Meeting minutes
<TallTed> fantasai -- see https://
Introductions
plh: Philippe Le Hégaret, W3C Team, co-chair of Process CG
fantasai: Elika Etemad, Invited Expert -> Apple
<Param> Congratulations Elika!
TallTed: Working for OpenLink Software since 2000, many W3 groups since 2002, involving linked data / RDF /etc, I'm probably involved
… I do a lot of nitpicking on PRs
plh: thanks fo ryour nitpicking, much appreciated
fantasai: +1
joshco: Josh Cohen, random community person, from ?? and previously IETF
florian: Florian, French living in Japan, Invited Expert primarily CSSWG and also a bit i18nWG
… also Advisory Board (AB) member and liaison to W3C Board
baram: [missed name], making accessible applications, most recently life insurance company
… all lifecycles of software development
… role here is education of importance of accessibility
… recently senior at IEEE, this is my first time in this group
plh: Thanks, fantasai, for actually sending an agenda (unlike your other co-chair...)
Candidates for a possible 2023.1 release
plh: When we talked about commentary from the AC during AC review of previous process
… there were a few substantive changes that we asked the Director to consider
… and Director declined to do at the last minute, because concerned about possible controversy
… idea was that some of these changes, we might need before end of the year
… primarily motivated by participation in the TAG
… There was a request to increase the number of the TAG
… there is an election for the TAG at the end of the year; we need to know by November how many seats would be available
… so that would be one of our motivations to move fast
florian: That's basically question, there are a few things to consider:
… 1. do nothing
… 2. do only TAG size increase
… 3. do all three changes considered
… 4. also do editorial fixes
… my recommendation would be to do TAG change only
<plh>
florian: I think throwing anything else into cycle would slow things down
… so if we want to do a quick cycle before the election and have as narrow as possible of a scope
plh: So I think we are considering 3 pull requests (not necessary to take as-is)
plh: #761, #760, #733
plh: I think 733 is hardest
florian: I think the opposite, a change in number of seats on TAG, whether we want it is interesting, but if we decide we want it
… then doing it is very simple
… the other two are more open-ended
plh: In the reaction that I've seen, e.g. from mnot
… some uneasiness in increasing the number of the TAG
plh: One of concerned is balance, in Council we have TAG + AB
… and Team appoints to the TAG after elections
… with approval of 2/3 of TAG and AB
… nominated seats was to improve diversity (of various axes)
plh: Now that we approved Process, we have consecutive term limits
… do we count from the past? or do we start counting when the Process was adopted?
… my thinking was we start counting in the past
<plh> fantasai: we would start counting from the day the process came into effect
<plh> ... that's my recollection
<joshco> "grandfather" clause
<joshco> or "grandmother"
plh: I'm pretty sure that was our plan
florian: I agree with fantasai, though I'm not sure where the minutes are
florian: I've also heard from the people concerned that they are thinking to go into election ahead of time
… which is another reason to increase elected seats
… because elected seats are already a crowded field
… this makes the transition harder
… so giving some breathing room so that they can run into an election without feeling like good people, possibly themselves, will lose just because not enough seats, would help
plh: To be clear, the two people under question here are the co-chairs of the TAG
… Team has been re-appointing them consecutively
… though I agree it would be good if they run for election
joshco: so 2 consecutive terms, so if necessary they can take a year off and be nominated again?
plh: not like US president
… can run for election
florian: or sit one out
plh: Underlying goal of Team is to address diversity and fill the gaps we have in the TAG
(in expertise, etc.)
joshco: So the point of the Team nomination is diversity? Is the assumption that marginalized communities can't win election?
florian: not quite it
… there are many facets, marginalized communities is only one
… but TAG specifically has a broad technical role for which we think it's important that every technical aspect is represent
<plh>
<plh> The Team should use its appointments to support a diverse and well-balanced TAG, including diversity of technical background, knowledge, and skill sets.
florian: so if we miss important communities in terms of individual characteristics, that is indeed an aspect of diversity to consider
… but also, consider e.g. "oh, this TAG has nobody who knows about security", or "we have no expertise in rendering"
… or we "oh, we have nobody with skills to chair"
… need to fill the gaps
… that's why we moved appointment to after election rather than before, so we can look and see what's missing
joshco: It's a bit weird for a group to appoint its own...
florian: There was some consideration of that, and we added some guardrails, e.g. AB has to ratify also
… but existance of appointment isn't new
… what we're changing is the timing and constraints
… and also now we're discussing adding people
… that suggestion came so late in the cycle that the Director felt it was too late to take it in
florian: Since we previously thought it's a good idea, then it's probably still a good idea
… but if we wait until next cycle, then we miss the next election
joshco: The decision to do Director-free, was that in a different body than this?
… and Process CG has a mandate to figure that out?
florian: Yes. The AB is in charge of evolving the Process
… but in order to broader participation
… decided to have work done in CG
… but the AB has oversight
… However in the end the AB doesn't decide, the membership at large votes through AC, but AB sets the direction
joshco: So this question of whether to expand elected/appointed seats
… should that have a decision from AB?
fantasai: The AB already resolved to make this change
joshco: so membership decided we want this, and this group figures out how to do it
florian: There's a twist, because the comment came in right at the end of approving Process 2023
… AB's suggestion was to fold it in
… so in the past; but Director didn't approve
… so now we need to consider for the future
… AB will probably review in August
florian: Our job is to say if there's anything else we need to fold in...
plh: [outlines timeline]
… Terms start Feb 1 2024, but election season starts in mid-October
… so we would need the new Process to get adopted October 1st
… which means AC review needs to finish mid-September
… so we need to start mid-August the AC Review of 2023.1
… so we have a month and a half to come up with that
… my recommendation is we propose new Process to AC before mid-August
… we discussed the three PRs in the past and were supportive
… so that would be the only way to get it done in time
florian: I agree with you, except I would only take the one PR
… since we can do the other 2, even if we don't land them
… it makes the rules mandatory, but we could do them voluntarily
… I want to reduce the topics about which there might be objections
… so smallest change possible seems advisable
<plh> fantasai: +1 to limit the scope
florian: if we increase the scope, then people might ask for more things to be addressed
… though I'm ok to add some simple editorial fixes
plh: Agree. Nothing complicated, or that people might say "maybe that's substantive, unclear"
joshco: Yeah, don't want to say it's open season
… so be specific and say "this is the only thing we're going to do"
… want to be able to say in a binary and non-personal way, this is the only thing we're doing
florian: So I suggest an action item on the editors to prepare all the material
… e.g. changelog, description of what it is, etc.
… so we can go to the AB meeting and ask if we should launch the whole review
… so we don't have to start in mid-August
plh: would be more comfortable if we had +1 from Chris/Tzviya
florian: We can check with them, and if they think it's reasonable, can prepare
… next virtual call is July 6th, but I will miss them
… don't know agenda
plh: How about I'll call their attention to this, and secure their +1
… if they have a -1 it would be good to know!
plh: but I believe the resolution is to fold 733 into 2023.1
florian: Recommend to AB that we do the .1 release with that and that only
https://
plh: Anything else on Process 2023.1
PROPOSED: Propose Process 2023.1 to include only PR 733 and kick off AC Review no later than mid-August, spanning TPAC
plh: I think it would be better to end before or during TPAC, so that if someone FOs, we can discuss during TPAC
… if we can finish sooner, more power to us
RESOLUTION: Propose Process 2023.1 to include only PR 733 and kick off AC Review no later than mid-August
Process 2024
plh: What do we want to do for Process 2024?
florian: There's a pile of issues to pick from what we can deal with
… but wrt topics to priorities, I have 3:
… 1. Chartering. Chartering process is very free and open
… though we have many practices around it
… who chairs chartering, whether there's a DoC, none of this is formalized
… I think it would be good to be clear on these things
… I know that there's some discussion in the community that for a member-led Consortium, the chartering phase may be insufficiently Member-controlled ... unclear what to do about it, but that's an area
… 2. I heard many times that changes for REC maintenance in 2020 were too heavy and too complicated
… all of the complication is there for a reason
… but we should have another look and see if there's any way we can improve
… it's possible we can't simplify, but worth a look
… 3. idk if Process or not, but in P2020, a fastpath was introduced for publications
… and it seems to me this path is underused
… Is it a problem with Process, or is Process fine and we need tooling, or is it unnecessary?
… So should we improve tooling or remove this facility or what?
plh: On chartering, I fully agree. Right now plh has veto power on sending charters to AC
… we need to figure out how to fix that
… although in practice it's more complicate
… but I have two groups that want to be chartered and aren't yet
… one is the ? CG came with a proposed charter, and Team is blocking because they believe the charter is too vague or too open-ended
… so asking for it to be clarified
… other case is DID WG charter, where one member of that group objected to a chair decision, and has been asking for an appeal of the decision
… looking into it until recently
… so 2 charters not sent to AC by Team for two different reasons
… I'm not saying Team is wrong here, but the room for debate is not clear (even though most of these conversations are happening in GH)
plh: REC maintenance, too heavy too complicated, I agree
… every time I explain to WG how ti works, they look at me sadly
… part of it is tooling
… for legal reasons, we can't let them modify the actual text of the spec
… and we don't have good tooling to make this easy
… We need to make lawyers and AC happy by not changing normative text without normative review
plh: Wrt fastpath, the tooling was never done
… but on the other hand, I haven't heard any groups complaining
… very often we reject a publication it's because horizontal review wasn't properly done
… e.g. new a11y section in the spec, and no proof that the AP group looked at it
… so I don't think it's hurting in practice
… we're pretty good at dealing with transition requests as they come in now
plh: 4th point, you didn't mention the frontier between incubation and standardization
… 4 I's question is big one
… and that's partly what's blocking on ??? WG charter
… we want to have a group on privacy, but don't know how to move forward
… and they want a blank check to define their own scope
… and team believes that AC would not allow for this
… while the WG would really like to do that
<florian> fantasai: I think the reason we haven't used the fast path is both lack of tooling and because the normal path has got faster
<florian> fantasai: we might not need it anymore
<plh>
plh: People are starting to understand how transitions happen, and starting to comment on them
<Zakim> tzviya, you wanted to ask about simplification
tzviya: Wrt what to work on in P2024, I don't think this will be easy
… but people say Process is too long and too complicated
… nobody wants to even touch it
… I think wendyreid suggested an overhaul
… but when Wendy was doing a transition, she had to read the process 3 times to consider the possibilities for the standard
… that's a problem
… she had some proposals
florian: There is a general length problem, but Wendy was specifically talking about #2, REC maintenance
… it's very complicated. We should try to do something about it, but I don't know what
<tzviya> /github.com/w3c/w3process/issues/429///github.com/w3c/w3process/issues/429
plh: Problem is chairs only look at this at best once a year
… and it's a long list
… so every time they do a transition request, hard to remember what's involved
… so our standards have been getting higher about what we want from the standard
… e.g. we got an FO against an implementation report, because it wasn't explicit enough
… so I have a long list of recommendations for implementationr reports
… but this will increase requirements, and make transitions harder
florian: So in general, if we find an opportunity for simplification
… a couple years ago there was a trend by Wendy Seltzer to shorten (by lenght)
… but just shortening isn't the goal imho
… making shorter and harder to understand, not an improvement
plh: Process is one single document, but it covers everything about Consortium. Defines groups, defines liaisons, defines REC track, etc.
florian: Yes, though in P2021, we did refactor it
… so each chapter is largely self-contained
… Not quite CSS modularization with different documents, but now each chapter is more self-contained
tzviya: yes, it's chunks, but individual sections are incomprehensible
… trying to find the issue wendy filed
… but ppl not involved in Process, it's impenetrable
… nobody reads unless they have to
florian: I don't disagree, but it's also not actionable
joshco: I agree. I'm new to W3C, and I founded the Process community in DMTF
… org was evolving, becoming more international, a lot of process was not really documented, and working on easy to read document
… but trying to read the document here, a lot of time if walls of text/code that's hard to epnetrate
… some can be helped by formatting
… e.g. list the requirement, and then separate paragraph for explanatory text
… in different sections, you have buried a normative rule in a paragraph, and then go into edge cases in the next paragraph
… but reorganizing the text a bit and reformatting...
… ppl who are new will skim the document to find the piece they're looking for, and will be helped by framework of high-level points
… and then go deeper
… way it's formatted now, the minutiae are mixed in with the important points
plh: Even though Wendy didn't submit an AC review comment, I did hear from people that Process is getting more complex
… it's getting super complicated
… our trend is to make more complicated
… because trying to address every loophole
… and Director-Free really added a lot
florian: One of the sections that has worst usefulness to complexity ration is Member Submission section
… it's used increasingly infrequently, and it's convoluted
… we could possibly drop the whole thing
tzviya: not the section ppl are looking at
… focus on the progressing a spec section
plh: OK, so that's a start for P2024
… we do have a lot of open issues
… I guess we need to do a triage session?
florian: This is big goals, we'll catch a bunch of small things too
… and we'll start receiving more issues about Councils as we run more of them
End
plh: Next meeting is 12th, do we want to keep it or cancel? I'll be away
<tzviya> @fantasai and @plh here is wendy's issue W3C/