CFV: New OpenJDK Members Group Member: Simon Tooke

[David Holmes]

Hi Andrew,

On 28/05/2020 6:06 pm, Andrew Hughes wrote:
> On 28/05/2020 06:10, David Holmes wrote:
>> Vote: Veto
>>
>> "An OpenJDK Member is a Contributor who has demonstrated a history of
>> significant contributions to the Community as recognized by a vote of
>> the existing OpenJDK Members."
>>
>> Simon has only a handful of contributions to date.
>>
>> David
>>
> 
> Hi David,
> 
> Thanks for raising this. I don't think it's ever been clearly clarified
> what one is meant to achieve to obtain OpenJDK membership.
> 
> The wording of 'significant contributions' on [0] is (I believe
> intentionally) ambiguous. Commits are also not the best measure of this;
> not everyone's contribution to OpenJDK takes the form of regular
> commits. There's also the case that authorship is often lost. due to the
> real author not having OpenJDK authorship, or backport work being
> credited to the original author, not the backporter.
> 
> I nominated Simon & Alex because both have been working on OpenJDK for
> well over the year mentioned on that same page. In terms of commits, I
> believe both have sufficient numbers for committer status (Simon has
> this, voting on Alex's is underway).
> 
> Is being a committer not enough to be considered a member of the
> community? I'm a little concerned that the barrier is being set too high
> here and, given the main purpose to being an OpenJDK member is to gain
> voting rights, we risk disenfranchising a significant proportion of
> those working on OpenJDK. Looking through other recent membership votes,
> there are some that seem to have taken place relatively late, unless we
> expect all members to have hundreds of changesets.

I would state most definitely that merely being a committer (which is 
itself a very low bar) is nowhere near enough to be considered a Member 
(capital 'M') of the OpenJDK Group. Kim addressed many of your points 
and I will reiterate his words. There is a significant distinction 
between being a Member of the OpenJDK Group and being a member of the 
OpenJDK community.

"As a rough guide, a candidate should be a Member of at least one
Group, or a Reviewer or Project Lead of at least one Project, and make
regular and significant contributions to that Group or Project for at
least one year before being nominated to be an OpenJDK Member."

Your candidates do not meet any of these criteria at this time, and so 
my veto stands.

> I could accept the need for a minimum status of reviewer status for
> membership, if that was clearly stated and consistently applied. As it
> stands, the initial OpenJDK members group was populated not from the
> pool of reviewers, but from those with group membership [1]. I remember
> this clearly myself, because I recall transitioning to the bylaws with
> reviewer status, but not as a member of this group. I instead had to
> find someone who was and get them to nominate me, which is perhaps why
> it has always seemed a little peculiar to me going forward.
> 
> The set of groups seems to have changed little since the early days of
> OpenJDK and their main role seems to be sponsoring projects. I see
> plenty of people listed who have achieved reviewer status, but have no
> group membership. My only group membership was this one until the
> creation of the vulnerability group recently.

The set of groups was chosen to cover the various areas in OpenJDK so 
there has not been much reason to expand the set of groups.

"Groups are meant to capture the slowly-evolving social structure of the 
Community. They are usually longer-lived than Projects and therefore 
more difficult to create and dissolve. Generally speaking, a new Group 
should be proposed only after it has become clear that an existing 
nontrivial set of regular Participants who are already collaborating 
together would benefit from its creation. These are not meant to be 
lightweight "interest groups".

The CSR group and Vulnerability group, are examples of more recent groups.

Any Reviewer who is not a member of the associated Groups for their 
primary areas of involvement should seek a nomination to become part of 
the group. This has happened many times.

> I seem to remember that the initial groups and their members were based
> on the setup inside Sun at the time, so, if that in turn was used to
> onboard the initial OpenJDK members, then there will be some who became
> OpenJDK members by virtue of being in the right place at the right time,
> rather than having to meet the requirements of this voting process.

Everything has to be bootstrapped somehow.

> In short, it would be good to finally get this clarified and agree on
> clear criteria for OpenJDK members. I'll be happy to withdraw these
> votes at that time, if the criteria agreed upon is not met by Alex & Simon.

By all means seek to initiate a process to get clarification. I don't 
need further clarification for the current votes as it seems quite clear 
cut to me.

> Incidentally, the bylaws also still dictate automatic expiration after a
> year [2] so, in theory, none of us are OpenJDK members any more and
> can't propose anyone to become one... ;)

Again Kim addressed this.

Cheers,
David

> [0] https://openjdk.java.net/groups/members/
> [1] https://openjdk.java.net/groups/gb/bylaws/draft-openjdk-bylaws-10#_B
> [2] https://openjdk.java.net/bylaws#expiration
>