WebPlatform Web Components F2F 2018

[TakayoshiKochi]

At TPAC, we agreed to have another F2F around February 2018, at somewhere not in the US, so everyone can join. Let's plan the next F2F! (last update Feb. 20, 2017)

Date: Mar. 5-6 (Mon-Tue), 2018 10am - 6pm
Venue: Google Tokyo Office (Roppongi, Tokyo, Japan)
For detailed logistics see w3c's meetings page.

• Agenda (1.5hour x 4 slots / day)
  • Day 1
    • Agenda building + unconference planning (30min) / Go through existing issues (1)
    • Go through existing issues (2)
    • Template instantiation
    • Unconference 1
  • Day 2
    • HTML modules
    • Scoped Custom Elements registries / Declarative Shadow DOM
    • Unconference 2
    • Unconference 3

[annevk]

Some thoughts:

• Multiple days seems good.
• Let's decide well in advance which days to focus on existing issues versus new functionality. That makes it easier to plan for folks with less time/budget.
• Let's make sure to extend invitations to React, Ember, Vue.js, ? well in advance as well. Representation from Angular has historically not been a problem, but to succeed we'll need more folks to weigh in.
• If we manage to get the various frameworks to attend I'd like us to have some unconference sessions for problems with the web platform in general, not restricted to web components.

[annevk]

Since email addresses are hard to come by: @yyx990803 @wycats @tomdale @sophiebits @gaearon @sebmarkbage would you all (or someone you know who can represent the JavaScript framework you work on) be interested in meeting up in Toronto next year February primarily to discuss extensions to the DOM and HTML in browser implementations ("Web Components"), but not necessarily limited to that.

Our rough plan is to have a day to go through existing issues, and one or more days to discuss new ideas, such as HTML modules and HTML Template Instantiation, and learn from you all what the pain points are with the platform.

[TakayoshiKochi]

@justinfagnani @azakus @kevinpschaaf from the Polymer team (they were at TPAC) may also be interested in this.

[yyx990803]

For the record, I am particularly interested in HTML modules and HTML Template Instantiation and would be willing to attend the days that are focusing on these.

[tomalec]

Could we bring up customized built-ins/extending native elements/custom elements with different parsing contexts to the agenda?

I would say, >430 comments, hundreds of thumbs-ups, many participants in the single issue and even more if we count related issues, proves the need for it.

[annevk]

Pointer? Is there a concrete proposal with some implementer support?

[domenic]

My understanding is that customized built-ins, as specced, have 3/4 implementer support. So they may not be a fruitful topic for discussion; instead web developers can bring up their concerns with their Apple representatives directly at another venue (e.g. in said centithread).

[tomdale]

I would definitely be interested in attending. My email address, if it makes things easier: [email protected].

[TakayoshiKochi]

Updated the agenda in OP.
Tentatively set the schedule for 3 days - which means it starts on Monday, Tuesday, or Wednesday
(to Wednesday, Thursday, or Friday, respectively).
Also I'd like to ping @travisleithead and @rniwa if anyone from MS and Apple can join.

[rniwa]

Do we really need three days? I don't think I can quite afford to be there for three days.

[hober]

Two days sounds good. I think 2/19–20 work for me.

[JanMiksovsky]

2/19/2018 is President's Day, a federal holiday in the U.S. Many people have that day off, and some people stretch the holiday weekend by a day or so.

As it turns out, in the corner of the country where I live (Seattle), many schools have that entire week off as a mid-winter break. My children happen to attend two such schools, so while I'd really like to make it to this meeting, I couldn't attend a meeting held the week of 2/19.

Obviously any given time will be inconvenient for some people, so the group should decide what works well for a quorum of people. But if this could be deferred to the following week or early March, that'd help people taking time off around that national holiday.

[annevk]

@rniwa @hober my idea was that we'd have two structured days focused around existing issues and new proposals, and one day that's a bit more exploratory. If we stick to that then I think those that want to take less time can figure out for themselves which days they're interested in. And hopefully nobody falls in the category of wanting to skip new proposals...

FWIW, I can do a week later (26 - 28, 27 - 1, or 28 - 2) in theory.

[rniwa]

If it's on the week of Feb 26th, then I can't do Monday, Tuesday, or Wednesday.

[annevk]

Here's a doodle: https://doodle.com/poll/nry9ut6n3g7x2cge. I suggest we pick based on the result of that doodle at the end of next week (so please please fill it in when you read this message) so everyone has plenty of time to make arrangements.

[TakayoshiKochi]

Thanks @annevk for setting up the doodle!
I'm okay for any of these 3 weeks, but considering many are from the US and the holiday, the week of Feb. 19 was not a good option, and considering @rniwa is a must for the meeting, Mar. 1-2 or
the week of Mar.5 would currently looks the stronger candidate.

If we were to squeeze the schedule into 2 days, maybe we can prepare for going through the existing bugs beforehand, and shorten what need to be done in F2F, then we can still have time for discussing new stuff? E.g. On each day if we have 4 slots (1.5-2hrs?), how about doing like this?

1st day

• contentious bugs resolution
• template instantiation
• template processing
• buffer / unconf 1

2nd day

• HTML modules
• HTML modules (cont.)
• unconf 2
• buffer / remaining stuff

We need to have pre-meeting prep for identifying really contentious parts, and bring concrete proposals or use cases beforehand, but it would make the meeting more productive, I believe.

[rniwa]

Yeah, something like for two days sounds good. I also like the first week of March.

[domenic]

I've filled out the Doodle and have a conflicting event on the 1st and 2nd. First week of March is sounding more promising.

I would prefer to avoid meetings on Mondays or Fridays since they would then require traveling on a weekend.

[rniwa]

I'd prefer not having this meeting in the middle of a week since I have way too many other commitments, and flying during a week day would mean that I'd have to work over weekends to compensate for it.

[TakayoshiKochi]

Thanks for putting your dates on doodle.
As @annevk already said, let's close this at the end of this week (Sat. Dec. 16) to decide the date.
Note that you don't have to put your name/date if you are flexible, putting your name is not mandatory for joining the meeting.

So far people are good for 2 days conf, but I see some schedule conflict.
I'd try best to accommodate all preferences, so please speak up!

[justinfagnani]

I think first week of March works for me too, I'm checking.

I'd like to throw scoped custom element registrations on the agenda, if possible: #716

@TakayoshiKochi is there a difference between "template instantiation" and "template processing"? By "template processing" do you mean just focusing on the built-in template processor?

[TakayoshiKochi]

is there a difference between "template instantiation" and "template processing"? By "template processing" do you mean just focusing on the built-in template processor?

No, I just split the HTML Template Instantiation into 2 parts somewhat arbitrarily, guessing that it would need time for 2 slots.

[travisleithead]

Also I'd like to ping @travisleithead and @rniwa if anyone from MS and Apple can join.

We'll try to have someone from Microsoft be there regardless of the dates--either myself or (hopefully) one or two of my colleagues; haven't figured that out just yet...

[robwormald]

I’ll attend from the Angular team, and am flexible on dates.

[developit]

I'll be happy to attend for Preact!

[rniwa]

By the way, while Toronto is a lovely city there might be a snow storm & long delays for flights in February & March so I would throw in Tokyo & London as possible alternatives (assuming Google / Mozilla's respective offices can accommodate us) although that might make traveling harder for some U.S.-based individuals.

[sophiebits]

Closing the loop from the React side: I don't think we'll make it to this unless there's something specific you'd like input on – we're not actively working on anything in this space. Thanks for thinking of us!

[TakayoshiKochi]

I have never been to Toronto so I have no idea how terrible to be there at the season... but I believe Ryosuke for his experience. @annevk @smaug---- do you have any idea where it would be in Europe?
If it will be in Tokyo, I'm happy to work on securing meeting rooms at Google Tokyo.
Let's decide on location asap for everyone to schedule.

[smaug----]

Mozilla's office in London may not have good meeting space. The office in Berlin would be better.

[prushforth]

Could we bring up customized built-ins/extending native elements/custom elements with different parsing contexts to the agenda?

I second this. I would like to add that there is a general need for the ability to create 'speculative polyfills' for unknown elements, which autonomous elements don't support. Example <foo is="new-foo">

I would say, >430 comments, hundreds of thumbs-ups, many participants in the single issue and even more if we count related issues, proves the need for it.

+1

My understanding is that customized built-ins, as specced, have 3/4 implementer support. So they may not be a fruitful topic for discussion; instead web developers can bring up their concerns with their Apple representatives directly at another venue (e.g. in said centithread).

Is Chrome waiting on Safari to implement this? If so they will be waiting a long time, I think. Better to follow Mozilla's example and implement based on about as good a consensus as you are likely to get.
Because the webcomponents speculative polyfill for built-ins are blocked on Chrome, apparently, so Web developers can't even vote with their feet.

At the very least, I think there's something to discuss. My .02 cents.

[caridy]

Paris, Tue-Wed (Mar. 6-7) - We have booked a big room as a backup plan!

[binoculars]

I'm really happy to see that there is formal discussion around template instantiation and html modules. Thank you all for pushing these forward!

[rniwa]

Yeah, I gave my preference to Toronto but Mon-Tue in Tokyo is fine for me.

[TakayoshiKochi]

Thanks all, let's fix the place and date: Tokyo, Mon-Tue (Mar. 5-6)!

[chaals]

Meeting page is now up.

Please add yourself by PR if you are attending / have agenda requests. We will scrape this issue for the agenda too.

[GarrettS]

[CMN: The content of this was inappropriate - an unprofessional ad hominem attack on an individual - and has been deleted]

[rniwa]

In the future, please be reminded of Code of Ethics and Professional Conduct.

I understand your sentiment but that's not a productive feedback. If you have specific concerns or complaints about web components, please file a new issue.

[treshugart]

Would it be worth discussing declarative shadow DOM on one of the days?

[tomalec]

As Anne asked for a proposal. I'll publish my strawman proposal today, maybe someone would be interested in reviewing it.

[tomalec]

My slightly more structured proposal for Declarative Shadow DOM is at
https://github.com/w3c/webcomponents/blob/gh-pages/proposals/Declarative-Shadow-DOM.md (#731)

[treshugart]

Nice one, @tomalec. I raised an issue with some of my ideas awhile back in whatwg/dom#531. I've also published some similar musings in https://github.com/treshugart/react-shade which uses React portals to implement a declarative API. I'd be interested in using your proposal as a base to discuss related ideas.

[treshugart]

I'll be attending on behalf of SkateJS / Atlassian.

[travisleithead]

Per my previous note:

We'll try to have someone from Microsoft be there regardless of the dates--either myself or (hopefully) one or two of my colleagues; haven't figured that out just yet...

Looks like @patrickkettner will be able to be there in person. Thanks Patrick!

[annevk]

I think we should do some in-person triaging for these issues:

• Needs implementer interest: https://github.com/w3c/webcomponents/issues?q=is%3Aissue+is%3Aopen+label%3A%22needs+implementer+interest%22
• Needs consensus: https://github.com/w3c/webcomponents/issues?q=is%3Aopen+is%3Aissue+label%3A%22needs+consensus%22

Reducing the number of these issues by stating implementer interest or agreement early is welcome. Then we can use the face-to-face time for other things. (I've mostly removed the v2 label except for things we have clear agreement on. That makes it clearer where the various issues are at in the process.)

Another thing we should discuss is how to address the remaining v1 and upstream issues. In particular there's quite a lot of UI/focus/editing stuff that hasn't really been getting attention while being somewhat essential for the success of shadow trees.

[rniwa]

I just reviewed & commented on most issues.

[TakayoshiKochi]

The meeting is only 1 week away!

If you attend the meeting but have not put your name on w3c's minutes page, please register yourself by the end of Tuesday this week (Feb. 27) (at your timezone).

We (host at Google Tokyo office) are collecting the list for distributing security badges for the building and the office for the guests, and we need your names for it.

If you have any last minute change on your travel after that, don't worry, please let me ([email protected]) know. We have some room for additional people.

[annevk]

@TakayoshiKochi it doesn't seem like https://github.com/w3c/WebPlatformWG/blob/gh-pages/meetings/18-03-Web-components.md#agenda aligns with #713 (comment). I sure hope we keep the unconference aspects to some extent. We could potentially replace it by just asking everyone at the start of the meeting what they think requires discussion most urgently and adjust the agenda based on that. In general we want to keep it pretty flexible I think.

[TakayoshiKochi]

@annevk yeah, will update it (I am thinking of this thread is the canonical discussion for the agenda).

[diervo]

Would it be possible have a slot to talk about the progress made on the AOM proposal?

[annevk]

@diervo yeah, we should definitely discuss accessibility. I think the best thing as I said in an earlier comment is that we each state briefly at the beginning what we want to discuss and make up the agenda from that. And also do some (quick) issue triage together to cover the various proposals that have been made over the years.

[TakayoshiKochi]

Thanks all for the productive discussions!

The actual agenda used was:
https://public.etherpad-mozilla.org/p/web_components_agenda

@chaals would you post the meeting minutes at https://github.com/w3c/WebPlatformWG/blob/gh-pages/Meetings.md ?

[TakayoshiKochi]

The minutes are posted, thanks @chaals and @diervo !

Monday: https://www.w3.org/2018/03/05-webplat-minutes.html
Tuesday: https://www.w3.org/2018/03/06-webplat-minutes.html

(edit: the second link was wrong, fixed now)

[rniwa]

See #786 for 2019's spring F2F.