You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucy.apache.org by Marvin Humphrey <ma...@rectangular.com> on 2012/02/03 01:31:04 UTC
[lucy-dev] Thanks for voting!
On Mon, Jan 30, 2012 at 10:10:06PM -0600, Peter Karman wrote:
> Peter Karman did not vote.
Heh.
We should update the email template in release_commands.pl so that the
dev-list email includes a line like this...
Here's my +1.
This assumes that the RM isn't trolling the community with a bogus release. ;)
> The tally should have been:
>
> +1 Marvin Humphrey *+
> +1 Chris Mattman *+
> +1 Chris Hostetter *+
> +1 David E. Wheeler *
> +1 Nick Wellnhofer *
> +1 Logan Bell *
Thanks to everyone who voted, and to Peter for serving as Release Manager!
This VOTE took longer than most, and it's worth reviewing why that happened
and what we can do to prevent such delays in the future.
Quoting from a blog entry by Apache Board member Bernard Delacretaz:
https://blogs.apache.org/comdev/entry/what_makes_apache_projects_different
A formal PMC vote is required to publish a release. By voting to accept
the release, the PMC makes the release an act of the foundation, as
opposed to a personal action of the the release manager. This is a very
important distinction should any legal issues arise.
So, if e.g. somebody wants to sue over an ASF release, they can only go after
the Apache Software Foundation itself -- not individual committers[1]. Of
course legal complications are not commonplace, but Apache projects can be
pretty high profile and stuff happens -- the now-retired Apache Harmony
project, for instance, is tangled up in the Google/Oracle spat.
IANAL, but if I understand correctly, the ASF's legal indemnification
mechanism works something like this:
* The Members own the Foundation.
* The Board is elected by the Members to represent them.
* The Board, acting by voting to approve a Board resolution, establishes a
PMC and directs it to make software for the public good.
* The PMC is fulfilling the Board's directive when it VOTEs to accept a
release.
At this time, though, Lucy doesn't have a PMC. It has a PPMC -- "Podling
Project Management Committee" which has not received a directive from the
Board and thus does not have the authority to accept a release on behalf of
the Foundation.
The Incubator PMC, or IPMC, in contrast, is an official PMC. In order to make
a Lucy incubating release official, we need members of the IPMC to VOTE. If
we get a majority vote of IPMC members with a quorum of three +1s, our Lucy
release becomes an act of the Incubator PMC and by extension an "act of the
Foundation".
Lucy has four Mentors assigned to it, all of whom are members of the IMPC:
* Chris Hostetter
* Chris Mattmann
* Joe Schaefer
* Upayavira
An Incubator podling's Mentors are normally expected to provide the binding
IPMC votes which make its releases official. However, any member of the IPMC
can also cast a binding vote. I joined the IPMC about a year ago, so that
includes me.
Doing the math... we were trying to get 3 IPMC votes out of a pool of 5
people. That didn't work so well this time. I voted right away, but we
haven't heard from Upayavira in a long time and our three other Mentors all
just happened to be very busy.
There's not much we can do about this while we're still in the Incubator and
dependent on IPMC votes. Once we graduate from the Incubator and persuade the
Board to approve a resolution establishing the Lucy top-level project[2],
though, our PPMC will become a PMC, and the votes of Apache Lucy PMC members
will be binding. Releasing still won't be instantaneous, but VOTEs will
typically last 72 hours rather than 2+ weeks.
We just have to remember this lesson on what happens when you have a small
pool of potential votes to draw from. An Apache PMC that falls below 3 people
can't make a release at all. An Apache PMC that is small or that has a large
number of inactive members who don't vote may experience procedural problems
and delays along the lines of what we just witnessed.
Therefore, it's good to have a PMC of decent size (some IPMC members won't
vote to graduate a PMC with fewer than five people on it) and for lots of
people in the community to get in the habit of voting.
Did I remember to thank the people who voted? Hey, thanks for voting, y'all!
Marvin Humphrey
[1] More info on indemnification:
http://www.apache.org/foundation/faq.html#how
[2] Board resolutions establishing top-level-projects:
http://incubator.apache.org/guides/graduation.html#tlp-resolution
Re: [lucy-dev] Thanks for voting!
Posted by Marvin Humphrey <ma...@rectangular.com>.
On Thu, Feb 02, 2012 at 08:02:53PM -0800, Joe Schaefer wrote:
> FTR Marvin, the reason I didn't vote this time
> round is because I read your own vote on the
> candidate and was utterly perplexed by it, so
> I was expecting to see another one come down
> the pike. Apologies.
Having IPMC members vote to bless incubating releases is an imperfect system.
As fellow members of the IPMC and the ASF, you, me, Mattmann and others are
working to improve matters so that future podlings experience fewer hiccups --
and for those of us who that know the history, there can be no doubt about
your dedication to solving this problem!
Marvin Humphrey
Re: [lucy-dev] Thanks for voting!
Posted by Joe Schaefer <jo...@yahoo.com>.
FTR Marvin, the reason I didn't vote this time
round is because I read your own vote on the
candidate and was utterly perplexed by it, so
I was expecting to see another one come down
the pike. Apologies.
>________________________________
> From: Marvin Humphrey <ma...@rectangular.com>
>To: lucy-dev@incubator.apache.org; peter@peknet.com
>Sent: Thursday, February 2, 2012 7:31 PM
>Subject: [lucy-dev] Thanks for voting!
>
>On Mon, Jan 30, 2012 at 10:10:06PM -0600, Peter Karman wrote:
>> Peter Karman did not vote.
>
>Heh.
>
>We should update the email template in release_commands.pl so that the
>dev-list email includes a line like this...
>
> Here's my +1.
>
>This assumes that the RM isn't trolling the community with a bogus release. ;)
>
>> The tally should have been:
>>
>> +1 Marvin Humphrey *+
>> +1 Chris Mattman *+
>> +1 Chris Hostetter *+
>> +1 David E. Wheeler *
>> +1 Nick Wellnhofer *
>> +1 Logan Bell *
>
>Thanks to everyone who voted, and to Peter for serving as Release Manager!
>
>This VOTE took longer than most, and it's worth reviewing why that happened
>and what we can do to prevent such delays in the future.
>
>Quoting from a blog entry by Apache Board member Bernard Delacretaz:
>
> https://blogs.apache.org/comdev/entry/what_makes_apache_projects_different
>
> A formal PMC vote is required to publish a release. By voting to accept
> the release, the PMC makes the release an act of the foundation, as
> opposed to a personal action of the the release manager. This is a very
> important distinction should any legal issues arise.
>
>So, if e.g. somebody wants to sue over an ASF release, they can only go after
>the Apache Software Foundation itself -- not individual committers[1]. Of
>course legal complications are not commonplace, but Apache projects can be
>pretty high profile and stuff happens -- the now-retired Apache Harmony
>project, for instance, is tangled up in the Google/Oracle spat.
>
>IANAL, but if I understand correctly, the ASF's legal indemnification
>mechanism works something like this:
>
> * The Members own the Foundation.
> * The Board is elected by the Members to represent them.
> * The Board, acting by voting to approve a Board resolution, establishes a
> PMC and directs it to make software for the public good.
> * The PMC is fulfilling the Board's directive when it VOTEs to accept a
> release.
>
>At this time, though, Lucy doesn't have a PMC. It has a PPMC -- "Podling
>Project Management Committee" which has not received a directive from the
>Board and thus does not have the authority to accept a release on behalf of
>the Foundation.
>
>The Incubator PMC, or IPMC, in contrast, is an official PMC. In order to make
>a Lucy incubating release official, we need members of the IPMC to VOTE. If
>we get a majority vote of IPMC members with a quorum of three +1s, our Lucy
>release becomes an act of the Incubator PMC and by extension an "act of the
>Foundation".
>
>Lucy has four Mentors assigned to it, all of whom are members of the IMPC:
>
> * Chris Hostetter
> * Chris Mattmann
> * Joe Schaefer
> * Upayavira
>
>An Incubator podling's Mentors are normally expected to provide the binding
>IPMC votes which make its releases official. However, any member of the IPMC
>can also cast a binding vote. I joined the IPMC about a year ago, so that
>includes me.
>
>Doing the math... we were trying to get 3 IPMC votes out of a pool of 5
>people. That didn't work so well this time. I voted right away, but we
>haven't heard from Upayavira in a long time and our three other Mentors all
>just happened to be very busy.
>
>There's not much we can do about this while we're still in the Incubator and
>dependent on IPMC votes. Once we graduate from the Incubator and persuade the
>Board to approve a resolution establishing the Lucy top-level project[2],
>though, our PPMC will become a PMC, and the votes of Apache Lucy PMC members
>will be binding. Releasing still won't be instantaneous, but VOTEs will
>typically last 72 hours rather than 2+ weeks.
>
>We just have to remember this lesson on what happens when you have a small
>pool of potential votes to draw from. An Apache PMC that falls below 3 people
>can't make a release at all. An Apache PMC that is small or that has a large
>number of inactive members who don't vote may experience procedural problems
>and delays along the lines of what we just witnessed.
>
>Therefore, it's good to have a PMC of decent size (some IPMC members won't
>vote to graduate a PMC with fewer than five people on it) and for lots of
>people in the community to get in the habit of voting.
>
>Did I remember to thank the people who voted? Hey, thanks for voting, y'all!
>
>Marvin Humphrey
>
>[1] More info on indemnification:
> http://www.apache.org/foundation/faq.html#how
>
>[2] Board resolutions establishing top-level-projects:
> http://incubator.apache.org/guides/graduation.html#tlp-resolution
>
>
>
>
[lucy-dev] Re: Thanks for voting!
Posted by Peter Karman <pe...@peknet.com>.
Marvin Humphrey wrote on 2/2/12 6:31 PM:
> On Mon, Jan 30, 2012 at 10:10:06PM -0600, Peter Karman wrote:
>> Peter Karman did not vote.
>
> Heh.
>
> We should update the email template in release_commands.pl so that the
> dev-list email includes a line like this...
>
> Here's my +1.
>
great idea. committed in r1239990
>
>> The tally should have been:
>>
>> +1 Marvin Humphrey *+
>> +1 Chris Mattman *+
>> +1 Chris Hostetter *+
>> +1 David E. Wheeler *
>> +1 Nick Wellnhofer *
>> +1 Logan Bell *
>
> Thanks to everyone who voted, and to Peter for serving as Release Manager!
>
/me doffs cap
> This VOTE took longer than most, and it's worth reviewing why that happened
> and what we can do to prevent such delays in the future.
>
[big snip]
thanks for that thorough explanation of the little-p politics involved in
getting a release out the door. It is said that software is never truly
released, but rather it escapes.
--
Peter Karman . http://peknet.com/ . peter@peknet.com