Re: [VOTE] Move Chukwa to incubator
> I meant, purely in terms of Hadoop putting forward the TLP motion to the > board, with incubator's endorsement. But let's let options play out and > find out what incubator recommends first. I wasn't suggesting that Hadoop > would have some additional responsibilities here, in terms of mentoring or > otherwise. Oh I see; sorry, I hadn't parsed that context correctly. As the Incubator makes its decision, I just want it to be clear that the Hadoop PMC is no better equipped to make decisions about Chukwa's leadership or membership. All parties are trying to recover from this oversight by recommending that Chukwa spend time in the Incubator, to assemble a community that can discharge that responsibility. It will likely be a brief stay; all the reasons ant cites are true, with minor qualifications already raised. The TLP alternative proposed (1)- adding PMC members not involved in Chukwa's development, but influencing the project direction and membership- seems contrary to the spirit the Incubator endeavors to instill. It will have more autonomy in incubation and an unencumbered life as a TLP, rather than a protracted semi-incubation period as a "supervised TLP." -C - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
On 6/25/2010 12:40 PM, Chris Douglas wrote: >> >> But the Incubator doesn't just say yes/no. We can refer this back to Hadoop >> proposing this as a TLP, and even offer the list of mentors as observers, or >> members of the initial PMC. > > The Hadoop PMC is wholly unqualified to manage Chukwa. It voted to > recommend it to the Incubator so that it can gather a community > invested in its future and capable of meaningful consensus on the > challenges before it. Given that the very direction of the project has > been raised as an open issue, the benign neglect of the Hadoop PMC is > not a viable solution. -C I meant, purely in terms of Hadoop putting forward the TLP motion to the board, with incubator's endorsement. But let's let options play out and find out what incubator recommends first. I wasn't suggesting that Hadoop would have some additional responsibilities here, in terms of mentoring or otherwise. Thanks for all the rest of your observations. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
I agree with the analysis from Jerome and Greg. All three of Chukwa's current committers are its original contributors. An important- and often difficult- part of Apache community is growing and managing the developers working on the project. Some experience adding people in the Incubator would be valuable to members (both old and new) and would result in a quorum less volatile than three. Given that the team is small and renaming will be expensive and tedious, that requirement seems purely punitive. No positive reason for it has been raised. So I am +1 (non-binding) for option (2), incubation with TLP class naming, but incubation release naming. > Yes, it seems that Hadoop PMC should supervise the vote for the chair, with > a slightly frustrating cc list of all of the actual committers to chukwa. > That way, nominations and votes for chair are archived on a private list. > > But the Incubator doesn't just say yes/no. We can refer this back to Hadoop > proposing this as a TLP, and even offer the list of mentors as observers, or > members of the initial PMC. The Hadoop PMC is wholly unqualified to manage Chukwa. It voted to recommend it to the Incubator so that it can gather a community invested in its future and capable of meaningful consensus on the challenges before it. Given that the very direction of the project has been raised as an open issue, the benign neglect of the Hadoop PMC is not a viable solution. -C - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
Hi, As one of the initial contributor, I still think that the incubator road is the good one even if it's just for a short period of time. The incubator will gives us the ability to get some committers outside of the initial group of people and to mature the project. When we will have reach that point then TLP will make sense but not before that also the incubator road in my mind will help us in getting a better understanding of what our users really need; Is it a end-to-end product or more an SDK where user can take some part and build their own tools/products? I'm in favor of the SDK but this still need to be discussed/implemented. For all of this I vote for the incubator but in order to simplify the transition and for our current users I would like to keep the current naming convention. +1 incubator but with Chukwa's current naming convention. /Jerome, On 6/25/10 1:55 AM, "Bernd Fondermann" wrote: On Thu, Jun 24, 2010 at 21:21, William A. Rowe Jr. wrote: > On 6/23/2010 8:12 AM, Bernd Fondermann wrote: >> On Wed, Jun 23, 2010 at 14:45, ant elder wrote: >> >>> IMHO we should insist on using the incubator naming for the Chukwa >>> website/svn/MLs because I think Chukwa should just go directly to a >>> TLP and if they have to use the incubator naming it may help them >>> decide that the direct to TLP route really is better ;-) >> >> I see you blinking here, so I guess this is not just for putting up a >> strawman ;-) > > Well folks, it's a fun debate and all, but it isn't helping bring this > vote to a conclusion :) Wasn't the debate started just for the sake of hijacking the vote thread? ;-) Seriously, [DISCUSS] before [VOTE] is always recommended. > Is anyone in agreement with ant? Otherwise we should just move ahead > and can hold a separate vote on allowing tlp resource creation at this > time. > > If the proposers want (Eric?) a three choice vote, 1. recommend TLP with > guides to help the initial pmc, 2. accept incubating with tlp resource > naming, but -incubating release naming, or 3. accept incubating requiring > all incubator naming conventions, that might help the incubator simplify > this decision. I don't understand. The Hadoop community released a subproject for Incubation. The Incubator accepts or denies the proposal. In case of denial, the ball is in the Hadoop field again isn't it? > At this point, I personally guess that 1. might be the most sensible in > terms of resource creation and management; it would simply require the > group to vote for an initial chair/VP. Who is "the group"? The list of initial committers? This PMC? The Hadoop PMC? > If they are unsure of their group > yet, perhaps one of the other mentors would offer to serve as their chair > for the first six months, if they rather would do that? I'm still +1 to do proper Incubation. Bernd - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
On Fri, Jun 25, 2010 at 10:34 AM, William A. Rowe Jr. wrote: > Yes, it seems that Hadoop PMC should supervise the vote for the chair, with > a slightly frustrating cc list of all of the actual committers to chukwa. > That way, nominations and votes for chair are archived on a private list. > > But the Incubator doesn't just say yes/no. We can refer this back to Hadoop > proposing this as a TLP, and even offer the list of mentors as observers, or > members of the initial PMC. What do the originators of the proposal think of this direction? It sounds like they want to be a TLP, but are not sure they are yet ready. That leaves a burning question in my mind: Why do they feel they need the Incubator, rather than the Hadoop PMC, to help them mature into a TLP? I supported the Incubator proposal as it was written because I'm not sure Chukwa is getting the mentoring they need in their current home. I figure the Incubator is a good place to do that. I also support using the TLP naming because I think the Incubator can give them the mentoring they need and set them on their way in good time, and I'd like to avoid an extra renaming step. I would support what Bill has suggested if it was put to a vote, and honestly, I can't really say if one direction is preferable over the other. So, for now, I still support the proposal that has been put to a vote. Greg - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
On 6/25/2010 3:55 AM, Bernd Fondermann wrote: > On Thu, Jun 24, 2010 at 21:21, William A. Rowe Jr. wrote: > >> Is anyone in agreement with ant? Otherwise we should just move ahead >> and can hold a separate vote on allowing tlp resource creation at this >> time. >> >> If the proposers want (Eric?) a three choice vote, 1. recommend TLP with >> guides to help the initial pmc, 2. accept incubating with tlp resource >> naming, but -incubating release naming, or 3. accept incubating requiring >> all incubator naming conventions, that might help the incubator simplify >> this decision. > > I don't understand. The Hadoop community released a subproject for > Incubation. The Incubator accepts or denies the proposal. > In case of denial, the ball is in the Hadoop field again isn't it? Yes, it seems that Hadoop PMC should supervise the vote for the chair, with a slightly frustrating cc list of all of the actual committers to chukwa. That way, nominations and votes for chair are archived on a private list. But the Incubator doesn't just say yes/no. We can refer this back to Hadoop proposing this as a TLP, and even offer the list of mentors as observers, or members of the initial PMC. >> At this point, I personally guess that 1. might be the most sensible in >> terms of resource creation and management; it would simply require the >> group to vote for an initial chair/VP. > > Who is "the group"? The list of initial committers? This PMC? The Hadoop PMC? Group of proposed committers/actual contributors (it is existing code, there is svn history of who committed). >> If they are unsure of their group >> yet, perhaps one of the other mentors would offer to serve as their chair >> for the first six months, if they rather would do that? > > I'm still +1 to do proper Incubation. If that's the incubator's preference, I don't object, but I'm waiting for more than five people to be heard on this point of contention. You strongly want one option, ant strongly wants another, and we've heard few other voices. - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
On Thu, Jun 24, 2010 at 21:21, William A. Rowe Jr. wrote: > On 6/23/2010 8:12 AM, Bernd Fondermann wrote: >> On Wed, Jun 23, 2010 at 14:45, ant elder wrote: >> >>> IMHO we should insist on using the incubator naming for the Chukwa >>> website/svn/MLs because I think Chukwa should just go directly to a >>> TLP and if they have to use the incubator naming it may help them >>> decide that the direct to TLP route really is better ;-) >> >> I see you blinking here, so I guess this is not just for putting up a >> strawman ;-) > > Well folks, it's a fun debate and all, but it isn't helping bring this > vote to a conclusion :) Wasn't the debate started just for the sake of hijacking the vote thread? ;-) Seriously, [DISCUSS] before [VOTE] is always recommended. > Is anyone in agreement with ant? Otherwise we should just move ahead > and can hold a separate vote on allowing tlp resource creation at this > time. > > If the proposers want (Eric?) a three choice vote, 1. recommend TLP with > guides to help the initial pmc, 2. accept incubating with tlp resource > naming, but -incubating release naming, or 3. accept incubating requiring > all incubator naming conventions, that might help the incubator simplify > this decision. I don't understand. The Hadoop community released a subproject for Incubation. The Incubator accepts or denies the proposal. In case of denial, the ball is in the Hadoop field again isn't it? > At this point, I personally guess that 1. might be the most sensible in > terms of resource creation and management; it would simply require the > group to vote for an initial chair/VP. Who is "the group"? The list of initial committers? This PMC? The Hadoop PMC? > If they are unsure of their group > yet, perhaps one of the other mentors would offer to serve as their chair > for the first six months, if they rather would do that? I'm still +1 to do proper Incubation. Bernd - To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
Re: [VOTE] Move Chukwa to incubator
+1 for 1. Regards, Eric On 6/24/10 12:21 PM, "William A. Rowe Jr." wrote: > On 6/23/2010 8:12 AM, Bernd Fondermann wrote: >> On Wed, Jun 23, 2010 at 14:45, ant elder wrote: >> >>> IMHO we should insist on using the incubator naming for the Chukwa >>> website/svn/MLs because I think Chukwa should just go directly to a >>> TLP and if they have to use the incubator naming it may help them >>> decide that the direct to TLP route really is better ;-) >> >> I see you blinking here, so I guess this is not just for putting up a >> strawman ;-) > > Well folks, it's a fun debate and all, but it isn't helping bring this > vote to a conclusion :) > > Is anyone in agreement with ant? Otherwise we should just move ahead > and can hold a separate vote on allowing tlp resource creation at this > time. > > If the proposers want (Eric?) a three choice vote, 1. recommend TLP with > guides to help the initial pmc, 2. accept incubating with tlp resource > naming, but -incubating release naming, or 3. accept incubating requiring > all incubator naming conventions, that might help the incubator simplify > this decision. > > At this point, I personally guess that 1. might be the most sensible in > terms of resource creation and management; it would simply require the > group to vote for an initial chair/VP. If they are unsure of their group > yet, perhaps one of the other mentors would offer to serve as their chair > for the first six months, if they rather would do that? > > - > 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
