Minutes: June 25 ARIA Task Force Meeting



http://www.w3.org/2015/06/25-aria-minutes.html


Text:

Protocols and Formats Working Group Teleconference
25 Jun 2015


See also: IRC log


Attendees
Present
      janina, Joseph_Scheuhammer, Joanmarie_Diggs, Bryan_Garaventa,
      James_Nurthen, Stefan, JonGund, JF
Regrets
Chair
      Rich
Scribe
      jamesn
Contents
      Topics
         1. aria-describedat
         2. keyboard shortcuts
         3. discuss a freeze date for the heartbeat draft
         4. annotationsd
         5. ACTION-1591
      Summary of Action Items



<trackbot> Date: 25 June 2015


<richardschwerdtfeger> Meeting: W3C WAI-PF ARIA Caucus


<richardschwerdtfeger>
https://lists.w3.org/Archives/Public/public-pfwg/2015Jun/0168.html



<clown> https://rawgit.com/w3c/aria/master/aria/aria.html#attrs_widgets



<clown> "This section contains attributes specific to common user interface
elements found on GUI"


<scribe> scribe: jamesn


aria-describedat


RS: have told the DPUB folks that by mid August need something


JS: lets see if they can make a Thursday
... they understand they need to come on this meeting to discuss. Also will
send a formal mail about describedat to this group in the next week


RS: would like to make 1 point they talked about on the call
... 1 think they said and must be cogniscent of this. Know how hard it is
to get product teams to do a11y.
... they said if we pull this now could be a problem. They have p[eoples
attention at the moment
... we have to think about the strategy


JS: notes that they are aware of annotations but that is not a rewality yet
... they are looking for something in the next several months
... the fact that it can pertain to any element makes it more useful


<Zakim> joanie, you wanted to suggest we schedule something so James Craig
can be present


RS: hoping will get something set up soon


keyboard shortcuts


RS: for some reason people getting confused with accesskey. (1) it is for
activation. (2) as a spec we don't define user agent behaviour
... if the UA host language wants to do more with it then they can but we
don't require them to do it
... accesskey module has sat for 6 years. Can't say need to synchornize
with accesskey
... other than getting vocab correct don't see how this should be confused
with accesskey


JN: like the keyboardshortcuts proposal. It tries to allow an author to
state what is happening today programiatically. It is not trying to solve
the bigger issue of how to specify shortcuts across platforms etc.


discuss a freeze date for the heartbeat draft


RS: would like to get it out
... what needs to be merged


JD: dont think i haver any not merged stuff


<richardschwerdtfeger> Action-1493?


<trackbot> Action-1493 -- Joseph Scheuhammer to Propose spec text allowing
¨implicit value for role¨ to provide default values for required properties
-- due 2014-10-06 -- OPEN


<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1493



clown: one of mine i want to get in


<richardschwerdtfeger> Action-1606?


<trackbot> Action-1606 -- Joseph Scheuhammer to Create proposal to modify
aria-valuemin, aria-valuemax, and aria-valuenow to have reasonable default
values when they are missing, similar to HTML5 spec on <input type="range">
-- due 2015-06-11 -- OPEN


<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1606



JG: how would shortcut keys be mapped to accessibility APIs


JD: ATK has an action interface
... if IA2 follows that there would be a way there
... there is no such thing as a keyboard interface


<joanie>
https://developer.gnome.org/atk/unstable/AtkAction.html#atk-action-get-keybinding



<clown>
http://accessibility.linuxfoundation.org/a11yspecs/ia2/docs/html/_accessible_action_8idl.html



<joanie>
http://accessibility.linuxfoundation.org/a11yspecs/ia2/docs/html/interface_i_accessible_action.html



JD: it is the same one


RS: might need to clarify for JC some of the vocabulary in there
... dont know anbout mac


<richardschwerdtfeger> s/anabout/about/


RS: one of the things JC was concerned about was the vocab. s/about/about/
... freeze date
... can you have 1606 ready for next thursday


clown: ready now


RS: lets see what else we need for a heartbeat and lets freeze it


JD: strictly for this hearbeat freeze?


RS: yes


MC: I was targetting the 14th so July 6 is fine


RS: trying to get 1.1 out of the door


annotationsd


<clown> issue-655


<trackbot> issue-655 -- Consider creating annotations roles for comments,
spell-check errors, etc -- open


<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/655



<clown> issue-670


<trackbot> issue-670 -- Need to support remote annotation -- raised


<trackbot> https://www.w3.org/WAI/PF/Group/track/issues/670



RS: the person who requested it has left MSFT. getting no response on
questions
... this is not under the scope of aria 1.1
... could be better to look at it in the ARIA 2.0 timeframe
... don't want to lose them but feel they should be ARIA 2.0


JN: +1 to that.


<jongund> JG: +1


RS: status is that are waiting for a response from CS before can do
anything


<joanie> JD: +1


RS: I would rather us deal with more serious issues like error handling


clown: do error messages for invalid input belong in annotations?


RS: have a concern that this is a big problem. all forms have this issue
today
... there is also a remote annotation that is brought up


<Zakim> joanie, you wanted to explicitly state that annotations seems like
it's meaty enough that we don't want to rush it


RS: i feel the same way


<JF> +1 to Joanie


<richardschwerdtfeger> Proposal: Move Office Suite Annotation Issues and
Actions to ARIA 2.0 to provide more time to develop


<richardschwerdtfeger> Rich: This would cover issues 655, 670, and actions
1455 1n 1454


<bgaraventa1979> +1


<richardschwerdtfeger> +1


<clown> +1


<joanie> +1


<jongund> +1


+1


<JF> +1


<Stefan> +1


<richardschwerdtfeger> RESOLUTION: Move issues 655, 670, and Actions 1454
and 1455 to ARIA 2.0


<clown> \o/


ACTION-1591


<clown> action-1591


<trackbot> action-1591 -- Joanmarie Diggs to checkbox role doesn't list
aria-readonly -- due 2015-03-04 -- OPEN


<trackbot> https://www.w3.org/WAI/PF/Group/track/actions/1591



<clown>
http://www.w3.org/html/wg/drafts/html/master/semantics.html#the-readonly-attribute



clown: if you read the html5 spec on its readonly property. It lists that
only text controls can be made readonly


<Stefan> +q


<richardschwerdtfeger> JN: One use case where, based on a role that someone
has assigned, they don’t have the permissions to modify something but it is
not disabled


<richardschwerdtfeger> JN: Where they have disabled it is based on
something you can do to enable it. Where you have aria-readonly it is based
on a privilege


<richardschwerdtfeger> JN: On a transactional flow where you have a sumary
screen, you have a wizard, and you have a summary of what has been done but
they are not disabled and they are read-only


<richardschwerdtfeger> JS: Visually it is greyed out


<richardschwerdtfeger> JD: Readonly controls are still focusable where
disabled controls are not


JD: if you can tab to a checkbox but not change its state then it is
readonly


<Stefan> -q


<richardschwerdtfeger> Rich: I have seen controls where items are disabled
but are focusable


JD: if there is a use case. JN was describing conceptual reasons where
something is changeable
... screen readers have commands to move between different control types


<richardschwerdtfeger> jD: If you press X in Orca you go to the next check
box


<richardschwerdtfeger> JD: If the checkbox is not something you can tab to
you should skip it


<richardschwerdtfeger> JS: you are on a summary page and you are about to
submit but you can’t change them


<richardschwerdtfeger> Rich: disabled items in menus are often in the
navigation order but are marked disabled


<Stefan>
https://sapui5.hana.ondemand.com/sdk/test-resources/sap/ui/commons/demokit/CheckBox.html



JD: what i am trying to suggest is that yes we do not to want to be the
same as HTML. I would go further in that HTML has a bug...


<Stefan> +q


JD: I think it does make sense to allow focus to these things


<Zakim> joanie, you wanted to ask about focusable versus not


clown: about menu item example. on the mac it skips disabled menu items. if
you have VO turned on you can arrow to them


<Stefan>
https://sapui5.hana.ondemand.com/sdk/test-resources/sap/ui/commons/demokit/CheckBox.html



<Zakim> clown, you wanted to mention how VO handles disabled menu items.


Stefan: we have other states for checkboxes too.


JN: A disabled state indicates to the user that they can do something to
enable it


<Stefan> -q


JN: a read-only checkbox indicates that the state it out of their immediate
control


BG: I don't see the point really. When something is disabled then i know I
can't interact with it.


JD: can imagine a case where there is a menu item where you can't change it
... I agree with the notion that checkboxes should be able to be readonly
or disabled
... the text that Joseph pasted from the HTML spec is why we need to
explain it


RS: a checkbox in a menu I can see read-only. I don't think it is terrible
that it inherits
... if we are ok with the inheritance I dont think it is terrible


clown: doesn't inherit into treeitem


<richardschwerdtfeger>
http://rawgit.com/w3c/aria/master/aria/aria.html#treeitem



<clown> https://rawgit.com/w3c/aria/master/aria/aria.html#option



JN: need radio buttons too


JD: think they inherit from checkboxes
... switch is fine. Think it is a silly UI


JN: +1 :)


<clown> it goes checkbox -> radio, menumitemcheckbox, switch.


<clown> and menuitemcheckbox -> menuitemradio


<richardschwerdtfeger> Proposal: Add aria-readonly to roles checkbox which
will inherit into, radio, menuitemradio, menuitemcheckbox, and switch


<Zakim> JF, you wanted to ask about adding it to fieldset?


JF: my experience that generally in a fieldset.... (or group)


RS: do people want to take it off radio?


JD: we have this right now with the mixed state where switch and radio
don't inherit checked=mixed
... have ugly notes
... are you ok with the situation right now where mixed is inherited by
radio


I see Stefan has a readonly radio button
https://sapui5.hana.ondemand.com/sdk/#test-resources/sap/ui/commons/demokit/RadioButton.html



JD: I will come up with a proposal


<joanie> Propsal: aria-readonly should apply to checkbox, check menu item,
switch, and radio menu item. It should not apply to radio. This will
require a corresponding taxonomy change.


Summary of Action Items
[End of minutes]


Rich Schwerdtfeger

Received on Thursday, 25 June 2015 18:04:54 UTC