Re: Release vote thresholds

[Dave Fisher <da...@comcast.net>]

On Nov 27, 2013, at 2:41 PM, Marvin Humphrey wrote:

> On Fri, Nov 22, 2013 at 4:45 PM, Marvin Humphrey <ma...@rectangular.com> wrote:
>> On Fri, Nov 22, 2013 at 4:20 PM, Dave Fisher <da...@comcast.net> wrote:
>> 
>>> I am not sure that this is a step in the correct direction.
>> 
>> Dave, I'm sure you realize that your opposition is deeply disapointing to a
>> number of us, when we seem so close.
>> 
>> I will work with you to see if we can build what you are currently calling
>> the "Incubator Committer" solution into something which can achieve
>> consensus.  It is not necessarily going to be easy to get that proposal as
>> close as this one has come, but let's try.
> 
> It's been 5 days since we last hear from Dave.  It's possible that the
> Thanksgiving holiday is a busy time, but we need to move forward.
> 
> We're working to modify the proposal for PPMC binding votes to address
> concerns about providing adequate oversight.  In the absence of a viable
> competing proposal, I expect to initiate a PROPOSAL followed by a VOTE once
> the revised version is sufficiently mature -- possibly in a week or so.

I am here now after the holiday and time away from most email.

I have a suggestion - inline.

On Nov 25, 2013, at 12:46 AM, Bertrand Delacretaz wrote:

> On Thu, Nov 21, 2013 at 7:09 PM, Marvin Humphrey <ma...@rectangular.com> wrote:
>> ...Ant, you proposed a minimum of one IPMC vote for each release:
> ...
>>    How about simply changing the rules for Incubator releases so that they
>>    don't require at least three binding votes, but instead make it at least
>>    three votes only one of which must be binding...
> 
> I would agree with that provided that we require other voters to
> document what they have actually reviewed in the release, something
> like:
> 
> -PPMC members review their release and create a text file in the
> incubator's svn repository, based on a template that indicates what
> needs to be reviewed in a release
> 
> -At least one Incubator PMC member reviews the release, updates the
> file accordingly, and their +1 vote along with the usual PPMC vote
> validates the release

c/one Incubator PMC member/one podling Mentor who is an Apache Member/

> 
> -For the first release we require at least three IPMC members to
> participate, instead of just one.
> 
> This would provide good oversight, while helping educate the PPMC
> members and keep a good record of the release requirements that other
> podlings can use as examples. And make sure people's +1s are based on
> something tangible and documented.

I am only comfortable allowing singular IPMC votes from members of the ASF. I think the IPMC owes this much to the other members of the Foundation.

Regards,
Dave




> 
> Marvin Humphrey
> 
> ---------------------------------------------------------------------
> 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: Release vote thresholds

[Dave Fisher <da...@comcast.net>]

On Dec 5, 2013, at 7:26 PM, Marvin Humphrey wrote:

> On Sun, Dec 1, 2013 at 12:41 PM, Dave Fisher <da...@comcast.net> wrote:
>> I am only comfortable allowing singular IPMC votes from members of the ASF.
>> I think the IPMC owes this much to the other members of the Foundation.
> 
> I've now contemplated this condition for a few days and while I'm prepared to
> accept it for the sake of compromise, I'm not in favor of it.
> 
> As I look over the IPMC roster, there may be people who are not the strongest
> with regards to reviewing releases (they are often strong in other areas),
> but I don't see a correlation between those people and whether or not they
> were elected onto the IPMC.  If anything, it's the opposite -- the people who
> routinely miss errors, cast bare +1 votes with no explanations of what they
> reviewed, or fail to vote at all, are most often ASF Members who exercised
> their prerogative to join the IPMC by request.
> 
> In contrast, the people who got onto the IPMC by demonstrating Incubator merit
> and getting elected tend to be more conscientious.  If any group stands out as
> particularly competent, it is those who were elected onto the IPMC first and
> subsequently became ASF Members.  But in my estimation, the group which would
> be excluded under this proposal -- people who were elected onto the IPMC but
> who are not ASF Members (yet) -- is on average, considerably above the level
> of the IPMC as a whole[1].
> 
> So... I question whether this provision will succeed at guaranteeing that solo
> IPMC votes come from someone highly competent.  It complicates the release
> process by stratifying the IPMC.  And it doesn't jibe with my sense of
> meritocracy.

So, do you agree to the rule of three IPMC? Any Member can be on the IPMC. Your argument above can be expanded to support the status quo. Can we drop the VXQuery experiment notion? I put the "comfort" level and Member vs. IPMC as more of a thought experiment. 

> 
> I'd rather that we go with Bertrand's proposal unmodified, which takes a
> different tack: striving to improve the quality of each vote cast.
> 
> What do others think?

I am not an other, but...

I like where Bertrand's proposal is going and it is something that can be used to assess all podlings great and small. My concerns and lack of comfort with this experiment were more to do with notion that not being able to provide oversight was leading to an experiment with institutionalizing less oversight. My argument about ASF members is that it is The ASF members who are responsible - we elect the Board. My one was asking what would be the bare minimum, it was more of a reductio ad absurdum.

I think that we all are looking for Incubation to be a natural growth of a community's culture and not a process that involves unnecessarily high barriers. We are looking to create the flattest petri dish possible.

What I like about Joe's experiment that the Incubator is the place to find people who are member material as this is more visible to more of the membership than people who gain merit within an individual TLP. You certainly agree.

My other point was that the notion of the VXQuery experiment was going to be an invalid experiment because that podling has already had someone moved into the IPMC using the Joe experiment. So, even if the release threshold experiment was worth pursuing in this case the results would be hard to assess one way or another.

I think that the situation is very positive for VXQuery. The podling has shown sustainability as it has refreshed the community with people who earned merit in the community while original members have moved on. Next question is if they have enough numbers. Here we will run into the 3 PPMC vs. 5 PPMC discussion.

Regards,
Dave

> 
> Marvin Humphrey
> 
> [1] Let's avoid mentioning names on the public list.
> 
> ---------------------------------------------------------------------
> 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: Release vote thresholds

[Marvin Humphrey <ma...@rectangular.com>]

On Sun, Dec 1, 2013 at 12:41 PM, Dave Fisher <da...@comcast.net> wrote:
> I am only comfortable allowing singular IPMC votes from members of the ASF.
> I think the IPMC owes this much to the other members of the Foundation.

I've now contemplated this condition for a few days and while I'm prepared to
accept it for the sake of compromise, I'm not in favor of it.

As I look over the IPMC roster, there may be people who are not the strongest
with regards to reviewing releases (they are often strong in other areas),
but I don't see a correlation between those people and whether or not they
were elected onto the IPMC.  If anything, it's the opposite -- the people who
routinely miss errors, cast bare +1 votes with no explanations of what they
reviewed, or fail to vote at all, are most often ASF Members who exercised
their prerogative to join the IPMC by request.

In contrast, the people who got onto the IPMC by demonstrating Incubator merit
and getting elected tend to be more conscientious.  If any group stands out as
particularly competent, it is those who were elected onto the IPMC first and
subsequently became ASF Members.  But in my estimation, the group which would
be excluded under this proposal -- people who were elected onto the IPMC but
who are not ASF Members (yet) -- is on average, considerably above the level
of the IPMC as a whole[1].

So... I question whether this provision will succeed at guaranteeing that solo
IPMC votes come from someone highly competent.  It complicates the release
process by stratifying the IPMC.  And it doesn't jibe with my sense of
meritocracy.

I'd rather that we go with Bertrand's proposal unmodified, which takes a
different tack: striving to improve the quality of each vote cast.

What do others think?

Marvin Humphrey

[1] Let's avoid mentioning names on the public list.

---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org