Meeting minutes
Agenda bash
fantasai: looks good to me
jeff: someone recommended that we need to draw a line between things with PRs and things without
fantasai: not today. the registry stuff still needs sorting. maybe end of month.
david: we're targeting Spring AC
… informal review in March?
… so AB approval in March
… we'll start drawing lines by the end of the month
… refine the work in February
#461 - Elevating a TAG document to W3C Approved Status
GitHub: https://
David: +1 to not require Notes not to contain implementation technology
… [reading his notes on GH]
fantasai: CSS working group produced a Note that was an interesting proposal informing others
… we should clarify that Notes don't receive IP commitments
plh: example of MSE, where we published a tech mapping
… as a note. That had implementable tech
<fantasai> proposal: W3C NOTE SHOULD NOT specify implementable tech, transition request MUST include rationale for not being on REC track
plh: +1
Resolution: : W3C NOTE SHOULD NOT specify implementable tech, transition request MUST include rationale for not being on REC track
fantasai: what about the naming?
fantasai: "W3C Memorandum"
david: ok with me
<wseltzer> ask the TAG, who wants to produce the things?
jeff: at some point, we should send this to the TAG for them to comments
wseltzer: "memorandum" feels legal to me :(
fantasai: it is used in various context. UK definition for non-legal usage: "an official report about a particular subject that is written for a company, organization, or government to consider"
<wseltzer> "working paper"
<weiler> [and memorandum seems.... not-self-explanatory. it's clearly new, but it doesn't help the outsider]
<wseltzer> "paper"
<wseltzer> "whitepaper"
david: we can use "note", "report", "memorandum", "whitepaper"
fantasai: having separate names would help depending on the level of consensus
david: ok, we'll do a poll in the pull request
… I'll go with memorandum for now
Resolution: Go with Memorandum in the PR for now, bikeshed the name later
#356 - Streamlining the process document
GitHub: https://
David: how can we achieve that?
wseltzer: tried to and it's a big task
… maybe folks want to join me with a big chunk of time to attempt to do this
david: we could have a dedicated meeting for this. I still see this as a serious issue
fantasai: this will have to be done through individuals pull requests
… rather than having a meeting
… it would help if we split the document
… into separate files
… we don't have support for multi-pages documents in bikeshed
tantek: the reason a meeting would be useful is that some of the assertions are debatable.
… we should do it at the highest level rather than doing per section
fantasai: if you have a suggestion, please post it
wseltzer: less formal meeting and more folks sitting with a whiteboard on what needs to be rewritten
david: we can remove repetition and guidance
wseltzer: I doubt we'll achieve this in 2021
tantek: +1 to wendy. any large rewrite takes a while. maybe next year.
<Zakim> tantek, you wanted to note a meeting could ask/discuss what are the goals of the process document, who is the audience (today, not 20 years ago)
tantek: to help the meeting, we can start by looking at the goals and who the audience is
… it's been a while we looked into that
david: I'll encourage folks to look for redundancies and related in the meantime
#130 - Enumerate the requirements for wide review
GitHub: https://
fantasai: we need a specific proposal for text in the process. the revision of /Guide helps a lot
… we seem fine for now
… maybe we can add something next year if needed
david: ok. I updated the issue
<tantek> I like Leonie's comments in the issue
plh: Pointed to /Guide document. I don't think it answers all the questions, but agree with fantasai that I don't see what can be added to the Process document at this time
tantek: was to address a problem or make things easier?
fantasai: HRGs were getting called in very late in the game just before CR transition, which was stressful and difficult for them, wanted clearer guidance on getting involved earlier
… the process is at high level but, to make things more specific, the new guide document helps
#466 - Registries Survey
Github: https://
david: the survey needs to be as helpful as possible
tantek: I like the questions in general. little confused about how many applied to creating vs updating a registry
… 2 different stages
… having troubles to answer the survey as it is
fantasai: the maintenance of the registry is the proposal
… the process to update a registry has to be documented
… the process document does not impose any requirement on that process must be
fantasai: I'll update the proposal to include "for the registry definitions..."
jeff: i'd like to move fast on this since it's getting in the way for 2021
… we should empower a small set to finish the wording
<tantek> I would agree with that, registry table(s) should not necessarily be in /TR
david: if something is on /TR and operates under the Director, should the registry tables be under /TR?
david: we can have a w3.org URL but /TR?
fantasai: we conclude it's possible to publish the tables and the registry rules in the same document
<tantek> I'm more interested in seeing how much this process allows for non-w3.org URLs to registry tables
<fantasai> it doesn't
<fantasai> that's intentional
<fantasai> If someone wants to maintain a registry table somewhere else, they can already do that
plh: I'm fine with the tables being under /TR under the authority of the groups
david: do we need the last question about separate process tracks?
fantasai: I agree that the question doesn't need to be asked to the community
… happy to push it to the AB
david: don't feel strongly about it, just concerned
<Zakim> jeff, you wanted to comment on the length of the process document
jeff: not worried about the length of the process document. complexity/verbiage is more of a concern
… if it's a separate piece, that's fine
<Zakim> tantek, you wanted to give example of HTML and microformats rel registry
tantek: separate registries: HTML refers to the microformat rel registry
tantek: maintaining the flexibility of having the tables outside w3.org should be kept
<jrosewell> Observation: this solution (https://
david: it could be a redirect
fantasai: we need a w3c url to guarantee long time maintenance of the URL
… otherwise it's not a w3c registry
<jrosewell> I see the reason to have on W3C. It's about maintaining a controlled archive. Don't we have a similar issue with use of GitHub?
<fantasai> jrosewell, yes, we do. Which is why I keep complaining about groups publishing their work at github URLs.
fantasai: we need to conclude on the survey.
<tantek> I'm ok with "registry on w3.org" meaning ok that it's a redirect or archived copy on w3.org while the actual maintained / edited registry is elsewhere (on a separate domain etc.)
david: doesn't sound like we can get rid of the last question
<tantek> yes, common practice is tables are separate from registry definitions, AND on a separate domain
david: it's a mistake to bar the common industry practices. how to verify that the registry process is being followed
<fantasai> tantek - sure, but is that because it's a good idea or because of technical / organizational limitations
<fantasai> tantek - we don't have such limitations here
<tantek> because it's what individual groups/communities want, and therefore it's what they think is a good idea. it's about agency and adoption
fantasai: goal of the survey is to get opinions on what people needs
<tantek> also, don't break existing registry processes
<jrosewell> Could we ask IANA for a copy of iana.org/protocols solution and then host it on registries.w3.org ? Then adapt that once it's up and running to meet specific W3C needs.
<tantek> also this is about re-use of existing registries
<fantasai> we're not changing any existing requirements
fantasai: the underlying question is "if you're making a registry at W3C, what do you need from us?"
<fantasai> jrosewell, if you want to host at IANA, host at IANA and reference it. Nothing preventing you
tantek: there are a lot of existing registries. you want to make it easy to reference a registry
… with heavy wording on must be on w3c, this seems a bad idea
… the primary use case is to reference registries
fantasai: we're targeting audience that want to publish registries on w3c
tantek: we should drive to use existing registries
david: our primary audience is the folks who lack a registry process at w3c
fantasai: sure. we're looking for feedback
tantek: the survey is about forking or creating rather than reusing registries
… which is harmful
david: can we improve the first question?
… I'll rephrase the third question
fantasai: we need the feedback asap
<tantek> Does your group need to define a registry or preferably re-use an existing registry?
<tantek> ^ this is the kind of wording change we need
<fantasai> This isn't about using registries, it's about maintaining them.
david: ok, I want to comments before end of this week and I need permission to send this out
<fantasai> You can reference any registry you want
<fantasai> nothing about what we're doing is changing that
<tantek> it doesn't read that way
<fantasai> "Does your Working Group use a registry of some kind?"
<tantek> it fails the clear case of what happened in HTML
david: I'll work with plh to setup the survey based on comments this week
<tantek> the WG did not maintain the registry, merely used it by reference
<fantasai> tantek: it can continue to do so
<fantasai> tantek: it doesn't need to be a W3C Registry for a W3C REC to reference it
<weiler> do your WG's specs create or use registries?
<tantek> biasing towards moving the locus of control (maintenance) of a registry to a WG may be harmful
<fantasai> tantek: Referencing registries is a solved problem and not what we're about here.
#435 - Tooling
Github: https://
[out of time. to be continued]
Next meeting
David: 20th, 7am PT
tantek: ical for this meeting?
david: not yet
jeff: we're working on it actively
[adjourned]