W3C

- DRAFT -

Accessibility Guidelines Working Group Teleconference

04 Jun 2019

Attendees

Present
AWK, Raf, Laura, alastairc, Detlev, bruce_bailey, Chuck, JustineP, stevelee, Fazio, SteveRepsher, MichaelC, david-macdonald, JF, johnkirkwood, mbgower, MarcJohlic, Rachael
Regrets
Nicaise, Jake
Chair
SV_MEETING_CHAIR
Scribe
Laura, david-macdonald

Contents


<AWK> Zakim agenda order is 1,2,3,5,4

<laura> Scribe: Laura

<stevelee> preset+

TPAC registration https://www.w3.org/2019/09/TPAC/ (Early Bird pricing open until June 21)

AC: reminder early pricing is good for only a week or so more.
... 9 people registered so far.

Charter process discussion

AC: my first charter process.
... need a charter by october.

<alastairc> https://raw.githack.com/w3c/wcag/charter-2019/charter.html

AC: we have a draft.
... will be similar to our previous charter.

<scribe> …new dates

bruce: will charter state 2.2 can have normative changes from 2.1.

<Zakim> bruce_bailey, you wanted to ask if charter mentions permission to deviate from 2.0?

MC: won’t get to into that detail.

<Zakim> AWK, you wanted to speak to timeline and process a bit

awk: talk about that separte from charter
... we have to get charter to management by end of june
... not in a panic mode to get this done.
... will be responsible about of time needed.

ac: let chairs know if you have any comments.

<JF> Q

bruce: some normative changes from 2.1 to 2.2?

ac: been discussing color and 4.11
... looking at significant changes.

awk: we need to address the issue in some way.
... we are working on proposals for 2.2 and then hammer out backwards compat and changes.
... all tbd

jf: color and 4.11 we haven't made a decision.
... lessen is okay. I think I would have an issue with making it stronger.
... leave contrast alone

ac: add a sentence for focus visible?

jf: if adding breaks tool chains then no.

<bruce_bailey> @JF if you (or anyone) thinks TT says something this group questions, please raise the alarm!

jf: concerned about conflicts

<bruce_bailey> +1 on the color contrast discussion getting dense !

jf: we need a new contrast SC not fix the old SC.

<bruce_bailey> To be clear, I am not proposing that 2.2 revisit 1.3.1

jf: regarding 3.3.1 and landmarks: could add a new sc but not change 3.3.1

<Zakim> bruce_bailey, you wanted to say i do not have large ambitions, but some phrasing in 2.0 is not consistant

<AWK> we probably shouldn't get into discussing the merits of any specific changes to 2.0/2.1 now

bruce: if TT says something this group questions, please raise a stink.
... group is open to feedback.

<bruce_bailey> The one thing I would like for us to consider is changing 1.2.3

bruce: if we are open to normative changes, would like a conversation on 1.2.3
... some places where phrasing should be updated.

ac: add an issue in GitHub. with wcag 2.2 label.

david: regarding contrast discussion greg explained how algorithm was created.

<bruce_bailey> @JF, yes 1.2.5 superceeding 1.2.3 is terribly confusing to people

<Zakim> JF, you wanted to say that 1.2.5 at AA supersedes 1.2.3

jf: 1.2.5 at AA. It supersedes 1.2.3
... it is an education problem.

<bruce_bailey> @JF back in 2008, i thought teaching 1.2.5 supercedes 1.2.3 would be easy

jf: could update the understanding doc. but not change normative language.

<bruce_bailey> it has not been easy

ACT Survey: https://www.w3.org/2002/09/wbs/35422/ACT-rules-process/

ac: ACT survey
... rules are like wcag failures

<alastairc> https://www.w3.org/2002/09/wbs/35422/ACT-rules-process/results

ac: 4 answers replied to the survey.

<alastairc> Process: https://github.com/w3c/wcag-act/issues/353

ac: thought is very clear and good.
... focus on top of the issue.

<Zakim> AWK, you wanted to say that the content is non-normative so if we find problems in the process we can easily change it

ac: need to look at how to hook into understanding

awk: the content is non-normative so if we find problems in the process we can change it

wilco: heard from shadi. pubish rules separately since a redesign is underway.

ac: for discussion. some issues in Github.

New techniques & issues survey: https://www.w3.org/2002/09/wbs/35422/Tech_Und_survey/

WCAG 2.1 techniques update: https://docs.google.com/spreadsheets/d/15idlBl1qQTNr2SIi4Drzk1Q1vnWAi5L26GCCv6mjD2g/edit#gid=0

ac: some progress since last week.

<alastairc> DAvid's updates: https://github.com/w3c/wcag/pull/583

some things to go into a survey for next week.

<alastairc> https://github.com/w3c/wcag/pull/669

ac: Micheal and detlev should take a look at Rachael’s Technique for device motion actuation

mg: did some followups on label in name. but not seeing in editor’s draft.

awk: take a look now.

ac: did merge some PRs.
... still looking for orientation and gestures techniques.

New techniques & issues survey: https://www.w3.org/2002/09/wbs/35422/Tech_Und_survey/

<mbgower> https://github.com/w3c/wcag/pull/732

<alastairc> https://www.w3.org/2002/09/wbs/35422/Tech_Und_survey/results

Path Based Gestures

ac: trying to pin things down.
... stemmed from an issue detlev created.
... some changes to understanding doc.

bruce: definition changed from option 1 to 3

ac: things in option 1 included in 2 and 3
... Option 2 The medium option is to base the 'path based gestures' aspect on whether it is a pre-set direction (or directions). If a movement has to follow a pre-determined direction it is covered. If it is unrestricted it is not covered. This is covered in the updated pull request.
... all of the above examples would be in scope (fail without an alternative), as anything that is not required to be free-form must have an alternative.

<alastairc> https://docs.google.com/spreadsheets/d/1NGxqbwbcaG8dwSs5H2eTRy0gDELTdSuAi1ad0le4fDU/edit#gid=0

ac: option 2 focused on what is a path based gesture and how it differs from drag and drop.
... made a spreadsheet with lots of examples.

<bruce_bailey> i feel like "drag and drop" (without alternatives) might sometimes be permissible under 2.1.1 -- but not with any of the examples in the survey

<Zakim> bruce_bailey, you wanted to ask if "excluded" means allowed?

<Zakim> JF, you wanted to ask about signatures - in scope or not?

jf: don’t see signatures in spreadsheet

ac: Drawing, e.g. a signature or freehand picture out of scope.

<AWK> Signatures are path-based. As a path-based gestures there needs to be a way to accomplish the same task with a single pointed unless the path based gesture is essential.

jf: we should be explicit.
... one is creative and one is legal.

<bruce_bailey> we have dodged the question about drawing

ac: using starting and ending points is problematic. as soon as you lift your finger it doesn’t matter.

<Zakim> bruce_bailey, you wanted to say we address "painting" but not "drawing"

bruce: is freeform drawing in or out of scope? painting out of scope.
... not time based.

<alastairc> "An exception is made for functionality that is inherently and necessarily based on complex paths or multipoint gestures. For example, entering one's signature may be inherently path-based (although acknowledging something or confirming one's identity need not be)."

ac: understanding doc explains.

awk: last week we resolved “The WG intended that path-based gestures in 2.5.1 does not include gestures that only depend on the start and end points.”
... if we have a path based gesture that relies on 2 points it is not a gestures.
... needs to have at least 3 points.

email example is an example of a path based gesture.

<bruce_bailey> hmm, not sure i agree about the email example

<mbgower> +1 to the 3-point method of defining 'path-based' in relation to gestures.

chuck: start point “relative” to end point.
... need to have “relative” in the definition.

<mbgower> +q to say we're obviously stretching, but 'direction' seems to me to be the ultimate determinant for path-based in relation to gestures

ac: tried to take out free form in the understanding doc.

<AWK> -1 to the relative approach. You can leave a slider in the same place if you want, but this also loses the connection to the use of path in keyboard SC

<Zakim> mbgower, you wanted to say we're obviously stretching, but 'direction' seems to me to be the ultimate determinant for path-based in relation to gestures

mc: we are trying to reverse engineer the SC.
... user is required to initiate
... initiate a direction

ac: need to consider a slider…may not require a direction.

<mbgower> is that really true, Alastair?

ac: 3 point would help with that use case.

<alastairc> https://shopify.github.io/draggable/examples/drag-events.html

has an example in spreadsheet.

scribe change?

<Zakim> bruce_bailey, you wanted to say that mousedown+move-a-little is just a selection

bruce: disagree with the email example.
... like 3 point definition.

ac: added another column to the spreadsheet.

david: if you can’t do it with a single pointer then its okay.

ac: need to be careful with that.
... depends on definition.

<mbgower> "relative direction"?

chuck: pitch again for using relative

<Zakim> AWK, you wanted to say that the slider example for Alastair is implementation specific

awk: sider is a good example. can be coded correctly or incorrectly.

<AWK> https://developer.mozilla.org/en-US/docs/Web/HTML/Element/input/range

awk: 3 point model makes it clear.

scirbe change?

ac: 3 point model scopes in everything we need and scopes out drag and drop.

<david-macdonald> scribe: david-macdonald

<alastairc> https://github.com/w3c/wcag/pull/403

Detlev: the original intent of the mobile task force was to ensure that any type of path based gesture would be covered in the success criterion. We scoped out drag-and-drop because we felt it was just too hard to do

I haven't thought long and hard about the three point and we will consider it not on this call. The problem was scoping out dragging, means that will scope out almost anything that we intended with the success criterion

Strictly speaking any type of content slider would be out of scope which means there would be very little left that would fail in terms of path based gestures which I think would not a nine or our original intention

<alastairc> Swipe, strict (e.g. next); Swipe, open (e.g. mail); Swipe, multi-value; Full page swipe (open); Slider (strict); Slider (open); Horizontal scrolling box (overflow-x).

Alastair: I'm pulling up my list now of things we want to put in and out of scope

These are the things that I think are covered

For instance the horizontal contents letter that you mentioned. I think we have a difference of terminology regarding what year talking about a swipe gesture versus a drag

putting a figure down moving at one direction and letting it go. I think that's covered

Detlev: content sliders are sometimes supplemented using drag-and-drop. If you put your finger down and wait just a fraction of a section longer you find you have no directionality enforced

An author wants to find a loophole would be able to argue that there is directionality involved

And then they can justify that they don't have to put in an alternative

<alastairc> Then excluded: Sortable list; Drag-drop to pre-defined places; Drag-drop to anywhere; Resize-drag of box corner; Panning Map; Drawing; Helicopter game

Alastair: we need some examples in this and I think we are in a relatively good position to do that.

Michael: if there is another method of swiping to carry something out then that's not a loophole then that's another method of interacting. I understand your point where you believe that mobile task force initiated something. We can say that for every SC. What often where we end up as light different than the task force attention

<AWK> +1 to Mike

If somebody comes up with a way of activating something with a swipe of our press and hold and go whatever direction they want it's no longer a requirement of direction so as not a loophole it's just two methods of operating the same thing.

<Zakim> AWK, you wanted to mention http://w3c.github.io/Mobile-A11y-TF-Note/#touchscreen-gestures

Andrew: I was going to point out to people that task force note around touchscreen should provide some additional clarity. But it does not really because as a best practice it talks about allowing users to either use a simple path or swipe gestures.

Yes we should talk about this for 2.2, but the original intent of the mobile task force was to have it to be a single tap and I believe that's what it was at one point. If it evolved and we think there's still gaps there then we need to address that in 2.2

Detlev: I think the normative text as it stands does not rule out that we cover dragging as well as swiping. I don't see that it's absolutely necessary to draw the distinction because it depends on our definition of path based and we haven't made that definition at so it depends on how we define path based to divine what's in scope without a scope. And we had a survey about that and we have six people in favour of keeping both of these things [CUT]

So there is no consensus right now that we would draw the distinction and exclude dragging

Michael: what is the definition about base that doesn't allow tracking.

<bruce_bailey> From 2.1.1: except where the underlying function requires input that depends on the path of the user's movement and not just the endpoints

Detlev: the language that we have previously but the start point in the endpoint ignore that there's something else that's important which is the ability to drag something across the screen and that is why we have the continued reference to the exception 2.1 is not helpful because now are talking with operational gestures that have an intent to operate on screen making a swipe or dragging a certain direction would indicate that
... I think we are free to define past based to cover the substance of operational path based gestures and how we do that is that we don't refer to the old exception language.

<Zakim> JF, you wanted to note that it's more than just "your finger" - you can be using a stylus as well

Alastair: I tend to agree that to 2.1.1 references not useful.

<laura> s/pubish rules separately since a redesign is underway/Publish rules separately since a redesign is underway./

JF: it's not just the finger it's also stylus. Thinking about project silver I could see the supply to touchpad interfaces where you sign your name. And asked me to sign on the screen there and the stylus is provided. I will be careful that we don't get so caught and that about swiping on the screen that were talking about any touch enabled interface where there is a pointer interaction. Whether it's a physical pointer are finger or something else.

Alastair: anyone else?

This was kicked off by Detlev's original issue. Were trying to find what's scope by trying to find Pat-based gestures. I'm just somewhat struggling to do that in a way that makes sense for all these different interactions.

If you essentially scope out drag-and-drop as in the free-form drag-and-drop, and things like drawing, then it's very difficult at not have that apply to things like sliders and certain types of swipe as well

So far the three point definition or method if you like seems to hit what I thought we were looking for. I can appreciate that sortable list should also be accessible but it comes down to should whether it should be accessible through the success criterion

Which I think is a slightly different question

So in terms of the examples I put above, which is above 1211, includes all kinds of swipes, sliders, certainly in terms of the mobile context if you have to go horizontally instead of vertically. An horizontally scrolling box which Detlev defined as a content slider. I think those are all in scope. The excluded ones are drag-and-drop, signatures, helicopter game and sortable list

<Detlev> +1 it has my support

Are those examples of in scope and out of scope what we would like to support has path based gestures?

JF: can any of these things be completed using the keyboard alone. I have left and right and up and down arrows, so I don't see how panning around map would be very difficult.

Alastair: we have keyboard separately

JF: the slider can be operated with the keyboard arrow keys.

<Detlev> @JF we are talking situations where pointer users have no kb

Alastair: the slider example at the top of the scroll or the bottom would now be in scope and so that would change.

<Zakim> mbgower, you wanted to say I think we need to focus on interpretation of movement

Alastair: there is a difference between making a required by keyboard but the question is about whether we need to make a work with a button

Michael: on the map if you want to think about a gesture is an interpretation of a movement versus moving left right up down. If you put your finger down and just repositioning the map. But unlike a yahoo email swipe move gesture left in be interpreted as initiation to interpret a message.

Alastair: I think we've gone as far as we can with this. We have a plus one from Detlev which I find heartening

<AWK> We are going to need to undo the resolution from last week in order to accept this

<AWK> Last week: The WG intended that path-based gestures in 2.5.1 does not include gestures that only depend on the start and end points.

<alastairc> Poll - do you agree that the 3 point method is the right approach to defining path based GESTURES?

Alastair: I think we just refining it. We are just adding the three-point method to bring in and scope

<bruce_bailey> please add example of painting (in contrast to drawing)

<AWK> +1 to 3 point

<bruce_bailey> please add at least one more game example to spreadsheet

<AWK> Pac Man!

Bruce: if were not making a distinction between drawing and painting that were missing something because painting includes speed pressure how thick is your style this. Where's drawing doesn't matter what I used to create it with.

<JF> +1 to Bruce

<bruce_bailey> i know of switch based interfaces for drawing

<bruce_bailey> i do NOT know of how to switch enable painting really

Alastair: or the ends of path based gestures was to put in anything that transfers the path to the content such as drawing. Painting is a superset of drawing which includes more characteristics.

<bruce_bailey> drawing is not usually how hard or how fast

<Detlev> +1

+1

<laura> +1

<mbgower> +1

<Chuck> +1

<JF> +1 to 3 point method

<MarcJohlic> +1

<bruce_bailey> paint effect usually depend on speed, and may depend on pressure

Alastair: I definition is the next thing to do. I did have a sort of draft but I need to update that based on this three-point method because I think it'll be a bit neater.

JF: I'd like to see this definition before we close the issue.

Alastair: will have to get the text in and then I'll agree that that's a we intended.

RESOLUTION: the approach is agreed-upon and it will be written up

Understanding update: Reflow doc referencing 1.4.4 (PR 715)

Alastair: I don't think I can face number two right now. It comes back to path based gestures.

<Detlev> let me just say we have added buttons to the slider

Alastair: so go to the next one: it is about resize text. The proposal is to discuss how to we get to 200% when something is magnified to 400%

RESOLUTION: accept pull requests 715

PR: 692

<alastairc> https://github.com/w3c/wcag/pull/692/

DETLEV: I question whether just saying "Hey Kim" would have such an effect.

Michael: yes if the focus is not an input than the speech recognition will type in the words by their letters, and any associated character key shortcut will be triggered if it's not an input

RESOULTION: accept PR 692

Motion Actuation (PR 693)

<alastairc> https://github.com/w3c/wcag/pull/693

RESOLUTION: accept PR 693

Non-text contrast (PR 701)

<alastairc> https://github.com/w3c/wcag/pull/701

<JF> +1

RESOLUTION: accept PR 701

Should non-drafted techniques be included?

<alastairc> https://github.com/w3c/wcag/issues/369

JF: I don't want to lose any of these ideas even though they have not been fully baked. I think if nothing else we stand those off to the people of Silver because they may some thoughts resources around that. I agree however with Bruce that if we had a list of what the gaps really are can we make it bit of an effort to create them.

All of a sudden we were working on 2.1 and we had taskforces that were exploring the edges and the gaps in our requirements. But I don't think we should just abandon them. They were there for a reason and we should at least make sure that we have addressed that reason.

If we don't have techniques in our we asking for compliance?

<alastairc> Survey results https://www.w3.org/2002/09/wbs/35422/Tech_Und_survey/results#xq19

Michael: there is an implied promise when we say the word future technique. It might be better if we just said not yet published. Perhaps finding something that's not quite so promising about the future might be a little less embarrassing to hang around for 10 years. I agree with John it's good to keep the intentions because somebody could be motivated later on.

JF: I'd go further and I agree with the notion of a promise. I think we should put some effort into seeing how many of those techniques we could write up. Changing the name is like rearranging deck chairs but I think would be better spent trying to get some techniques in there.

Alastair: the quick reference draws in anything that has a title,

MichaelC: an update of the techniques in the quick reference is on the way right now with Eric and myself.

Alastair: is it possible to pull out a list of the things that got removed

Michael C: we could pull out a diff document between what was pulled out and what was there before

Alastair: sounds like the consensus is that we should keep them but were not quite sure how to at this point.

Some people suggested that we ask for volunteers to write them up, or have somebody to analyse them and make suggestions to the group and what to do, and send them over to Jeanne and Shawn over at the task force.

s/sean/Shaun

Alastair: we'll get that list together and it will probably come down to people working through the understanding Dobbin to decide on which ones should go in.

<JustineP> Muted in a loud environment but +1 to this approach

RESOLUTION: Michael Cooper to compile a list of the techniques that were deleted.

Bypass blocks (PR 721)

<scribe> ACTION: Michael Cooper to compile a list of the techniques that were deleted

<trackbot> 'Michael' is an ambiguous username. Please try a different identifier, such as family name or username (e.g., cooper, melledge, mdgilbert, mbgower).

CTION: MichaelC_ Cooper to compile a list of the techniques that were deleted during the empty future technique clearing

RESOLUTION: Accept PR 721

<laura> bye

<alastairc> trackbot end meeting

Summary of Action Items

[NEW] ACTION: Michael Cooper to compile a list of the techniques that were deleted
 

Summary of Resolutions

  1. the approach is agreed-upon and it will be written up
  2. accept pull requests 715
  3. accept PR 693
  4. accept PR 701
  5. Michael Cooper to compile a list of the techniques that were deleted.
  6. Accept PR 721
[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/06/04 16:59:39 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/charis/chairs/
Succeeded: s/2.21/2.1/
Succeeded: s/4.4.1/4.11/
Succeeded: s/thing/think/
Succeeded: s/1.31/1.3.1/
Succeeded: s/algorthen/algorithm/
Succeeded: s/superceeds/supercedes/
Succeeded: s/in the spread sheet/in the survey/
Succeeded: s/simlar /similar /
Succeeded: s/chater/charter/
Succeeded: s/separte/separate/
Succeeded: s/managment/management/
Succeeded: s/panick/panic/
Succeeded: s/responsible abount of time though/will be responsible about of time needed/
Succeeded: s/ignificant/significant/
Succeeded: s/havent/haven't/
Succeeded: s/sentecne for focus visisble/sentence for focus visible/
Succeeded: s/contast/contrast/
Succeeded: s/is was/is/
FAILED: s/pubish rules separtely since a redesign is underway/Publish rules separately since a redesign is underway./
Succeeded: s/orentation/orientation/
Succeeded: s/stemed/stemmed/
Succeeded: s/incuded/included/
Succeeded: s/expicit/explicit/
Succeeded: s/legial/legal/
Succeeded: s/explanes/explains/
Succeeded: s/emaile/email/
Succeeded: s/with a a single /with a single /
Succeeded: s/incorrrectly/incorrectly/
Succeeded: s/in every thing /in everything /
Succeeded: s/separte/separate/
Succeeded: s/desion/decision/
Succeeded: s/yahoo thing/yahoo email/
Succeeded: s/i do know of how to switch enable painting really/i do NOT know of how to switch enable painting really/
Succeeded: s/please add at least one more game/please add at least one more game example to spreadsheet/
Succeeded: s/except/accept/
Succeeded: s/Sean/Shawn/
FAILED: s/sean/Shaun/
Default Present: AWK, Raf, Laura, alastairc, Detlev, bruce_bailey, Chuck, JustineP, stevelee, Fazio, SteveRepsher, MichaelC, david-macdonald, JF, johnkirkwood, mbgower, MarcJohlic, Rachael
Present: AWK Raf Laura alastairc Detlev bruce_bailey Chuck JustineP stevelee Fazio SteveRepsher MichaelC david-macdonald JF johnkirkwood mbgower MarcJohlic Rachael
Regrets: Nicaise Jake
Found Scribe: Laura
Inferring ScribeNick: laura
Found Scribe: david-macdonald
Inferring ScribeNick: david-macdonald
Scribes: Laura, david-macdonald
ScribeNicks: laura, david-macdonald

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 04 Jun 2019
People with action items: cooper michael

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)


[End of scribe.perl diagnostic output]