Accessible Rich Internet Applications Working Group Teleconference -- 05 May 2016

<Rich> https://lists.w3.org/Archives/Public/public-aria/2016May/0014.html

<jongund> what is the webex id (9 digit number?) for the aria room?

<Rich> scribe: Jemma

make log world

<scribe> scribe:JaEunKU

<scribe> chair:Rich

TOPIC

<MichaelC> scribeNick: jemma

CFC

<Rich> https://lists.w3.org/Archives/Public/public-aria-admin/2016Apr/0033.html

<jongund> what is the Password?

<joanie> s/What is the Password?//

rich: Joanie will close the agenda item

Cynthia: We have UIA mappings?

rich: joseph: we didnot have any entry yet

<joanie> Joanie: I am still considering how to map this in ATK and AT-SPI2.

joseph will create aria-keyshortcut action item

<joanie> Joanie: The problem is that the way all keyboard shortcuts are exposed is through the Action interface.

<joanie> Joanie: But real actions have actions which can be invoked.

<joanie> Joanie: This is not the case for our new ARIA property.

Password role

Rich: how are we doing with voiceover, Joanie?

<Rich> ACTION: Joanie: Speak with Apple about VoiceOver rendering, rendered text for passwords [recorded in http://www.w3.org/2016/05/05-aria-minutes.html#action01]

<trackbot> Created ACTION-2061 - Speak with apple about voiceover rendering, rendered text for passwords [on Joanmarie Diggs - due 2016-05-12].

<Zakim> JF, you wanted to ask about some of the other comments from the secureity folks

Rich: the browser manufactures, especially firefox indicated that they dont have security measure/
... Cynthia, Ms do you indicate that it is insecure in the page?

cyn: it is in browser, not in the page

s/MS/microsoft

cyn: in ie, if there is a bug, it kept saying "security alert' repeatedly so we don't want to do that.

rich: questions is more about narrative

cyn: screenreader need to know where to go and see the status

jf: browser know the relationship

<JF> +1 to the feture request

jf: is there any way that we can have any non normattive way to talk about pw? some note for creating custom pw ?

cyn: usual pattern is ui login is in inscure page and after login it is in secure page.

rich: I can put a note about it.

<jamesn> jamessn has joined #aria

cyn: we need some suggestion form developer about where to put pw role, for example, before form action or not.

jn: this is security requirement, not a11y requirement

joanie: I agree with Jame

rich: I don't disagree with it

<Zakim> JF, you wanted to ask about author guidance here.

<JF> “My concern is more aligning a user context, and the behavior of the browser/tool : aka taking special care of the data. It seems to me that in terms of security, the creation of the role, password, is valuable if you have a secure expected usage associated. By secure, I mean, making sure the content is protected from direct reading on the screen, special protection against (including unexpected cloning or storage) *and* that the user is getting a sense of the sens

rich: it is aboout "expected secure usage"

fred: it is all applied to credential management
... where the pw is stored in the browser so on..

rich: it should be indicated somewhere, the question is where
... how do we wire that- credential managemeent- to aria rule?

Fred: AAM is tracking this and we pass that to COGA

rich: who will take this agenda? fred?

fred: you can give me an action on that

joanie: give actions to APA?

<Rich> ACTION: Fred work with APA to address password roles/elements/credential management in a secure environment for TPAC [recorded in http://www.w3.org/2016/05/05-aria-minutes.html#action02]

<trackbot> Created ACTION-2062 - Work with apa to address password roles/elements/credential management in a secure environment for tpac [on Fred Esch - due 2016-05-12].

active decendants

<Zakim> clown, you wanted to suggest the crendential management spec needs an accessibility section.

<Rich> https://rawgit.com/w3c/aria/action-1743/aria/aria.html#aria-activedescendant

<mck> https://rawgit.com/w3c/aria/action-1743/aria/aria.html#aria-activedescendant

matt: I used the branch michale created to work on active decendants
... in the branch, michale already has active decendants property concerning with application role. I merged that recent master and made more edits.
... trying to simplify the language

rich: where there is underline in the widget?

mattt: widget is for glossary reference

matt: composite is role referece

rich: editorial suggestion

glossary reference out of "widget"

matt: will remove glossary reference from "widget"
... group here is strange
... 3 subclass of group

would like to take out of group from the section, "used in roles"

matt: row is also subclass of widget

rich: "group" does not damage anything.

matt: will add textbox to first paragraph

*trying to save type :-)

matt: can we say "interactive element"

rich: how about "designated"?

<Rich> The aria-activedescendant property provides an alternative method of managing focus interactive elements, such as menus and grids, that may have multiple focusable descendants.

joseph: it should "should", not "must" because it has exceptions

jn: how about making bulleted list?
... make them as three separate bullets

fred: aria control does not have decendant while aria-owns has decendant

js: aria control does not define children

jn: aria active descendants can be determined by aria-own or by dom tree

matt: active decendant can point anything in the dom

Joseph: CoreAM we do force to have correct/direct decendants, not any dom as decendants

Josephwe need this for focus management

Joshep: there is difference in focus management for dom decendant- it is easy or active decendants

matt: API perspective, when do we care about when it is focusable or not.

Josephie, presentational item is not focusable

<Zakim> clown, you wanted to point out that RFC MUST is exceptionless, use SHOULD for exceptions

matt: I can change it back as "should"

rich: I like making it as bulleted list.

matt: I would like to make a change about "Some user agents might use.."

Cynthia: if we make a decision this week, I can take it to the development

*joanie can you fill the question mark

cyn: it may difficult to do the way IA2 did. I remember there is some difficulty before.

joanie: let's don't take out, there must be a reason for the existence of "Some user agents might use.."
... I would be happy to dig out why that sentence is there if someone give me an action

<joanie> joanie to research why it says "user agents might use aria-activedescendant..."

<joanie> action

<trackbot> Created ACTION-2063 - Research why it says "user agents might use aria-activedescendant..." [on Joanmarie Diggs - due 2016-05-12].

Action 2059

<joanie> action-2059

<trackbot> action-2059 -- Joanmarie Diggs to Add aria-posinset, aria-setsize as supported states and properties to roles: menutitem, menuitemcheckbox, and menuitemradio -- due 2016-05-05 -- OPEN

<trackbot> http://www.w3.org/WAI/ARIA/track/actions/2059

Add aria-posinset and aria-setsize to roles menuitem, menitemcheckbox, and menuitemradio

<joanie> Just supporting it on menuitemradio is confusing for all the reasons treating grouped menuitems as separate sets is confusing. So we do need to resolve this.

<joanie> We should explicitly state in the spec that authors SHOULD treat the set as the full menu, with the setsize being all menuitems (and their subclasses, regardless of type or the presence of visual separators).

joanie is coping the conclusions

<joanie> The APG should provide sufficient examples to make this clear, and perhaps explain some of the rationale behind the normative statement we put in the spec.

joanie has three conclusions from her research

matt: how screen readers know about grouping menu?
... did not realize semantic value but we still recommend to put those.

joanie: I agree with what Matt said
... would like to get consensus on conclusion #2 We should explicitly state in the spec that authors SHOULD treat the set as the full menu, with the setsize being all menuitems (and their subclasses, regardless of type or the presence of visual separators).

<Rich> https://rawgit.com/w3c/aria/action-1743/aria/aria.html#menu

<joanie> https://rawgit.com/w3c/aria/master/aria/aria.html#menuitem

joanie: how about taking out " group" from context in menuitem?
... keeping group as required context group role. another is using group to make things to clear to screen users.

matt: we need to make label the group to distinguish sub groups.

rich: should we include "level" for this?

joanie: no

joannie: I will do draft for existing action.

Action-1489

<Rich> https://rawgit.com/w3c/aria/ACTION-1489/aria/aria.html#host_general_conflict

michale: we just need review from the group
... change the note to spec text
... it is section7.5
... added aria-label. it is bulleted list but it does not look like one here

rich: I am ok with this
... can we put it as cfc?

joseph: aria checked and html checked has direct conflict
... overriding cases

matt: typo for "intrinsic"

RESOLUTION: Accept Michael’s proposal for Action 1489 pertaining to Native Host language semantics

- DRAFT -

Accessible Rich Internet Applications Working Group Teleconference

05 May 2016

Attendees

Contents

CFC

Password role

active decendants

Action 2059

Add aria-posinset and aria-setsize to roles menuitem, menitemcheckbox, and menuitemradio

Action-1489

Summary of Action Items

Summary of Resolutions

Scribe.perl diagnostic output