HTML F2F meeting

21 Jun 2018


Chaals, LĂ©onie, Xiaoqian, Scott, Sangwhan, Steve, Job, Mallory, Bruce, PLH, Michiel


LW: For the record, yesterday was spent reviewing the differences between the HTML5.3 and HTML standard specifications.
... This morning was spent working on issues.


LW: PLH, can you tell us about the negotiations between W3C and WHATWG, and the plan for HTML5.3.

PLH: In December W3C was contacted by the WHATWG.
... They wanted to meet, to tell us some news.
... Jeff attended the meeting.
... I was there remotely.
... They told us they had adopted a new working mode.
... Also that Microsoft was going to join the WHATWG.
... Our response was to try and find a way to collaborate with them.
... Since December we have been working on that agreement.
... We wanted to stablilise the HTML document, but with the possibility of publishing a Recommendation.
... We have looked at things like how will we raise issues.
... What will the final specification document look like.
... How will we deal with IP matter.
... These things have yet to be answered.
... In May at the AC meeting in Berlin, we had a panel with members of the WHATWG Steering Group (SG).
... We wanted to hear from members.
... Unsurprisingly, there was some concern.
... There was concern about the WHATWG being just four US based browser companies.
... Overall, the AC was encouraging aout trying to find a collaboration though.
... We have not yet presented a collaboration agreement to the AC.
... We have been working through issues raised by the AC, by others.
... We need to understand what it would mean for soeone from W3C to contribute in WHATWG for exmple.
... We expect to have a draft agreement in front of the AC no later than TPAC.
... Negotiations have been at a slow pace.
... There are difficult questions to answer.
... But W3C thinks we should do our best to try and make a collaboration work.
... The WebPlat co-chairs asked what this agreement could mean for HTML5.3.
... Ralph met with the co-chairs last week.
... W3C's current thinking is that a lot of good work has gone into HTML5.3.
... We don't want to lose that work.
... However, W3C needs to be sensible about this.
... It's likely that publishing a Recommendation in the current climate would be difficult..
... One suggestion, made at the AC meeting, was to publish HTML5.3 as a Note.
... Then raise issues against the WHATWG HTML standard going forward.
... When we raise issues, we will see whether our good faith is rewarded.
... There has been bad blood between the WHATWG and W3C.
... HTML5.3 is a WD at the moment.
... We want to understand how the W3C editors feel about the agreement, and a possible future working mode.
... ... Do you think it would be difficult for you to contribute?
... Is there a way to triage issues from W3C before they reach the WHATWG?

SF: I see potential problems.
... The W3C HTML specification has been developed over the last five years, at least, to include a lot of information (authoring conformance information, informative information) that is not present in the WHATWG HTML standard.
... The WHATWG has indicated through previous issues/discussions, that they are not interested in the developer/author audience.
... So if we mode to work in WHATWG we will lose all that work.

PLH: Good point.
... I didn't mention HTML extensions.
... We have WG working on different HTML extension specifications.
... The oal is to continue producing HTML extensions at W3C.
... In my experience the WHATWG is sensitive to implementation information, and less sensitive to authoring information.

SF: I've actively dealt with the WHATWG community.
... To some extent they are less sensitive, but as far as authoring conformance requirements are concerned, and the semantics of HTML, they are still controlling.
... So I still regard this as problematic because W3c will have no say.
... For example, authoring conformance guidance that was developed with involvement with the web community.
... Things like changes to the blockquote, cite, and address elements.
... These changes were taken to the WHATWG, things that had been widely accepted and/or adopted by the web community, and still rejected by the WHATWG.
... So what will happen, is that all those good things will be overwritten essentially.
... So I see it as a problem.

PLH: What would be helpful Steve, is for you to put together a list of such things.
... Being able to take those to the WHATWG to see how they react will help us get a sense of their attitude.

CMN: Two things
... We've been looking at things in the W3C spec that are not in the WHATWG spec and vice versa.
... We have also been looking at things we've consciously chosen to do differently.
... In several cases we have filed issues on the WHATWG HTML repo.
... In some cases we have asked the TAG for their help.
... Question: have you talked to them about their developer edition, and the possibility of W3C working on that.

PLH: That is not part of the agreement negotiation itself.
... It has been mentioned in the conversations though.
... We will no stop working on ARIA in HTML for example.
... For developers there is MDN though.

SF: The problem with MDN is that it draws all its authority from the WHATWG.
... So if it isn't in the WHATWG HTML standard, it isn't in MDn.

LW: MDN is also not a consolidated document that can be used by organisations as part of their policies for authors.

PLH: We can use extensions to create new attriutes.
... We might want to produce an extension that provides authoring conformance information that is in harmony with the WHATWG HTML standard.
... If we did build an authors edition on top of the HTML standard at WHATWG, we would need to address questions about how/where the ARIA in HTML specification might fit it.

CMN: To follow up, the WHATWG HTML covers a pile of specifications that the W3C has developed.
... What is the scope of this agreement?

PLH: The scope is the HTML and DOM specifications.
... On our side it is HTML and DOM also.
... HTML at W3C is split into different specifications, like Web Workers etc.
... The WHATWG HTML standard includes some of those.
... So those specifications from W3C would also be folded.
... The Encoding specification is also included.

CMN: Is there a defined set of specifications?
... Also, is there an expectation about absorption of future specifications?

PLH: Yes, the approach is that the community will incubate ideas in Community Groups (CG).
... If members want to incubate an idea and create a Working Group (WG) to do the work, we would consider it.
... We would then do that.
... If the WHATWG also wanted to fold that specification into the HTML standard, we would try to find an agreement.

CMN: is there a mechanism for doing that?

PLH: The process is that we would sit down and talk with them.
... There is also a clause in the agreement for what wil happen if things go wrong.

LW: Can you describe the work mode for HTML that is being proposed?

PLH: We will have a WG on the W3C side to help the W3C community review the specification, raise issues etc.
... It isn't a requirement to go through the W3C WG, but it will be there if help is wanted.
... The current draft is that issues will be filed on the WHATWG repo.
... If there is a problem it will first be escalated to the WHATWG SC.
... Then it may be escalated to the W3c WG, and possibly to the TAG also.
... Ultimtely, the Director will have the last call on what goes into the W3C Recommendation.

LW: What may not be clear to everyone here, is that the proposal is to publish a W3C and WHATWG Recomendation, yes?

PLH: Correct.
... The WHATWG will publish a Review Draft, and we will take it through to CR/PR/Recommendation.

PLH: There is some room for differences between the WHATWG HTML standard and the Recommendation.
... But the emphasis is to be on finding consensus with the WHATWG as the priority.
... There are some differences that are fundamental though.
... For example W3c requires implementation experience (at least two implementations for each feature).
... Whereas the WHATWG only requires implementation commitment.

CMN: When do you think this will be an operable agreement?
... The W3C Policy is fairly clear.

PLH: I believe that if there is a wilingness on both sides, we can make it work.

CMN: No, the question is when, not whether.

PLH: I mentioned earlier that we expect to have a final draft to the AC by TPAC.

MVA: Is there a Plan B, if this doesn't work?
... What are the plans to make sure the disaster isn't too big.

PLH: We would need to fork the specification again.
... That would be seen as bad news.

CMN: It's taken us three years to start rebuilding the TML community at W3C, and it is now having an effect.
... How do we protect that community, in the event this agreement doesn't work out?

PLH: If it doesn't work out, and we're unable to build trust, we would essentially need to fork the specification back again.
... We reserve the right to publish our own Recommendation if the agreement doesn't work out.

CMN: I'm not concerned about that.
... Agree that the idea is to make this work.
... But the question is that in the pursuit of trying to make this work, are we throwing away our capability and community?

PLH: Yes, that is a risk.
... We will keep working on HTML extensions.
... If people want to stay ... If people want to stay involved in HTML they'll need to remain in the W3C community.

CMN: I'm skeptical.

JVA: +1

LW: What is the role of our editors under this agreement?

PLH: I want to understand from the editors what they're looking for.
... What you want to see in the collaboration agreement?
... We should not wait for the agreement to begin filing issues on the WHATWG.

MVA: What is the plan when there is a stalemate?

<chaals2> [we started that process... i can find half a dozen by mail if you like]

LW: I asked Jeff to propose the W3C Editors being WHATWG editors, has anything come from that yet?

PLH: Nothing yet.

SF: Has it been proposed to the WHATWG?

PLH: I can't answer that.
... We have perhaps 30 or 40 issues under discussion.

CMN: In the scenario that we get to TPAC and still don't have a draft agreement, what do we do?

PLH: I can't speak for Jeff.

CMN: But if we all believe an agreement can be reached, but recognise it could take another six months, what happens?
... I can't speak for the other editors, but as a contributer, there is a time boxed expectation that sometime we will have something to show for our current work.
... It seems that getting the editors due recognition i the WHATWG has not been raised, and that worries me.

PLH: I understand.

LW: When the co-chairs met with you and Ralph last week you told us not to transition HTML5.3 to CR.

PLH: We didn't tell you not to transition to CR, we recommended that you keep working on a WD.
... If you feel you have a strong case to move to CR, you should do so.

CMN: I'd like to ask the editors where they see themselves fitting into this world.

SOH: I don't know. I'd like to continue contributing, but I'm concerned with how that's actually going to work out.
... I've seen editorial or accessibility concerns raised with in the WHATWG space, and they were not taken up and dealt with. (at least not after serious,
... prolonged debate, where the community was asking for a change that the editors were not willing to make)

MVA: It would be good if there was a clear conflict resolution path.
... That might help people feel more confident about contributing in the WHATWG space.

SOH: To reiterate that point
... I hear a lot about raising issues in WHATWG space, asking them to consider things we've done in the W3C specification, and see how it goes.
... But it feels like it's all us being asked, to re-do our work, and for them to only consider those changes.
... I don't see what the WHATWG is conceding.

SF: I don't think they are.
... We will have to re-open and/or re-commit all the issues we've dealt with over the past N years.
... My fear is that the sort of issues we've worked on, and that are important to the W3C community/audience that we provide for, will not be iportant to the WHATWG.

XW: Part of the agreement is that Mike Smith and I can spend a few months looking at the differences between the two specifications.

CMN: Do you have editing authority to include them in the WHATWG HTML standard?

XW: No.

CMN: Right, we're already doing that.

SF: The outline algorithm for example.

XW: Please give us examples.

SF: There are many exaSF: In this case the W3C is accurate for authoring conformance information, it is better and accurate.
... But the WHATWG guidance has not changed in five years, and it is still fictional.
... All this authoring conformance/dvice will not be a priority for the WHATWG and I don't see that changing.
... Unless they accept it is important.

PLH: We have been working on an escalation path.
... We had feedback from the AC that the proposed path may not b the best one, and we've been tweaking it.
... The expectation is that there will be an escalation path though.
... Regarind raising issues, Xiaoqian already mentioned that there will be help filing issues.

SF: It's not filing them that's the problem it's the fact they are not resolved.

PLH: Part of the agreement is that we'll expect issues to be addressed in a timely fashion.

LW: But "addressed" could mean blocked because the WHATWG editors do not understand or agree with the issue.

PLH: That's what the escalation path is for.

BL: I agree with Steve and Scott.
... My interest in editing HTMl is to have a specification that defines reality, and that vives authors useful advice on accessibility and everything else.
... I also think that having two specifications is not a good thing.
... The work done in WHATWG has done a lot of good, perhaps even saved the web in my opinion.
... But perhaps there is amechanism by which the W3C editors could be given power/capability to add the conformance/authoring advice that will be missing.
... So we can really have one true specification.

SM: The WHATWG does have a developer edition.

LW: But it's a stripped down version of the full specification, not a deliberately tailored set of content.

BL: Right, but if they enabled the W3C editors to work on that edition, because this is where our strength lies.

SM: I've been told they will welcome PRs.

SF: We've raised issues with regard to providing authoring conformance content and it's received pushback.
... The trouble is that they don't want the authoring conformance advice.

MVA: Authors go to scattered places for guidance.

SF: What authoring guidance there is in the WHATWG spec is different, often out of date, and now we're going to lose the more up to date, community driven, content that it's in the W3C spec.
... There are no non-browser stakeholders in the WHATWG SG, or amongst the editors.
... That is not the case with W3C.
... There are no advocates for user groups, like people with disabilities for example, anywhere in the WHATWG SG/Editors.

MB: Do they use the constituency of priorities?

SF: No
... The browsers control the mplementation, and the authoring guidance is layered on top.
... It's the cream, and they want the cream as well as the milk.

CMN: If we are trying to move to a collaboration agreement, there are two things we need to focus on
... First is Plan B, what happens when this doesn't work.
... We could walk away.
... We could keep doing what we're doing, recognise there is both cost and benefit to doing that.
... Or the agreement will work.
... We're being told that the WHATWG editors are listening in a way they have previously not have done.
... There is mistrust, but we are also filing issues to see if it can work.
... We can try doing this more over the coming months and see how it goes.
... But we should also look at milestones and steps.

MVA: What does the WHATWG want?

CMN: PLH? You sat on the negotiations.

PLH: What they want is to get contributions from the W3C to their document.
... To Sangwhan's comments ^^, I described an early version of the escalation path.

MVA: What about a trial period?

XW: New ideas are incubated in the WICG.

PLH: Domenic is an HTML standard editor, but he is also involved in other W3C activities like Web Payments.

CMn: You said that what the WHATWG wnats is contributions.
... What is our motivation to do that?

PLH: Our interest is to have one web for all.
... Having two specifications is not a good thing.
... It isn't beneficial to the community.
... If there is a way to work on one specification we sould try to do it.


CMN: I agree.
... But the fact remains that you're asking a group of people who've done a pile of work in the W3C's name, to start re-doing that work for the WHATWG.
... So, in the new scenario, what is their to motivate the existing W3C editors, beyond that there is one spec.
... On the flip side, understanding what motivates the editors to work on HTML at W3C now.

PLH: We want to make the web work, to move the web forward.

CMN: Tere are about 47 different things I could work on to move the web forwad.
... I could work on accessibility or privacy or something else.
... What I'm trying to get at, is where you go from being a contributor in W3C, where something obviously appeals
... To being a contributor in WHATWG, where the idea of filing an issue based on solid data gives me the willies and takes months to get resolution.

PLH: Desire on both sides.
... I can't guarantee what will happen tomorrow? No.
... That's why we want an agreement in place.

SOh: This is where my original question came from.
... I contribute to the W3C specification because I feel I can.
... I'm an editor of that specification.
... I can collaborate with my colleagues/other editors.
... It's a cordial environment. We're all equals in our roles.
... I don't feel like the WHATWG will accept me as an equal.
... What I'm hearing is that they want us to file issues, but they don't want us as editors.

SF: No, they don't.

PLH: They say they will accept PRs.

LW: PLH, come on!

SOH: Some agreement that we get to contribute in a meaningful way, perhaps on the developer edition. That'd show some good faith.

PLH: Thank you, that is useful feedback.

MVA: I keep hearing about PRs.
... There is a negative raction.
... Is there a process by whic PRs are accepted/denied in WHATWG?

SF: The WHATWG editors knowledge/opinion.

MVA: So semi-arbitrarily?

CMN: Well, we do that to some extent to.
... The thing is what are the goals, on what basis are those decisions taken on.
... Part of the issue is that the basis for their decisions is oriented towards what browser vendors care about.
... That is a genuine concern for some in W3C.

JVA: So do they have domain experts for different parts of the spec in WHATWG?

LW: No

PLH: The WHATWG process gives a lot of weight to the editor.
... If there is disagreement, there is the escalation path.
... The only prblem is that the escalation path, which is also composed of the same four browser companies.

LW: We've been going for two hours, suggest we wrap this up.

PLH: On our side the negotiation team is Jeff jaffe, Wendy Seltzer, Ralph Swick, and myself.
... If you have concerns, I encourage you to reach out.

CMn: Your AC rep will see it when the time comes.

PLH: We havent shared the current draft even with the AB yet.

LW: Can ask you to share the draft with this group of people.
... Not all of them have AC reps, but it's a critical document for all of them.

PLH: Yes, I will.

CMN: Thank you PLH.

Summary of Action Items

Summary of Resolutions

    [End of minutes]

    Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
    $Date: 2018/07/03 12:23:29 $