Meeting minutes
Slideset: https://
IanJ reviews the slides
IJ: 6 CGs maintaining RECs
IJ: Sometimes work starts in a WG, establish twinned CG
IJ: But WGs at W3C aren't the only destination, sometimes they go to WHATWG, schema.org, IETF, ECMA
IJ: Given all of our experience, and evolution of W3C, how do we more systematically align the CG program with our mission
IJ: Setting aside CGs that don't produce specifications
IJ: "As CG specifications gain traction on the way to standardization, the enhancements should increase alignment with W3C mission"
IJ: Some assumptions
… we want CGs to be lightweight, no cost to participate, easy to propose or join groups
… especially during early i ncubation phases
… some ppl thought maybe we could have WGs do specs and CGs to do non-specs; but we would like CGs to continue to incubate specs
… but we don't want CGs to become like WGs. We want WGs to fulfill their purpose, CGs different purpose
… just want to fine-tune for mission-alignment
… some discussion of single process for incubation + standardization; but don't think we should do that
… don't want to create a new standards program
IJ: Some concerns
… risks from CG spec that has too much traction, delayed from becoming a standard with guarantees
… there's then a misperception that because it has traction, has W3C standards-track guarantees
… but not true
… Issue is big delay between big delay and advancement to WG
IJ: Guarantees: mission-alignment, IPR guarantees -- stronger ones in WG
… quality controls e.g. wide review, testing, etc.
… implementation experience
… and process guarantees -- was it produced in an environment where all participants have a voice, appeals process, etc. Standards-track
… and lastly broad stakeholder awareness and involvement
… Lack of these guarantees expected in CGs, but when something gainst traction needs those; and problems if there's a long delay before standardization
IJ: obstacles to participation for contributing non-Members ...
… as we said, REC track is how you get guarantees at W3C
… but ppl reticent to transition to WG, and that doesn't get those guarantees
… so we want to facilitate continued participation in a WG by non-Members who have made significant contriutions to a CG specification
IJ: So we might need to upgrade CG
… review program and make sure it's up to expectation
… also we have lots of experience with CG now,
… and standardization ecosystem changes
IJ: Want to help ppl understand guarantees of CGs vs WGs
… support chairs and participatns so they can do good work
… and be inclusive and aligned to the mission
… and conducive to productivity
… if a group is floundering, either help it or stop it
IJ: We want to increase guarantees as traction increases
… but don't want to increase burden on early phases of incubation
… reduce friction to WG transition
… crisper "move up or move out" expectation; don't want CGs to languish forever
… clearer expectation of their standardization plans
… no rush, through discussion, but are you going to WG/ECMA? not go anywhere stay CG? abandon work?
IJ: One concept important here is notion of guarantees. Another concept is traction.
… we care less about specs without traction, except we prune if not going anywhere
… but when spec has traction, need to take action
… So early traction:
… intent to prototype / impement / ship from browsers
… references from other REC-track deliverables
… references other SDOs
… mentions by regulators
… expressions of interest from relevant stakeholders beyond CG members
… pilots
… multiple contirbutors to a spec
IJ: For significant traction, e.g. multiple shipping features, significant adoption in the world even if only one implementation
… or two or more implementations or announcements of intent
… so may want to take different steps at different periods of traction
… not one-size-fits-all
IJ: Started to log topics so ppl can contribute via GH issues
Additional decision / appeal requirements for specifications with traction
Set "Move up or move out" expectation
How do we establish that a CG Specification has traction?
What is the full lifecyle of a CG specification?
CG Specifications with traction that do not advance to a WG may weaken or harm the W3C mission
Facilitating continued participation in a Working Group
IJ: Looking for feedback on these
dom: Lots of successes, looking to where to improve where CGs introduce ambiguity or gaps
… gap between traction and specification
… e.g. specs that have 3 implementatoins, but still in CG report
… coming from a lot of feedback from community wrt ambiguity or lack of clarity wrt CG program
[slides 15]
florian_irc: I like the direction. Want to add an extra concern.
… you mention some of the time successful work graduates to places other than W3C WGs
… in principle that's great, helping the ecosystem
… but IPR concerns with this, because have coupling with our IPR policy but not other SDOs
… wrt move up or out -- if something that has traction and we push it out of the patent policy that's not great
… idk what we can do, but let's not amplify cases where we have holes in patent policy
IJ: I proposed to the AB an approach for that situation, where we work organizations -- especially ones where we expect work to go with them -- where we come up with an MOU
… and we not change the CLA, because we want lightweight for early work
… but when there's an emerging destination and there's support in the ecosystem, create a Final Specification Agreement coupled with MOU
… negotiate each outcome with each organizatin
… doesn't cover case of a new organization, but I think it's handle-able
<Zakim> dsinger, you wanted to react to florian_irc
dsinger: Florian said there's a good match between CG and WG IPR policies, but actually I think we have a problem.
… because we have problem with getting ppl to sign a Final Specification Agreement
… if only some members of the CG migrate to WG, then you have concerns around whether the commitments of non-transitioned members move with the spec to the WG
… so if another member of the CG contributes an idea, but I have a patent on that idea, there's no license for that patent
… without an FSA
… so we have some serious mismatch for CG -> Wg transition
IJ: We didn't have in-scope for this project revisiting the CLA
… in a WG there are a lot of tools to manage the ecosystem of ppl who might have essential claims who are not in the WG
… e.g. reciprocal grant provision
… I believe in a court of law, it would look bad for someone to assert...
… I can pull my RF grant away from someone who won't grant themselves
… but if we can find a better way to get FSAs
dsinger: in matters of the law and money, shame is a very thin defense
… I'm not saying the CLA needs revision
… but we have an issue here.
IJ: Making it easier to get into a WG from non-members is important
Dom: Part of what we look at is operationally improving transitions
… e.g. seeing what's the gap between FSA and CG membership, etc.
cwilso: You mentioned in the notes, do you intend the Community Council CG to be a more active discussion place for this?
dom: yes, please join
cwilso: Agree with dsinger with some caveat. Totally agree that FSA is nearly impossible to get, especially from large companies who have large patent portfolios
… not much upside, lots of potential downside
… a lot easier to do when we charter a WG
cwilso: I want to say, having looked at this recently, for a CG to WG transition I'm comfortable with what we do today
… operationally not quite implemented in practice as idea, but it works and locks things in
… however, this does not work when the spec goes outside of W3C
… that is a primary concern, particularly when things go to WHATWG
… it was covered when we were republishing WHATWG specs, but not anymore
… I don't think we can fix that problem without changing the CLA in some way
… but 100% agree with dsinger that "it would look bad for you" is not a sufficient defense
<tzviya8> +1 to magic IPR
cwilso: we must set things up as a waterfall that automatically happens when you sign up for a group
… really difficult to get signed agreements
IJ: It may also be difficult ... let's say CLA said, this is how transition to W3C works, and here's how transition to WHATWG works, and here's how transition to ECMA works, etc.
… would it be harder for lawyers to sign that?
cwilso: harder, but that difficulty is warranted, but that's probably (off the cuff) the most promising approach
wendyreid: Noticed in discussions, things in CGs that have traction, where ppl asking "why isn't this on the REC track"
… just this week talked about web speech API and scroll-to-text-fragment
… Why aren't these in WGs?
… and follow-up is, who is taking that work?
… we make mistake that 1 CG = 1 WG, but that's not always the case
… some don't need a WG of their own, could fit into workload of an existing WG
… but I see challenges there
… 1. how to identify the group?
… 2. charter might not include this work
… So creating a pathway for addition that is lighter-weight than a recharter
fantasai: we do have a charter amendment process
Dom: Part of program enhancements is to help CGs to have a plan
… realize that when you start a spec, question will come up of where does it go
… TAG will ask CG reports, what is your standardization target
… want to make that more systematic
… So both setting expectation, and also increasing training
… When a spec has early traction, and we have early signs of what WG it would target, then we can start the chartering conversations earlier
… rather than 3 implementations and no chartered group and nobody knows what to do
… for existing situations, need some clean-up. E.g. web speech API
… has been implemented for years, still CG spec; that's not right
… not sure what's the right solution, but it's a problem
… but staff will be there askign the hard quesitons and starting the discussions
… that's goal of enhancements
<csarven> Is the queue for remote participants taken into account or only IRC?
tzviya8: Also want to talk about transition from CG to WG
… Ian and I working on this
… very different when ppl who know W3C well and when don't
… cost of W3C...
… We have a very unusual culture. If you're in CG, you can do whatever
… can take minutes in Google docs
… Hear all the time about better unified tooling, but when transition to WG and they have to change verything
… that's uncomfortable
… Either need to start CG on our tooling, or have really good training and support for the transition
… same with GH repos. People don't always use W3C repo, transitioning is annoying
… other aspect is, at least in some CGs there's a real sense of community
… and I know you try to help people join as IEs
… but sometimes keep CG just for the community to hang out
<csarven> dom , not sure if you can see but there are hands raised on Zoom
<Zakim> tzviya, you wanted to talk about transition to membership, tooling, community
fantasai: +1 to Chris on needing to solve the IPR transition
fantasai: for a company like Apple, we need to know very specifically the scope of a CG before we can join it
… especially if that scope is malleable
… this needs very careful and usually ends up having us having very limited ability to contribute directly to CG (vs WG/WHATWG)
… changing the output target of a CG to e.g. WHATWG is probably less of a concern
… +1 to Wendy and Tzviya on improving transitions; +1 to having staff support the CG identify the path forward and help with the transition
… I don't want us to recreate even a small portion of the W3C Process in the CG context - it should happen in WGs
<cwilso> +1
<cwilso> +1 on scope changing/malleability
IJ: would tighter scope help?
fantasai: yes, but also the malleability of the scope of a charter is an issue
florian: this leads to getting any contribution reviewed by lawyers
Rahul: I've been contributing to SOLID CG for 4 years, and we just transitioned to Web Storage WG
… during the transition we had a long process where we were publishing our specs under MIT license
… then drew up a charter, and we transitioned W3C Software and Document License
… when transition to WG happened, decided that no you can only publish your specs under a CLA
… troube with CLA, it asks me to give up all my rights, but CG doesn't have to give up its rights
… so it went from open spec to all rights reserved
… I had a long back-and-forth about this
… but said relicensing spec would create problem
… either process or CLA is flawed
… it said that [missed]
… it's going to discourage people
… I will have to withdraw the spec that I authored instead of signing FSA
… feel that my contributions are not being valued
IJ: IANAL, so reluctant to say anything else, but I don't think it's accurate to say that the CG owns the spec
… the contributors own their contributions, and they agree to license them under CLA
… the set of contributors owns the specification
Rahul: Then why does it say published by CG?
IJ: You can publish in the sense of "make available as an aggregation, where each contribution is owned by its contributor"
… CLA says licensed by the contributors under the CLA
… so the CG is just creating an aggregation
Rahul: if even that language change was instead of "published by" instead of "available by"
… but when we discussed license, we should have been told that we published under the CLA
… TimBL asked for an exception
IJ: When CG process says specifications published by CGs use these licenses, that's known and part of the process
… I don't recall any conversation with Tim, but let's review that offline if you like
Rahul: Sure, but this kind of back and forth is impediment to transition
IJ: Coversation of improving support for transition, building a shared understanding of licensing is something we can improve
EgeKorkan: Web of Things WG and twin CG. We did the reverse transition, actually, where we revived CG
… but want to keep using tools we're used to
… as mentioned by Tzviya, impossible because tools owned by staff
… so if you don't have staff then the minuting tools, zoom, etc don't work
… procedurally
… so if we want CG to use same tools in the beginning, then need associated
Dom: Thanks, that's twice we've heard that tooling transition shoudl be part of our exploration
Sarven: Acknowledge some issues that are heart of this conversation
… one is facilitating continued participation
… but some of the concrete things include acknowledging the contributions of the CG contributors
… worked on this, not just signing off, but can join the Wg
… and if in good graces with chair can join as IE
… if there's no assurance about continuity is in place
… from key contributors to specifications that are used as input to the WG
… then it comes across as, there's a body of individuals contirubting their time and labor into advancing the specification
… but not ultimately part of that whole story
… so incubation period, but not so much being part of REC track
… appreciate some issues updating charter draft etc, want to see that continue
… good signal to continue that existing contributors are part of that story
… perhaps more fit to continue working on the specs as opposed to random participants
… because in WG becoming editors, authors
… issue 18 on the CG Council seems adequate, but stronger language or more assurance along those lines would help
Dom: Yes, that is what we want to address. If you invested time and resources in an effort, don't feel that it takes away from you.
… yes, moving to a WG is a change, differnet processes, new people
… but don't want to feel excluded
IJ: One of the things tied into earlier comments, want to make sure IPR commitments are in place
… even if lower fee or zero feel or temporary zero fee or whatever
… we would want an organization patent commitment even as a non-member
<Zakim> cwilso, you wanted to talk about friction increase
IJ: I think very much in scope to make that clearer
cwilso: Wanted to riff on sudden friction in work of CG vs WG
… in a lot of ways this is natural because lots of things we have in WGs that aren't required in CGs
… don't have horizontal review requirements
… don't have ability for other ppl to add stop energy, can't formally object to a CG's work
… also CGs don't have to define their space very well
… and can compete, addressing same space with their solutions
… but WGs not so much
… but the friction shouldn't be so sudden
… shouldn't move something into WG that has a standing start
… shifting things, increasing consensus and traction, put to AC vote to transition to WG
… Have a tendency to go on too long before saying we have enough momentum for WG
<Zakim> dsinger, you wanted to mention branding and identity
cwilso: Need to transition even if work isn't *done*
dsinger: the amount of oversight is something to reconsider
… especially where work is intended to move to WG
… if something is inappropriate for W3C mission
dsinger: Publications of W3C are not distinguished visually very well
… things appear to be W3C specifications, but they're a CG report
… I know that's been on the radar a long time, but gently urge to take action
Dom: definitely part of this program
… if CG report, needs to be clear
… which specs have which commitments
… adopt new branding
dsinger: they shouldn't even look similar
anusha: Very unrelated, but circle back to last slide calling out 3 implementations of storage access without WG
… seems you have a lot of mitigations to avoid getting into this situation
… but what do we do with such a case?
dom: Ideal outcome of this program is to avoid such situations
… but part of it is working with existing CGs to address those
… bring clarity to the environment
IJ: Wrapping up
… lots of comments in line with what we're intending to do
… one thing not on our radar was patent commitment timing
… but transfer part is on our radar
… thank you and please continue to make comments in the council.
Meeting closed.