Accessibility Guidelines Working Group Teleconference -- 14 Dec 2017

<AWK> +AWK

<alastairc> scribe: alastairc

Contextual information: https://www.w3.org/2002/09/wbs/35422/Contextual_Information_review/results

AWK: Not too many responses, 2 happy, 3 with things that need addressing.
... (Summarises Jason's comment), anything else to add?

Jason: I tried to make the comment as clear as I could, and compared with 1.3.1 and point out that this proposal doesn't have the criteria that 1.3.1 does.
... there you have the touchstone of presentation provided, but here you don't have an indication of how-much is required.

<Lisa> +1 to jason suggested

Jason: but having this kind of data available is un-scoped. Generally support the idea of increasing the richness of formal semantics, so agree with goal, but not the proposal.

Lisa: Happy with the suggestion, so 'some contextual information' is provided, could live with that. Would make it testable, something that clarifies content's meaning. At this point most important thing is to get it in, and get in the technques.
... I did create a definition, hopefully that helps.

<Lisa> Contextual Information: Information which clarifies the meaning of an object, the purpose of an object and the object’s relationship with other objects and processes, including identifying the position in a multi step process such as may be identified as breadcrumbs.

<Lisa> publicly available vocabulary : a taxonomy, ontology or controlled vocabulary that is available to the public on the internet such that each entry can be referenced in metadata, such as WAI-ARIA, wordnet and Schema.org

<Lisa> region: A perceivable section containing content that is relevant to a specific, author-specified purpose and sufficiently important that users may likely want to be able to navigate to the section easily and to have it listed in a summary of the page or may want it hidden for simplified versions. Any area that would require a landmark role would be a region.

Lisa: Bit worried about defining regions as it is used in many places already, we might not want to define that.

<Zakim> AWK, you wanted to ask about "some"

Lisa: As this is a complete locked-in situation for AAC users, who are locked into whatever their symbol-set provider has included. It is the be-all / end-all for these groups. Having this in is very important.

<Lisa> bruse , yes it is an example,

AWK: What would be the ideal if someone publishes a webpage, with home & contact us, paragraphs of content. Is the author supposed to markup the home & contact-us links (covered under 1.3.4), but for paragraph content do these tools recognise language and replace it with symbols?

Lisa: Not about translating a language, no requirement on general text. If you have made content full of symbols, then they should be able to transpose those symbols with another set.
... should be interoperable, so they can be translated to the (symbol) language that the user knows.
... for controls we're asking for all user-interface controls, even though you can always add the meta-data from wordnet (which has every known noun), it is AAA. right now everyone has different symbols sets.

AWK: You said "some contextual information".

Lisa: Not some on the page, on every control.

AWK: I worry that we would still need to clarify contrextual information, region, symbols... I get the use case, it's important for users, but we need to be able to specifiy in a way that people can follow and understand when they have met it.

<Zakim> bruce_bailey, you wanted to ask if "such as may be identified as breadcrumbs" is an example?

Bruce_Bailey: Previously when I worked with students, they understood English and read ok, but the writing was a problem. I'm not seeing the utility of this.
... the definitions are ok, but the examples would be better as notes.

Jason: One of my comments was about "symbol" not being defined, it really needs clarification & definition. On Bruce's point, I'm concerned about the utility of this. If, in order to implement, and the author has to implement 3.3.5, then we need to note with a cross reference. Again, I'm concnerned this is not going to work on general websites. Not sure where to go with it, open to discussion, but signifificant issues with lack of tech to make it

all work.

<Lisa> defintion of A symbol is a mark, sign, or word that indicates, signifies, or is understood as representing an idea, object, or relationship. Symbols allow people to go beyond what is known or seen by creating linkages between otherwise very different concepts and experiences.

scribe: not sure what it adds beyond the accessibility supported requirement, may not need the public vocab part? Range of problems with the proposal, like the broad semantics, but need methods with concrete benefits.

Lisa: This group of AAC users, the world is so inaccessible, it is a worst-case in terms of how we have constructed society, the world is not accessible/usable for them.
... prevalent on us to try and reach them. It is AAA, where people have to learn new symbols it is very burdensome. It would enable some people to use github who couldn't use it before.
... Most sites are not adding sign language to their site, so it doesn't apply. Triple-A is often used by people targeting people with disabilities. It would be wonderful for every site to use it, but not needed for AAA.
... If we had a definition of symbol, what is the definition we're happy with.

<Lisa> A symbol is a image that indicates, signifies, or is understood as representing an idea, object, or relationship.

Lisa: Think it needs to be an image rather than words.

AWK: We also need to respond to comments, either with changes or responses.

<Lisa> can you ptu the link into the wiki for the coments

<Detlev> Jake Abma has offered help

<Detlev> in a mail

AWK: Great, thanks Jake. Let's leave this open, Lisa has to go.

Lisa: Would be hapy to have a call on monday.

RESOLUTION: Leave open

Resolving issues: https://www.w3.org/2002/09/wbs/35422/resolving_issues_1/

AWK: Next survey, we have a number where everyone is in agreement. Can we accept the ones where we have unanimous acceptance?
... Any objection to accepting those as agreed?
... No? good

RESOLUTION: Accept 602, 608, 616, 617, 618 as proposed

601

AWK: Core issue was that they saw the old version, then someone else added more comments.
... We can defer until the larger discussion, but we could make a quick decision on this one.

Detlev: Thought this one was related to section 7.1 common control purposes?

AWK: Yep, the commenter was Dec 4th, change in section on Dec 7th. So we could add it (?) to the section.

Detlev: The response doesn't quite explain that context.

AWK: Yep, good point.

<Detlev> you know the sighted kb user argument...

<Zakim> JF, you wanted to ask Jason why one method is preferable to another - isn't outcomes the goal?

Jason: We have the HTML main element and ARIA regions that are more effective than skip to main content links. We should encourage those methods rather than having meta-data to markup their purpose as well. Given that we scope this to markup, we have ways of satisfying the underlying need.

JF: I agree there are mutliple methods, not sure why we should encourage one method over another? More interested in the functional outcome.

Jason: The point was that with ARIA/markup methods, you satisfy the navigation requirement and this new requirement.

JF: It isn't the most elegant technique, but don't care how the developer gets there, choice is good.

AWK: I do think there are some relevent points for this, there is a long list of common link types from the COGA group. This wasn't one of them.
... also concern that skip links are a bit of a hack (some think). I question whether it is an important common purpose? This perhaps isn't worth adding.

<Glenda> I’m neutral

<gowerm> I think skip to links is pretty old school

<ryladog> +1 to Detlev

<Joshue> +1 to Detlev

<JF> +1 to Detlev

<bruce_bailey> +1 to Detlev

<laura> no harm

<Glenda> I still personally recommend use of skip links for keyboard users…so I lean towards detlev

Detlev: Think we should add, skip links still help keyboard users, no point not adding it.

<KimD> I don't think we need to add it to the list

AWK: Adding to the list may open flood gates to more?

<gowerm> If it is hidden from the viewer until tabbed to, a common technique, would adding a purpose make it confusing?

<Detlev> don't advertise that it has been augmented :)

+1 to Detlev, but not hugely concerned either way.

<Zakim> steverep, you wanted to ask about the review implications of adding more items

AWK: We've got more people saying it's worth adding than not.

Steverep: What is the review implications of adding this now, given the public can't review it until CR?

MichaelC: I was expecting that the whole list to be marked 'at risk', but perhaps that wouldn't be suitable.

AWK: If this were the only change then we probably wouldn't change the at-risk status, but if we add 10 it would be a different story.

MichaelC: So long as the whole list isn't removed then the SC wouldn't need to be at risk.

AWK: Comments could cause the SC to be at risk.

<gowerm> The list is normative though, isn't it?

MichaelC: Adding to the list is likely to be less of an issue for reviewing than subtracting.

AWK: Anyone else? Shall we accept the comment, add the item? Or defer response until we have the entire list?

Jason: There were wider issues about how generic the list should be, and it isn't clear to me that we've resolved all those. If we tackle those later we'd have to revisit this one, unless we're confidence the list stays the same.

AWK: Anyone else? We can send it through now, or we can wait until we finish 1.3.4.

MichaelC: Short of a wholesale change of the list, I doubt we'd go back to this one.

AWK: (reads proposed response), and we'll add: Working group will add an item to represent skip-to links.

<AWK> Thank you for your observation. We believe that you refer to section 7.1 Common Control Purposes" https://www.w3.org/TR/WCAG21/#control-purposes. The Working Group will add a item to represent "skip to" links.

<JF> +1 to accept

AWK: Any objections to accepting that as ammended?

<Glenda> +1 to accept

<Mike_Pluke_> +1

RESOLUTION: Accepted as amended.

609

<Detlev> I can scribe

<Detlev> scribe: Detlev

topic is issue 609

<gowerm> None changed the meaning

AWK: Relates Gregg's concerns about which / that discussion (make it clear that the clauses are restrictive)
... will do a pull request to implement changes suggested (changing whoch to that and removing comma)

goverm: another issue where the change is important has now a separate issue

AWK: has been closed (608 redundant)

<Glenda> no objection

AWK> Any objections?

RESOLUTION: 609 accepted as amended

Contextual information: https://www.w3.org/2002/09/wbs/35422/Contextual_Information_review/results

(Josh now chairing telco)

Animation from Interactions: https://www.w3.org/2002/09/wbs/35422/animation_from_interaction_review/results

Josh: Survey question was to provide suggestions to SC and understanding text
... most think it can be made testable
... Laura suggested text change (see survey)
... AWK had concerns

<Joshue> "Animations triggered by a user action have a mechanism to be disabled while still performing the action, unless the animation is essential to function or to the information being conveyed."

Josh: What does group think about Laura's suggestion?

Alastair: Happy with Laura's change

<kirkwood> +1 to laura’s change

Alastair: Interested in alternatives to 'user action' - you need to do soemthing to load a page

<Zakim> gowerm, you wanted to say that we might be able to put in a user agent clause

goverm: user agent animations are not controled by the author and therefore exempted

<JF> +1 to Mike's point

goverm: that could also cover page load animations

Alastair: thinks it may work - concerned it may add exceptions - will consider

<AWK> AWK is sure we've discussed it but still thinks it isn't clear how this differs from 2.2.2 Pause, Stop, Hide

Jason: comfortable if UA animations are excluded

<alastairc> AWK - animation based on scrolling doesn't get caught by 2.2.2

Katie: Aggrees exclusion of UA is good - should be 'must have a mechanism'

<gowerm> Animations triggered by a user action have a mechanism to be disabled while still performing the action, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed.

Josh: clarifies meaning of 'mechanism disabled'

<Zakim> steverep, you wanted to suggest using "functionality"

Steve: use 'functionality' instead of talking about 'user actions'

Alastair: Would cause overlap with 2.2.2

<david-macdonald> Patrick presented the following concern https://github.com/w3c/wcag21/issues/18#issuecomment-320538900

Josh: Steve can you propose a wording - struggling how an author would parse that..

Steve: Alastair, anything but parallax scrolling covered?

Alastair: mouse movements doing things to background

goverm: mentions page load animations

Josh: Steve, please propose alternative text

<gowerm> Animations triggered by a user action can be disabled while still performing the action, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed.

Jason: Animations triggered by user interface components might work

Alastair: Doesn't cover the main problem of animation caused by scrolling

<Zakim> gowerm, you wanted to say Jason's wording is not addressing the same issue, I think

goverm: SC addresses vestibular disorders caused by scolling - Jason's wording is not addressing the primary concern

<laura> parallax scrolling is the primary use case

<alastairc> Example: If you have sight and plenty of bandwidth, try https://www.apple.com/mac-pro/

goverm: If the author is not adding animations then the SC is satisfied

<Glenda> +1 to exempting the user agent..and focusing on author content that makes people dizzy

goverm: transitional animations can be removed without changing functionality - so that should be possible to disable

<laura> +1 to MG’s text: “Animations triggered by a user action can be disabled while still performing the action, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed.”

David: Early concern was fading type animations that are not a problem - has an amendment been agreed to focus on motion?

<Joshue> https://github.com/w3c/wcag21/issues/18#issuecomment-320538900

<alastairc> That come to: "For motion or scaling animations triggered by a user action that are not an essential part of the action, there is a mechanism for the user to pause, stop or hide the animations while still performing the same action."

<david-macdonald> https://github.com/w3c/wcag21/issues/18#issuecomment-320605052

Alastair: has proposed motion or scaling animations - but this is not part of the current language

Josh: David do you suggest to add that?

<steverep> I now remember a long mailing list thread on that too

David: yes, scope to movement and scaling

<gowerm> For motion or scaling animations triggered by a user action that are not an essential part of the action, there is a mechanism for the user to pause, stop or hide the animations while still performing the same action.

Josh: agrees with suggestion

Jason: a way to exclude UA animations would be to say "animations provided by the content" - different ways to handle the edge cases

<Zakim> Joshue, you wanted to say why not remove the 'triggered by user action" bit

Josh: removing "triggered by user action" might work and remove ambiguity, also scoping to motion and scaling type animations

<laura> "Motion or scaling animations triggered by a user action and provided by the content can be disabled while still performing the action, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed."

Alastair: Change would mean significant overlap with 2.2.2 Pause Stop Hide
... no intention to prevent dropdowns from working, just ensuring that animations around that can be turned off

<gowerm> I can live with Laura's latest posted version at 9:47

Alastair: Like what is possible if you disable that in iOS

<Joshue> For non-essential animations such as motion and scaling triggered by a user action, there is a mechanism to disable the animations yet still perform the action.

Josh: Better to make it clear what the main thrust of the SC is

Alastair: Text nees to be rearranged but all points are there

Detlev (off scribe) too long!

<gowerm> you can remove "and provided by the content"

<alastairc> 'provided by the content' and 'not modified by the author' are not both needed

Josh: not sure about "provided by the content"

<gowerm> +1 to Alastair

Josh: seems consensus to include the scope on motion and scaling

<alastairc> "Motion or scaling animations triggered by a user action and provided by the content can be disabled while still performing the action, unless the animation is essential to function or to the information being conveyed."

Katie: likes 'provided by content', makes it clearer

<gowerm> "Motion or scaling animations triggered by a user action can be disabled while still performing the action, unless the animation is entirely determined by the user agent or is essential to function or to the information being conveyed."

Josh (reformulating)

<Joshue> For non-essential animations such as motion and scaling provided by the content, there is a mechanism to disable the animations yet still perform the action.

Josh: When animations are part of the chrome?

Alastair: not if its UA-provided

<david-macdonald> user action and provided by the content can be operated without Motion or scaling animations, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed."

Katie: would specify 'non user agent animations'

<Glenda> I think “understanding” can help with people groking it

<david-macdonald> user actions provided by the content can be operated without Motion or scaling animations, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed."

Jason: Same thing, different expression

Alastair: 'user action' makes no sens in this context

<ryladog> non user agent generated animations

Josh: wants it terse yet clear

<Zakim> gowerm, you wanted to say 'Motion or scaling animations triggered by a user action can be disabled while still performing the action, unless the animation is entirely determined by

<gowerm> Motion or scaling animations triggered by a user action can be disabled while still performing the action, unless the animation is entirely determined by the user agent or is essential to the function or information being conveyed.

David: why not 'functionality provided by the content'?

Can you past that, Mike?

<david-macdonald> "content can be operated without Motion or scaling animations, unless the animation is determined by the user agent and not modified by the author or is essential to function or to the information being conveyed."

<alastairc> works for me

<Glenda> +1 to gowerm

<laura> +1

<marcjohlic> +1

<Joshue> Motion or scaling animations triggered by a user action can be disabled while still performing the action, unless the animation is entirely determined by user agent

<Glenda> gowerm’s text at 11:55am

<Glenda> “gowerm: Motion or scaling animations triggered by a user action can be disabled while still performing the action, unless the animation is entirely determined by the user agent or is essential to the function or information being conveyed.”

<marcjohlic> +1 to gowerm's text

<JF> Definition of "scaling animations" - do we have one?

<KimD> *have to run

Mike: confused about 'essential to function' - verb or noun?

Mike: change to 'the function'?

<alastairc> do we need anything after essential?

JF: do we have a def of scaling animations? Term not familiar, should be defined

<gowerm> Detlev, we need to allow things like drag and drop

Alastair: Dont have one right now but James Craig has posted something, so can be done

<gowerm> We can come up with a defintion of motion or scaling animations

Alastair: 'changes shape or moves across the page' (quoting)

<JF> @Alastair - fair enough, we just need a glossary definition as part of the SC

David: disabling 'while using it' may not be the only option - can be doen beforehand

<ryladog> Motion or scaling animations triggered by non user agent generated content can be disabled while still performing an action, unless the animation is entirely is essential to the function or information being conveyed.

<gowerm> the "while" was bothering me too

Josh: Someone has to put together the different suggestions to get the definitive wording - Alastair?

Alastair: will do.

<gowerm> Motion or scaling animations triggered by a user action can be disabled without affecting the action, unless the animation is entirely determined by the user agent or is essential to function or to the information being conveyed.

RESOLUTION: Leave open, Alastair will propose amended text

Josh: Please pick up issues on github and formulate responses

<Joshue> Please look at GH https://github.com/w3c/wcag21/issues

Josh: Do not respond directly in Github, formulate response in Wiki

<laura> https://www.w3.org/WAI/GL/wiki/Draft_Responses_to_Dec_WD_Issues

<Joshue> trackbot, end meeting

- DRAFT -

Accessibility Guidelines Working Group Teleconference

14 Dec 2017

Attendees

Contents

Contextual information: https://www.w3.org/2002/09/wbs/35422/Contextual_Information_review/results

Resolving issues: https://www.w3.org/2002/09/wbs/35422/resolving_issues_1/

601

609

Contextual information: https://www.w3.org/2002/09/wbs/35422/Contextual_Information_review/results

Animation from Interactions: https://www.w3.org/2002/09/wbs/35422/animation_from_interaction_review/results

Summary of Action Items

Summary of Resolutions

Scribe.perl diagnostic output