forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Crossley <cross...@apache.org>
Subject Re: [Vote] remove active/inactive concept
Date Thu, 11 May 2006 08:07:52 GMT
Gav.... wrote:
> This vote mentions both PMC Members and Committers as though they
> Are separate entities.

I cannot see where. Part B was just talking about
the rearrangement of the who.html part to get rid
of the unworkable "active" distinction regarding
committers. Part A is the main vote.

> Please remind me, I thought it was decided
> (unanimously amongst those present) that there was no longer a
> Distinction , you can not be a committer without being also a
> PMC Member. I did query the statement in /guidelines.html which
> Reads :-
> 
> " ... However, there may be extraordinary cases where we want limited
> work-related commit access (not also a PMC member)... "
> 
> IIRC the reply was that this is no longer the case.

You asked two slightly different questions in
the above paragraphs. I will try to answer them both
separately.

"Extraordinary" was intended to leave a mechanism for
something else to happen. See the Proposal thread
and the mention of GSoC for example.

We always want new committers to also be PMC members.
Otherwise they do not have a binding vote and so we
will create classes of committers. We don't want that.
However, it is up to them. When we invite a new committer
we do let them choose not to be on the PMC, though we
do not encourage that. I will try to enhance
http://forrest.apache.org/guidelines.html#elect

> If this stands then there is more that needs doing to clarify this new
> Stance in electing Devs to become PMC Members (becoming committers as
> A consequence and vice-versa). 

Primarily they are elected as a Committer
and we also want them to be a PMC member.

However that is not the subject of this vote.
That has been clarified many times in the past
and i am concerned that it is being raised again.
We need better docs obviously and cannot expect
new people to fully understand these issues.

> Such as :-
> 
> Update the Meritocracy of Roles and Responsibilities to remove
> Committer as though it was a separate role from PMC Member.
> 
> Where it is sometimes mentioned 'PMC Member' and sometimes
> Mentioned 'Committer' needs to be decided if using just
> One of these terms throughout the documentation to avoid
> Confusion.

No need. It *is* a separate role.

For example, Stefano is still a committer but is
not a PMC member because he was not around when
the project gained a PMC. We would be delighted
if any of those old people came back to participate
and commit again. We would probably ask them to become
a PMC member.

Not removing committers when they seem to be inactive
is very important. They can then fix a bug at any
time and not be held up by the patch process.
We trusted them before, and we trust them still.

I can see what you mean about using the terms
correctly. "Committer" is probably the term that
should be used most of the time. Use "PMC member"
when it comes to issues about project management.

> More ...
> 
> 
> > -----Original Message-----
> > From: David Crossley
> > Sent: Wednesday, 10 May 2006 11:27 AM
> > To: dev@forrest.apache.org
> > Subject: Re: [Vote] remove active/inactive concept
> > 
> > Reminder that there is about 24 hours remaining
> > for this vote.
> > 
> > -David
> > 
> > David Crossley wrote:
> > > David Crossley wrote:
> > > > The proposal addresses two separate issues:
> > > >
> > > > ------------------------
> > > > A) Remove each mention of "Active" from guidelines.html
> > > >
> > > > So this will mean that the people who have binding votes
> > > > are "PMC Members".
> > > >
> > > > Also change the definitions of "Unanimous consensus" and
> > > > "2/3 majority" to stop saying "All voters with binding
> > > > votes must vote" and instead refer to number of "votes cast".
> 
> Hmm, are you sure you mean 'Unanimous Consensus' to be included
> Here? 

Yes. See the discussion in the proposal leading
up to the vote. "Unanimous Consensus" is impossible
to achieve if it applies to "all PMC members".

> What about important decisions such as electing a new PMC Member
> Or even more importantly removing a PMC Member - you are saying
> That if at the time say only 4 PMC Members are currently active,
> Then only these 4 PMC members need vote on such an important
> Decision? It's not a likely scenario or the project would be in
> Dire Straits but I'm emphasising my point.

Yes. For most important situations "Consensus approval"
is used: At least 3 +1 and no -1 votes.

See the next note about quorum. If the chair thinks
that there is too much lack of interest, then they
will stop or extend the vote.

You are correct. If this situation occurs too often
then the chair will be reporting to the board that
the project is not healthy.

> > > > Add a note about the chair deciding whether a quorum
> > > > has been reached.
> > > >
> > > > Add a note to reinforce that PMC members can be as active
> > > > as they choose, with no pressure from the project.
> > > > People can be quiet and speak up occasionally when they
> > > > see a topic that motivates them enough to contribute
> > > > to the discussion or to cast a vote.
> 
> So are you removing the current policy of 'automatic removal
> If inactive after 6 months' ?

Yes. See the Proposal for the reasons.
"Activity" is impossible to enforce.

> I have to be careful in what I say here, it will more than likely
> Show my misinterpretation, then again it is good so you guys
> Will correct it for me. 

Thanks for being brave and speaking your mind.

> Devs are expected to attain a certain level of lets say
> 'Achievement' within the project, in terms of contributions
> Which could be documentation (which is highly regarded), in
> Terms of code (the blood of the project), or in terms of
> Activity in the discussion lists either helping new users
> Or suggesting new ways etc etc. 

Community is the blood actually, code will be
a consequence.

> This is expected to be kept up for quite some time so that
> PMC Members may then notice and discuss possible reasons
> For them to be included in the PMC. 
> Lets face it, someone may be the bees knees with coding, contribute
> fantastically and regularly for 6 months or so - and then nothing.

Correct.

> So will not even get to the stage of being discussed on the 
> PMC list.

Yes they will. We might decide to let them go on
for a bit longer and we would propose them again later.

> In other words, PMC Members have shown outstanding
> Abilities in whatever contributions and have maintained this
> Consistently over a long period of time in order to be
> Recognised and eventually voted in. Of course you already
> Know all this; I'm getting to the point...
> 
> Recapping what it states on the who.html page:-
> 
> " ... The Forrest Project operates on a meritocracy: the more you do, the
> more responsibility you will obtain. This page lists all of the people who
> have gone the extra mile and are Committers ... "
> 
> Ok, the point? Can I repeat a part of this vote proposal:-
> 
> " Add a note to reinforce that PMC members can be as active
> as they choose, with no pressure from the project.
> People can be quiet and speak up occasionally when they
> see a topic that motivates them enough to contribute
> to the discussion or to cast a vote. "
> 
> So, I am confused and concerned, someone makes all that effort to eventually
> Become PMC Member whom you voted in BECAUSE OF THEIR CONTRIBUTIONS AND
> EFFORT, then you say basically you don't care how much or how little they do
> They can stay PMC Members until they themselves decide whether or not they
> want to continue. If I misunderstood all of this and the 6 month rule still
> Applies, please clarify for me, thanks. In the meantime, my vote on this
> Section is based on what I understand so far.

We trust these people. They demonstrated to the rest
of the project that they are committed. They have
demonstrated they can work in a collaborative way.
We are pleased with whatever level of work that they
can manage. We don't want to put any performance
pressure onto them. We value their opinion whenever
they speak up.

Yes the six month thing has gone.

> > > > PMC members can still declare themselves "emeritus"
> > > > if they so choose. Nicola Ken was an example. In such
> > > > a case we notify the board and remove them from the
> > > > list of PMC members.
> 
> See above, and to add, a Project needs continually active PMC members
> Who can commit and apply time regularly - Once a month say I would
> Consider minimum to be ok.

Please see the Proposal and prior discussion.
It is impossible to enforce that. This is one
reason that we ended up here.

> Going back to the flip-side again, would you
> Vote someone in a PMC Member who contributed once in a while when they
> feel like it ? 

I, for one, would not.

> Important decisions (i.e. that need Unanimous Consensus) need to be done
> By ALL PMC Members, holidays and normal delays accounted for,

This is impossible to achieve.

> if a PMC
> Member can not be bothered to vote on something, then they surely have lost
> Interest in the project. You can't have PMC Members waltzing in and out of 
> The project whenever they feel like, They
> Have responsilities and should be made to uphold those in the interests of
> The community and project as a whole. Don't be soft on inactive PMC Members,
> They got there by shear hard work and dedication of which they should be
> Extremely proud and glowing in the achievement (oh how I am going to regret
> Re-visiting this post in the archives) to which they have attained, but this
> Must/should(?) continue at an acceptable rate.

If we put that sort of pressure on, then people will
run away or never come. One would hope that people
do act as you say but we cannot force them.

> > > > ------------------------
> > > > B) Remove the separation of "active" committers from who.html
> > > >
> > > > There will be one list. The PMC members will be marked.
> > > > Emeritus PMC members will be marked in a different way.
> > > >
> > > > There will be explanations of how to see who is "active"
> > > > for those who care, e.g. search the mail archives; look
> > > > at changes.html; etc.
> 
> Hmm, yes Changes.html, not changed in a while has it.

That is why i am pushing so hard on FOR-865. If there
was a level above Blocker, then i would put it there.

> > > > Add a note to explain the evolution of Forrest to
> > > > a top-level project and hence the introduction of
> > > > a PMC which did not include the old absent committers.
> > > >
> > > > ------------------------
> > >
> > > Anyone can vote. It is good to hear the opinions of developers.
> 
> You sure about that :)

Yes.

Next time please raise the issues as soon as possible,
preferably during the Proposal leading up to the vote.
But yes, never too late. If you must, then do it during
the vote. However doing it at the last minute does
not give people sufficient time to challenge you, or
perhaps your opinions might cause them to change theirs.

> > > As usual the binding votes are those of PMC members.
> > > Using "Consensus approval":
> > > http://forrest.apache.org/guidelines.html#approvals
> > > The vote will run for the normal one week and so finish at
> > > midnight UTC on 2005-05-10
> > >
> > http://www.timeanddate.com/counters/customcounter.html?year=2006&month=05&
> > day=11
> > >
> > > Please vote on these two proposals:
> > >
> > > A) Remove each mention of "Active" from guidelines.html
> 
> -1 (non-binding)

Does your -1 still stand, after the explanations from
me and other PMC members? You would need to summarise
the reasons.

-David

> > > B) Remove the separation of "active" committers from who.html
> 
> +1 (non-binding)
> 
> Gav...

Mime
View raw message