Meeting minutes
Presenter: Aki Rose Braun
aki: Today we are going to converse about ECMAScript outside the main language, less visible stuff that doesn't always require the browser
aki: I am just going to mention my personal pref is to use ECMAScript when talking about the core standard and Javascript when I want to include the web apis (non normative just me)
aki: TC39 - not webIDL, not Web Platform
aki: We also abide, a golden rule for us, don't break the web. Browsers are not our only host they are our largest host and it is important to keep it working correctly.
[side] technical committees and task groups
TC39 is broken into 5 task groups, the first one is the least important for this session. ECMA262 and then 404 is the interchange system - it is JSON it is unchanging and sacrosanct
JSON standard is brought up every meeting but we keep it stable, no changes
Since the core language is not our topic today we won't go deep into it today.
TC39 does contribute to the web platform tests test262.fyi - gets run once nightly.
One of the conversations I got into about the web platform tests and why we can't do more there - I know we have and can contribute to the tests
Is anyone familiar enough with how we could get the tests better talking to each other (web platform and 262)
Phillipe was very proud that they got the harness for these tests set up and then groups came in and added more to it. They slowly got all the working groups contributing to the test and there is only a little bit of tests working
This is one of the things on my mind as something that could be done better that I'd like to see improved.
I want to make it easy for the web platform to potentially run the 262 tests in an automated fashion. It would be good to see all the results real close together. I don't think they belong in the web platform tests though.
TC39-TG2 - responsible for ECMA 402 internationalization
[slide] TG2's Programme of Work
TG2 develops the spec, standardizes i18n behavior, maintains the spec and tests, communicates often with w3c and all their proposals and PRs require tests be added.
ECMA wants to work more closely with w3c specifically. We have our ISO fast track and a handful of other standards bodies but a close relationship makes sense
Westin's session was talking about country specific data and Emile and I were able to talk about how this came up in W3C and the ECMA taskgroup and the shortcomings etc...
???1: There is a proposal around smart unit preferences
aki: as it turns out this is not easy to figure out. Ireland has eircodes (sp?) and that seems to work well.
The smart unit preferences proposal, I like it, it is very informal. You could use the API to say where you are and it can describe measurements in the most appropriate way for the location
We saw a presentation from Canada that was very complex in how different measurements are made
[slide] Where is the overlap?
aki: I want to come up with a list of topics that ECMA might encounter and w3c intl working group might encounter and figure out when we encounter these topics it should just trigger lets have a joint meeting to talk about it.
but I don't know enough about the w3c working group myself to confidently say lets do a joint meeting myself.
We can do this without running afoul of IPR concerns I think
Intl just feels like the most obvious convergence.
Jason: Yes Intl seems like a good place to start.
aki: the three other task groups - security (not responsible for publishing standards) they have technical documents they are working on.
They are working on a formal technical report of the security model that goes through similar model for standards
they are also looking at a security and privacy guide, one of several inspirations is the security and privacy questionnaire that w3c produced.
I look forward to seeing something like that on the ECMA side and it can only be good to do more of it
TG3 is also responsible for dealing with security disclosures, including going off the record during a plenary to discuss.
Sometimes TG3 keeps an eye on things and can bump them back to W3C where it makes sense to discuss.
An example being the seeking Trusted Types feedback on Array.isTemplateObject (shown on slide)
[slide] security team-up
aki: I'm trying to figure out what is missing here. ECMA seems pretty good at brining these to w3c. Should we have a regular meeting? Do we just trust it will be dealt with when they pop up
AramZS: We have a security staff person who leads security concerns for the w3c
there are also TAG reviews that handle security examinations
aki: I have time and will bring that up with them
[slide] TG4
aki: TG4 is doing the first ever source map specification. Sourcemaps have been around for around 15 years but were basically described in a google doc for that whole time
I'm glad people put them together and I'm glad they exist but it is impressive that they never just fell down. TG4 is putting together the formal standard. They have meetings, tests, debugging features, and name mapping and ignoreList
Then it will be formatted and published once it is complete. It is really exciting.
It is a lot of people's first experience in standards. Green fielding a standard is really impressive. I'm very impressed with their work
They have a super scaled back process document, not dissimilar to the rest of TC39 but much less of a size which is reasonable.
Source maps will also cover CSS and I will be interested in seeing if there end up being any other text files that show up at the door and say we want to work at the door as well.
JaseW: very exciting! I think the tests were taking from Mozilla's code base I think. Someone went around and looked at all the tests that existed and formalized them.
aki: I imagine the code standard is a night mare though
JaseW: yeah, some tests might have even contridicted each other.
aki: this is going to cover such a breadth of code. How do we find out what edge cases are missing?
AramZS: does 'covering CSS' include SCSS and SASS?
aki: yes, the superscripts are expected to produce sourcemaps that match the spec
it should make it almost a non event for those superscripts that they can continue to match what we've seen in the wild while making the standard.
[slide] TG5
aki: TG5 is experiments in programming languages standardization.
Provides a forum for research and experiments on ECMAScript and related technologies. This is new.
Big thinking is happening here.
University students and people who have been doing standards their whole career talking about what is the past and future of standards
they are researching what other standard bodies to work with and how standards can be expanded in use.
Going hard on being standards merds.
aki: recently they demoed for me meta programming system (MPS) used to make a programming language. Meta programming about meta programming.
[slide] A domain specific language for creating programing languages. Talk about meta programming. meta meta programming in Jetbrains.
aki: if you could aim a standards research arm at a problem where would you aim them?
There is activity in universities and I'm looking forward to hearing what all the university students ask me
JaseW: The question is like if you could / how could they university students be more involved in what they can bring to standards?
aki: more than that, they are doing standards and user research, looking at the realities of what has come up and how people actually use the features and what we can do on top of it
A slides deck in their next meeting starts, in detail, in the 19th century.
AramZS: will that presentation be available?
aki: yeah it will wind up on Github somewhere in TG39 TG5 repo I think (missed link description further)
JaseW: Students are getting involved in proposals as well. Is TG5 going to universities and trying to get students more involved?
aki: Julia(sp?) is getting more involved in that. I think she's a driving force there.
aki: one big difference between ECMA and W3C is that non-profits and universities pay 0 dues to ECMA. Makes it easy for universities to become members and a few delegates have started as university students and remained on as invited experts to handle their proposals.
aki: we have someone who is a practicing medical doctor but stayed on.
aki: something to think about and keep thinking about is that you can aim the university students as a problem and they'll just do research on it. It's very cool and we just need to think up more problems.
aki: experiments don't have to be tied to an existing language like the programming language that is about making a programming language. It has relevance in that it gave me the ability to track all the proposals, meetings, attendees, IPRs, all built in to this system.
aki: He is demoing at the next TG5 meeting and I'm looking forward to it
the other TC is TC53 working on ECMAScript modules for invented systems.
Trying to work on ECMAScripts that use the least amount of storage and processing power possible for very small devices.
They published ECMA-419 which is for embedded systems along with TR/110 and other highly specialized stuff.
aki: this is not dissimilar to the work we do with w3c in collaboration. w3c's drafts for behavior and communication for interacting with these things and 419 is more like how it actually works.
it is exciting that interop is the main goal.
[slide] more additive overlap
aki: it is constant references in the spec to how it interacts with w3c specs
That is all the non 262 stuff that EXMA is working on.
[slide] A whole lot of links to stuff discussed.
aki: there are certainly members of both groups who bounce between but it is harder.
it is good having people who work together but represent certain interests and can effectively hash out who is best to handle it
Right now the Winter CG (sp?) want to publish standards but it doesn't exactly make sense in w3c and ECMA has IPR issues.
aki: The groups are different. ECMA's TCs look at the highly technical details of what makes a good programming language etc.. but w3c takes a look from a more humane angle. It isn't that ECMA doesn't think about that, but the order in which they come up is different.
That's my impression.
JaseW: Yeah I think that's fair. I am doing both. TC39 can be hyper focused on some of the proposals going quite deep and the question of educational aspect and how it looks to beginners comes a bit later on.
aki
aki: it isn't that it doesn't come it is just the order is a bit different.
JaseW: people feel they have to iron out a lot of edge cases before they get to stage 1
aki: which is funny because all you need is a problem statement for stage 1.
Pete: there can be deeper consequences with a programming language.
aki: Yeah, I think it makes sense for the groups to operate where they do.
and how they do
I have nothing more.
we don't have a consistent meeting process, and that's what I'm trying to iron out. Is there something that should trigger joint meetings.
Pete: it can be hard to know all the things that are going on in w3c - so many github repos all over the place.
there is likely not that much communication between different parts. We may need a mechanism to trigger that.
aki: Horizontal reviews are a good topic to think about this. Philipe's breakout session is on this and it might be good as a way to get more people involved. This may be a way to work together with ECMA where they could be part of the horizontal review. The Intl groups in both orgs could review with each other.
AramZS: yeah horizontal reviews seem like the right place, the groups are getting used to activating horizontal review processes once standards get to a particular point.
JaseW: Are there specific areas to join that you are targeting?
aki: I'm just going to sort of get a sense of things now and then we'll start figuring out how to talk to each other.
I think I'll be consulting for ECMA through the end of the year on this.
rssagent, please draft the minutes
RSSAgent, please draft the minutes