APA FtF Day 2

30 Oct 2015


See also: IRC log


Ted_O´Connor, Joanmarie_Diggs, Peter_Linss, James_Nurthen, Bert_Boss, Can_Wang, Michael_Cooper, Florian_Rivoal, Takeshi_Kurosawa, Matt_King, Jesse_Beach, Dave_Kramer, Alan_Stearns, Takashi_Yamanishi, Steve_Zilles, Janina_Sajka, Wei_Wang, Taketoshi, Yamane, Dave, Cramer, hober, Cynthia_Shelly, Markku_Hakkinen, Jason_White, Bert_Bos, fantasai, Rossen_Atanassov, rossen, LJWatson, JF, Liam_Quin, ShaneM
MichaelC, JF, joanie, liam


<Can> q_

<MichaelC> scribe: MichaelC

mk: we chatted yesterday

not all people understand the challenges of reordering the dom

some concerns about true in practice, not just true in theory

I´m coming to understand that some things will be inherently inconvenient by the nature of the functionality

and that there isn´t an actual problem

cyns said the horse is out of the barn and it´s unreasonable to ask people to reorder the dom

now I wonder if we should challenge that, rather than challenge the css wg

<JF> Preseent+ JF

which I see as fantasai´s view

cw: @@ we should leave to AT to recognize inherent order, other structure in the content

should not put too many constraints on css

we should expand capability of at

<JF> Scribe: JF

JN: don't see many places where you wouldn't want to re-order the tab order to match screen layout

can't see a circumstance where you wouldn't want to maintain this

MCK: sort of agree - is changing the dom order very difficult?

JN: why? Changing the dom order negattes the layout

MCK" maybe the purpose is when you *want* the visual order and reading order to be different than layout order

Florian: think one difficulty is that visual layout is 2-d mapping, which makes a difference in the liniar layout of reading order

when you use flexbox, it's a 2-d layout

you place the content in the source-order according to importance, but the layout can change that

CSS WG have focused on - if you use the DOM correctly, you place things according to importance

there are ways of authors to do it right, and ways to do it wrong

cannot create a system that disallows authors to use it improperly

have similar problems with floats, etc. - not a new problem

if we have to solve this meta-problem, we need to fix it in more than flex-box

if authors use meaningless DOM order, and then try to fix it via CSS, then this is the wrong approach, but cannot forbid authros fromdoing tis

cyns: would like to see the "Must not change tab order" provision

different ways of "reading" a page - touch, screen magnifyer (which could have the magnifyer jump around)

ditto with keyboard navigation - and this will also touch on reading and learning disabilities - these all rely on a crrelation between visual and logic order

cyns: think that this should be fixed in user-agent layer

different needs for different user-stories

MCK: don't want browser to remove things that aren't specified behavior?

<discussion o which WGs would be involved here>

cyns: user interface

Florian: when dealing with visual layout, there is one way of doing it

<fantasai> Florian was explaining that in 2D layouts, the visual reader will read things in an order that is not necessarily ltr-ttb.

<fantasai> Based on other visual cues provided by the author.

when we talk about different types of user agents, then there are different ways of doing things - if we have to standardize on each combination of input and output devices, then standarizing this is very hard

<fantasai> The author is responsible for providing a DOM order that matches the preferred reading order.

cyns: would reverse recommendation from MUST NOT to Should or May

mck: I disagree strongly with that (changing tab order)

JS: don't have clear concensus in our house, so we should go back and continue to discuss - suggest we move to other topics

JN: reading pointer out loud

MCK: sounds like an sauthor statement, not a UA statement

cyns: getting visual order and dom order out os sync is a problem

mck: disagrees

florian: not primary goal of flex-box

cyns: order matters, and having order match matters


MCK: the author statement sounds like not what we want

cyns: essentially, getting layout order and dom order out of syn is generally bad

<joanie> Statement being discussed can be found in https://drafts.csswg.org/css-flexbox/#order-accessibility

fantasai: for visual design, L-R, T-B is not always the order, it can be many variants of that visually

visual order can also be dictated by things like size, color, imagery, etc. - many other things that go into visual perception

if you say that the page must read L-R, T-B then that may not actually make sense

<joanie> And/or may be here: https://drafts.csswg.org/css-flexbox/#flow-order

Fantasai: @@ the order that you read things in visually is going to be similar in most cases, but may have variants

and that can change as well based on device and screen size

so you should allow the user to explore the page in a spacial manner

<Zakim> janina_, you wanted to say some of the problem arises from touch interfaces--explore vs swipe

<fantasai> but the page should be in the logical reading order

<fantasai> which is not necessarily the same as its spatial layout, ltr-ttb

JS: number of usecases exploded with new interfaces, especially touch

so yes, for magnification etc. this is a problem, ditto for when using sip/puff and focus jumps around

but when you explore via touch this gets even more complex

where they want to explore the layout as well as the DOM order

and it may not be a logical reading order

this manifests on touch - when you explore it may be out of sync, but swyping will give you a DOM order - there is a big disconnect here based on how the non-sighted user explores the page

cyns: depends n what kinds of tech and AT you are using

there is also another issue - when interacting with a web app

getting frmo "text to send button is a logical experience

recalls a thing where they looked at nested grides (etc.) - made sense from a code perspective, but tab order was totally messed up

MCK: agree that the goal is to have a logical tab order that is generally speaking in the traditional L-R pattern, and taht we'd like a common experience when exploring with different models
... @@

use flexbox when youneed to to get the layout and DOm order

JW: predictability for the author is fundemental, so that the DOm has a logical reading order

<fantasai> mck (earlier): There have been cases where to get the right layout, the author has changed the DOM order from its logical order to get that effect, and in Flexbox we have the tools to do that while preserving the correct source order, and correct tab order. Flexbox allows us to fix this kind of problem. Follow the MUST NOT recommendation to a T and use the right tool for the right job.

cyns: proposal that this should not be in the standards layer, its a UA layer things

<Zakim> cyns, you wanted to say that moving things in the DOM is much less performant and harder programming than using CSS.

cyns: having different things based on different devices is UA issue, not spec issues

having a universal thing that does it all reasonably is not realistic

JN: the issue with CSS is an algorythm is how to define how things are laid out - that's what CSS is
... want to echo Cyns point - devs will break things

JS: think we understand the issues better - so thanks

JN: read from issue list noted Thursday

JS: starting at bottom - where are we with system colors and base fonts

CYNS: want to be able to say color: system button color

<astearns> https://drafts.csswg.org/css-color/#system-colors

fanasai - this has been deprecated however, but it must still be supported

but it is archaic - does not reflect modern browser interfaces and modern OSes

<astearns> (at least 8 different *text colors)

Cyns: what replaces it?

fantasai: nothing today - if you are trying to make a button look like a system button there isn't a way today

JN: notes that the deprecated issue replace by CSS apperance property, but..

fantasai, that's being dropped too

florian: we are still looking at this, and there is brainstorming going on

this is a thing that affects more than a11y - authors want this for other purposes too

fantasai: form input styling was implemented by Moz, and found that authors were breaking things

<Zakim> JF, you wanted to ask if this impacts high-contrast, and is that even an issue today?

Florian: looking at this from 2 angles - system color issue, and also whether or not the OS should fix th epage, or whether the page fix itself

general approach is to do this with media queries - that the page makes the statement and then it fixes itself

then there is also the situation where you want to go to inverse colors - you want to do that for text etc., but not images for example

so this is where the discussion is, but not there yet

cyns: 2 scenarios - good devs doong good jub, and then what happens when they do a bad job? make taht a UA issue?

Florian: similar here is when user wants to print, and you want to turn off images - if the user knows a better way to do things they should be empowered to do so

Rossen: sounds like interaction between content layer and host layer (browser/OS)

browsers do thigs differently - addressing this is too hard

but the other issue is who should be in power when user wants to make a change - one school of thought is that the browser has control

or we can say leave it in the content layer, and instead make a polite request

wth media query approach, we can advise overall system colors have been changed - we can communicate that browser has "pulled the plug" too bad, or the other option is

to be more polite and give the choice

so if the dev wants to respect "theming", because the host is not communicating any useful data to make logical changes [sic]

the struggle comes when you need a uniform presentation system across the board, and if we give authors the ability to participate in changing that, then they need full participation

cyns: if users don't know what they are doing, then browsers should make reasonable suggestions

but if author wants to specifically provide a high-contrast mode, then everyhting needs to be stylable

MC: base-fonts - was in HTML

were the author specifies something about the font based upon an uderstanding

with CSS now, there is a system level setting for pefered typeface, color, weight, etc.

and authors could express things relative to *that* - for example authors like to make things smaller, but user has a "larger font setting"

fantasai: we hve the REM unit for that today

a lot of authors take up font-size @@

and don't take advantage of this, but if authors leave the root alone, they can take advantage of this

<astearns> https://drafts.csswg.org/css-values/#font-relative-lengths

fantasai: REM units can be used on all kinds of things, margins, spacing, headings,

there was a discussion around this and media queries, which looked at layout issues around this

EM units were hared due to math, but REM units are easier - but lack of author awareness today

Rossen: funny enough, we were talking about he education problem, and arrived at that - there is an action item to write good documentation and best practices

fantasai: we need to ensure that it is accessible, and we have taken this as an action item

<Zakim> MichaelC, you wanted to mention wcag techniques, and best practices

<fantasai> Best practices for a11y is a core part of best practices for CSS

<fantasai> JF: Conceptually, we're looking to be more pro-active rather than reactive, and getting people embedded in groups to improve things

<fantasai> JF: Want to get some of our ppl more directly embedded in groups

MC: notes that WCAG WG working on similar issues

JF: discussion about embedded SMEs in other WGs

hober: there are a bunch of other values you can have above system font sizes

sounds like perhaps yo need to file a bug

JS: introducing braille CSS

JW: perhaps not best person to talk about this, but there were ast efforts to develop braille style sheets

there was work done many years ago, but now D-Pub re-surfacing this

looing at Daisy, who are using CSS to impact their print layouts

<fantasai> http://www.w3.org/TR/CSS2/media.html#media-types

<hober> I was referring to various values that can be set on the font shorthand (e.g. font: caption); see https://drafts.csswg.org/css-fonts/#font-prop for the whole list

Florian: so you are talking about braille-print?

JW: differences between braille output and print, yes

JS: another issue is COGA

COGA doing gap analysis and hve been working on this - how to address those types of issues

they are almost complete with gap analysis - loads of information, but it needs to be distilled into an executive summary

<MichaelC> scribe: MichaelC

jf: there´s a need to distinguish literal from non-literal language uses

e.g., ¨it´s raining cats and dogs¨ is non-literal

could embed literal alternate content like ¨it´s raining very hard¨

and user could prefer which they want

fantasai said that might be doable with ruby

does that work for this?

fantasai: ruby annotates base text

jf: but this is a replacement

fantasai: you can replace the base text with the annotation

jf: maybe that would work

<mhakkinen__> +q

but anyway for now we wanted to let you know we´re exploring this kind of thing

the coga group has put together a lot of use cases

though it´s a big doc, not expecting you to read that just yet

we´ll try to distill it

ra: is this alternate or supplementary?

jf: alternate in markup, but might be replaced

mh: this is big in assessment

provide simplified text for language users

we need to do both text replacement and text supplement

dk: sounds like conditional text, from earlier XML efforts

so it´s a general problem

jf: has it been discussed recently?

dk: no but would be interesting

<JF> JW: moving onto media queries

<janina_> scribe: JF

<MichaelC> jw: a11y has been evolving towards presenting content customized to needs of user

a shift towards more responsive and personalized interface based on user preferences

<MichaelC> based on a profile expressed as preferences

as opposed to "the blind view" [sic]

question now is how to bring those preferences to the UAs

JW: what we expted to see is that ARIA WG will exand on this, but it seems that media queries will play a major role

so there may be an architectual discussion that will need to happen if there is also a need for an API

there is work happening outside of W3C in this area as well

so another heads up to CSS WG taht there may be anothder discussion

JN: want to note that there are possible security issues here, as well as the issue with testing concerns

Florian: recall seeing this before - for example can you express if a Screen Reader is present

interest there as well, but there were also concerns around privacy, and concluded that CSS WG is not the place to work that out

hober: we are interested in this as well
... the existing system preferences differ greatly on platofrms as well

think that its important to figure out how to express these preferences in an independant way

other issues: there are many more user preferences than there are system settings - need to think about if we re trying to make things to adapt to the user config, or to expose preferences beyond different scenarios

second one sound scary-hage

hober: in Indie UI there was some work done there...

JS: sounds like we need to carefully scope this to esnure we have a clear understanding of all of the issues


<hober> joint meeting with dpub is in #dpub

<joanie> scribe: joanie

Research Task Force

MH: Shadi is the staff contact for the task force.
... Previous incarnation of R&D group is gone. It did not get rechartered.
... The APA charter includes reference to a research task force.
... What is the name of this task force? What is it trying to achieve?
... We want to make it clear what this task force is trying to do so.
... We're trying to identify gaps in accessibility that we can then research in order to fill the gaps.
... We also want to share the results of the research with groups which are not part of the W3C.

JB: An alternate view on the naming question: I've spent a lot of time talking to those who commented on the charters so that I understood the reasons for their concerns.
... Several organizations objected to folding the scope of any R&D work into APA.
... After divesting APA from its ARIA work, which was seen as a distraction in the past, there was concern that R&D might prove to be another/new distraction.
... For R&D there were a series of online webinars.
... So Mark and I have been discussing names so that it's clear that we're still focusing on spec reviews.
... With the reframing and follow-up discussion, the concerns about this work seem to have been addressed.
... And the relevance to accessibility seems to be better understood.

MH: We need to start a process forward for this group operating in the context of APA.
... Do we need a separate mailing list? Do we start looking for participants? What are our priorities?
... Can we help the COGA process in any way?
... What can we possibly do regarding role proliferation (e.g. with DPUB)?
... RDWG partipation in the past was good. Do we approach those people?
... In the past, we generated symposia and associated reports.
... That is not our present model or what we will be doing.

JW: One of the opportunities which this task force creates: APA, with its own resources and collaboration with other working groups, is identifying other issues for research.
... Of course, the task force can do this autonomously.
... But collaborating with other groups with interest in this area would be helpful towards addressing architectural issues with respect to W3C technologies.

JB: I don't know how much impact this group will be able to have on W3C as a whole.
... For many member organizations, connecting with research is not a top priority.

<mhakkinen> +q

JB: These organizations often prefer to develop and examine the results.
... My guess is that in order to attract people to the group, you'll need a clear picture about how to conduct this work.

<shadi> [[Accessibility Research Tracking (ART) TF]]

JB: Scientific panels and the like probably won't be a good fit.
... By the same token, we don't want to discard that approach.
... Some organizations like, with light overhead, to bring people together to dig into a topic. For instance, for a day or so.
... Such a mode might be a sweet spot of one organizational approach.
... And then say, "this is the fishnet we want to capture your questions with."
... With a spec review, you are sometimes lacking sufficient time and just want to identify any possible issues with that spec.
... And you don't have time to dig in or write a paper.
... So maybe a combination of these two approaches would be good.

MH: I wanted to echo the point about the symposia.
... I do think we have a heavy model before.
... I'm in favor of a lighter-weight mini symposia.
... I'm also interested in participating in accessibility-related conferences (e.g. CSUN and ICCHP and ASSETS)
... I would like to be more proactive in that regard.

<Zakim> shadi, you wanted to ask about direction of interaction

Shadi: I had a similar question about the workflow.
... If APA is looking at, for example, Web Of Things. Would this task force go out and look at what already exists? Or would it put the questions out to the research and/or accessibility communities? Or maybe both.
... There's going to be a lag in response time.
... How do we see the process with respect to identifying the issues and then interacting with the communities.

MH: We have the task force members themselves with domain expertise.
... But as part of our work, we'd utilize our task force as a means to solicit input.
... Finally we'd reach out to a broader (beyond W3C) group.

JB: My first response is: Why would it be an either or? It depends.
... The kinds of questions identified will have an extraordinarily broad range.
... A particular feature, if tweaked, might go from a narrow scope to something with much broader applicability to disability groups.
... But you have to first prove that such an approach is useful to the secondary needs.
... For instance, something useful for individuals who are blind/visually impaired might prove useful for individuals with cognitivie impairments. But initially, that would be speculation.
... In some other case, we might conclude it will take a couple of years to work out.
... In some cases, we might make a checklist and then identify a subset of items to take on.

RS: Two questions: First one: When doing research, whom are you going to engage? Pure researchers? Members of W3C who are not researchers?
... Second question: Have you thought about a process for prioritizing things?

MH: With respect to the first question, I think we're going to be fairly open in trying to identify the best people to address these questions or develop them further.
... I'm not going to pre-define whom we go to.
... I think we need to be open.
... To the second point, I've not had time yet to identify processes for prioritization.
... We're not going to set our own agenda. We have to be in consensus with others.
... Let's get the questions defined and then move forward.

Shadi: I appreciate those answers. And I agree that we'll have different types of questions with different types of timelines.
... I'm just wondering if we'll be going out now and inviting people to participate in this taskforce.
... I'm not sure if the charter is clear in this regard. And I don't know if that is the place to do so.
... We can then go out and solicit participation.

MH: I've not been involved in the start up of a task force before.
... What is the proces for task force formation?

JB: The nuts and bolts are: You'll ned a task force work statement, and at least one task force facilitator.
... There's a process involving the chair and consensus of the group.
... You'll also want to get in touch with your domain lead.
... You'll also need to set expectations for participation.
... Michael is very good and knowledgable about those.
... I think what's important is to take the transformation that I think Shadi has already started doing with the original task force scope. And then transmogrify it into the spirit of the altered/clarified scope that is in the APA charter.
... Almost start fresh and do a good cleaning on it. And then get feedback on it from the working group and also from some of the non-current participants in the working group with a strong research interest.
... I would also solicit feedback from those who had concerns.
... Then if you work out some talking points with everyone in the working group.
... My hope is that you'l have significant overlap with the other working group activities and members.
... We're also really hoping that you (Mark) can help with your connections to external research organizations.
... In the past, something we don't want to repeat, is drawing from regional research and non-profit organizations that are outside of those driving work within the W3C.
... That resulted in a disconnect between the group and members in W3C.

MH: During the rechartering process, I was really looking for ways to get more participation from member organizations, in particular industrial organizations.
... This is definitely on my list.

JS: I guess my concern is that we don't see this cross participation between core APA people and people interested in research (other than Mark and Jason).
... I don't know if we're going to need to recruit for that.
... I think what I'm hearing from APA core members is that they agree this work needs doing, but might not be able to participate.
... How we solve it, I don't know. But we do need to solve it.
... One thing we're hoping to do is to have subject area expertise as part of core APA so that we have people we can get input from as we review specs.
... We want to find a set of people involved with both the review activities and the task force.

JB: Given that it's late on Friday, I would not be discouraged by the smaller number of people in the room at this time.
... But I do agree with your concerns, Janina.
... Getting clear and predictable ways to involve people in this process would be beneficial.
... Janina, as the APA chair, if you could identify those tasks/questions the group could take as well as the more challenging task/question ("hyena") for the task force.
... In other words, if you have predictable planned steps (id tagging and a handoff).
... The clearer the process, the more likely people will be to partipate.

RS: I think that the thing is.... I cannot emphasize this enough: Whatever you do for research, in order to get more people involved in the discussion, I think it's important to sit down and think of the business value to companies.
... This doesn't happen enough.
... One example is more focus on cognitive accessibility.
... You can consider the work currently going on as research.
... But one business use case is the aging population.
... There are huge business opportunities in this space.
... It's the type of thing that will cause people to stop and think, "while I'm really busy in these other areas, I'll make time because there are important business opportunities."

MH: I agree.
... I don't think it's appropriate at this point to say that we're conducting research.
... We're identifying issues.
... But other than that I agree.
... Through collaboration with member organizations and also outside organizations, we can look at these questions and then try to get them answered. But our group itself won't be doing the answering.

JW: We're talking about identifying research that bears on important questions that arise in the development of accessible technologies through specification review.
... And it's not just identifying the issues, but also informing our review process, if I'm understanding correctly.
... I think the specific processes for doing this will evolve over time.

JS: We might also want to look at other task force work statements as examples for ours.

JB: I'd really like to see this task force take off.
... Let's identify who is doing what.

<Judy> ...by when.

MH: At the next APA meeting, can we summarize what we've discussed today....

JB: I tend to be really detailed with action lists.
... Things to do:
... 1) Getting us time on the upcoming APA agenda.
... 2) Get the facilitator-elect/designee to come up with four or five bullets/talking points.

<MichaelC> s/1 pm/12 pm/

JB: 3) A few folks ought to look at the last task force work statement version and pretend you've never seen it before and compare it to the APA charter and also the concerns received.
... On item 3, Judy, Jason, and Mark.
... 4) Propose at least three methods of work.
... This can be super simple, like capturing things from workgroup discussion by such-and-such method.
... Spinning up some topics in the task force, selecting some for longer discussions, and selecting from amongst five different options to send to the group (e.g. a straw person)
... In other words, clear workflow flavor with examples.
... 5) Draft two sample outcomes (a light one and a more detailed one).
... Keep it simple at first.
... 6) Come up with a straw recruitment list. Just a sample to give people an idea. (e.g. these five research labs, APA WG, and other W3C groups)
... Just to give people a flavor of what we're thinking.
... But also include a few people for APA, HTML5/Web Platforms, etc.
... They can discuss and then we can go from there.

<Judy> Judy thanks Joanie's good minuting!!!

<liam> scribe: liam

<scribe> scribenick: liam

HTML Accessibility Task Force

<scribe> chair: JS

JS: This session is about the future of the HTML Accessibility task force

and the reelationship between APA and the Web Platform

Do we value this thursday check-in time, the TF meeting?

LW: Right now we have 2 main work items

first is draft proposal for transcript

e.g. transcribing video

There's also a spec for panel and panel set

for elements to be used e.g. for carousels, tabs, etc

probably will come under Web Platform, taking up HTML

transcriptions probably will go to timed media

JS: also we're still working on edits to what the HTML spec says about alt

LW: and that will result in bugs files against HTML, so Web Platform

Charles: I don't object to the meetings on thursdays. As often as not it's a pain to have a meeting. I'm not a big teleconference fan.
... More generally what does concern me is, if the WP group has to have some a11y stuff in an over-there place that's als osaying it's not in the places where it really needs to happen!

I'd rather have the stuff that HTML Accessibility TF was doing being done in the Web Platform WG.

scribe: It's very important in that case that APA gets a feed of what's coming down the pipe in the Web platform.

LW has been nominated to keep APA informed in this regard.

E.g. I [Charles] am likely to take some a11y-related HTML bugs directly into HTML

You will want to know about that, I presume! "we did this and this, it's in".

How do we handle the transition?

Technically there's a Web Platform WG; what it does is still up for grabs. Right now you can't really edit the HTML spec.

We plan to change that so we can fix bugs... and then you can go to the Web Platform WG & do stuff.

LW: OF the 4 cochairs of WP, 2 of them are really invested in a11y, sends out a message.

We and the other co-chairs really want to being as much a11y expertise & knowledge into the WG as we can.

Cynthia: in Web Apps does the bulk of the work happen in git?

Charles: Yes, git.
... specs go into git, issues are tracked in github tracker.

There's no requirement that everything be done that way; e.g. there's an editing tf with its own mailing list & meetings;

scribe: a web components group, we don't have formal substructures [so some use mailing list & some use github]

Charles: Web Components had 3 f2f meetings this year so far.
... There are some unsolved ussues in Web Components; people want to solve them but not yet got a way to do it; probably will be an f2f topic. Dmitry Glaskov in particular,

they say, "here's what you have to read/prepare to be at this meeting"

so the browser people show up ready to engage.

Most Web Components meetings are in California.

Cynthia: we are obviously more distributed geographically.

Charles: if the editing tf has a f2f it's likely to be in Berlin.

<Zakim> JF, you wanted to repeat my concern about documentation and moving to new pltforms/tools

JF: Cynthia asked about docs; there's a problem at W3C, in helping people use github and use these newer systems.

Jason: In general I agree with LW that we've been saying for a long time that a11y needs to be taken into account throughout the development of new specifications

Ideally a11y expertise should be engaged as part of the group making the spec

<JF> slight correction to the minutes - there is a need for good documentation to aid the transition to these newer tools

scribe: I suppose the only issue I'd have there would be aboue formal participation requirements

LW: to take an example, James Nurthen and I just talked with Johannes about editing rich content in HTML, he was looking for a11y input
... so as a first step I agreed to have a conversation with them and also to take this back to APA.

So that's one way we can keep you guys up to date.

Charles: John, need for documentation noted. Github is unintuitive.

We have some level of documentation about our workflow, the things we expect.

Participation requirements are fairly straightforward. You join in and participate. We expect to behave like you know how to be nice to people.

Not a lot of formal requirements

We haven't had a good standing rule for about 8 years.

No required telecon. We have a f2f at TPAC.

For a number of years we've hard a f2f around California in the early part of the year, and more recently moved to smaller & more focused f2f meetings.

There are a number of mailing lists. We don't like flame wars or people being rude to each other.

They're very big groups but seem to be mostly workable.

Cynthia: so, how do people wanna do?

Charles: given that description of a group that you (PF/APA) have to engage with, I'm interested to hear [response]

Cynthia: I know a lot of people don't like time-synchronous stuff.

Charles: I don't expect to have a regular Web Platforms call.

There's more willingness to have a phone call or several phone calls now than 5 years ago, to help us through a situation

and for other issues, github issues + f2f seem to be the dominant modes.

<LJWatson> +Q+

Jason: as to my personal involvement I'm interested, issues relating to HTML and APIs I'm interested in. I participate in APA.
... I want to engage in issues that would involve the WP WG whenever I can make a contribution to the specific issue on a specific spec.

Whether that involves becoming a participant in the WP WG would have to be decided, and also by my AC Rep, so I can't make a commitment about that,

scribe: but I've occasionally been involved in meeting with the task force because of specific issues.

<chaals> ack +

JS: I'm also one of those people that prefers talking to writing

I don't think any one modality covers all people, there's always things left out, the precision of writing things down certainly isn't there in a phone conversation.

On the other hand it's easier to pick up if people are upset on a phone call.

scribe: I'd like to have a mechanism available. How much are we needing a mechanism with ongoing work? We've needed it in the past, when we had a packed agenda.

There were as many as 3 or 4 subteams working too.

The MAUR is now [mostly] done.

The canvas 2d spec is done.

So I don't know if some of these new things would benefit from that kind of regular participation.

Clearly we'll need to communicate. I'm very happy that LJWatson is going to be participating directly in WP.

So depending on how much we need to talk to each other on a regular basis, there really wasn't much time in the Wednesday PF call to talk about HTML TF stuff.

Recently it's not been unusual for a Thursday call to end early.

So if we can schedule when we have specific issues going on, that might be fine.

I'm expecting we'll have a good working relation with HTML, as we've had in recent months, recent years, unlike at the start.

So i'm inclined to keep a mechanism open but don't know how much need there will be for it.

People in the TF but not in PF would be asked to join APA.

ljw: yes, that door will always be open.

I should also have mentioned new work, ideas for new specs will be incubated in the community group [WICG]

If there's enough interest they can bring it to a certain point of maturity & then complete an Intent to Migrate document.

In that document you also document accessibility impact.

So the expectation is that a11y is an intrinsic part of anything new.

Could the alt work be moved to PF/APA?

JS: The other day we decided that the APA calls would stay at 60 minutes but we'd allow ourselves to extend them.

<Zakim> Judy, you wanted to suggest a way to bridge lack of shared experiences and understanding at the kick-off stage of new work areas.

Judy: Through some recent conversations I've become generally leaning towards support of phasing out the TF, but I wanted to list some concerns and also some suggestions

One is a concern just about the volume of APA work, and so the HTML WG itself was pretty reliable in creating a large volume of things that needed significant attention from a bunch of a11y experts

scribe: and APA doesn't just have to worry about Web Platform but also all other W3C work and cultures.

But LJWatson, if you coordinate well with janina_ it may work really well

scribe: So if you say, these are the topics, maybe you take 25% of an hour's APA call. We should be vigilant, since w're dropping an hour a week & merging 2 hours into one, that we don't start dropping things on the floor.

Going to more esoteric concerns about how people work together well, to get accessible technologies working requiers a significant presence from different stake-holder communities,

scribe: and a shared experience. With the volume of stuff in the HTML work it's not doing it just once but many times,
... and a11y folks need to adjust to different dynamics and buzzwords of each area.

My guess is it'd be useful to have a joint kick-off meeting at the start of each new spec / work item.

scribe: So that'd be a case where it might be unfortunate to have dropped the TF. But if we remember to do it for each new area it might help.

JS: +1

One of the things I worry about - the climate of HTML and WP is great compared to what is was. A11y is sometimes on the edge & it's ideal to have a lot of different perceptions - what can we learn about the best way to bring people in and have it work well?

I like that the 2 cochairs in the room are committed ot having that work well.

It may be harder than expected for pepole from disability organziations to come to the table and contribute.

I'd love to know how that engagement & pulling in can happen.

<Zakim> liam, you wanted to note the reason for the lull in work

<joanie> scribenick: joanie

Liam: You observed that there has been less work taking place on the taskforce.
... That is in large part due to the rechartering.
... And might not reflect the amount of work

<liam> scribenick: liam

Charles: I strongly believe that people who have sat down together work better together than people who have never met.

As far as I'm concerned people with interest in various different areas with a11y, e.g.publishing, regional, are amongst the stakeholders we have to go and find.

I hope we get some reasonable number of people from different disability areas directly participating in work.

scribe: And yet I think it'll be important to maintain the connection back to APA. So we still want to make sure you're looking at what we're doing.

One problem with kick-off meetings for new work is that we get a reasonably large number of specs coming in, browsers start implementing, then after a year or 18 months they discover it doesn't work, throw it away.

scribe: It has happened repeatedly, e.g. file system access in Web Apps, others, had a proposal, done a pile of work, then said no.

WebSQL was another one.

People come to e.g. TPAC just to get to know each other.

I don't think this is going to be perfect - not underestimating the difficulty. A11Y is just one of the pieces.

One obvious bug of medium size, some of them, are a11y-related. Not controversial, but expecting to use them as one of the tools to draw people in.

<chaals> s/one obvious bug/Medium size bugs/

Jason: On the formalities of this... When I became a partcipant in the HTML TF there was an expectation of also becoming a participant in the HTML WG, but that wasn't a burden; seems it'll be same with WP... but there were valid reasons behind that expectation of being
... a participant in both groups, so a determination should be made about this case, e.g if you were involved in work that engages with HTML or any of the platform APIs of the WP WG in relation to a11y, is there going to be a participation that you'd become a WP participant as well?

And I think there were IP requirements.

I'd be willing to be in both groups & to commit time on an ad hoc basis to work on issues.

plh: Github - well, we are switching to github and are trying to switch as many WGs as possible to github.

We have more documents, the Web is getting larger, so I'm trying to scale with the resources I have.

scribe: I think it's important that we find ways to monitor the incubator group. E.g. there was a proposal about when an element comes into a viewport.

Having a f2f meeting is difficult, we need to find the right target ,get the right people round the room.

I don't think the WP WG had enough preparation for their first f2f, but it was their first!

Liam said earlier the was a lull because of transition; the WP WG also doesn't know what they want to do yet either.

<Zakim> Judy, you wanted to (1) talk about leaving room for experimentation; to (2) suggest a threshold for needing a kick-off meeting; and to (3) suggest maybe some deliberate balancing

Judy: (1) when you're getting people to work together well it can be extraordinarily helpful to set & manage expectations
... So charles, what you were saying earlier about a failed example, that's a normal part of development, and it'd be good to tell some of the a11y experts up front that hey, this sometimes happens

Some people will get very frustrated, and may be gong through contortions to get a11y features into something that goes away

So it might be useful to say upfront, this particular group is going to be doing a lot of experimentation.

Whatever you do at requierments layer will still be valuable

2. for a bunch of areas it may be the same people working together so you may not need a kick-off meeting.

3. I'm wondering when it comes to integrating a11y experts with very different levels of standards work experience

maybe we can liaise back from WP WG to APA

e.g. with check points to check with a broader circle of review

<Zakim> chaals, you wanted to talk about IPR commitments

Charles: If you're in WP filinf a single bug, anyone in the world can do that. If you start suggesting significant changes, we'd like you to formally join the WG and make the IPR commitment.

That's easy on the W3C side, you might have to convince your management.

E.g. if you're suggesting substantive changes for a11y, join the WG or there's another way to give IPR comitments.

Re. balancing, I want a reasonably continuous flow of information to APA, so APA knows what we're doing. The

scribe: great hope is that WP is flooded with the best experts the world has to offer. But my hopes and dreams don't always turn out. So we'll see.

Any group that's been working together builds up its own culture and shorthands and it's more difficult to engage when you come to that group,

scribe: so there are always friction points, culture clashes, cream puffs. We'll do our best as chairs to minimize them,
... sort out how we get the work though and not spend time fighting.

I have a strong preference for meetings that don't go over an hour.

The WP meetings I expect will continue to follow the pattern of we have an hour, then a break, then an hour, then a break.

Charles: for short term we'll run the calls for a little while longer at least. If there's a call for 2 people it's going to end.

[James leaves the room]

scribe: and if there's a community that needs a telecon regularly, that's what we'll do.

<JF> +1 to staying flexible

Judy: Sometimes it takes a while for pepole to set up a task force when it's needed quickly, so might want a provisional or informal or standing arrangement

LJWatson: that's like the subgroups

Judy: yes, also in HTML A11Y TF.

Janina: Adjourned.

We'll play it by ear.

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/10/30 08:11:17 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.140  of Date: 2014-11-06 18:16:30  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/cler/clear/
Succeeded: s/(changing/mck: I disagree strongly with that (changing/
Succeeded: s/authro/author/
Succeeded: s/sue/use/
Succeeded: s/@@/CSS apperance property/
Succeeded: s/droped/dropped/
Succeeded: s/ost/host/
Succeeded: s/nees/needs/
Succeeded: s/q?/scribe: JF/
Succeeded: s/Indue UI/Indie UI/
Succeeded: s/discusion/discussion/
Succeeded: s/@@@@/ICCHP and ASSETS/
Succeeded: s/Getting us time/1) Getting us time/
FAILED: s/1 pm/12 pm/
Succeeded: s/research labs/research labs, APA WG, and other W3C groups/
Succeeded: s/yy/LW/
Succeeded: s/we'er/we're/
Succeeded: s/Charles;/Charles:/
Succeeded: s/yy/Cynthia/
Succeeded: s/James Snedders/James Nurthen/
Succeeded: s/MAWR/MAUR/
Succeeded: s/hTML/HTML/
Succeeded: s/; I like/: I recognize/
FAILED: s/one obvious bug/Medium size bugs/
Succeeded: s/recognize the idea/strongly believe/
Succeeded: s/enoug/enough/
Succeeded: s/yuo/you/
Found embedded ScribeOptions:  -final


FAILED: s/1 pm/12 pm/
FAILED: s/one obvious bug/Medium size bugs/
Found Scribe: MichaelC
Inferring ScribeNick: MichaelC
Found Scribe: JF
Inferring ScribeNick: JF
Found Scribe: MichaelC
Inferring ScribeNick: MichaelC
Found Scribe: JF
Inferring ScribeNick: JF
Found Scribe: joanie
Inferring ScribeNick: joanie
Found Scribe: liam
Inferring ScribeNick: liam
Found ScribeNick: liam
Found ScribeNick: joanie
Found ScribeNick: liam
Scribes: MichaelC, JF, joanie, liam
ScribeNicks: MichaelC, JF, joanie, liam
Present: Ted_O´Connor Joanmarie_Diggs Peter_Linss James_Nurthen Bert_Boss Can_Wang Michael_Cooper Florian_Rivoal Takeshi_Kurosawa Matt_King Jesse_Beach Dave_Kramer Alan_Stearns Takashi_Yamanishi Steve_Zilles Janina_Sajka Wei_Wang Taketoshi Yamane Dave Cramer hober Cynthia_Shelly Markku_Hakkinen Jason_White Bert_Bos fantasai Rossen_Atanassov rossen LJWatson JF Liam_Quin ShaneM
Agenda: https://www.w3.org/WAI/PF/wiki/Meetings/TPAC2015/APA
Got date from IRC log name: 30 Oct 2015
Guessing minutes URL: http://www.w3.org/2015/10/30-apa-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]