AB Role in TAG Appointment

[martinthomson]

Decisions about TAG appointments are currently ratified by a two-thirds majority of both the TAG and the AB. I do not see a reason for the AB to be involved in this decision at that level.

For boards and groups operating at the level of the TAG, an appointment process is a good way to ensure that there is sufficient diversity of members available. Particularly when elections are used to fill seats, the process of filling seats can leave gaps in expertise or other qualities that would diminish the capacity of the group to function effectively.

A more regular process would be to give the group itself the ability to fill a small number of positions. For some reason, the W3C process gives this responsibility to the Team. That is somewhat irregular, but not really the subject of this issue.

The decision by the Team then needs to be approved by the TAG with a two-thirds majority, which is entirely appropriate. If the TAG believes that they cannot work with an appointed person, or the appointments do not address a diversity shortfall, then the TAG needs to be able to have a say.

The AB is given the same approval power, something that seems poorly justified. Obviously, the AB and TAG frequently work together, but while the two functions are complementary, the AB has no significant stake in the composition of the TAG or the technical work that the TAG does. The function of the AB is to advise the team, yes, but that is generally limited to procedural and operational matters.

Giving the AB input into TAG appointments at the same level as the TAG is unnecessary. I see no substantive reason that the AB needs a role that is significantly greater than any other consortium member.

If the goal of having the AB vote on appointments is to provide better accountability to the members of the organization, the formal objection process is a more appropriate way to ensure that membership has a way to object to particular appointments. That option is available to the AB just like any community member.

I propose that the process be amended to remove the AB from this process.

[frivoal]

Just tying things together, not arguing one way or the other. The AB's participation in the ratification came from:

• issue: https://github.com/w3c/w3process/issues/715
• AB discussion: https://www.w3.org/2023/03/16-ab-minutes.html#t07
• PR: https://github.com/w3c/w3process/pull/718

[frivoal]

I agree with the issue as raised, and am unconvinced that the AB having a say in this is necessary or helpful, at least in the current form.

I do think that having the TAG ratify the Team's appointment is an effective way to ensure that the Team did have the necessary consultation with the TAG, so I think that's important to keep, but under the current format, I'm not sure the AB's involvement does anything meaningful. My support in the past for what has now become the Process came not so much from including the AB, as from keeping the TAG in the loop. For the AB's inclusion, I was, and remain, unconvinced.

Maybe, if we want to preserve both the ability of the TAG to push back in case the Team's consultation is insufficient, AND have some mechanism to prevent the TAG from being self-serving, a possible process could be:

• The Team consults with various people, TAG included (unchanged)
• The Team proposes the appointments (unchanged)
• The TAG (alone) ratifies those appointments. If they do, process ends here.
• If the TAG fails to ratify the Teams appointment, then we go to the AB to settle the dispute. After looking into it, they can side with the TAG and block the appointments if they feel the Team indeed didn't do an appropriate job, or ratify the seats over the TAG's wishes if they find that it is the TAG that is being unreasonable / self serving…

[martinthomson]

That sounds like a reasonable process, though I'd like to poke on the conflict resolution piece a little.

I don't see a particular problem in the case of a failed TAG approval if the Team wants to go back and propose an alternative option. Maybe the TAG says "no" and then the Team and TAG talk and find out what resulted in that outcome. Maybe they can agree on a do-over.

Obviously, if there is a real disagreement, having an independent arbiter is useful. The AB seems like a reasonable choice for serving that function.

The other option on the table is not making appointments. It is implied in policy, but not explicit, that the Team has no choice in this. That is, the Team MUST make three and exactly three appointments. I'm not sure that I like that. Not just because a looser "MAY appoint up to three" would be a better way to ensure that the TAG membership is accountable to membership, but because it creates another alternative outcome in the case of conflict. (Maybe that's a separate issue though, as the decision to not appoint is also a decision that would need ratification/approval.)

[frivoal]

It is implied in policy, but not explicit, that the Team has no choice in this. That is, the Team MUST make three and exactly three appointments.

I'm not sure. I could easily read the text as a SHOULD as well.

https://www.w3.org/2023/Process-20231103/#TAG-appointments

[martinthomson]

Wonderful. The use of "SHOULD" there definitely makes it unclear, but my initial read was that those uses were to recommend how the Team was expected to undertake the job, not that the appointments were discretionary. To the new issue page!

[cwilso]

I disagree. There was a thorough discussion of this in the AB at the request of the Process CG (Florian linked to the minutes above), and there was a clear reasoning - to avoid insiderism on the TAG. Although it might be appropriate to review this in a few years, the TAG is only now considering the first appointments it is involved in. I don't think it is appropriate to consider jettisoning this at this point, as we don't have any new information about how this might play out.

At the same time, I would point out it's incorrect to say the AB has the same level of input into TAG appointments as the TAG itself. The Process is that the Team works together with the TAG to determine appointees - that is then ratified by the TAG and the AB.

Likewise I would claim it's incorrect that "the AB has no significant stake in the composition of the TAG or the technical work that the TAG does... [the AB's advice] is generally limited to procedural and operational matters." The AB has a broad charter, and with the addition of the Board, operational matters are largely out of the AB's purview; it is expected that the AB will be involved with the technical agenda of the W3C (at least, I've heard that explicitly started by Board members). But at any rate, this was meant as a high-level guardrail against insiderism, and I don't understand what has changed since that decision was made.

[frivoal]

there was a clear reasoning - to avoid insiderism on the TAG

I'm unsure how well this works. With the current ratification system, the TAG would have no difficulty in rejecting outsiders, which largely boils down to a similar thing.

I don't think it is appropriate to consider jettisoning this at this point, as we don't have any new information about how this might play out.

It is true that we have no experience with this new system, and we're likely to learn from using it. I'm not in a particular rush, and I don't think it is especially likely to cause problems. I'm just unconvinced that it is apt at solving the kind of problem it was intended for.

[martinthomson]

I don't see any particular urgency either. With #809, which would let the process run longer, I would suggest replacing the AB vote with a membership vote instead.

I agree with @frivoal as far as the "insiderism" thing.

[dwsinger]

I agree, I see neither urgency to change nor forceful reasons for the current text. If the AB did not have a formal vote, it still has influence: talk to the TAG, talk to the team. We could even say that the AB must be advised at the same time as the TAG is scheduled to approve (as at that point the appointments might not be public).

[plehegar]

As a side, we'll need a post-mortem on the Process once the TAG appointments are announced. I expect the Team and the TAG will want to provide input on how the Process worked for them.

[nigelmegitt]

@plehegar flagging use of phrase "post-mortem" - many folk don't like this for various reasons.

[TallTed]

@nigelmegitt — "Flagging use of [a] phrase" without offering alternatives is not nearly as effective nor helpful as offering some alternative, even if that then spawns a bike-shedding cycle.

@plehegar — In the spirit of progress, a couple of alternatives that come to my mind, that might be put to good use when you open a followup issue to track your wish, are "review of lessons learned" and "retrospective review".

[nigelmegitt]

Thanks @TallTed quite right. I expect @plehegar is able to generate alternatives easily enough; your suggestions are good ones.