w3process
w3process copied to clipboard
w3c
Is a Community Group the right place to develop our own Process?
I suspect we adopted the CG structure because it was easy, lightweight, and allowed for public visibility. But it seems strange to be using the vehicle designed to support public exploration of web-related ideas, as the vehicle to develop the Process that applies to members.
Should this be an IG with public visibility, ability to invite experts, but making it clear that the W3C are the ones with some authority, and public guests' input is appreciated and welcome but not binding?
Rather than addressing this question in general, I would like to comment on Process 2022 specifically.
For Process 2022 there is a great deal of momentum to revamp the process so it no longer revolves so much around the Director.
For these changes we need to build a larger community of participants (starting with our Members - but maybe not exclusively) that actively participate. We should strive to demonstrate how the revised approach has widespread support.
So, if I wanted to, tomorrow I could start a CG called "Process2" and it would have just as much weight as this CG does. As a participant in this CG, it's not at all clear to me how decisions are made (in the details, and especially when there's contention), how I would object to a decision before it was taken to the AC, and so on. Despite that, the AC for some reason treats the output of this CG as legitimate, usually accepting whatever it recommends. I suspect getting the AC and especially the Team to pay attention to a fictional Process2 CG would be an uphill battle, at the least.
That seems like a bad governance smell to me. If power is going to reside in this group, it should be formalised and appropriately managed.
Developing Director-Free in an accountable, well-understood and rules-based fora is the best way to demonstrate widespread support. Non-members can and do participate in WG discussions, so that won't be a barrier.
So, if I wanted to, tomorrow I could start a CG called "Process2" and it would have just as much weight as this CG does.
Not quite: the AB chose to delegate discussion here, but still holds the strings. Without the AB listening, or the ability to get community consensus and Director approval, another CG would be whistlin' in the wind.
As a participant in this CG, it's not at all clear to me how decisions are made (in the details, and especially when there's contention), how I would object to a decision before it was taken to the AC, and so on.
At the moment, we seek community agreement on multiple levels: detailed work is done in this CG, points of principle and policy are discussed in the AB, and once the AB is happy with a text, it's sent to the AC on a dual-cycle (informal review, formal vote), and W3M for comment, and finally we hope the Director approves it.
As far as I am concerned, D-F means that AC ratification is the actual last step, not an advisory one.
I'd prefer an IG as it would make clear that only members have formal influence; outside voices are welcome for comments and advice.
I'd prefer an IG as it would make clear that only members have formal influence; outside voices are welcome for comments and advice.
I don't have a strong opinion either way. Mark makes some good points about it being good governance to have a formalized process to change the rules. But maintaining the Process is the AB's main formal responsibility,
it was the AB that chose to do much of the work in public, partly to get the input of people who have distanced themselves from W3C so they would tell the AB what needs to be done to win them back. That made a lot of sense back in the days of strong disagreement with WHATWG folks about process issues such as "living standards." Maybe that's not necessary any more?
Also, I wonder whether there is a real problem here worth investing the time (and opportunity cost) to fix. There are only a handful of people who care enough about the formal Process to engage in detailed discussions of how to improve it.. I agree with @jeffjaffe that the more pressing challenge is to get a critical mass of people who will actively participate in generating ideas and building consensus for them. Whether that happens in a CG, IG, WG, or simply the AB+invitees is less important than making it happen.
update: the level of participation in the CG remains low. TPAC2022 was an opportunity to send a call for participation to the community (breakout, chairs, AC, etc.). No one at TPAC2022 has suggested that the CG is the wrong vehicle to make progress.
Not sure what you're saying there @plehegar -- do you mean TPAC was a lost opportunity, or did one go out (I didn't see it)? It's hard to give weight to what people don't suggest here - we should be doing the right thing regardless of whether someone brings it up.
I'm not aware of complaint, I do worry about risk and being excessively public about something that is quintessentially a member matter. We lay ourselves open to being interfered with, and to 'washing our laundry in public'.
If a CG is deemed inappropriate, then we need a new and special charter for, and conversion to, a long-lived and/or easily-rechartered WG, and right quick.
Also, if a main concern is to keep this "quintessential member matter" in our own laundry room, then all current and future Process work should be done in a member-only space, including the members-only WG mailing list, and extending to a restricted-access GitHub repo, etc.
I'm not sure we need a WG (they are usually used to develop rec-track specs with patent commitments); an IG may suffice. I'm not saying that everything has to be member-only, just that in a CG is is all public (there's no discretion), and public and members of the consortium are roughly on an equal footing.
In general, W3C's preference has been to work in the public.
Specifically, for our process evolution, we have operated in public for many years. The results have been (imho) quite successful. We have enhanced our process on an almost annual basis. We have introduced huge change: dropping Last Call, adding CEPC, living standards, registries. We have good momentum on Director-free.
My preference would be to continue what has worked well, unless there has been some specific problem identified.
I'm having a hard time thinking of a case where equal-basis public involvement and exposure was actively helpful. Can anyone?
I'm having a hard time thinking of a case where equal-basis public involvement and exposure was actively helpful. Can anyone?
Not at this level of detail for this kind of work.
(Because others may have the same mental hiccup, IG above now clearly was meant to mean Interest Group, as opposed to the long-ago Incubation Group which was mostly subsumed into today's concept of a Community Group.)
I agree with @dwsinger that this should be chartered into an Interest Group rather than a CG, and also that it is not urgent to do so.
We don't need to make it a Member-restricted IG, it can be a publicly-visible IG that continues to use this existing repository, and I think that's fine. (In fact, unless we fix the problem that most Invited Experts do not have Member access, I would formally object to making it a Member-only IG.) But I think it makes sense to be formally chartered and to draw official participants from the membership of W3C rather than the public at large.
That seems reasonable, and I agree that it's not urgent. Depending on the timelines, it might be nice to get in before director-free ships, since that's the biggest change on the horizon.
Now that we're (hopefully) shipping a new Process, perhaps it's time to reconsider this?
I agree that a CG is a little peculiar (though lightweight and convenient), but this isn't really acting like an ordinary CG, since it is acting under delegation from the AB. I'm not against a little more formalism, but I am unsure that an IG is the right approach: it would be atypical for a IG to have a charter making it subordinate to the AB, which is how this group operates. A taskforce of the AB, open to broad participation, would seem more appropriate to me.
Whether IG or CG or Taskforce AB, I do think having a clear charter wouldn't hurt. Then it'd be clear why this group (and not some other) is empowered to do its work on the Process, and how it operates.
+1 to that -- the exact form is unimportant; it just needs to be properly constituted.
I also think it peculiar that we use a structure open to the public. A TF of the AB, open to members, would seem much better.