You are viewing a plain text version of this content. The canonical link for it is here.
Posted to general@incubator.apache.org by "Roy T. Fielding" <fi...@gbiv.com> on 2006/07/20 00:09:17 UTC
piling on
I believe that it is a bad idea to allow people to add themselves
to a proposal as committers without first obtaining the consent of
the person(s) making the proposal. Being a committer in the incubator
is giving a person the right to veto code changes based on whatever
technical reason they deem significant. We should not hand out that
right like candy to anyone who happens to edit a wiki page.
Podlings must be open to new contributors and should add contributors
to the list of committers fairly quickly, at least when compared to
more established Apache projects. However, the core team must be
allowed to select other core members based on mutual consensus,
since consensus is how code changes are approved. How a community
exercises that consensus is one of the primary determinants for
graduation.
In contrast, letting anyone "pile on" to a podling while it is at
the proposal stage is placing an unequal burden on a new podling
that we would never place on a full project. If the community is
not cohesive, no consensus will be possible and we effectively
hamstring the podling before it is even started.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Paul Fremantle <pz...@gmail.com>.
>
> Sure, and that is up to the proposer. If the proposal does not gain
> sufficient support from Apache because of that fact, that's life.
> Nevertheless, it is wrong for us to force a new podling to accept
> arbitrary committers just because they happen to have been proposed
> as an incubator podling.
+1. I've never seen that happen. Admittedly I'm fairly new round here
but I've never seen a podling forced to accept committers.
>
> > 2) Apache Committers, from whatever different project, have at some
> > point proved their worth in some way. In general I would expect them
> > to be a welcome addition to an incubator project. If nothing else they
> > have some idea of the Apache model and approach.
>
> Sorry, not a chance in hell. Each project at Apache has its own sense
> of what makes for a good contributor.
True, and its up to the podling to figure out their own approach.
Having some people from other projects has been known to help.
> Some of them hand out commit
> access to anyone with a vague idea of contributing.
And some incubation proposals hand out "proposed committers" to anyone
with a vague idea of contributing.
> Some of those
> "committers" have proven to have no clue whatsoever when it comes to
> handling Apache-style decision making.
Just like some of the new committers.
> > So I favour the active approach.
>
> I do not. Each community needs to construct itself and nobody here,
> including me, has the right to declare themselves part of a new
> community without being invited first.
Its a wiki. Its designed to be edited. That's how it works. Adding
your name to a pre-proposal that is being edited isn't "declaring
yourself part of a community". Its volunteering. It ought be matched
with a note to the mailing list, and if the proposer isn't comfortable
he should edit it out.
The final proposal that is voted on is whatever the proposer sends
out. Anyone trying to force the proposer to do something they aren't
comfortable with is out of order. I've never seen that happen.
Paul
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 20, 2006, at 2:54 PM, Paul Fremantle wrote:
> On 7/19/06, Roy T. Fielding <fi...@gbiv.com> wrote:
>> I believe that it is a bad idea to allow people to add themselves
>> to a proposal as committers without first obtaining the consent of
>> the person(s) making the proposal. Being a committer in the
>> incubator
>> is giving a person the right to veto code changes based on whatever
>> technical reason they deem significant. We should not hand out that
>> right like candy to anyone who happens to edit a wiki page.
>
> I agree. But I also think that making it very easy for existing Apache
> Committers to join an project is valuable.
It is very easy -- ask the proposer.
> 1) Some incubator proposals include only the existing core development
> team on a project. However, some other incubator proposals include
> many "new" proposed committers, some of whom have never submitted code
> to the project. I don't have anything against this, but I think it is
> an important point that the proposed committers are not necessarily
> existing coders on that project.
Sure, and that is up to the proposer. If the proposal does not gain
sufficient support from Apache because of that fact, that's life.
Nevertheless, it is wrong for us to force a new podling to accept
arbitrary committers just because they happen to have been proposed
as an incubator podling.
> 2) Apache Committers, from whatever different project, have at some
> point proved their worth in some way. In general I would expect them
> to be a welcome addition to an incubator project. If nothing else they
> have some idea of the Apache model and approach.
Sorry, not a chance in hell. Each project at Apache has its own sense
of what makes for a good contributor. Some of them hand out commit
access to anyone with a vague idea of contributing. Some of those
"committers" have proven to have no clue whatsoever when it comes to
handling Apache-style decision making.
I personally believe that it is good for Apache to have multiple
competing projects trying to address the same technology space,
particularly when their strategies appeal to different sets of users.
Among other things, it allows communities to fork *within* Apache
when a fundamental design choice cannot reach consensus, or when
the burdens of backwards compatibility impede significant progress.
> So I fully agree it shouldn't be a matter of "piling on" or just
> editing the wiki page. On the other hand, sending a note to the
> mailing list, *as well* as editing the wiki page is not a crime. The
> wiki page can always been edited back again (and in fact I've had that
> happen to me when I've offered to mentor). The first few time I
> volunteered on an Incubator project, no-one bothered to add me to the
> wiki, simply through inaction. So I favour the active approach.
I do not. Each community needs to construct itself and nobody here,
including me, has the right to declare themselves part of a new
community without being invited first.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Paul Fremantle <pz...@gmail.com>.
Roy
On 7/19/06, Roy T. Fielding <fi...@gbiv.com> wrote:
> I believe that it is a bad idea to allow people to add themselves
> to a proposal as committers without first obtaining the consent of
> the person(s) making the proposal. Being a committer in the incubator
> is giving a person the right to veto code changes based on whatever
> technical reason they deem significant. We should not hand out that
> right like candy to anyone who happens to edit a wiki page.
I agree. But I also think that making it very easy for existing Apache
Committers to join an project is valuable.
1) Some incubator proposals include only the existing core development
team on a project. However, some other incubator proposals include
many "new" proposed committers, some of whom have never submitted code
to the project. I don't have anything against this, but I think it is
an important point that the proposed committers are not necessarily
existing coders on that project.
2) Apache Committers, from whatever different project, have at some
point proved their worth in some way. In general I would expect them
to be a welcome addition to an incubator project. If nothing else they
have some idea of the Apache model and approach.
So I fully agree it shouldn't be a matter of "piling on" or just
editing the wiki page. On the other hand, sending a note to the
mailing list, *as well* as editing the wiki page is not a crime. The
wiki page can always been edited back again (and in fact I've had that
happen to me when I've offered to mentor). The first few time I
volunteered on an Incubator project, no-one bothered to add me to the
wiki, simply through inaction. So I favour the active approach.
I believe that sending a note to general@incubator saying that you are
a committer on X & Y project and some of your background, and you are
a willing participant is a reasonable approach and should be looked on
favourably by the proposers of new projects to incubation.
Paul
--
Paul Fremantle
VP/Technology, WSO2 and OASIS WS-RX TC Co-chair
http://bloglines.com/blog/paulfremantle
paul@wso2.com
"Oxygenating the Web Service Platform", www.wso2.com
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 7/19/06, Noel J. Bergman <no...@devtech.com> wrote:
> Garrett Rooney wrote:
>
> > Sure, but I think it should be made clear to mentors that the fact
> > that nothing is preventing them from committing changes to the project
> > they mentor doesn't mean that they should do so with abandon.
>
> Does my response to Martin sufficiently reflect your view?
Yes, we are 100% in agreement as far as I can tell.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Garrett Rooney wrote:
> Sure, but I think it should be made clear to mentors that the fact
> that nothing is preventing them from committing changes to the project
> they mentor doesn't mean that they should do so with abandon.
Does my response to Martin sufficiently reflect your view?
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 7/19/06, Noel J. Bergman <no...@devtech.com> wrote:
> Garrett Rooney wrote:
>
> > Personally, yes, I feel this should apply to mentors as well.
>
> The Mentors are all Incubator PMC members, and the Incubator PMC as a whole
> oversees all Incubator projects. Every Incubator PMC member has an equal
> binding vote on every Incubator matter, and ought to be able to act in an
> appropriate and responsible manner. Yes, if we create our own discord,
> we'll have to deal with it amongst ourselves, but we're not going to start
> by limiting the scope of our PMC members. Especially not without some
> justification that it is necessary.
Sure, but I think it should be made clear to mentors that the fact
that nothing is preventing them from committing changes to the project
they mentor doesn't mean that they should do so with abandon. If they
want to contribute as a committer they should earn that right as any
other contributor would. Of course, if they need to change something
because the podling has screwed something up from a
legal/procedural/whatever standpoint, that's another issue entirely,
and they shouldn't hesitate to do so.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Garrett Rooney wrote:
> Personally, yes, I feel this should apply to mentors as well.
The Mentors are all Incubator PMC members, and the Incubator PMC as a whole
oversees all Incubator projects. Every Incubator PMC member has an equal
binding vote on every Incubator matter, and ought to be able to act in an
appropriate and responsible manner. Yes, if we create our own discord,
we'll have to deal with it amongst ourselves, but we're not going to start
by limiting the scope of our PMC members. Especially not without some
justification that it is necessary.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/19/06, Noel J. Bergman <no...@devtech.com> wrote:
>
> Ian Holsman wrote:
>
> > should the podlings have a say on who gets to mentor them?
> > for example I could become a mentor of Blaize (I actually like the
> > project) but shouldn't the proposer have a say ?
>
> If you accept that a Mentor is just a name for an Incubator PMC member who
> is active in the project community --- and what else does it mean,
> really --- then the answer is no. All Incubator projects are governed by
> the Incubator PMC, and only the ASF Board (in a somewhat tenuous use of
> the
> word "establish", but we all accept that interpretation of 6.3 as
> operational) determines whom is on that roster.
>
> So, no, projects do not get to determine their own oversight. They
> haven't
> earned that right, and won't until and unless they graduate to TLP status.
+1
But in terms of the leadership, or even respect, facet of Mentor semantics,
> Mentors do not become leaders by fiat. They have to earn it and maintain
> it.
+1
- robert
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 20, 2006, at 7:56 PM, Sanjiva Weerawarana wrote:
> On Thu, 2006-07-20 at 15:00 -0700, Roy T. Fielding wrote:
>> I think so -- an unwelcome mentor is a waste of everyone's time.
>>
>> I also think mentors need commit access, since I don't believe it is
> ^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> This is the documented practice; see
> http://incubator.apache.org/incubation/
> Roles_and_Responsibilities.html:
>
> "The candidate shall declare an initial set of committers. On
> acceptance
> of a candidate project, the assigned Mentor shall be given access
> to the
> Podling's cvs repository for the duration of the incubation process.
> This is to allow the Mentor to perform their incubation duties, and is
> for administrative purposes only. To be given full committer
> privileges,
> such as the right to add new code to the repository, the Mentor must
> earn them as would any other potential new committer. In some
> cases, the
> Mentor may be part of the initial set of declared committers, but this
> is not a requirement of the Incubation process."
>
> However, note the emphasis on "administrative purposes" .. you seem to
> indicate more access than that:
No, I didn't. I indicated that the mentor needs to help do the work.
Administrative work is a huge part of getting a project going. I
include
in that things like setting up accounts, helping infrastructure, making
accurate infra requests, following up with root responses, setting up
subversion directories, and developing a process for publishing a
website.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "William A. Rowe, Jr." <wr...@rowe-clan.net>.
Sanjiva Weerawarana wrote:
> On Thu, 2006-07-20 at 15:00 -0700, Roy T. Fielding wrote:
>
>> It is kind of like getting advice at a barn-raising from some bystander
>> who isn't willing to lend a hand. The advice will be heard for about
>> five minutes, after which the people doing the work will simply ignore
>> the bystander (even if the advice is good). Community is more than
>> just being in the same place at the same time.
>
> I agree, except that my impression so far has been that a mentor does
> not have to be someone who's a technical person in the area or at least
> necessarily someone who writes code for that project. Mentoring is not
> about how to do the code but rather about getting to the "Apache
> Way" (which is of course a moving target in its own right).
Sanjiva - I totally concur with you. There is nothing wrong with a project
mentored by one of the committers, in fact Roy's points are good. BUT that
doesn't mean someone with a little less vested interest is a bad idea,
either. Someone for whom it's possible to stand back and say, "Hey, yes
that's technologically good for this application, but a bad social practice
here at the ASF". A deeply vested mentor can actually undermine things in
that sort of a scenario, for the long-term.
By that I mean we will have mentors who 1) are very competent ASF members
and otherwise make very good decisions, and 2) have a deeply vested interest
in the technology advanced by a project, and 3) happen to be employed by
a company leveraging the technology. When those three things are true, I'd
suggest it's actually not an "issue", but that additional mentors to provide
a sanity check on the project become more than a "nice to have", it would
ultimately become necessary. (And, said "very competent ASF member" would
see full well they need to ask for another mentor to avoid any future claim
of impropriety, from ever arising.)
So Roy's implication that every mentor should wade into the code just doesn't
jive with me. The stdcxx project hasn't 'required' me to, the 'lokahi' project
hasn't required me to. The stdcxx project is close to graduation once we can
attract more folks who aren't employed by the original contributor, and the
lokahi project is close to floating their first 'incubating release' to let
more folks dig into the code. I really don't think either project would fault
me for not committing willy-nilly into those two repositories ;-)
Bill
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Sanjiva Weerawarana <sa...@opensource.lk>.
On Thu, 2006-07-20 at 15:00 -0700, Roy T. Fielding wrote:
> I think so -- an unwelcome mentor is a waste of everyone's time.
>
> I also think mentors need commit access, since I don't believe it is
^^^^^^^^^^^^^^^^^^^^^^^^^^
This is the documented practice; see
http://incubator.apache.org/incubation/Roles_and_Responsibilities.html:
"The candidate shall declare an initial set of committers. On acceptance
of a candidate project, the assigned Mentor shall be given access to the
Podling's cvs repository for the duration of the incubation process.
This is to allow the Mentor to perform their incubation duties, and is
for administrative purposes only. To be given full committer privileges,
such as the right to add new code to the repository, the Mentor must
earn them as would any other potential new committer. In some cases, the
Mentor may be part of the initial set of declared committers, but this
is not a requirement of the Incubation process."
However, note the emphasis on "administrative purposes" .. you seem to
indicate more access than that:
> possible to mentor a collaborative project without showing by example.
> It is kind of like getting advice at a barn-raising from some bystander
> who isn't willing to lend a hand. The advice will be heard for about
> five minutes, after which the people doing the work will simply ignore
> the bystander (even if the advice is good). Community is more than
> just being in the same place at the same time.
I agree, except that my impression so far has been that a mentor does
not have to be someone who's a technical person in the area or at least
necessarily someone who writes code for that project. Mentoring is not
about how to do the code but rather about getting to the "Apache
Way" (which is of course a moving target in its own right).
Is that not the case any more? If so I need to get out of mentoring at
least one project that I volunteered to mentor.
Sanjiva.
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 19, 2006, at 3:34 PM, Ian Holsman wrote:
> I was more thinking of how mentors volunteer to guide the project
>
> should the podlings have a say on who gets to mentor them?
>
> for example I could become a mentor of Blaize (I actually like the
> project)
> but shouldn't the proposer have a say ?
I think so -- an unwelcome mentor is a waste of everyone's time.
I also think mentors need commit access, since I don't believe it is
possible to mentor a collaborative project without showing by example.
It is kind of like getting advice at a barn-raising from some bystander
who isn't willing to lend a hand. The advice will be heard for about
five minutes, after which the people doing the work will simply ignore
the bystander (even if the advice is good). Community is more than
just being in the same place at the same time.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Ian Holsman wrote:
> should the podlings have a say on who gets to mentor them?
> for example I could become a mentor of Blaize (I actually like the
> project) but shouldn't the proposer have a say ?
If you accept that a Mentor is just a name for an Incubator PMC member who
is active in the project community --- and what else does it mean,
really --- then the answer is no. All Incubator projects are governed by
the Incubator PMC, and only the ASF Board (in a somewhat tenuous use of the
word "establish", but we all accept that interpretation of 6.3 as
operational) determines whom is on that roster.
So, no, projects do not get to determine their own oversight. They haven't
earned that right, and won't until and unless they graduate to TLP status.
But in terms of the leadership, or even respect, facet of Mentor semantics,
Mentors do not become leaders by fiat. They have to earn it and maintain
it.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Ian Holsman <li...@holsman.net>.
I was more thinking of how mentors volunteer to guide the project
should the podlings have a say on who gets to mentor them?
for example I could become a mentor of Blaize (I actually like the
project)
but shouldn't the proposer have a say ?
On 20/07/2006, at 8:17 AM, Garrett Rooney wrote:
> On 7/19/06, Ian Holsman <li...@holsman.net> wrote:
>> are you referring to mentors as well?
>
> Personally, yes, I feel this should apply to mentors as well. While
> there are cases where a mentor needs commit access for some sort of
> procedural issues (maintaining STATUS files, helping to fix up
> licensing issues, and whatnot), they should absolutely not have
> general commit access to the codebase until they have earned it like
> any other developer.
>
> Note that this is exactly what I did with the Abdera podling, despite
> the fact that I'm acting as a mentor I sent in patches and waited for
> them to be applied like any other contributor would have to.
>
> -garrett
>
--
Ian Holsman
Ian@Holsman.net
join http://gyspsyjobs.com the marketplace for django developers
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/20/06, Noel J. Bergman <no...@devtech.com> wrote:
>
> Martin Sebor wrote:
>
> > one of the checkboxes on the status page says:
> > Give all Mentors access to all incubator SVN modules (to be
> > done by PMC chair).
> > so it seems they are required to have access whether the rest
> > of the committers like it or not.
>
> Yes, they must have it to enable oversight, but they do not have to use
> it.
> One can exercise restraint and earn one's merit, unless abnormal
> circumstances demand action from an oversight role.
+1
<snip>
Likewise, I would expect any Incubator PMC member to demonstrate
> self-restraint, and to behave in accordance with the project's practices.
+1
a mechanism exists to remove any mentor unable to demonstrate the required
qualities: the incubator PMC determines the mentors. any mentor can be added
or removed by a PMC VOTE.
- robert
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Martin Sebor wrote:
> one of the checkboxes on the status page says:
> Give all Mentors access to all incubator SVN modules (to be
> done by PMC chair).
> so it seems they are required to have access whether the rest
> of the committers like it or not.
Yes, they must have it to enable oversight, but they do not have to use it.
One can exercise restraint and earn one's merit, unless abnormal
circumstances demand action from an oversight role.
There is an interesting relationship between earned (contrasted with
unearned) ability and restraint. Consider that I can, as a purely technical
matter, modify any file anywhere in the ASF, in source control or otherwise.
If there was the slightest chance that I would actually (ab)use that
ability, I would not have it. Nor would I ever do so. At the least, the
repercussions would be disastrous.
Likewise, I would expect any Incubator PMC member to demonstrate
self-restraint, and to behave in accordance with the project's practices.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Martin Sebor <se...@roguewave.com>.
Garrett Rooney wrote:
> On 7/19/06, Ian Holsman <li...@holsman.net> wrote:
>
>> are you referring to mentors as well?
>
>
> Personally, yes, I feel this should apply to mentors as well. While
> there are cases where a mentor needs commit access for some sort of
> procedural issues (maintaining STATUS files, helping to fix up
> licensing issues, and whatnot), they should absolutely not have
> general commit access to the codebase until they have earned it like
> any other developer.
FWIW, one of the checkboxes on the status page says:
Give all Mentors access to all incubator SVN modules (to be
done by PMC chair).
so it seems they are required to have access whether the rest
of the committers like it or not.
Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 7/19/06, Ian Holsman <li...@holsman.net> wrote:
> are you referring to mentors as well?
Personally, yes, I feel this should apply to mentors as well. While
there are cases where a mentor needs commit access for some sort of
procedural issues (maintaining STATUS files, helping to fix up
licensing issues, and whatnot), they should absolutely not have
general commit access to the codebase until they have earned it like
any other developer.
Note that this is exactly what I did with the Abdera podling, despite
the fact that I'm acting as a mentor I sent in patches and waited for
them to be applied like any other contributor would have to.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Ian Holsman <li...@holsman.net>.
are you referring to mentors as well?
On 20/07/2006, at 8:09 AM, Roy T. Fielding wrote:
> I believe that it is a bad idea to allow people to add themselves
> to a proposal as committers without first obtaining the consent of
> the person(s) making the proposal. Being a committer in the incubator
> is giving a person the right to veto code changes based on whatever
> technical reason they deem significant. We should not hand out that
> right like candy to anyone who happens to edit a wiki page.
>
> Podlings must be open to new contributors and should add contributors
> to the list of committers fairly quickly, at least when compared to
> more established Apache projects. However, the core team must be
> allowed to select other core members based on mutual consensus,
> since consensus is how code changes are approved. How a community
> exercises that consensus is one of the primary determinants for
> graduation.
>
> In contrast, letting anyone "pile on" to a podling while it is at
> the proposal stage is placing an unequal burden on a new podling
> that we would never place on a full project. If the community is
> not cohesive, no consensus will be possible and we effectively
> hamstring the podling before it is even started.
>
> ....Roy
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
--
Ian Holsman
Ian@Holsman.net
http://economychat.com discussion on the economy and economics in
general
Re: piling on
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 7/19/06, Roy T. Fielding <fi...@gbiv.com> wrote:
> I believe that it is a bad idea to allow people to add themselves
> to a proposal as committers without first obtaining the consent of
> the person(s) making the proposal. Being a committer in the incubator
> is giving a person the right to veto code changes based on whatever
> technical reason they deem significant. We should not hand out that
> right like candy to anyone who happens to edit a wiki page.
>
> Podlings must be open to new contributors and should add contributors
> to the list of committers fairly quickly, at least when compared to
> more established Apache projects. However, the core team must be
> allowed to select other core members based on mutual consensus,
> since consensus is how code changes are approved. How a community
> exercises that consensus is one of the primary determinants for
> graduation.
>
> In contrast, letting anyone "pile on" to a podling while it is at
> the proposal stage is placing an unequal burden on a new podling
> that we would never place on a full project. If the community is
> not cohesive, no consensus will be possible and we effectively
> hamstring the podling before it is even started.
+1, I would much prefer to see interested parties starting to
contribute in the usual way (patches sent to a development mailing
list) and get their commit access once the project has decided they
are ready for it.
Of course, this assumes that the project members are on the lookout
for new blood and are actively noticing worthy new contributors and
voting them in as committers, but a new podling should be doing that
anyway.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Fwd: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
Begin forwarded message:
From: "Roy T. Fielding" <fi...@gbiv.com>
Subject: piling on
Date: July 19, 2006 3:09:17 PM PDT
To: general@incubator.apache.org
Reply-To: general@incubator.apache.org
I believe that it is a bad idea to allow people to add themselves
to a proposal as committers without first obtaining the consent of
the person(s) making the proposal. Being a committer in the incubator
is giving a person the right to veto code changes based on whatever
technical reason they deem significant. We should not hand out that
right like candy to anyone who happens to edit a wiki page.
Podlings must be open to new contributors and should add contributors
to the list of committers fairly quickly, at least when compared to
more established Apache projects. However, the core team must be
allowed to select other core members based on mutual consensus,
since consensus is how code changes are approved. How a community
exercises that consensus is one of the primary determinants for
graduation.
In contrast, letting anyone "pile on" to a podling while it is at
the proposal stage is placing an unequal burden on a new podling
that we would never place on a full project. If the community is
not cohesive, no consensus will be possible and we effectively
hamstring the podling before it is even started.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Matthias Wessendorf <ma...@apache.org>.
also +1 on Roy,
On 7/19/06, Garrett Rooney <ro...@electricjellyfish.net> wrote:
> On 7/19/06, Andrus Adamchik <an...@objectstyle.org> wrote:
> > +1.
> >
> > As a member of an incubating project I totally agree with this.
> >
> > What I didn't know is that there was a policy that allowed anyone to
> > just declare themselves committers without consensus of the existing
> > project participants?
asking me the same :)
> I don't believe there's actually a policy that actively allows it, but
> I have seen it happen with several projects, and it certainly seemed
> odd to me.
in adffaces podling we don't. we prefer the *regular way*
-Matthias
> -garrett
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
--
Matthias Wessendorf
further stuff:
blog: http://jroller.com/page/mwessendorf
mail: mwessendorf-at-gmail-dot-com
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "Jean T. Anderson" <jt...@bristowhill.com>.
Andrus Adamchik wrote:
...
> [...] Probably makes sense to mention in the incubator docs [1] that
> new committers are to be voted in following the normal Apache procedures.
Actually it's the incubator ppmc doc that discusses voting in new
committers:
http://incubator.apache.org/guides/ppmc.html
If that wording isn't clear, please post.
-jean
> I am +0 to extend this policy to the mentors as well. From Cayenne
> experience I don't recall a mentor ever having to commit code *as a
> mentor*. Although we have a mentor who was already a committer, but
> that's a totally different story.
>
> [1] http://incubator.apache.org/guides/committer.html
>
> Andrus
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Andrus Adamchik <an...@objectstyle.org>.
On Jul 19, 2006, at 6:19 PM, Garrett Rooney wrote:
> On 7/19/06, Andrus Adamchik <an...@objectstyle.org> wrote:
>> +1.
>>
>> As a member of an incubating project I totally agree with this.
>>
>> What I didn't know is that there was a policy that allowed anyone to
>> just declare themselves committers without consensus of the existing
>> project participants?
>
> I don't believe there's actually a policy that actively allows it, but
> I have seen it happen with several projects, and it certainly seemed
> odd to me.
>
> -garrett
Ah, ok. Probably makes sense to mention in the incubator docs [1]
that new committers are to be voted in following the normal Apache
procedures.
I am +0 to extend this policy to the mentors as well. From Cayenne
experience I don't recall a mentor ever having to commit code *as a
mentor*. Although we have a mentor who was already a committer, but
that's a totally different story.
[1] http://incubator.apache.org/guides/committer.html
Andrus
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 7/19/06, Andrus Adamchik <an...@objectstyle.org> wrote:
> +1.
>
> As a member of an incubating project I totally agree with this.
>
> What I didn't know is that there was a policy that allowed anyone to
> just declare themselves committers without consensus of the existing
> project participants?
I don't believe there's actually a policy that actively allows it, but
I have seen it happen with several projects, and it certainly seemed
odd to me.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Andrus Adamchik <an...@objectstyle.org>.
+1.
As a member of an incubating project I totally agree with this.
What I didn't know is that there was a policy that allowed anyone to
just declare themselves committers without consensus of the existing
project participants?
Andrus
On Jul 19, 2006, at 6:09 PM, Roy T. Fielding wrote:
> I believe that it is a bad idea to allow people to add themselves
> to a proposal as committers without first obtaining the consent of
> the person(s) making the proposal. Being a committer in the incubator
> is giving a person the right to veto code changes based on whatever
> technical reason they deem significant. We should not hand out that
> right like candy to anyone who happens to edit a wiki page.
>
> Podlings must be open to new contributors and should add contributors
> to the list of committers fairly quickly, at least when compared to
> more established Apache projects. However, the core team must be
> allowed to select other core members based on mutual consensus,
> since consensus is how code changes are approved. How a community
> exercises that consensus is one of the primary determinants for
> graduation.
>
> In contrast, letting anyone "pile on" to a podling while it is at
> the proposal stage is placing an unequal burden on a new podling
> that we would never place on a full project. If the community is
> not cohesive, no consensus will be possible and we effectively
> hamstring the podling before it is even started.
>
> ....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Andrus Adamchik <an...@objectstyle.org>.
I guess I misread the proposed bootstrap procedure and its
ramifications.
> - bootstrap the PPMC from the PMC (assigning Mentors)
> - election by the PPMC of project contributors to the PPMC
> - election by the PPMC of Committers
1. I was wondering why incubating project committers who are not PPMC
members can't vote for new committers. Of course they can still
propose the candidates and leave the vote up to PPMC. Fair enough.
2. Ok, so ASF existing committers who sign up as committers (not
mentors) on the initial proposal would have to go through a PPMC vote
as well. That's a good solution.
I am all for implementing this procedure then.
Andrus
On Jul 19, 2006, at 7:30 PM, Noel J. Bergman wrote:
> Andrus Adamchik wrote:
>
>> PPMC can oversee the process and should be able to veto proposed
>> committers without sufficient earned karma, but I don't see the
>> downsides of self-government of the incubating project.
>
> The PPMC *is* the self-governing body for the Incubating project.
> Which is
> why I keep saying that the recommendation should be to have at
> least three
> (3) Mentors on the project, thus enabling the PPMC to actually hold
> binding
> votes without needing to wait on the rest of the PMC (although we
> should
> still be included). Without at least three mentors, a PPMC lacks the
> necessary minimum number of votes, and must wait for other PMC
> members to
> vote on each issue.
>
> As for the question of whether or not a candidate has "sufficient
> earned
> karma", the PPMC, following the explanation above, has the ability
> to make
> the decision for each situation.
>
> --- Noel
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Andrus Adamchik wrote:
> PPMC can oversee the process and should be able to veto proposed
> committers without sufficient earned karma, but I don't see the
> downsides of self-government of the incubating project.
The PPMC *is* the self-governing body for the Incubating project. Which is
why I keep saying that the recommendation should be to have at least three
(3) Mentors on the project, thus enabling the PPMC to actually hold binding
votes without needing to wait on the rest of the PMC (although we should
still be included). Without at least three mentors, a PPMC lacks the
necessary minimum number of votes, and must wait for other PMC members to
vote on each issue.
As for the question of whether or not a candidate has "sufficient earned
karma", the PPMC, following the explanation above, has the ability to make
the decision for each situation.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Andrus Adamchik <an...@objectstyle.org>.
As always there is another dimension to every problem... Noel brings
a good point about single company contributions. The solution seems
wrong though.
PPMC can oversee the process and should be able to veto proposed
committers without sufficient earned karma, but I don't see the
downsides of self-government of the incubating project. IIRC there
are rules requiring committer diversity on graduation. So if the
project committers turn down existing ASF committers *who earned
karma*, the project is not going to graduate anyways. And if outside
people, who did not earn karma, formally get on board and do not
contribute, what good does it do to anybody?
Andrus
On Jul 19, 2006, at 6:39 PM, Noel J. Bergman wrote:
> Roy,
>
> This piling on behavior seems to have come from the notion that if
> you get
> on the initial vote, you're in, but otherwise you have to earn
> committership. And the justification for the first part seemed to
> be making
> sure that a company could not start with a lot of its own people,
> and keep
> out existing ASF committers and other interested parties.
>
> My long standing proposed solution for this is embedded in my
> description of
> project startup:
>
> - bootstrap the PPMC from the PMC (assigning Mentors)
> - election by the PPMC of project contributors to the PPMC
> - election by the PPMC of Committers
>
> So the Mentors first task is to formally bring on board the key
> outside
> players, and then that PPMC elects Committers, both to start and going
> foward. This addresses your point that:
>
>> Podlings must be open to new contributors and should add contributors
>> to the list of committers fairly quickly, at least when compared to
>> more established Apache projects. However, the core team must be
>> allowed to select other core members based on mutual consensus,
>> since consensus is how code changes are approved. How a community
>> exercises that consensus is one of the primary determinants for
>> graduation.
>
> --- Noel
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Sanjiva Weerawarana <sa...@opensource.lk>.
On Sun, 2006-07-23 at 15:52 +0100, robert burrell donkin wrote:
>
> let's be honest - the policy document is a mess.
>
> i can't even get 3 +1's to remove commented out material from the policy
> document.
+1 from me for removing cruft ;-).
Sanjiva.
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
sorry posted the wrong email :-/
On 7/23/06, robert burrell donkin <ro...@gmail.com> wrote:
>
> On 7/23/06, Noel J. Bergman <no...@devtech.com> wrote:
>
> > Roy T. Fielding wrote:
> >
> > > Robert Burrell Donkin wrote:
> >
> > > > no change is necessary - the current policy is sufficient.
> >
> > > Well, no, the expectation is clearly being ...
> >
> > I agree with Roy. I'm not sure if he and I yet agree or disagree on
> > what
> > the policy should be, but there are clearly conflicting views and
> > "traditions" in place. So we should formally (re?)establish the
> > consensus
> > by vote, and then make sure that what we have written everywhere is in
> > line
> > with that consensus.
>
>
> let's be honest - the policy document is a mess.
>
> i can't even get 3 +1's to remove commented out material from the policy
> document.
>
given that we wouldn't be changing the current policy, wouldn't it be
cleaner and clearer just to pass a PMC resolution noting that the proposal
has the final say?
- robert
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/23/06, Noel J. Bergman <no...@devtech.com> wrote:
>
> Roy T. Fielding wrote:
>
> > Robert Burrell Donkin wrote:
>
> > > no change is necessary - the current policy is sufficient.
>
> > Well, no, the expectation is clearly being ...
>
> I agree with Roy. I'm not sure if he and I yet agree or disagree on what
> the policy should be, but there are clearly conflicting views and
> "traditions" in place. So we should formally (re?)establish the consensus
> by vote, and then make sure that what we have written everywhere is in
> line
> with that consensus.
let's be honest - the policy document is a mess.
i can't even get 3 +1's to remove commented out material from the policy
document.
- robert
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/23/06, robert burrell donkin <ro...@gmail.com> wrote:
> On 7/22/06, Roy T. Fielding <fi...@gbiv.com> wrote:
>
> > On Jul 22, 2006, at 1:50 AM, robert burrell donkin wrote:
> >
> > > no change is necessary - the current policy is sufficient.
>
>
> (no change in policy was meant - hopefully that was clear from the context)
>
> > Well, no, the expectation is clearly being set that anyone can add
> > themselves to the proposal on the wiki, and I for one vote to approve
> > a proposal based on both the wiki and the emailed content. If the
> > emailed version is different from the wiki, then that will get a -1
> > from me simply because they are inconsistent.
>
>
> i agree that the problem is the expectation and not the policy
>
> so let's fix the expectation
>
> > > ATM the proposal on the wiki is not normative. the sponsor votes on
> > > the
> > > proposal as submitted to the list. this proposal contains a number
> > > of names
> > > proposed as initial committers. the sponsor votes to accept the
> > > proposal
> > > submitted including the list of initial committers. so, the list of
> > > committers is specified by the proposer and approved by the sponsor.
> >
> > In theory, yes, In practice, no. A proposer should not be placed in
> > the position of fighting a wiki-edit war for consistency when it is far
> > easier for us to tell volunteers to be polite by asking the proposer
> > for permission before editing *their* proposal.
>
>
> +1
>
> but again, this a problem with practice not policy.
>
>
> > > what other goals would any new policy have in the area?
> >
> > Just a small bit of documentation for people preparing a proposal to
> > inform them that they don't have to accept additional committers during
> > the proposal process.
>
>
> that would be my preferred solution. i just don't see why it needs to be
> formal policy.
>
> > There is a serious social disconnect here: people
> > who are making a proposal are extremely sensitive about pissing us off,
> > and will tend not to reject an added committer even when they know that
> > person is not qualified or is deliberately attempting to steer the
> > project in a direction that it may not want to go.
> > We need the policy
> > to protect new proposals from being unduly influenced by our already
> > established mindset.
>
> we don't need new policy which just re-iterates the old. what we need is
> appropriate documentation describing how the process should work. (we also
> need to clean out the cruft so that people can see what policy we actual
> have but that's a slightly different issue.)
i've added some more documentation that i hope will help to address this issue.
http://incubator.apache.org/guides/proposal.html#template-initial-committers
http://incubator.apache.org/guides/participation.html (draft guide to
incubator netiquette)
http://incubator.apache.org/guides/proposal.html#developing
please review and suggest or patch improvements
if anyone feels that a policy reaffirmation is still necessary, then
please create a patch in JIRA and start a VOTE.
- robert
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/22/06, Roy T. Fielding <fi...@gbiv.com> wrote:
>
> On Jul 22, 2006, at 1:50 AM, robert burrell donkin wrote:
>
> > no change is necessary - the current policy is sufficient.
(no change in policy was meant - hopefully that was clear from the context)
Well, no, the expectation is clearly being set that anyone can add
> themselves to the proposal on the wiki, and I for one vote to approve
> a proposal based on both the wiki and the emailed content. If the
> emailed version is different from the wiki, then that will get a -1
> from me simply because they are inconsistent.
i agree that the problem is the expectation and not the policy
so let's fix the expectation
> ATM the proposal on the wiki is not normative. the sponsor votes on
> > the
> > proposal as submitted to the list. this proposal contains a number
> > of names
> > proposed as initial committers. the sponsor votes to accept the
> > proposal
> > submitted including the list of initial committers. so, the list of
> > committers is specified by the proposer and approved by the sponsor.
>
> In theory, yes, In practice, no. A proposer should not be placed in
> the position of fighting a wiki-edit war for consistency when it is far
> easier for us to tell volunteers to be polite by asking the proposer
> for permission before editing *their* proposal.
+1
but again, this a problem with practice not policy.
> what other goals would any new policy have in the area?
>
> Just a small bit of documentation for people preparing a proposal to
> inform them that they don't have to accept additional committers during
> the proposal process.
that would be my preferred solution. i just don't see why it needs to be
formal policy.
There is a serious social disconnect here: people
> who are making a proposal are extremely sensitive about pissing us off,
> and will tend not to reject an added committer even when they know that
> person is not qualified or is deliberately attempting to steer the
> project in a direction that it may not want to go.
We need the policy
> to protect new proposals from being unduly influenced by our already
> established mindset.
we don't need new policy which just re-iterates the old. what we need is
appropriate documentation describing how the process should work. (we also
need to clean out the cruft so that people can see what policy we actual
have but that's a slightly different issue.)
i think that the right way to address this issue would be in the
documentation aimed at proposers. we just explain to them that they are free
to submit whatever list they want. we also explain to those wanting to add
their names that they should be polite enough to ask on the list first. most
likely, all that it would take to change attitudes is for a few people to
jump on the next few people who pile in without asking.
it'd make sense to add a new section in proposals called interested
developers where people can add themselves. the proposer could then upgrade
any they wanted to the initial committer list.
- robert
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Roy T. Fielding wrote:
> Robert Burrell Donkin wrote:
> > no change is necessary - the current policy is sufficient.
> Well, no, the expectation is clearly being ...
I agree with Roy. I'm not sure if he and I yet agree or disagree on what
the policy should be, but there are clearly conflicting views and
"traditions" in place. So we should formally (re?)establish the consensus
by vote, and then make sure that what we have written everywhere is in line
with that consensus.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 22, 2006, at 1:50 AM, robert burrell donkin wrote:
> no change is necessary - the current policy is sufficient.
Well, no, the expectation is clearly being set that anyone can add
themselves to the proposal on the wiki, and I for one vote to approve
a proposal based on both the wiki and the emailed content. If the
emailed version is different from the wiki, then that will get a -1
from me simply because they are inconsistent.
> ATM the proposal on the wiki is not normative. the sponsor votes on
> the
> proposal as submitted to the list. this proposal contains a number
> of names
> proposed as initial committers. the sponsor votes to accept the
> proposal
> submitted including the list of initial committers. so, the list of
> committers is specified by the proposer and approved by the sponsor.
In theory, yes, In practice, no. A proposer should not be placed in
the position of fighting a wiki-edit war for consistency when it is far
easier for us to tell volunteers to be polite by asking the proposer
for permission before editing *their* proposal.
> what other goals would any new policy have in the area?
Just a small bit of documentation for people preparing a proposal to
inform them that they don't have to accept additional committers during
the proposal process. There is a serious social disconnect here: people
who are making a proposal are extremely sensitive about pissing us off,
and will tend not to reject an added committer even when they know that
person is not qualified or is deliberately attempting to steer the
project in a direction that it may not want to go. We need the policy
to protect new proposals from being unduly influenced by our already
established mindset.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by robert burrell donkin <ro...@gmail.com>.
On 7/21/06, Roy T. Fielding <fi...@gbiv.com> wrote:
>
> On Jul 20, 2006, at 7:30 PM, Sanjiva Weerawarana wrote:
<snip>
> "Policies" (for whatever level they exist here) have been set up
> > established practice- and as you know the mode of operation here so
> > far
> > has been "add yourself to the wiki". So for all practical purposes, it
> > *was* the policy here until your email.
>
> No. Policies are set by agreement of the incubator PMC. What people
> do in the *absence* of policy is not policy.
+1
> I have no problem with changing it but don't try to make it look like
> > there's "piling on" when established practice has is followed. I'm
> > absolutely +1 for changing the process and having the proposer control
> > who joins the project as you recommend.
>
> The change would be adding policy where none existed before.
>
no change is necessary - the current policy is sufficient.
ATM the proposal on the wiki is not normative. the sponsor votes on the
proposal as submitted to the list. this proposal contains a number of names
proposed as initial committers. the sponsor votes to accept the proposal
submitted including the list of initial committers. so, the list of
committers is specified by the proposer and approved by the sponsor.
what other goals would any new policy have in the area?
- robert
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 20, 2006, at 7:30 PM, Sanjiva Weerawarana wrote:
> On Thu, 2006-07-20 at 14:54 -0700, Roy T. Fielding wrote:
>> On Jul 19, 2006, at 3:39 PM, Noel J. Bergman wrote:
>>
>>> This piling on behavior seems to have come from the notion that if
>>> you get
>>> on the initial vote, you're in, but otherwise you have to earn
>>> committership. And the justification for the first part seemed to
>>> be making
>>> sure that a company could not start with a lot of its own people,
>>> and keep
>>> out existing ASF committers and other interested parties.
>>
>> I think people just got in the habit and assumed it was policy.
>
> Please remember Roy that your mails do not make policy either.
Which is, of course, why I said "I think".
> "Policies" (for whatever level they exist here) have been set up
> established practice- and as you know the mode of operation here so
> far
> has been "add yourself to the wiki". So for all practical purposes, it
> *was* the policy here until your email.
No. Policies are set by agreement of the incubator PMC. What people
do in the *absence* of policy is not policy.
> I have no problem with changing it but don't try to make it look like
> there's "piling on" when established practice has is followed. I'm
> absolutely +1 for changing the process and having the proposer control
> who joins the project as you recommend.
The change would be adding policy where none existed before.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by Sanjiva Weerawarana <sa...@opensource.lk>.
On Thu, 2006-07-20 at 14:54 -0700, Roy T. Fielding wrote:
> On Jul 19, 2006, at 3:39 PM, Noel J. Bergman wrote:
>
> > This piling on behavior seems to have come from the notion that if
> > you get
> > on the initial vote, you're in, but otherwise you have to earn
> > committership. And the justification for the first part seemed to
> > be making
> > sure that a company could not start with a lot of its own people,
> > and keep
> > out existing ASF committers and other interested parties.
>
> I think people just got in the habit and assumed it was policy.
Please remember Roy that your mails do not make policy either.
"Policies" (for whatever level they exist here) have been set up
established practice- and as you know the mode of operation here so far
has been "add yourself to the wiki". So for all practical purposes, it
*was* the policy here until your email.
I have no problem with changing it but don't try to make it look like
there's "piling on" when established practice has is followed. I'm
absolutely +1 for changing the process and having the proposer control
who joins the project as you recommend.
Sanjiva.
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Re: piling on
Posted by "Roy T. Fielding" <fi...@gbiv.com>.
On Jul 19, 2006, at 3:39 PM, Noel J. Bergman wrote:
> This piling on behavior seems to have come from the notion that if
> you get
> on the initial vote, you're in, but otherwise you have to earn
> committership. And the justification for the first part seemed to
> be making
> sure that a company could not start with a lot of its own people,
> and keep
> out existing ASF committers and other interested parties.
I think people just got in the habit and assumed it was policy.
There is nothing wrong with the proposer asking for and accepting
additional committers from the wide world of ASF. I did that for
Jackrabbit, for example, specifically because I wanted a lot of
experienced ASF folks to help mentor the project (even though I was
the only "official" Mentor). However, that is significantly different
from any wiki reader being able to add themselves just because they
(or their boss) thinks it might be worth getting in on the ground
floor of a project.
> My long standing proposed solution for this is embedded in my
> description of
> project startup:
>
> - bootstrap the PPMC from the PMC (assigning Mentors)
> - election by the PPMC of project contributors to the PPMC
> - election by the PPMC of Committers
>
> So the Mentors first task is to formally bring on board the key
> outside
> players, and then that PPMC elects Committers, both to start and going
> foward.
I think that is okay when you have three Mentors who are currently
active on the proposal. I think it is a waste of time if they are
not already active, since the proto-PPMC you mention hasn't even
earned that right yet. Sure, they have the responsibility for it as
delegated by the ASF, but when an existing code base is involved the
only person(s) who really know the community are the ones writing the
proposal. It saves a lot of time for the Incubator PMC to approve
both the proposal and the initial set of committers in one act.
I don't think either method is perfect, so I prefer the one with less
of an entry barrier.
....Roy
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
RE: piling on
Posted by "Noel J. Bergman" <no...@devtech.com>.
Roy,
This piling on behavior seems to have come from the notion that if you get
on the initial vote, you're in, but otherwise you have to earn
committership. And the justification for the first part seemed to be making
sure that a company could not start with a lot of its own people, and keep
out existing ASF committers and other interested parties.
My long standing proposed solution for this is embedded in my description of
project startup:
- bootstrap the PPMC from the PMC (assigning Mentors)
- election by the PPMC of project contributors to the PPMC
- election by the PPMC of Committers
So the Mentors first task is to formally bring on board the key outside
players, and then that PPMC elects Committers, both to start and going
foward. This addresses your point that:
> Podlings must be open to new contributors and should add contributors
> to the list of committers fairly quickly, at least when compared to
> more established Apache projects. However, the core team must be
> allowed to select other core members based on mutual consensus,
> since consensus is how code changes are approved. How a community
> exercises that consensus is one of the primary determinants for
> graduation.
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org