W3C

- DRAFT -

Accessibility Guidelines Working Group Teleconference

10 Oct 2017

See also: IRC log

Attendees

Present
AWK, JakeAbma, lisa, jasonjgw, Joshue108, alastairc, steverep, AndyHeath, Roy, Brooks, allanj, Melanie_Philipp, Detlev, Alex, Kathy, MichaelC, Mike_Elledge, marcjohlic
Regrets
Chair
AWK
Scribe
melledge, allanj

Contents


<AWK> +AWK

<AWK> Survey on Essential cases: https://www.w3.org/2002/09/wbs/35422/essential_breakout/results (only respond to 6, 8, and 10)

<AWK> Implementation process (https://www.w3.org/2002/09/wbs/35422/WCAG21_impl/results) - this was already surveyed but people wanted more time.

<AWK> Github practices (linking and replying)

<AWK> Discussion on Failure techniques

<AWK> AGWG Work Items progress check in and sign-ups: https://github.com/w3c/wcag21/issues?q=is%3Aissue+is%3Aopen+label%3A%22AGWG+Work+item%22

<interaccess> trackbot, start meeting

<trackbot> Meeting: Accessibility Guidelines Working Group Teleconference

<trackbot> Date: 10 October 2017

<interaccess> Chair: AWK

<AWK> https://www.w3.org/WAI/GL/wiki/Scribe_List

<AWK> Scribe:melledge

Professional expectations

AK: Addition topic. Lots of emails. Everyone is dedicated and passionate. Sometimes we don't agree. Generally good about staying on topic and being civil.

<david-macdonald> monitoring this meeting while at a conference

awk: Some people will read late and have to be deliberate about focusing on technical details and driving the discussion forward.
... If it isn't going that way, chairs may not have noticed, reach out to the chairs. Remember we're all doing this for the right reasons.
... Ask "is this what you mean?" All want best standards. :^)

Survey on Essential cases: https://www.w3.org/2002/09/wbs/35422/essential_breakout/ (only respond to 6, 8, and 10)

awk: Could spend multiple calls if we let us do that. Time restriction: 10 minutes for each. To see if we're near or at consensus. Keep comments brief, don't need to restate.
... keep comments to a minute, pls.

In content on hover or focus

awk: Change is proposed for SC uses the term essential. Essential content will not be obscured. Proposal to add "other than pure decoration" and remove "essential".
... Do have a definition of "pure decoration." Concern about that and "essential". Alister concern about restricitions on placement and Mark as well.
... Should include tool tips as well.

<allanj> In WCAG20 def pure decoration -serving only an aesthetic purpose, providing no information, and having no functionality. Note: Text is only purely decorative if the words can be rearranged or substituted without changing their purpose.

sr: I guess my thought that we'er not using, changing "essential" doesn't makes sense with current definition. Would have to re-phrase as an exception.

jason: Wondering when triggering is partially obscured that don't lose functionality. Should we focus on functionality of component? Agree that essential is not right term. Could do better.

awk: Instead of obscuring content, remove "access to functionality".

sr: Not talking about content here. Need to characterize what we don't want to obscure.

<Zakim> steverep, you wanted to say loss of visibility is the issue - if the border is what makes it clear, then don't block it or let me close it

alastair: It is quite narrow in scope, has to be triggered by hover or focus, non-decorative content within the trigger. What examples are there?

sr: We can't really talk about functionality. Trying to see the trigger. In the case of border, if pure decoration, then block it. If not, then don't block it.
... Two options: to put on side or ....very easy to solve.

? Missing some pattern that's being used. If you look away from it should go away. Something that pops up and goes away.

<Ryladog> +1 to Marc and Steve

awk: When a keyboard trigger causes content to disappear, have to make sure it can be repositioned.

<marcjohlic> "Either the additional content does not obscure any content within the triggering user interface component, or the additional content can be closed or repositioned by the user;"

<steverep> May I respond to that???

alex: location, location, location. This is about what is covered where there is no room. Can't go on other side of the screen. Different types of triggers need flexibility.

sr: Yes have to put it somewhere. If you have to obscure some part of trigger than give me option 2 to close it.

<marcjohlic> +1

awk: Need to identify if there are any examples to show where it would be a problem. Then is there a possibility to satisfy the visible bullet aspect bec content can't be closed.

katie: Happened to me this morning. Form, popup obscured part of form.

<steverep> Probably not.... imagine you were magnified

<marcjohlic> +1 hard for everyone - horrible design

Katie: Could see as problem for low vision person.

<alastairc> I don't think it would be covered unless it was triggered by hover/focus, and covered the trigger.

<Detlev> I accept whatever solution gets us out of this hole quickly...

awk: Over our ten minutes. Think we've got core question: not hearing support for pure decoration aspect. Seems to be whether essential should be there.
... Need some examples.

Katie: I'll put in url

<alastairc> Still can't see what is wrong with: "Either the additional content does not obscure  content within the triggering user interface component, or the additional content can be closed or repositioned by the user;"

<Ryladog> Is this an example of this: www.trainingcentertechnologies.com/MedicResponse/studentSignUp.aspx?PKID=-1&CTYID=302&PKORD=-19804

ja: there is an exception where user agent controls presentation. We're only talking about author here. If browser tool tip obscures, it's a browser issue or operating system. Need to consider whether it's an author issue. OS doesn't count.

awk: Need to find some realistic concerns.

<alastairc> Ryladog: no, that is persistent rather than triggered by hover/focus.

<Zakim> steverep, you wanted to say I don't think we should be deciding if there is consensus the world if flat, not round

awk: real word examples.

Implementation process (https://www.w3.org/2002/09/wbs/35422/WCAG21_impl/) - this was already surveyed but people wanted more time.

<Ryladog> The example aboce requires the user to either scroll or resize the page

<AWK> RESOULTION: Leave open

RESOLUTION: Leave open

<Ryladog> Thanks Alistair

in graphics contrast

awk: Started out with graphical contrast in content at 4.5:1. Replace "essential" with "necessary." 3 likes, 7 add'l discussion.
... Kathy: necessary seems to have same problem as essential. Couple of requests to explain why essential is wrong.

steve: So for this one, really difficult to substitute definition, because describing a graphical object. First part also doesn't seem to make sense. already saying essential for content.
... Then second part hard to see how it applies.
... Not sure it makes sense. Need a synonym, a carry over from defining a graphical object. Necessary first came. up.

<alastairc> As SC manager, happy with either "necessary" or "required".

alastair: More concrete examples? Icons that describe to us how does essential not work.

steve: don't want to discuss SC, but point is that you try to substitute essential how does it fit. "Can't be achieved in any other way."

alex: What part does not apply?
... Can't replace head part of icon? Put on some hair?

<alastairc> Please see the understanding doc, temp location here: https://alastairc.ac/tmp/graphics-contrast.html

steve: No, not saying that.

alex: No other way to draw figure? Don't understand?

<Detlev> replace term 'essential' with the definition in WCAG 2.0 - that concept is not clear to all right now

<allanj> WCAG20 Def - essential - if removed, would fundamentally change the information or functionality of the content, and information and functionality cannot be achieved in another way that would conform

alastair: Essential is defined if another way to meet than not essential. Doesn't apply in this case. Happy with acquired, necessary have the right meaning.

Alex: Another way to replace it?

Alastair: Can replace graphic with text. But probably wouldn't want to do that.

awk: Examples?

alastair: Include text as well, then graphic isn't required for understanding. for example want would be required for magnet icon. Would have to have contrast. don't think we need essential.

alex: acquired or necessary could be used. Do we want to be specific? Where pure decoration wouldn't need to have contrast ratio. Are there other categories where it wouldn't apply as well?
... more complex objects, not sure how it would work.

josh: Question about definition. A graphic that can be replaced by something else is not essential.

<Zakim> steverep, you wanted to agree with Jason, but an issue for another day

alastair: Required for understanding works just as well.

<alastairc> Current: "graphical objects that are essential for understanding the content or functionality..." works with required.

<Joshue108> thanks Alastair and Steve..

<alastairc> separate issue

steve: your point is direct. If we use essential, we have to talk about somehting that can only be replaced in a certain way.

awk: essential at the beginning?

steve: exception just adds a link. Essential for understanding is what we're discussing.

<Ryladog> +1 to Steve

awk: you're proposing adding link, that essential is the problem.

alex: current wording is "necessary."

<alastairc> Is anyone objecting?

alastair: I changed it.

awk: Were others misunderstanding as well?

<kirkwood> really oops yes

<kirkwood> sorry

<kirkwood> ;)

awk: In looking at this, compare this to 1.4.3 with exceptions that apply. So, this seems more palatable.

mg: Question of clarity. Confirm whether or not essential used in wcag 2.0, only use is going to be according to our definition.
... Can we use essential without it being our definition?

<steverep> +1 to AWK... no reason to take the risk

awk: Not sure. Up to lawyers. Shouldn't use it in two different ways and link to only one definition. Use a different word.

<steverep> Straw poss on "necessary" vs. "required"?

jason: Replace essential with required.

<Detlev> +1 for required

<Ryladog> +1 to required

awk: If changing in front have, would anyone have trouble replacing essential with required.

<alastairc> +1 for required, 1st part only.

alex: If decoration covers our needs, should use it. Otherwise be more precise than "required."

<Rachael> +1 for required as long as we use it as a consistent alternative

awk: Some subjectivity, but...

alastair: Understanding doc makes it hard to define(?)

awk: Put forward to the list. Not a lot of objections to change "essential" to "required." Objections to moving to call for consensus.

<Detlev> we HAVE a word

mike gower: Alastair should find word.

alex: My proposal to use decoration as criteria also objection to required. Excluding graphical object that doesn't fall into category.

<gowerm> +1. Jason's wording sounds fine to me

awk: Graphical objects that are not pure decoration. Don't feel strongly, but wanting to be more precise. Want to do analysis first.
... Wordsmithing...over target time.
... Jason and Alastair take on wording and sent out for comment. Headign in right direction.

alastair: Will do.

RESOLUTION: Alastair and Jason to wordsmith and send to list

awk: Number ten
... User interface contrast. Replacing essential. Most on board. One more discussion.
... Seems wide open if don't ahve soem aspects that aren't critical. Is there something that gives this wiggle room? If we have UI control that doesn't have more than 3 colors in it, virutally impossible to meet requirements.

alex: My issues i that the visual identifier is not being defined. Take out essential. Then apply definition to visual identifier. Need one so people will know what it covers.

<Zakim> steverep, you wanted to say if it's identifying then it's essential (that's already being covered in Understanding)

alastair: Taking out essential okay. Can add definition.

<allanj> 1.4.12 NOTE Examples of visual identifiers of user interface components may include (a border, edge, or icon), current value (such as non-text visual indication of aria-value now on a slider) and current state (such as selection indicator, focus indicator) or other visual indication (which do not rely on color alone). <-- could be used as a start for definition

steve: Think Jason's point has merit. My logic for removing it was if use "identifier", already is essential.

<alastairc> Create a new github issue for making the note into a definition?

awk: Similar to last one, except hinging on identifiers. If something is used to identify that control than needs to meet color contrast ratio.
... Does anyone think essential is essential, and why?

<allanj> Glenda is happy to not use word essential

<steverep> None - the proposal is to take it away

<alastairc> Worth noting that 'flat design' which has no visual identifiers will pass this.

? If we remove it?

awk: are there visual aspects that identify a "blue" button as a distinction?

steve: needs to be 4.5:1. not just a part.

awk: Printer icon for example. A bunch of colors involved, then question is which parts are the visual identifiers.

Mike Gower: For clarity there is no use of essential here. Don't think we need it. Visual indentifier needs to meet contrast.

<alastairc> Current (with essential): https://www.w3.org/TR/WCAG21/#user-interface-component-contrast-minimum

lisa: Don't want to make it harder for other groups to know what to design.

alastair: Issue: 1) whether there are visual identifiers, then 2) Whether it has contrast. Intentionally saying if you have them, they should have contrast.

<lisa> concern that people will get rid of borders now making it harder to use

<Zakim> steverep, you wanted to ask how removing essential changes that question?

jason: Note that follows the text. Does provide examples, does describe identifiers. Best to say visual identifiers without essential then link to defintion.]

<lisa> i think borders should have a minimum contrast. you shouldnt get out of them by removing them altogher

<Rachael> +1 to jason's suggestion

steve: What is a "visual identifier"? is the key question. Essential doesn't clarify.

<alastairc> lisa: then we would be heading into AAA territory as it impacts the design choices.

awk: I'm hearing that the confusion is on both. Agree doesn't matter if we say essential. Visual identifier si what we have to think about.

alex: MS Word. Ribbon where you decide how to highlight. Whatever color it is, it is essential. A user's choice. But the color is essential and a user decision. In that case
... might have to say 4.5:1 is the decision, but will ahve to meet it.

<steverep> No, it's not "essential" per the WCAG definition - that's the point.

awk: Highlights or text font?

<alastairc> but the indicator would be the line around the button, not the icon itself

alex: Yes. color is important, critical. other stuff not so much.

awk: Steve is saying that the color that the highlighter show is not critical it's the ?
... color is important in other way. Waht color you'll get when you click on it. So is down arrow that gives you color choices.

<Zakim> steverep, you wanted to offer a ribbon opinion

steve: Slipping bakc into dictionary definition, not WCAG defnition. Doesn't make sense with WCAG definition.

alex: How else to meet SC criteria.

alastair: If we're looking at ribbon, line around is what we're talking about for visual indicator and contrast.

alex: No contrast there.

awk: Need a battery of different images that are out there to see what this SC is talking about. There is a core problem with understandability and therefore implementability of SC.

<kirkwood> would this SC eliminate “flat design”

<alastairc> Glenda has already done that, at least to some extent

awk: Who would like to take this and find examples so we can grapple with it.

crickets

<alastairc> I've been gathering icons & examples for graphics contrast.

alex: Will do that.

<gowerm> +1 there's already an Understanding doc

awk: Will enable us to make sure we agree with SC.
... Who's been working on understanding part of doc.

alastair: Glenda

awk: Alex and Alastair look into it?

alex: Send a bunch of graphics to Glenda.

awk: And engage with her. lol

<allanj> scribe: allanj

RESOLUTION: Leave open Alex and Alastair to clarify.

RESOLUTION: Leave open Alex and Alastair to clarify.

<gowerm> +1 Get rid of essential

<melledge> thanks allan. Didn't realize we were switching.

awk: are there residual concerns about removing 'essential;

al: no comment till resolve image issue

RESOLUTION: leave open, discuss 'essential' later

<steverep> AWK, was there a resolution on Adapting Text?

brooks: ... graphic of ! with text of form error. this ! is not essential. how to parse 'essential' to rule out !

Github practices (linking and replying)

Implementation process (https://www.w3.org/2002/09/wbs/35422/WCAG21_impl/) - this was already surveyed but people wanted more time.

awk: surveyed in Aug. reviewing now
... as part of process, need 2 independent implementations of each SC

<AWK> Implementation process for WCAG 2.0: https://www.w3.org/WAI/GL/WCAG20/implementation-report/

<Joshue108> Implementation process

awk: do we want to adopt this... or do something else

sr:

https://www.w3.org/2002/09/wbs/35422/WCAG21_impl/results

mc: 2 is minimum. more is better. if we test 5 and only get 2, then need to see where the problem is

<steverep> I don't think that's what I meant.... checking my comments now

<Brooks> One context to consider when discussing color contrast for user interface controls and/or graphics, is when you have an exclamation mark icon with low color contrast paired with form validation error text. The icon is critical to some users to recognize the error, but it might not be considered "essential" when there is error text close by to supplement the meaning of the exclmation icon itself.

awk: in 2.0 we had 2 implementations and 10 illustrative websites with positive implementations

mc: only need to consider the new SC
... in CR conformance, we may need to consider that meeting a 2.1 SC makes it harder to meet a 2.0 SC

jason: in 2.1, use 2.0 content and test against 2.1. shouldn't be hard. need to be rigorous. ne

mc: agree, mostly. have a timeline. exercise the new SC as much as we can. need to get to REC. focus on CR 2 implementation bar.

khs: agree

awk: will discuss this on Thursday.
... other suggestions... remove 10 sites. focus on just 2 implementations.
... are other proposals for CR

Thursday will be 1 hr call from Noon eastern to 1 pm

<Joshue108> thanks all

trackbot, end meeting

Summary of Action Items

Summary of Resolutions

  1. Leave open
  2. Alastair and Jason to wordsmith and send to list
  3. Leave open Alex and Alastair to clarify.
  4. Leave open Alex and Alastair to clarify.
  5. leave open, discuss 'essential' later
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/10/10 16:32:09 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
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/Lots of things are triggers. Have to talk about things within the triggering mechansim. Is this in scope?/It is quite narrow in scope, has to be triggered by hover or focus, non-decorative content within the trigger. What examples are there?/
Succeeded: s/? /ja:/
Succeeded: s/ Issue: 1) if tehre are visual indentifiers, then 2) how to meet SC. Intentionally saying if you have them./ Issue: 1) whether there are visual identifiers, then 2) Whether it has contrast. Intentionally saying if you have them, they should have contrast./
Default Present: AWK, JakeAbma, lisa, jasonjgw, Joshue108, alastairc, steverep, AndyHeath, Roy, Brooks, allanj, Melanie_Philipp, Detlev, Alex, Kathy, MichaelC, Mike_Elledge, marcjohlic, david-macdonald, Pietro, Katie_Haritos-Shea, Makoto, MikeGower, kirkwood

WARNING: Replacing previous Present list. (Old list: (no, one))
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK


WARNING: Replacing previous Present list. (Old list: AWK, JakeAbma, lisa)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, JakeAbma


WARNING: Replacing previous Present list. (Old list: AWK, JakeAbma, lisa, jasonjgw, Joshue108, alastairc, steverep, AndyHeath, Roy, Brooks, allanj, Alex, Melanie_Philipp, Detlev, Kathy, MichaelC, Mike_Elledge, marcjohlic, david-macdonald, Pietro, Katie_Haritos-Shea, Makoto, MikeGower, kirkwood)
Use 'Present+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Present+ AWK, JakeAbma, lisa, jasonjgw, Joshue108, alastairc, steverep, AndyHeath, Roy, Brooks, allanj, Melanie_Philipp, Detlev, Alex, Kathy, MichaelC, Mike_Elledge, marcjohlic

Present: AWK JakeAbma lisa jasonjgw Joshue108 alastairc steverep AndyHeath Roy Brooks allanj Melanie_Philipp Detlev Alex Kathy MichaelC Mike_Elledge marcjohlic
Found Scribe: melledge
Inferring ScribeNick: melledge
Found Scribe: allanj
Inferring ScribeNick: allanj
Scribes: melledge, allanj
ScribeNicks: melledge, allanj
Found Date: 10 Oct 2017
Guessing minutes URL: http://www.w3.org/2017/10/10-ag-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]