let's find the holes and see if they are pluggable. Chris' model is similar in some ways as has been observed. I'm yet to see how the scale problem will be solved, but maybe I'm remembering the proposal incorrectly, In think a fundamental difference between Chris' radical model and mine is revolution vs evolution. Personally I think the current IPMC model works well 98% of the time, so evolution is appropriate, Just a thought of course, my solution is as flawed as anyone elses and I look to the IPMC Chair to find the good enough solution that will allow us to move on (sorry Benson). I look to all IPMC members. I meant Benson should coordinate, not dictate :-) Thanks for your useful critique. Ross Cheers Christian Ross On 27 March 2013 11:55, Benson Margulies bimargul...@gmail.com wrote: I suppose that as chair I ought to be heard from here. I've been off for Passover for a bit. In my view, the IPMC manifests two problems. I'd like to label them as 'operational' and 'decision-making'. This thread is about decision-making, but with some people seeing using terms like 'disfunctional', I think it's important to keep 'function' in context. Operationally, we 'started' 1.3 years ago with an acute problem of under-supervised and/or 'malingering' podlings. Under Jukka's leadership, we made a series of incremental changes that have considerably improved the situation. On the other hand, the recent influx of many new podlings worries me, because 'improved' is not the same as 'fixed'. And I'm not entirely sure that 'fixed' is possible. I'd like to see us find more incremental changes that help further, and I'd like them to scale via some mechanism other than my own personal time. I see this as a reason to put more thought into shepherds and champions. But I don't see this situation as 'disfunctional'. On the decision-making front, recent phenomena have demonstrated to me that this group is not succeeding in applying consensus process to decision making. I could write five paragraphs on what that process is and what it requires, but I'm not inclined to. I support the proposal here to apply majority rules to IPMC membership. When consensus process fails here, we have endless email threads. Many of us find these stressful, time-consuming, and disheartening. Under the proposal at hand, we'd still DISCUSS, and I'd hope that we would all try to be thoughtful and
Re: Vote on personal matters: majority vote vs consensus
The first thing I'd like to do, coordination-wise, is to call a vote on the proposal to decide things by majority. I think that this would help with some of the problems we hit, and we can meanwhile continue to discuss larger structural changes. On Wed, Mar 27, 2013 at 4:48 PM, Ross Gardler rgard...@opendirective.comwrote: Sent from a mobile device, please excuse mistakes and brevity On 27 Mar 2013 20:12, Christian Grobmeier grobme...@gmail.com wrote: Hi, this is a very interesting proposal. Let me ask a few questions. On Wed, Mar 27, 2013 at 5:18 PM, Ross Gardler rgard...@opendirective.com wrote: Why shouldn't the IPMC create an equivalent to the one item in the above governance structure that is missing today. That is why shouldn't it have an equivalent of ASF Members elect a board. It would something like IPMC elect 9-15 Shepherds. These Shepherds are responsible for ensuring that the IPMC membership is heard and that decisions are made for the good of the IPMC. They approve membership of the IPMC, they approve project entry/graduation/retirement but, and this is critical, they report to the IPMC. Most of the time their role is one of delegation to the PPMCs, occasionally their role is to break a deadlock by listening to the IPMC and making the best decision it can. it sounds a little bit as the Shepherds would be the true PMC of the IPMC. No. The IPMC would be the true IPMC, but they elect a representative body to make the IPMC function more efficiently. That body remains answerable to the IPMC. This is important because the mentors should be the ones making recommendations about graduation, report approval etc. More importantly only PMC members have binding votes on releases. The shepherds delegate all this to the PPMC (and its mentors). The shepherds only act to ensure the PPMC is capable, unhindered and healthy. Also if we follow this, Shepherds doesn't sound so nice. Actually it is a kind of Board. Yes it is. I avoided new names to prevent the false impression that this is adding new layers. The Shepherds already do almost everything I'm suggesting. The only addition is for them to be the ones who break consensus deadlocks. Why them? Because their role means they have more visibility into the breadth of the Incubator than many IPMC members. I don't think a chair should act with authority. Sometimes it is necessary, but I agree that it should be very rare. The problem with the IPMC is that it is needed too frequently. My proposal, as you observe, provides a representative group, answerable to the IPMC, to build consensus on these occasions. I am still not convinced if we need another layer of people - or if we just minimize the IPMC and give Mentors (= Committers) that binding vote. I have no doubt that my proposal is imperfect, let's find the holes and see if they are pluggable. Chris' model is similar in some ways as has been observed. I'm yet to see how the scale problem will be solved, but maybe I'm remembering the proposal incorrectly, In think a fundamental difference between Chris' radical model and mine is revolution vs evolution. Personally I think the current IPMC model works well 98% of the time, so evolution is appropriate, Just a thought of course, my solution is as flawed as anyone elses and I look to the IPMC Chair to find the good enough solution that will allow us to move on (sorry Benson). I look to all IPMC members. I meant Benson should coordinate, not dictate :-) Thanks for your useful critique. Ross Cheers Christian Ross On 27 March 2013 11:55, Benson Margulies bimargul...@gmail.com wrote: I suppose that as chair I ought to be heard from here. I've been off for Passover for a bit. In my view, the IPMC manifests two problems. I'd like to label them as 'operational' and 'decision-making'. This thread is about decision-making, but with some people seeing using terms like 'disfunctional', I think it's important to keep 'function' in context. Operationally, we 'started' 1.3 years ago with an acute problem of under-supervised and/or 'malingering' podlings. Under Jukka's leadership, we made a series of incremental changes that have considerably improved the situation. On the other hand, the recent influx of many new podlings worries me, because 'improved' is not the same as 'fixed'. And I'm not entirely sure that 'fixed' is possible. I'd like to see us find more incremental changes that help further, and I'd like them to scale via some mechanism other than my own personal time. I see this as a reason to put more thought into shepherds and champions. But I don't see this situation as 'disfunctional'. On the decision-making front, recent phenomena have demonstrated to me that this group is not succeeding in applying consensus process to
Re: Vote on personal matters: majority vote vs consensus
On Mon, Mar 25, 2013 at 12:12 AM, Roman Shaposhnik r...@apache.org wrote: On Sat, Mar 23, 2013 at 1:24 PM, Ted Dunning ted.dunn...@gmail.com wrote: One alternative to going for full-on majority voting is to recognize that a larger group is much more likely to have noisy vetoes by requiring that successful votes have n positive votes and m negative votes subject to some condition on n and m. Majority requires n m, strict Apache consensus requires n = 3 and m == 0. It is easy to imagine other conditions such as n = 4 and m = 2 which still have some of the flavor of consensus in that a minority can block a decision, but allow forward progress even with constant naysayers or occasional random vetoes. Personally, I'd suggest keeping these options in our backpocket and turning back to considering them in case a simple majority proposal runs into an opposition somehow. At this point, I'd rather try a simple solution first. I was in favour of simple majority - but a vote passing with, for example 9+1 and 8-1 is as bad IMO as a vote failing because of alot of +1 and only one -1. So I've changed my mind on this - I think it should be 3/4 majority. This avoids a small minority stopping something, but also doesn't completely throw out consensus. Niall Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
This whole exercise is pointless. Just drop the notion of vetoes for all IPMC votes and carry on as before. Sent from my iPhone On Mar 27, 2013, at 6:11 PM, Niall Pemberton niall.pember...@gmail.com wrote: On Mon, Mar 25, 2013 at 12:12 AM, Roman Shaposhnik r...@apache.org wrote: On Sat, Mar 23, 2013 at 1:24 PM, Ted Dunning ted.dunn...@gmail.com wrote: One alternative to going for full-on majority voting is to recognize that a larger group is much more likely to have noisy vetoes by requiring that successful votes have n positive votes and m negative votes subject to some condition on n and m. Majority requires n m, strict Apache consensus requires n = 3 and m == 0. It is easy to imagine other conditions such as n = 4 and m = 2 which still have some of the flavor of consensus in that a minority can block a decision, but allow forward progress even with constant naysayers or occasional random vetoes. Personally, I'd suggest keeping these options in our backpocket and turning back to considering them in case a simple majority proposal runs into an opposition somehow. At this point, I'd rather try a simple solution first. I was in favour of simple majority - but a vote passing with, for example 9+1 and 8-1 is as bad IMO as a vote failing because of alot of +1 and only one -1. So I've changed my mind on this - I think it should be 3/4 majority. This avoids a small minority stopping something, but also doesn't completely throw out consensus. Niall Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Wed, Mar 27, 2013 at 3:11 PM, Niall Pemberton niall.pember...@gmail.com wrote: I think it should be 3/4 majority. I agree that supermajority would be better than simple majority here. Moving to simple majority seems too radical. Over time it's more prone to building a PMC that cannot easily agree on things. If consensus has proven too difficult to reach for a group this large, then softening it a bit to supermajority seems like a better first step then moving all the way to simple majority. Doug - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Thu, Mar 28, 2013 at 12:11 AM, Niall Pemberton niall.pember...@gmail.com wrote: On Mon, Mar 25, 2013 at 12:12 AM, Roman Shaposhnik r...@apache.org wrote: On Sat, Mar 23, 2013 at 1:24 PM, Ted Dunning ted.dunn...@gmail.com wrote: One alternative to going for full-on majority voting is to recognize that a larger group is much more likely to have noisy vetoes by requiring that successful votes have n positive votes and m negative votes subject to some condition on n and m. Majority requires n m, strict Apache consensus requires n = 3 and m == 0. It is easy to imagine other conditions such as n = 4 and m = 2 which still have some of the flavor of consensus in that a minority can block a decision, but allow forward progress even with constant naysayers or occasional random vetoes. Personally, I'd suggest keeping these options in our backpocket and turning back to considering them in case a simple majority proposal runs into an opposition somehow. At this point, I'd rather try a simple solution first. I was in favour of simple majority - but a vote passing with, for example 9+1 and 8-1 is as bad IMO as a vote failing because of alot of +1 and only one -1. So I've changed my mind on this - I think it should be 3/4 majority. This avoids a small minority stopping something, but also doesn't completely throw out consensus. +1 - this sounds like the most reasonable proposal of all. -- Best Regards, -- Alex
Re: Vote on personal matters: majority vote vs consensus
On Sat, Mar 23, 2013 at 8:02 PM, Christian Grobmeier grobme...@gmail.com wrote: Hi, following a thread on private@, I would like to bring the discussion on how we vote on nominated IPMC members. We had the case were one person was nominated and received three +1. Another voter had concerns an voted -1. The vote has been marked as failed, because no consensus could be found. Now this was my understanding and I was surprised that the vote failed: Votes on procedural issues follow the common format of majority rule unless otherwise stated. http://www.apache.org/foundation/voting.html Joe brought this up before around 14 months: http://s.apache.org/majorityinipmc We have not found a consens, but one might highlight Roy Fieldings e-mail: http://s.apache.org/royCommitterVeto I still think like Joe and feel that consensus should not apply in the IPMC. We are way to different to normal PMCs. As IPMC members we have no code which we can veto. Its all about accepting podlings, discussing rules and mentoring. We also have 172 IPMC members to date (according committer index). Most of the people are not seen often; we have many awol mentors. Currently becoming an IPMC member is necessary to become a Mentor. It always felt wrong to me. I think one should be able to become a Mentor and finally be able to join the IPMC and discuss rules, when he has shown merit. With an IPMC of that size it becomes more and more easy to get a -1. Personally I would like to see the IPMC separating IPMC-ship and Mentor-ship. I have proposed this already, but it seems nobody else except me wants that. So I am proposing now to reconsider Joes original proposal and change our community voting to a majority voting unless we restructure the IPMC. I am sorry to bring this lengthy discussion up again, but from the original thread I have learned a couple of other IPMC members are thinking similar on majority / consensus. I would also like to suggest that this time we finish the discussion with a vote. Cheers Christian A concern with changing to majority votes instead of consensus is that it will get misunderstood and at some point someone will think its ok to do that in other projects too. Your second suggestion sounds like the thing to do to me - separating IPMC-ship and Mentor-ship - that would solve several of the problems we've being having including this one, it would open up a much bigger pool of potential mentors, and IPMC'ers would get much more visibility of people as they work here which should make the PMC voting easier. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Mon, Mar 25, 2013, at 07:52 AM, ant elder wrote: On Sat, Mar 23, 2013 at 8:02 PM, Christian Grobmeier grobme...@gmail.com wrote: Hi, following a thread on private@, I would like to bring the discussion on how we vote on nominated IPMC members. We had the case were one person was nominated and received three +1. Another voter had concerns an voted -1. The vote has been marked as failed, because no consensus could be found. Now this was my understanding and I was surprised that the vote failed: Votes on procedural issues follow the common format of majority rule unless otherwise stated. http://www.apache.org/foundation/voting.html Joe brought this up before around 14 months: http://s.apache.org/majorityinipmc We have not found a consens, but one might highlight Roy Fieldings e-mail: http://s.apache.org/royCommitterVeto I still think like Joe and feel that consensus should not apply in the IPMC. We are way to different to normal PMCs. As IPMC members we have no code which we can veto. Its all about accepting podlings, discussing rules and mentoring. We also have 172 IPMC members to date (according committer index). Most of the people are not seen often; we have many awol mentors. Currently becoming an IPMC member is necessary to become a Mentor. It always felt wrong to me. I think one should be able to become a Mentor and finally be able to join the IPMC and discuss rules, when he has shown merit. With an IPMC of that size it becomes more and more easy to get a -1. Personally I would like to see the IPMC separating IPMC-ship and Mentor-ship. I have proposed this already, but it seems nobody else except me wants that. So I am proposing now to reconsider Joes original proposal and change our community voting to a majority voting unless we restructure the IPMC. I am sorry to bring this lengthy discussion up again, but from the original thread I have learned a couple of other IPMC members are thinking similar on majority / consensus. I would also like to suggest that this time we finish the discussion with a vote. Cheers Christian A concern with changing to majority votes instead of consensus is that it will get misunderstood and at some point someone will think its ok to do that in other projects too. Your second suggestion sounds like the thing to do to me - separating IPMC-ship and Mentor-ship - that would solve several of the problems we've being having including this one, it would open up a much bigger pool of potential mentors, and IPMC'ers would get much more visibility of people as they work here which should make the PMC voting easier. The structural problem here is that at the ASF, it is only PMC members whose votes are binding. Without being on the incubator PMC, votes on releases are non-binding. Now, you might argue that mentoring is a lot more than voting, but we could create another bottleneck in getting release votes through, requiring votes from incubator PMC members who are not particularly focused on the podling. Solve that, and the idea has merit in my eyes. Upayavira - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Mon, Mar 25, 2013 at 8:36 AM, Upayavira u...@odoko.co.uk wrote: Now, you might argue that mentoring is a lot more than voting, but we could create another bottleneck in getting release votes through, requiring votes from incubator PMC members who are not particularly focused on the podling. Thats exactly what i would argue, mentoring is a lot more than voting on releases. Many (most?) poddlings don't get three mentor votes on releases anyway and have to come to general@ so changing to have non-PMC mentors isn't going to make that worse than it already is. ...ant - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
Hi, On Sat, Mar 23, 2013 at 9:02 PM, Christian Grobmeier grobme...@gmail.com wrote: ...We also have 172 IPMC members to date (according committer index). Most of the people are not seen often; we have many awol mentors. Currently becoming an IPMC member is necessary to become a Mentor. It always felt wrong to me. I think one should be able to become a Mentor and finally be able to join the IPMC and discuss rules, when he has shown merit The problem is that people must be members of the Incubator PMC to vote on releases. I don't have a problem with having 172 members on this PMC, it is clear that many of them are not currently active and that's fine. ...I am proposing now to reconsider Joes original proposal and change our community voting to a majority voting unless we restructure the IPMC What does that mean exactly? That we consider -1s in votes on IPMC membership to be just a -1 on a majority vote, as opposed to a veto? I'm in favor of that, and I don't think we need more rules. -Bertrand - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On 25/03/13 08:41, ant elder wrote: On Mon, Mar 25, 2013 at 8:36 AM, Upayavira u...@odoko.co.uk wrote: Now, you might argue that mentoring is a lot more than voting, but we could create another bottleneck in getting release votes through, requiring votes from incubator PMC members who are not particularly focused on the podling. Thats exactly what i would argue, mentoring is a lot more than voting on releases. Many (most?) poddlings don't get three mentor votes on releases anyway and have to come to general@ so changing to have non-PMC mentors isn't going to make that worse than it already is. ...ant I agree mentoring is a lot more than voting. One point though (not suggesting new process) - there is some value in coming to general@ at least once in the podlings incubation. It brings in a wider perspective and understanding because small-N mentors on a podling may not themselves have a complete awareness of everything. Andy - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Sat, Mar 23, 2013 at 1:02 PM, Christian Grobmeier grobme...@gmail.com wrote: So I am proposing now to reconsider Joes original proposal and change our community voting to a majority voting unless we restructure the IPMC. +1 for majority voting on personnel issues for the IPMC. I'm also fine with requiring a supermajority of 2/3 or 3/4 of those voting, though I favor plain majority voting for the sake of simplicity. I would also like to suggest that this time we finish the discussion with a vote. +1 It will be disappointing if we get sidetracked into debating more controversial proposals and miss another opportunity to make some meaningful incremental progress. Marvin Humphrey - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: Vote on personal matters: majority vote vs consensus
On Sat, Mar 23, 2013 at 1:02 PM, Christian Grobmeier grobme...@gmail.com wrote: We have not found a consens, but one might highlight Roy Fieldings e-mail: http://s.apache.org/royCommitterVeto I still think like Joe and feel that consensus should not apply in the IPMC. We are way to different to normal PMCs. As IPMC members we have no code which we can veto. Its all about accepting podlings, discussing rules and mentoring. I agree on this point. Code contributions are way stickier business compared to any kind of decision that incubator project is entrusted with. Incubator is way more into 'soft skills' if you ask me and hence a simple majority as a decision making tools makes much more sense. Personally I would like to see the IPMC separating IPMC-ship and Mentor-ship. I have proposed this already, but it seems nobody else except me wants that. So I am proposing now to reconsider Joes original proposal and change our community voting to a majority voting unless we restructure the IPMC. I think at this point I'd rather NOT conflate the 2 issues (although I do agree that both merit a discussion thread). Can we go ahead with a potential proposal for IPMC to adopt a simple majority rule? Personally, I've come to appreciate that the sheer size of IPMC could be an advantage here. Projects with a small size of PMC do have to worry about outliers in their voting - we, on the other hand, have way less chance of something outlandish passing just because there are a couple of dudes who went into the same bar on the same night. I would also like to suggest that this time we finish the discussion with a vote. Indeed! Thanks, Roman. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
