W3C

– DRAFT –
Revising W3C Process Community Group

25 August 2021

Attendees

Present
dsinger, fantasai, jrosewell, plh, wseltzer
Regrets
Jeff
Chair
-
Scribe
fantasai

Meeting minutes

wseltzer: Note that FOs are technically Member-confidential ...

jrosewell: I'm OK with having this discussion in Process CG

Results of AC Ballot

dsinger: we have 22 supporting adoption as-is, one suggesting changes, one FO suggesting changes which is jrosewell who is happily here

Process 2021, formal objections

florian: We defined a term that we don't use, and was wondering why that was

florian: I reached out to explain that the term will be used in Director-Free Process

florian: and we happened to land this bit first

florian: so that should be no problem

plh: Can we get that discussion archived?

florian: I'll ask

dsinger: Great, so consensus solution on that one

dsinger: Next issue was the FO from jrosewell, listed in the agenda

<plh> https://www.w3.org/Consortium/Process/Drafts/snapshots/2021-07-16#normative

dsinger: jrosewell, what documents that are referenced concern you?

jrosewell: all of them, because if they change they change the Membership Agreement

florian: Lots of references, but vast majority are non-normative

florian: CEPC and Patent Policy are normatively referenced

plh: Collaborator's Agreement, conflict policy, and collaborator's ???

jrosewell: My suggestion is to use dated references

<plh> CEPC, Collaborators, Pubrules

jrosewell: So e.g. if CEPC changes to say that anyone in a red shirt is inviolation in CEPC, it would be clear that it is required by Membership Agreement

<plh> Normative references

plh: undated references are CEPC, Collaborators, and Pubrules

plh: CEPC is updated by AC approval

dsinger: Same is true for Patent Policy

dsinger: so covered by going through same update policy as Process itself

plh: ...

plh: unsure if ever have a dated version

wseltzer: Currently-linked document was updated in 2015

<plh> https://www.w3.org/Consortium/Legal/2015/06-invited-expert.html

plh: normative reference is a redirect, so could change it to that version

dsinger: Do we have set of documents of concern?

florian: See the Normative References section

dsinger: Some of those just define terms, e.g. RFC2119

florian: CEPC goes through its own update cycle

florian: It means we can't update them until updating Process, but they go through ballot process on their own

florian: Collaborators Agreement is signed by the collaborators, so which version is relevant to the signer only

florian: wrt pubrules, the authority to judge whether a document is appropriate or not is delegated to the Team

florian: that they document it and use tools for it is part of their judgement

florian: ...

florian: document license

florian: Final point, note that FOs are technically resolved by Director. We can suggest changes to the Director, but can only suggest

dsinger: it is our job to try to find a consensus solution

dsinger: Document license is referenced by ...

wseltzer: The Document License has also been updated following a Member review process, so that is also subject to Member review

wseltzer: want to second what Florian said about the review cycle

wseltzer: What we've done rather than trying to synchronize everything is, we've given each the Member review and had it go into effect when approved by the Membership and Director

dsinger: there are 2 approaches

dsinger: either change to dated documents, and approval happens during approval of Process itself

dsinger: other is to have approval of each reference

dsinger: we've been taking the second approach

jrosewell: Trying to avoid ambiguity, articulated it nicely

jrosewell: if the latest version of document prevails, then simply one sentence is needed to say that

jrosewell: so that lawyer can be clear about the interpretation

jrosewell: if the other option is taken, which I prefer, with dated version used then there's no need for an explanation because linking to specific version

jrosewell: reason I prefer latter resolution because Members can see impact of changes to membership agreement once a year

jrosewell: rather than many times a year

jrosewell: I could go either way

jrosewell: but prefer that option

jrosewell: Main concern is to make sure it's clear what's meant

dsinger: Agree to be clear

dsinger: Need to be clear what set of agreements were in effect on a particular date

dsinger: in case of a dispute

dsinger: They all maintain a history trace, right? Is that true of all of them?

plh: I would be surprised if not the case

plh: even for documents only on our severs, we have version history

<wseltzer> wseltzer: Yes for Document license, Patent Policy

plh: can find the history from 10 or 20 years ago

plh: don't expose to outside, but it's there

dsinger: for Process there's a Previous Version link also

florian: Patent Policy etc. also have that

plh: the only one that might not have that is the Document License

wseltzer: we do in the header, just not in the same way

plh: Wanted to talk about pubrules

<Zakim> plh, you wanted to talk about pubrules

plh: I don't think we should link to specific version from Process

plh: that would be way too constraining

plh: we make changes to it on monthly basis, we make tweaks

plh: pubrules only affects what can be published on /TR

dsinger: reference is from Member Submission section, Team not required to publish document that doesnt' conform to pubrules, and noted that they can change from time to time

<Zakim> florian, you wanted to suggest we add the doc-license document to section 11

plh: can get the dated version if necessary

florian: further, pubrules doesn't apply to Members. It applies to documents that get published.

florian: Wendy pointed out when we did revise the Document License, we did go through ballot process

florian: Unlike CEPC and Patent Policy, Document License isn't rquired to do so

florian: Maybe we just do that

florian: so readers can trust that this will happen

dsinger: For everything we normatively reference, we should either use dated version or establish that it follows this process

<Zakim> wseltzer, you wanted to propose "then current"

dsinger: or that it's not material to conformance of Members

wseltzer: Agree to make it clear that we're referencing latest version approved by Membership and Director

plh: object to adding pubrules to that list

dsinger: let me rephrase, for documents that have impact on Member behavior, we need to make clear by using either dated version or applying approval process

jrosewell: Other comment is order of precedence

jrosewell: None of these documents explain how conflict among documents is resolved

jrosewell: without changing Membership Agreement, Process would be the right place to do it

fantasai: let's take that up separately

dsinger: I think I should have a conversation with wseltzer and Florian and draft up some text for jrosewell's review

dsinger: but see the way forward

<wseltzer> +1

plh: we will circulate to all the AC reps who responded

dsinger: of course

dsinger: Next question, do we need a precedence statement?

dsinger: Obviously we try to avoid conflicts, so haven't thought about it before, but is a reasonable question to ask

dsinger: I don't see a problem with such a statement

<plh> fantasai: agreed with adding the sentence but Patent policy prevails on the Process

fantasai: I'm fine with adding such a statement. I would give Patent Policy precedence over Process

florian: I was on the queue to say the same

dsinger: interesting, so not following the chain of inclusion

dsinger: anyone with a problem?

<wseltzer> [the proposal I heard was laws, member agreement, patent policy, process, linked documents]

dsinger: seems we have agreement, need to draft up the edits now

Open Pull Requests

plh: we should take the PR from wseltzer

<dsinger> https://github.com/w3c/w3process/pull/569

florian: There's a bit where the Process tries to describe the Patent Policy and does so incorrectly

dsinger: Does anyone object to incorporate this year?

plh: it's new text, and I made the mistake of incorporating into the new templates, and when Wendy reviewed that found the mistake

Resolution: incorporate wendy's fix to the PP description for P2021

dsinger: anything else?

florian: Don't want to land anything yet, because need to resolve some things for P2021 first

florian: but would like to draw attention to some open PRs on Director-Free

<plh> Director's free pull requests

florian: They are not the most important parts, they're the easiest parts

florian: want to clear the way for solving the harder problems

florian: hopefully by next time we can start incorporating them into P2022

dsinger: Yes, let's get started on P2022 as soon as we're done with P2021

P2022

dsinger characterizes some of the open issues

dsinger: We need to work on Director-free this year

dsinger: and establish our processes and practices for working without a Director

dsinger: didn't label as P2022 specifically, but that's the major thing for P2022

<dsinger> https://github.com/w3c/w3process/labels/P2022%20Proposed%20to%20address

dsinger: editors and I went through the list of issues, and labelled ones we believe are worthy of attention this year

dsinger: 2 questions

dsinger: are there issues we think we shouldn't be trying to land this year?

dsinger: and also, are there issues that should be on this list?

florian: clarification, doesn't mean we can't work on other issues, but that these are issues we are intending to close this year

wseltzer: proposed to address could include addressing as WONTFIX or addressing in other places than Process document

dsinger: we might decide not to do anything on these issues, to close with no action

dsinger: but we think these need conscious attention for the next year

Proposed to Close

dsinger: looked at issues that were not likely to make progress, or couldn't work out what change to Process could help

<dsinger> https://github.com/w3c/w3process/labels/P2022%20Proposed%20to%20close

dsinger: some we re-assigned to AB, because policy questions beyond our remit

dsinger: this is the list we thought we could close without further action

dsinger: one that's had a comment added since we proposed to close

dsinger: if we agree on closing, we'll close after this meeting

jrosewell: In relation to one we're proposing to close

jrosewell: many were raised by me following review last year of Process document

jrosewell: many of these become quite circular, go to a group and they say not our scope

jrosewell: People impacted don't feel that they can't get engaged on these issues

jrosewell: too public to be comfortable, or not have time

jrosewell: if I take one, Dan Appelquist raised issue about user research

jrosewell: I think that would be a big step forward

jrosewell: proposals to do work where there's no justification for the work being put forward

jrosewell: so seems like sensible thing to put into the Process

jrosewell: many issues being closed without a proper debate

jrosewell: maybe we should take these to TPAC sessions

jrosewell: ones I relate to are wrt competition

jrosewell: there was a suggestion of a workshop or something like that

jrosewell: the AB Member-only repo is a bit non-obvious for people, might not be aware

jrosewell: so rather than closing now, maybe taken to forum like that

jrosewell: rather than going through each one

fantasai: Suggest not to mass-close all of the issues

fantasai: but to go through a few each call to go through them individually and agree to close

plh: Agree, don't think we should close these all so quickly in August

florian: Agree

florian: that they deserve individual attention

florian: I agree to close, because part of group that did the triage, but should go through them one by one in this forum

florian: we can use TPAC etc to invite more discussion, but we should address them here

dsinger: Some are proposed to address elsewhere

dsinger: some we took to AB

dsinger: but some that need discussion in the broader community

dsinger: we don't have an issues repo for the Membership, just ac-forum

dsinger: should we have a community discussion repo?

plh: seems like a good idea...

plh: we have one for AB, could have similar one for AC

plh: unless AB says, no please use our repo as well

plh: interesting to see if AB is willing to open its repo to beyond AB participation

wseltzer: point of info, the AB Member-only repo is readable and writeable by the AC

dsinger: so we could use AB-memberonly repo

dsinger: but how would we highlight what needs community discussion vs AB work?

dsinger: so those of us on the AB should get this on the AB agenda

dsinger: and I will add a label, Proposed to Transfer

dsinger: and I'll ask if you add that, please comment where you propose to transfer it *to*

dsinger: that would enable James to suggest, e.g. if not discussing in Process repo because not a Process document issue, then move to some other forum so don't lose track

fantasai: Was going to suggest that AB-memberonly is the AC discussion repo

fantasai: AB made that repo intentionally open to AC participation

fantasai: so we should use it

dsinger: concern about another repo for AC, AC is not super active...

florian: in order to do anything, probably has to come to AB anyway, so it's an appropriate place

jrosewell: I would favor not having more GH repos to go to. Would like fewer.

jrosewell: easy to miss notifications, there are so many

jrosewell: more practically, the AB-memberonly repo and some of these discussions are not being picked up on by people who should pick up on

jrosewell: e.g. Jeff's comment about discussing these issues more with adtech

jrosewell: not a single member of adtech community picked up on that

jrosewell: I happened to attend the AC Office Hours discussion which mentioned it

jrosewell: some members tried to look, but got a 404 error

jrosewell: because didn't have their GH linked correctly

jrosewell: people we need aren't engaged on these forums

jrosewell: maybe need discussion at TPAC

dsinger: including in TPAC sessions is indeed appropriate

dsinger: agree, I get snowed by GH notification

dsinger: and don't want us to create a corner where nobody notices

dsinger: I'll raise the idea of Needs AC Input label in AB-memberonly repo

florian: maybe we can advertise that repo more to chairs and AC

dsinger: maybe advertise periodically somehow...

dsinger: so proposed resolution is to address these issues one by one

plh: remember that while AC can read AB repo, they cannot raise issues

plh: it's read-only, not write

florian: are you sure? It's meant to be writeable.

plh: other thing is, we need to make sure AC knows how to access those repositories

plh: every office hours I have to give access to another AC rep

<dsinger> (we agree to check that the AC can raise issues and have the access they need, and get notifications and are kept aware)

plh: it's not complicated, but requires some steps we cannot do for them

fantasai: For issues that should get broader discussion, someone should bring to attention of AC by posting to ac-forum

<florian> (there are issues in the AB member only repo that have been opened by people who're not on the AB. So I think it's writeable)

fantasai: also, could link to Needs AC Input issues from the AC Office Hours announcements

fantasai: because those go out periodically

jrosewell: One more issue, people who work for W3C Members and maybe don't even know they are,

jrosewell: have a session to help people link accounts etc.

fantasai: we should link to instructions for linking GH account alongside the link to the issues which aren't visible unless the account is linked

Scheduling

dsinger: I have a conflict two weeks from now

plh: Jeff is also unable to make Sept 8th

plh: I would vote to cancel

fantasai: could also reschedule

florian: I would prefer 1 week later

plh: it's our Geek Week week

dsinger: so we're canceling

<dsinger> we agree to cancel the sept 8th

dsinger: Thanks everyone

dsinger: and thanks jrosewell for your attention, appreciate it

dsinger: even if I don't always agree :)

Summary of resolutions

  1. incorporate wendy's fix to the PP description for P2021
Minutes manually created (not a transcript), formatted by scribe.perl version 136 (Thu May 27 13:50:24 2021 UTC).

Diagnostics

Succeeded: s/here/first/

Succeeded: s/??/Geek Week/

Maybe present: florian