<trackbot> Date: 12 June 2015

<richardschwerdtfeger> chair: Fred

<richardschwerdtfeger> RRSAgen, make log public

<richardschwerdtfeger> meeting: SVG Accessibility Task Force

<LJWatson> meeting: SVG A11y TF

<LJWatson> Chair: Fred Esch

<LJWatson> scribenick: LJWatson

The <use> element

RS: When you take <use> you can take a section of the DOM.
... Under the covers the shadow DOM needs to be mapped to the accessibility tree.
... Chrome is the only browser to implement tabindex support so far.

DS: In Describlr the navigation tree had to be created manually.
... This is about focus though, not <use>.

RS: No, it's bigger - you need to make sure things are exposed to the acc tree.

DS: I disagree. We discussed this at length last week. The <use> element isn't intended to be used for complex constructs - only single things.

RS: But you can can't you?

DS: It's unclear in the current spec.
... If there are two elements in an icon, and mousing over each caused something different to happen... different behaviour on each.
... If I put them in a group and re-used that, the behaviour would not persist.
... I don't believe the shadow DOM is exposed in this way, and I don't think it should be.

RS: Do we say that anywhere in the guidance?

DS: I don't know that we say it anywhere.

ABR: There are discrepancies between the spec and implementations, and between implementations.
... If you click on a complex icon you can figure out which thing was clicked on, but only the composite thing can react.

RS: If a person interacts with an SVG drawing and is given some functionality, and someone who is impaired does not get the same functionality, that's a failure.

ABR: SVG should adopt Web Components, but Web Components are not <use> elements.
... You use a template, but there is encapsulation. Each web component has its own DOM that can react differently from any other instance.

RS: Amelia can you take an action to look at the SVG spec to get this cleaned up?

<scribe> ACTION: Amelia to look at SVG spec for references to <use> element, to make sure there is no confusion about implementation/accessibility. [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action01]

<trackbot> Error finding 'Amelia'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.

RS: Can something copied with <use> be interacted with?

DS: If I copied a button, it would be interactive in the sense it could be clicked on to be activated, but I could assign different behaviour to the cloned button.

RS: I believe it's treated as a group by default.

AVR: New roles like icon and/or symbol will have fallback roles.

ABR: Looks like the spec has been cleaned up.

DS: Has it been cleaned up, or just it just say "this thing is opaque".
... Agree that <use> isn't consistently implemented.

ABR: We need to follow up with the main WG.

<richardschwerdtfeger> scribenick: Rich

<shepazu> (Created ACTION-3812 - Look at svg spec for references to <use> element, to make sure there is no confusion about implementation/accessibility. [on Amelia Bellamy-Royds - due 2015-06-19].)

<richardschwerdtfeger> ACTION: Rich make img the default role for <use> [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action02]

<trackbot> Created ACTION-1649 - Make img the default role for <use> [on Richard Schwerdtfeger - due 2015-06-19].

Role Taxonomy

<richardschwerdtfeger> Fred: When we left 2 weeks ago we had talked about icon and we were starting to talk about whether we would want symbol also

<richardschwerdtfeger> Amelia: I had initially proposed one role that would encompass a complex graphic that represents a simple concept in many cases.

<richardschwerdtfeger> Amelia: We discussed whether it made sense to use the same terminology to define symbols on a graphic vs. a chart.

<richardschwerdtfeger> Amelia: Even if common languages has different names for these, I think it is unnecessary to have 2 roles as we have to add terminology to define which context the role is valid in.

<richardschwerdtfeger> Amelia: The context should determine the context

<richardschwerdtfeger> Fred: I like the term symbol as it is used in the graphics domains all the time. When ever you google things you can google electical symbols vs. electrical icons

<richardschwerdtfeger> Fred: I definitely think there is value in telling ATs that there is value in telling the AT that these are symbols

<richardschwerdtfeger> Lioni: I see value in having both symbols and icons. There is quite a different to the 2 things in my mind

<richardschwerdtfeger> Fred: If you are trying to teach kids what the nomenclature is they can’t be referring to them as icons

<richardschwerdtfeger> Jason: There is interest in the role description property

<richardschwerdtfeger> Jason: I think it is an important issue

<richardschwerdtfeger> Jason: This is a larger ARIA issue

<richardschwerdtfeger> Doug: we can differentiate between Symbol and icon. Are we saying that a user agent or an AT uses the same word that we use for the role in order to impart the meaning to the user?

<richardschwerdtfeger> Doug: Finish is not the same as finnish

<richardschwerdtfeger> Doug: at some point I agree there is a distinction on what stands as a symbol vs. a functional thing

<richardschwerdtfeger> Doug: I don’t know that we should keep dwelling on this

<richardschwerdtfeger> Jason: there are important issues about using the ARIA role name

<richardschwerdtfeger> Jason: what we should distinguish here is that the different types can be treated differently.

<richardschwerdtfeger> Lioni: To the best of my knowledge screen readers will take guidance from the role attribute

<richardschwerdtfeger> Lioni: We need to make sure we correctly define what the thing is

<richardschwerdtfeger> Doug: I agree but the name that we choose for the role may not be what the AT uses

<richardschwerdtfeger> Doug: I can see where Apple says we have a better name for our users

<richardschwerdtfeger> Fred: Shere this is a domain for dpub or epub we can choose the name in the context of the name

<richardschwerdtfeger> Lioni: you tend to describe something by its role

<richardschwerdtfeger> Amelia: sounds reasonable. There are cases where people are already mapping to roles that are not in ARIA

<richardschwerdtfeger> Amelia: it is not guaranteed.

<richardschwerdtfeger> Amelia: my main concern is the distinction between symbol or icon

<richardschwerdtfeger> Amelia: if it is interactive it should have a widget role

<richardschwerdtfeger> Amelia: so, we should avoid the word icon and just use Symbol

<richardschwerdtfeger> Fred: a symbol you can put an action on

<richardschwerdtfeger> Fred: I would not prohibit doing an action on a symbole

<richardschwerdtfeger> Rich: only one role is mapped at a time

<richardschwerdtfeger> Rich: there is an aria-interactive role defined in ARIA 1.1

<richardschwerdtfeger> Doug: icon does not say what it is doing

<richardschwerdtfeger> Rich: we could subclass icon if we wanted to.

<richardschwerdtfeger> Lioni: Icon is a thing in digital interfaces

<richardschwerdtfeger> Amelia: perhaps having an icon button being a subclass of button as represented as a graphical display. I think this is worth having a role for

<richardschwerdtfeger> Doug: Right. But then we could have a symbol for a non-interactive graphic wether it is in a map or a chart

<richardschwerdtfeger> Doug: You could always use a different aria role for things

<richardschwerdtfeger> Jason: you could also use an aria role description for a button to say that it is an icon

<richardschwerdtfeger> Doug: can I try to persuade you for a moment?

<richardschwerdtfeger> Doug: when I say click on the trash icon do you look for the trash button?

<richardschwerdtfeger> Jason: yes

<richardschwerdtfeger> Doug: perhaps for the average user there is more clarity by calling it an icon

<LJWatson> +1 to clarity for users.

<richardschwerdtfeger> Jason: I think they get most of their information from interacting with it

<richardschwerdtfeger> Doug: May I suggest a path forward but we put an icon role out there asking for feedback

<richardschwerdtfeger> Fred: I agree

<richardschwerdtfeger> Leoni: adding a role for icon is always going to be the path of least resistance.

<richardschwerdtfeger> Leoni: making authors have to include a role of “buttion” with a role description of “icon” will not make them happy

<richardschwerdtfeger> Ameila: Is there anything else on that topic?

<richardschwerdtfeger> RESOLUTION: Use Icon as an Icon button Role which would be a subclass of the button role.

<richardschwerdtfeger> RESOLUTION: Amelia,Fred, will be initial editors of the document and we will add as we gain contributors

<richardschwerdtfeger> ACTION: Lioni set up github folder structure uner ARIA for the Graphics module [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action03]

<trackbot> Error finding 'Lioni'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.

<richardschwerdtfeger> ACTION: LJWatson set up github folder structure uner ARIA for the Graphics module [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action04]

<trackbot> Error finding 'LJWatson'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.

<richardschwerdtfeger> ACTION: Leeonie set up github folder structure uner ARIA for the Graphics module [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action05]

<trackbot> Error finding 'Leeonie'. You can review and register nicknames at <https://www.w3.org/WAI/PF/Group/track/users>.

<shepazu> ACTION: Leonie set up github folder structure uner ARIA for the Graphics module [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action06]

<trackbot> Created ACTION-1650 - set up github folder structure uner aria for the graphics module [on Léonie Watson - due 2015-06-19].

<richardschwerdtfeger> ACTION: Leonie set up github folder structure uner ARIA for the Graphics module [recorded in http://www.w3.org/2015/06/12-svg-a11y-minutes.html#action07]

<trackbot> Created ACTION-1651 - set up github folder structure uner aria for the graphics module [on Léonie Watson - due 2015-06-19].

<Zakim> AmeliaBR, you wanted to talk about symbol & role-description for shapes

Symbol

<richardschwerdtfeger> Amelia: Symbol is an atomic non-interactive graphic

<richardschwerdtfeger> Amelia: So a symbole will probably be the logical basic role for basic shapes

<richardschwerdtfeger> Amelia: this is a diamond, this is a heart

<richardschwerdtfeger> Amelia: The symbol should not be for any large graphic but that would include basic shapes

<richardschwerdtfeger> Amelia: It would follow the existing pattern

<richardschwerdtfeger> Amelia: it does not have a label on it you would not mention it at all

<richardschwerdtfeger> Amelia: if you have a label on that shape you would call it a symbol

<richardschwerdtfeger> Rich: so if you just have a circle with a label do you really want this to be a symbol?

<richardschwerdtfeger> Amelia: shape, ellipse, and path are details of the implementation

<richardschwerdtfeger> Amelia: a symbol can never have a child

<richardschwerdtfeger> Amelia: right now all these shapes will map to group

<richardschwerdtfeger> Amelia: the only things it can have in the tree are animate, ...

<richardschwerdtfeger> Amelia: there is no child text

<richardschwerdtfeger> Amelia: they can have implicit roles of role description circle

<richardschwerdtfeger> Rich: James Craig at Apple had wanted circle, ellipse, etc.

<richardschwerdtfeger> Jason: it seems reasonable to define roles for the pre-defined shapes

<richardschwerdtfeger> Jason: I would not be a proposed to a shape role either

<richardschwerdtfeger> Jason: I think we have several different options for treating the issues

<richardschwerdtfeger> Doug: So, Jason If you heard that something had the role of path what would you think?

<richardschwerdtfeger> Doug: 90% of all elements used in SVG are all paths

<richardschwerdtfeger> Doug: rectangles, circles, and ellipses are typically either a path or a polygon

<richardschwerdtfeger> Doug: if it has curves it is probably a path

<richardschwerdtfeger> Doug: most authoring tools use paths vs circle, etc.

<richardschwerdtfeger> Doug: Yoking it to the thing that it is a not very valuable

<richardschwerdtfeger> Doug: If we want people to know there is a circle on the screen it should probably be in shape or path with a roledescription that was circle

<richardschwerdtfeger> Fred: you could actually draw the entire symbol with a path and visually look the same.

<richardschwerdtfeger> Fred: when you are losing about basic shapes, … I don’t hear shapes used very often

<richardschwerdtfeger> Doug: other than when you are talking about a shape as a shape you need to be more explicit

<richardschwerdtfeger> Rich: Lets put symbol in the spec. and get feedback. Also state that we considered shape

<richardschwerdtfeger> RESOLUTION: We will have a symbol role that includes atomic non-interactive content that also encompasses the basic shapes

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/making them do/making authors have to include/
Found ScribeNick: LJWatson
Found ScribeNick: Rich
WARNING: No scribe lines found matching ScribeNick pattern: <Rich> ...
Inferring Scribes: LJWatson, Rich
Scribes: LJWatson, Rich
ScribeNicks: LJWatson, Rich
Present: Fred Esch Rich Doug Schepers LJWatson Amelia BR
Found Date: 12 Jun 2015
Guessing minutes URL: http://www.w3.org/2015/06/12-svg-a11y-minutes.html
People with action items: amelia leeonie leonie lioni ljwatson rich