Meeting minutes
[agenda bashing]
Progress on FO
dsinger: Should get an update from Team on where we are
plh: My understanding was she'd bring the issue to PSIG, and didn't see if anything came out of it
dsinger: I believe PSIG met last week, don't think they had comment on 3.1 and still thinking of 3.2 order of precedence and unlikely to conclude soon
<florian> fantasai: I was at the meeting, and for the question of precedence, they thought it needed a lot more review
<florian> fantasai: (by PSIG, the director, etc), which will take time
<florian> fantasai: so the suggestion is to approve the Process 2021 without the change, and deal with it in the next cycle
dsinger: Seems that all Process CG just needs to say that we're ready to land text for precedence order once the community has agreed what to do
florian: The comments are against what was in the Process already, not against what's new
florian: Is it OK to address later?
jrosewell: [clarifies concerns]
florian: My understanding was that PSIG wasn't concerned about the proposed order, just that it didn't have enough review
dsinger: We are not in the business of resolving FOs in the community itself
dsinger: But I think on the normative references that have implications on member behavior, we are agreed to pull changes that make them subject to the same review and approval as the Process itself or to reference a dated version
dsinger: I believe that's our consensus in the Process CG?
dsinger: so that changing the dated reference would require balloting as well
https://
jrosewell: Would like any conflicts to get addressed
florian: We're trying to address your FO with two PRs
florian: One is about what happens in cases of conflicts, which is awaiting PSIG review and we can't fix yet
florian: But the other part is ^
florian: Which we can address in this group
florian: for this cycle
dsinger: We can and would be willing to do for this year
dsinger: The second question, about order of precedence, we're happy in this group to insert such a statement
dsinger: but need PSIG, Director, Team, etc. to review it
jrosewell: don't understand PSIG's role
florian: PSIG is an informal gathering of lawyers
florian: If they don't approve, then AC will be concerned
dsinger: I can explain role of PSIG later
jrosewell: I'm concerned if it takes a lot of time for lawyers to resolve
dsinger: If you're aware of any actual conflicts among documents, now would be a good time to bring it up
dsinger: At the moment, we're only aware of this as a hypothetical concerned
fantasai: Issue is not just that PSIG and Director and Team need to sign off, but that this is a significant enough change that requires AC ballot as well
plh: what fantasai said, might have to go through AC for that kind of change
plh: but it is also PSIG's role to make sure our legal documents are in order
plh: Proposal is to adopt the Process as-is, and guarantee to fix it for 2022.
dsinger: Note that it is only convention that we change the Process once a year, could do another change as well
plh: to be realistic, I think it'll take PSIG a long time to resolve anyway
dsinger: So from Process CG's pov we'd like to solve 3.1 and would like to solve 3.2 as soon as we know what can be done
jrosewell: Our initial review was light, but as we participate becoming more aware of problems
plh: Other consideration is that Process 2021 includes a lot of other improvements, and you're proposing to hold those all back for the sake of this issue.
dsinger: Can you take that to an FO resolution meeting?
dsinger: That's outside the scope of this meeting.
florian: There is a fork for P2021, and top of tree is ready to take changes for 2022
florian: Only taking changes approved by Director for P2021
florian: As soon as P2021 takes effect ...
https://
dsinger: So CG recommends pulling this immediately
florian: My suggestion is to land it now, at least for P2022
florian: and if the Director decides for P2021, we can pull it for P2021
dsinger: But record that ProcessCG recommends landing for P2021
Resolution: Land normative referencing fix for P2022
Resolution: Recommend landing normative referencing fix for P2021 to Director
P2022 mechanics
florian: [reviews PRs that are open]
florian: All of these are fairly minor
florian: chaals pointed out that in one of the DF PRs, there's a tiny bit of missed dependency
florian: We change how MOUs are decided
florian: and in another part of document, describe how they're made but didn't update that
florian: that's a minor fix, can fix
dsinger: So you'll be updating CEO rather than Director for MOUs, right?
florian: Yes
florian: And the rest is ready to land
Topics: DF Pull Requests
<dsinger> https://
florian: With changes introduced in P2020, it's possible to update RECs to adopt new features
florian: We have a term defined in the specs for that kind of RECs
florian: PR is to fix to use vocabulary consistently
wfm
Resolution: Accept PR 567 fixing language about RECs that allow new features
florian: 563
florian: There's this strange notion that because comments by AC reps will be addressed by the Director, if you're in CR phase you *have* to address comments from everybody *except* AC reps
florian: There's no reason to make an exception for comments by AC reps
dsinger: Yes. Much easier to address comments earlier than later in Process
+1
Resolution: Adopt https://
florian: This next one is about ....
florian: Once you're at CR, there's path to switch to different stages, including WD
florian: but it is not possible to go from REC to WD directly
florian: even though you can do this in two steps, possibly even in the same day
florian: This allows that path
<Zakim> weiler, you wanted to discuss changelog/explanation (meta topic)
<florian> The PR currently discussed is https://
Resolution: Adopt the PR to allow this transition
weiler: Btw, how do we generate the list of changes?
florian: Manually
weiler: I was in a meeting where the Board forgot one major change in its list of changes
weiler: Maybe we should incrementally update the list of changes?
florian: We do multiple things
florian: There's a manually-curated changes list. It could be wrong, but I don't think it often is.
florian: We also provide a diff, which is automatically generated, which shows each of the changes
florian: We also link to the GitHub changelogs.
florian: And we *also* create a document listing all the issues that were addressed and how
weiler: ...
dsinger: We have multiple processes to manage this
plh: Let me rephrase Sam's suggestion. Why not each PR updates the Changes list?
florian: I'm not convinced that maintaining 5 changelogs is going to reduce the editors' amount of work
dsinger: Need to check again anyway, in case changes were overridden later in the cycle
florian: or if we accepted a PR that didn't include changelog
dsinger: It's a nice suggestion, but I delegate this kind of management to the editors
dsinger: I require them to produce the changelog, how they do it is up to them
dsinger: but thank you for the suggestion
dsinger: Any other PRs for today?
florian: Next ones are all about the Director
florian: These were made because they are easy
florian: We have harder ones coming later, but it will be easier to deal with complicated things if we do the clean-up things first
florian: SO I would like to request the group reviews all these so that group can land them
plh: Would be good to have Jeff around
florian: Let's make it clear that we will land them next time, unless some problem raised.
dsinger: Yes, I would like to start landing those pull requests
Resolution: Plan to pull the Director Free PRs next meeting unless reviewers raise problems with them
...
Exclusion opportunity when work leaves a Working Group
dsinger: For work transferred out of a WG, we don't have an exclusion opportunity.
<florian> github: https://
dsinger: Do we want to refer this issue to PSIG?
florian: It is a valid concern, and was brought to PSIG before, but they didn't have appetite to solve it
dsinger: Well, we can put it back on their plate
florian: If groups are diligent about what they do, can be worked around
florian: if we want it to be automatic, we need something in the process or the patent policy
fantasai: I did bring this up to PSIG last meeting
fantasai: I'm going to try to get them to address it
fantasai: I think they're more aware of the problem being a problem now than before, thanks to this issue
plh: So comment in issue that waiting on PSIG
P2022 Triage
dsinger: Florian and I labelled a bunch of issues for P2022
dsinger: Asked group to review
dsinger: Nobody seemed to add anything, should we move to P2022 milestone?
dsinger: OK
florian: I don't think we need both P2022 label and P2022 milestone
weiler: I'd like to propose all the Director-Free one
dsinger: Oh, those are included by implication. That's our top priority for P2022
plh: Do we have issues to track ??'s comments from the AC?
plh: I think we should. There's a question of P2021 vs P2022, can have both labels...
plh: unless already resolved and past the point of opening issues
dsinger: I think we already addressed normatively-reference member behavior, because already resolved that
dsinger: But maybe worth opening an issue on the precedence question, to get community input
florian: Would recommend leaving out the question of whether to address in P2021 or P2022, leave that to the Director
dsinger: We have too many issues in our repo, and got opposition to closing in bulk
dsinger: Open to suggestions of what to do with these one by one
Issues to Close
<dsinger> https://
dsinger: Propose to close 447 without action, because Process doesn't cover these types of groups
weiler: I disagree with rationale, but agree with the action
Resolution: Close 447 as out of scope
dsinger: 429
dsinger: We introduce a concept for W3C Statements to let any group to raise things to Statement level
dsinger: If we need another mechanism for political issues, I don't think we need to address here
Resolution: Close 429
dsinger: 356, streamlining the Process document
dsinger: We spent a lot of time on it last year
dsinger: And the issue is quite vague
florian: Agree to close it. We can address specific issues that are opened.
jrosewell: It is very difficult to follow in practice
jrosewell: I know not the intention
jrosewell: I wonder if next year as part of work for substantial changes
jrosewell: perhaps this issue should be addressed then
florian: What do you mean by addressing the issue?
jrosewell: It's not an issue that can be fixed by fixing a paragraph here and there
jrosewell: it's to achieve reduction and simplification
jrosewell: I don't know how that gets actioned, but it's not a single PR
jrosewell: it's a requirement of a new organization
jrosewell: I think it should be recognized as a different type of issue
jrosewell: so of course, close it if you wish
dsinger: Closing not because we disagree, but because we need specific requests for simplification
florian: Also we did do in P2021 a signification simplification of the document itself
florian: not of what it describes, but of the readability
florian: so as much as we can do of one shot
florian: If we want to simplify *W3C*, e.g. by saying "we no longer do X or Y" that's a different issue
jrosewell: If you say there are three things described over 2 pages each
jrosewell: and those can be consolidated to one thing that can be re-used
jrosewell: then that is a change to the Process which results in simplification of the document
jrosewell: That's the kind of thing I'd like to see it
dsinger: I'd love to see it, if anyone has a specific suggestion
fantasai: Wanted to say what Florian did: that we did a major editorial clean up last year
fantasai: and that we need specific suggestions if we want to do anything more
Resolution: Close issue about streamlining Process
dsinger: James, you raised a lot of issues. Would you be open to doing a triage together with myself, Florian, and Elika?
jrosewell: They fall into multiple categories
jrosewell: ...
jrosewell: Conducted a review of Process last year
jrosewell: tried to keep into discrete areas
jrosewell: very happy to do that
jrosewell: I think many of those issues came up via discussions in AB and Team
jrosewell: was referred to Process CG
jrosewell: so a bit circular
jrosewell: I'm happy to close if not relevant here, but where the issue still remains, would like to try to find a home for it
dsinger: OK, then I'll try to set up a call with the 4 of us
[ discussion about scheduling ]
Director-Free
dsinger: There are 35 open issues
dsinger: we need to start reducing this
dsinger: I'm not expecting to see action in this meeting, except to highlight that we should turn our attention to these issues
florian: I want to note while we're here, these are not 35 issues about the current text on Director Free
florian: Some are theoretical concerns that are not in the text we are considering
florian: We should not assume that the latest comments in any issue is current
florian: This is why I want to land the easy parts of DF first
florian: Then we can close some of those issues
dsinger: Anything you want to highlight as people to look at soon?
florian: Start with the 4 PRs and go at it bit by bit
<plh> https://
plh: Just wanted call attention to 522
plh: We're getting a lot of FOs that are touching on that particular issue
Single Implementations
plh: If Process CG could make progress on it would be appreciated
florian: Are we getting FOs that ...
plh: We're getting both
plh: We got pushback recently because multiple implementations
plh: Are they actually practical and useful implementations
plh: Comment from mchampion is relevant
dsinger: Added P2022 label
<florian> s/Are we getting FOs that …/Are we getting FOs that ask about 2 implementations to pass the CR criteria, or for 2 interested implementors before we can charter something/
florian: It's a political question, not just a Process question
florian: Exiting CR, we might be able to clarify
florian: but whether or not we should charter work based on whether there are implmentation commitments
florian: that's not something we can address here
dsinger: plh, can you take this question to the AC?
plh: I don't think I can commit to that today, especially bringing to the Council
[this will be brought to AB]
dsinger: broader question than Process CG can address
plh: yes W3M is bringing to Council
weiler: about tackling DF issues
weiler: and tying into formal objections for single implementations
weiler: We have objections on charters, and another issue on ...
weiler: I'm wondering about addressing the question of DF-ness
weiler: some approach for charter approval that involves taking the discussion to the community
weiler: observing the FOs getting to the Director or whatever replaces it
weiler: Give the community an opportunity to find its own consensus
weiler: So a more involved change
weiler: Not have the objection thrown to Director to resolve, but encouraging those disagreements to come out earlier
weiler: early enough for community to talk
weiler: That's what I want the most
florian: I don't know if it's specifically DF or not, because we'd have the same problem with or without Director
florian: But it's an interesting problem to solve
florian: an I have some thoughts on how, but that's for another day
dsinger: Yes, we need to find a way to better involve community
plh: open an issue?
florian: We have several
Scheduling
dsinger: Potential conflict for 13th
<dsinger> move the 13th to the 6th October?
Resolution: Next meeting 6th of October
dsinger: Thanks everyone. Adjourned
<dsinger> adjourned