Meeting minutes
Simplification ideas for the process
<brentz> w3c/
Brentz: AB also looking at a process refactoring; will interview groups during TPAC
Florian: Process 2020 includes a Rec Amendment process...proposed amendments are included as preview notes. Then there's a process for approving the amendments (using the full process).
… we also tried to make it easier (fewer stages to go through than the default set of stages)
… but we introduced some "inconsistencies" compared to the normal path
… it's kind of weird.
… in the meantime (more recently) we have simplified the "ordinary" rec track process (and in a less bizarre way)
… so one simplification is to recast the amendment process to align with the new simplified "ordinary" rec track process.
Florian: Another class of of simplification will be to review (and perhaps remove) some terminology.
hober: +1
<plh>
Florian: Tooling may be helpful. It would be good to have a way to do Rec maintenance without annotations and without dropping quality controls.
Ian: What should the CG be doing at this particular time in light of AB revisiting the process?
Brentz: The rough plan is that the Process CG cleans up the GitHub repo to see what issues are still outstanding and have a clean slate
… so that we have a starting point.
… that's what the AB wants the CG to start with.
… meanwhile the AB will formulate a questionnaire and plans to get feedback from the community
brentz: The AB will then collate information and that will inform AB plans for a more significant refactor.
… survey will inform how significant the refactoring should be
… in addition to cleaning up the GitHub repo, the AB would like to hear from the CG input on what should stay and what should be simplified.
hober: Regarding simplifying the amendment process: the 2020 process has been in place for about 5 years. How many times has someone tried to use this amendment process?
plh: 4 or 5
Florian: That's the number of times amended specs have graduated.
hober: Chicken and egg problem if the tooling is important to making this process easier.
… it would be good to know (but hard to measure) how many people were scared off before trying this process.
<brentz> anecdotally, at least two of my group haven't even attempted the amendment process when updating specs
hober: maybe we don't need the amendment process at all
plh: My advice would be to talk to the Chairs before doing anything.
… we are struggling to get specs maintained. It's hard to find people to maintain specifications; it's not the process.
<Zakim> brentz, you wanted to lay out my understanding of the plan
<hober> +10000 plh
plh: It's not a tooling issue. People just want to edit a single document and not have to deal with annotations. Before designing a solution, please wait for TPAC discussion. Need to find out how to make maintenance as low cost as possible.
tidoust: I see 4 ways to maintain a spec in the W3C process.
… (tidoust mentions all of them)
… some groups use different approaches (e.g., delta specs and CSS WG)
… my concern is that I don't know as TC what to recommend among the four approaches.
… the nuances are thin to me. It would be good if we could simplify that.
… or at least explain the differences.
<hober> +10000 tidoust
Florian: There are different approaches for different preferences. Where it gets tricky is that people want to get the Rec stamp without doing all the work.
… what we do depends on what we want (e.g., quality controls, patent policy, etc.)
… the area where we can simplify is amendments
Florian: In terms of effort, the thing I'm proposing is "medium effort"
… if we are planning a major revamp, then we should not do it yet.
TallTed: Tess asked about whether people have tried to do maintenance
… several groups in which I participated have tried to do the proposed changes w/ markup
… it's painful
TallTed: -1 to benign editor model. Editors are doing a difficult task but I think that some people do a better job than others representing the will of the group
TallTed: I've seen editors rewrite documents over a weekend, for example
TallTed: I think not all recs need to go through the full Rec track process.
… but many do
… there has to be a better way of tracking changes, flagging proposed changes with rationale; but manual markup is painful.
Florian: The WHATWG has a small number of very senior editors; W3C has more editors and this diversity suggests more guardrails.
Open PRs
<brentz> w3c/
<plh> +1 to Florian
<Ian> +1 to merge
RESOLUTION: Merge 1083
<florian> https://
Florian: A number of PRs depend on the AB (and/or TAG)
… I'm not comfortable landing anything prior to having AB direction.
brentz: In the absence of direction from the AB would be merge things into an AB/TAG branch that the AB can look at during its considerations.
hober: I'm fine with that
hober: I think the AB is unlikely to come back to some of these issues immediately.
<plh> +1 to have a separate branch
Florian: I'm generally fine with a separate branch, but it's very important to communicate clearly this is a separate branch that does not represent AB-approved content
Brentz: I will be the point person if concerns arise.
ACTION: Brent to ensure the AB is onboard with this plan to create a branch, and will proactively let the TAG chairs know the plan regarding a new branch
Ian: I propose we prioritize edits that help get work done, and deprioritize org topics
Hober: The AB is interested in both sides of the house: group work and org/governance
brentz: I think the AB is also looking into modularization
<brentz> https://
brentz: I will be looking at ll the PRs with the AB/TAG discipline topic
ACTION: Florian to merge those PRs and if issues with merging arise, reach out to the CG
<brentz> https://
<brentz> w3c/
Florian: I think that 1073 can be addressed in pubrules rather than in the process.
brentz: Reading it, I'm thinking similarly.
plh: Our registries document how they are to be modified.
… and the processes may be different.
… the process document does not (and should not) define a single registry update processes.
<brentz> discussing w3c/
Ian: Suggest we close 326 because we are working on this and it's very high level
(We add tab "propose to close')
Ian: Also suggest home work that people look at issues to close
Florian: I also suggest chatting with Brent to find issues to close