Veto by inaction

[mnot]

I thought this had been discussed before, but I don't see any applicable issue.

Right now, the Process says:

Any individual (regardless of whether they are associated with a Member) may appeal any decision made in connection with this Process (except those having a different appeal process) by registering a Formal Objection with the Team.

and further down:

When a W3C decision is made following an Advisory Committee review, Advisory Committee representatives may initiate an Advisory Committee Appeal.

These are our two appeal mechanisms. However, they do not address the failure to make a decision -- in particular, inaction by the Team in making a Team Decision or a W3C Decision.

For example, someone proposing a new charter could appeal a Team decision not to advance that charter for review, but if the Team were to just 'hold on' to that charter and refuse to make a decision, there is no recourse in the Process.

One could argue that this is a problem for Chair Decisions and Group Decisions too. However, they are visible -- everyone in that context knows that a decision is expected for a given issue, and if the chair fails to make a decision, they're under a fair amount of pressure to explain why.

That suggests one possible solution: making Team and W3C Decisions-to-be more transparent. This is discussed in eg #649 and #651. We'd need to go through the Process and consider how that would affect each instance of a Team or W3C Decision.

Another way of addressing this would be to go through each instance and add requirements for processing times, etc. However, that would likely be onerous, so I don't think it's viable.

Yet another approach would be to just allow appeal when a Team Decision or W3C Decision is not made. Because the appeal mechanisms require a certain amount of machinery, this would likely be self-limiting; it wouldn't be worth it to appeal three days after asking for a decision, for example, but if you're still waiting six months later and don't have a satisfying reason as to why, you might think of appealing.

[frivoal]

@mnot I had been thinking about that too.

Another way of addressing this would be to go through each instance and add requirements for processing times, etc. However, that would likely be onerous, so I don't think it's viable.

That was my first inclination, as quite a few things have deadlines already. Adding explicit limits (with reasonable values) for things that happen all the time might be worth it. But covering all cases might be tricky.

Yet another approach would be to just allow appeal when a Team Decision or W3C Decision is not made. Because the appeal mechanisms require a certain amount of machinery, this would likely be self-limiting; it wouldn't be worth it to appeal three days after asking for a decision, for example, but if you're still waiting six months later and don't have a satisfying reason as to why, you might think of appealing.

Right. And/or we could include a default duration. For instance 28 days for matters where the Process doesn't give a specific deadline.

However, I wonder we can use Formal Objections, at least as is, to deal with this. FOs are supposed to be asking the Council whether to overturn a decision or to let it stand. Which can be tricky if no decision has been made yet. If the question put to the Team is a yes/no question, I think the council could substitute, with current mechanisms largely unchanged. But if the question is not binary, I'm not sure that would be a good fit.

[fantasai]

I think for now, at least, the appropriate appeal mechanism for Team non-action is to appeal to the AB, which is chartered to advise on process and conflict resolution. The AB doesn't have any actual authority here, but they can put more direct pressure on the Team/CEO if necessary. In some cases, merely escalating to the AB might draw out enough attention and articulation on the problem that it gets solved without the AB's direct intervention. In others, triggering a discussion among the AB and the Team, possibly drawing in the TAG if necessary, could help work through any blockage. (And worst case if the Team is acting in bad faith, the AB can escalate that to the Board through their liaisons. But I don't expect that to happen.)

So before we add any rigid formal process here, I'd rather try using what we have.

[mnot]

I disagree that the AB is a viable solution here -- they're not chartered to do this, and they're not set up as an appeal-handling body. A Member who feels they're being ignored or brushed off by the Team needs a definitive answer, not an amorphous "these people might help you."

[dwsinger]

I sort-of agree with both. "Have you asked the AB for help, and if not, try that" is fine – but it's not a final stop-gap, definitive, "this is the way we handle it"

Maybe we need a mechanism to file for a formal request for decision, sort-of like an appeal? (And we should document the steps that should be taken first, such as informally asking, asking the AB's assistance, or whatever)

[nigelmegitt]

I think there's a framing problem here. Especially when it comes to WGs, Decisions are made when the Chair asserts that the group has consensus to adopt a Proposal. When there is no such consensus, no Decision is made, and the status quo ante applies. However, sometimes there was a formal assessment or moment in which the group decided (in common parlance) not to adopt the proposal, and effectively that proposal was removed from future consideration.

We need to be able to distinguish the scenarios:

1. A Proposal was assessed and rejected, so the group chose no Decision.
2. No formal assessment of the Proposal occurred, so the status remains open.

I'm not sure if there's a way for any member to force an assessment of a Proposal, so that scenario 2 becomes scenario 1, but my interpretation is that if we were to introduce such a mechanism, that would possibly resolve the issue.

[mnot]

I don't think anyone is suggesting we need to change WG decisionmaking -- it's Team and W3C Decisions (as defined in the process).

[dwsinger]

We need to be able to distinguish the scenarios:

1. A Proposal was assessed and rejected, so the group chose no Decision.
2. No formal assessment of the Proposal occurred, so the status remains open.

I think 1 is a decision, no? It's cases where someone feels that there is foot-dragging and their opponent is using lack of action as a way to say no without saying no.

The way out of this is usually to have mechanisms that anyone can force consideration of something, eventually. How that happens will vary based on venue.

[samuelweiler]

That suggests one possible solution: making Team and W3C Decisions-to-be more transparent. This is discussed in eg #649 and #651. We'd need to go through the Process and consider how that would affect each instance of a Team or W3C Decision.

The strategy pipeline exposes an index of these (potential) decisions, particularly for workshops and charters, which are the subjects of #649 and #651, respectively. The pipeline doesn't necessarily show the internal state of a team decision - it can be hard to tell the difference between inaction, slow progress, deadlock, and silent rejection.

... Yet another approach would be to just allow appeal when a Team Decision or W3C Decision is not made. Because the appeal mechanisms require a certain amount of machinery, this would likely be self-limiting; it wouldn't be worth it to appeal three days after asking for a decision, for example, but if you're still waiting six months later and don't have a satisfying reason as to why, you might think of appealing.

That seems like it would address the problem, which I concur is not about WG decision making.

[samuelweiler]

@dwsinger

The way out of this is usually to have mechanisms that anyone can force consideration of something, eventually. How that happens will vary based on venue.

I propose that we focus on the example in the opening message, namely chartering.

[nigelmegitt]

We need to be able to distinguish the scenarios:

1. A Proposal was assessed and rejected, so the group chose no Decision.
2. No formal assessment of the Proposal occurred, so the status remains open.

I think 1 is a decision, no? It's cases where someone feels that there is foot-dragging and their opponent is using lack of action as a way to say no without saying no.

It should be, but it is often hard or impossible to distinguish it from simple inaction. I cannot recall the last time I saw a group explicitly consider a "no action Proposal" as an alternative to some other Proposal, and log a Decision to adopt the "no action Proposal".

The way out of this is usually to have mechanisms that anyone can force consideration of something, eventually. How that happens will vary based on venue.

Right, yes. Is that anywhere in our Process?

[joshco]

How about a heartbeat that can be reset by the Team. Eg if a proposal from widgetsWG is N months old, the team must either make a decision to reset the timer (Giving another N months), or the proposal is decided by default. This gives widgetsWG a chance to show up and argue their case, and forces the Team to argue its case (either pass/fail/sleep).

Presumably the Team reviews its status periodically, and is implicitly or informally deciding (we're not ready to decide on WidgetsWG) anyway, so it's just adding visibility and transparency to what is probably already happening.

If the Team is taken over by aliens and keeps refusing to make a decision, then there is a visible and transparent audit trail.

[nigelmegitt]

It isn't obvious to me why giving the Team an action to deal with the Team's inaction would cause anything meaningful to happen, other than incrementing the count of actions assigned to the Team.

[dwsinger]

If we focus on the initial question as @samuelweiler suggests

I propose that we focus on the example in the opening message, namely chartering.

For existing groups, the group could hold a vote (or have consensus) to send a new charter for AC review. That should trigger a time-limit by when it will be sent. The team can (and should) explain to the AC their misgivings.

But that doesn't handle the question of new charters.

I wonder whether we need to open AC Appeal up and say an AC Rep may appeal the lack of an ability for the AC to decide something, i.e. appeal for an AC vote on something?

[joshco]

@nigelmegitt The point of my proposal was to provide a middle ground less than setting a fixed time-limit, since some things may genuinely require more time for consideration than others. It also assumes the Team is already reviewing status and implicitly taking a "not ready" action, and provides more transparency/a space to argue.

[fantasai]

For chartering specifically, I think we need to improve the structuring of the chartering process. Specifically:

• Require draft charters to pass through a central charter-drafting repo.
• Institute a clearly-tracked way to request a "last call" review on a charter once you're done drafting, and a clearly-tracked way for the charter drafters to request submitting it to the AC once they've handled all the comments (if any).

I think that should address a lot of the concerns around chartering, because it makes it clearer how it's done, what's in process, where it's stuck, etc. And if it's stuck on the Team, then we can know it and deal with it.

But that's part of a much broader discussion about Chartering, see https://github.com/w3c/w3process/issues?q=is%3Aissue+is%3Aopen+chartering

This issue, afaict, is about "what if the Team refuses to make a Team Decision"; which may apply to chartering but could equally apply to not approving a CRS or other things.

[plehegar]

[[ plh: this is about veto by inaction … this is not part of chartering; it's part of clarifying what happens when the Team basically makes decisions that are not recorded as Team Decisions. … there is a thread since July on this. Elika was the last one to comment. … this seems to be part of the chartering discussion. Is that the case?

florian: I suspect in theory it's broader than chartering, but in practice, 90% of what people care about is chartering. We should table this until we've dealt with chartering.

plh: any objection? … moving on... ]] https://www.w3.org/2023/09/27-w3process-minutes.html#t04

[frivoal]

Just wanted to point at two things:

• https://github.com/w3c/process/pull/851 (already adopted) addresses "veto by inaction" at chartering time
• https://github.com/w3c/process/pull/925 (under discussion) would address cases like the long-delayed Privacy WG or PAG WG charters, by clarifying deadlines and introducing a fallback mechanism when the deadlines are exceeded.

Though this doesn't plug everything (want a workshop and nobody is responding? want a new version of the Process and nobody is responding? want the CEO to take disciplinary action and nobody is responding?…) I think these will cover the most significant actual sources of concern with regards to veto by inaction.

[frivoal]

(This issue may also be related: https://github.com/w3c/process/issues/773)

[frivoal]

Given that the two main cases are being addressed (see https://github.com/w3c/process/issues/778#issuecomment-2398869866), I wonder if we should close this and ask for individual issues to be open about remaining problematic cases.

Do we need to keep this one open in order to pursue a generic solution? I'd say we don't.