W3C

- DRAFT -

User Agent Accessibility Guidelines Working Group Teleconference

01 May 2014

See also: IRC log

Attendees

Present
[Microsoft], Eric, Jeanne, Greg_Lowney, Jim_Allan, Kim_Patch, Jan, wuwei
Regrets
Chair
jimAllan, KellyFord
Scribe
allanj

Contents


<trackbot> Date: 01 May 2014

Main Comment Gateway - http://jspellman.github.io/UAAG-LC-Comment/

Latest editor's draft - http://www.w3.org/WAI/UA/UAAG20/

<scribe> scribe: allanj

open item 3

WebTV

ww: visiting scholar at MIT, working on WebTV
... second round of use cases, shaping requirements of TV
... considering a11y in WebTV, have some use cases

ja: need more use cases?

ww: yes, need more to help set technical requirements.

js: correction. have no a11y use cases, need some.

<wuwei> http://www.w3.org/2011/webtv/wiki/Media_APIs/Use_Cases

<jeanne> 1.1.1 has an example

<wuwei> http://www.w3.org/2011/webtv/wiki/New_Ideas

<jeanne> 1.1.6 has an example of positioning captions

ja: above line for format for use cases

<jeanne> 2.4.5 - alternative content search

ja: timeline for use cases?

ww: end of may deadline

ja: use cases - how user gets speech feedback, or change font size

kf: WebTV, on desktop there is an AT layer and the browser. not so on TV

gl: a closed system, no addition of external AT

<jeanne> SC 1.1.1 Alternative content - http://jspellman.github.io/UAAG/Implementing-UAAG20/#sc_111-e

kf: windows media center. works with nvda, mostly. can read, but not details...spell words

<jeanne> 1.1.6 Resize and reposition of media alternatives http://jspellman.github.io/UAAG/Implementing-UAAG20/#sc_116-e

jr: if no 3rd party, how to add scanning keyboard to a closed system if none is built in

kf: this may be beyond our scope

js: been dropping things in IRC of things that must be looked at

gl: we don't have TV examples

js: we have other projects that are more pressing

gl: WebTV should be able to generate use cases from UAAG, not us generating them

js: possible?

ww: will suggest to webtv to review UAAG to generate use cases
... don't think they will do it.

ja: UAWG will not pursue this as a group, however individuals can submit their own use cases if they choose to Jeanne

js: can submit the 4 we have discussed.
... has the webtv group looked at an underlying a11y api for basic OS a11y features

ww: looked at timed text, tuner control

js: not others - font size, captions placement, etc.
... will followup off line

zakim: close item 3

Jan - Action 973 - comment http://lists.w3.org/Archives/Public/w3c-wai-ua/2014AprJun/0014.html

<Jan> http://jspellman.github.io/UAAG-LC-Comment/

jr: orig. comment - remove web based user agent
... proposed response

In order to clarify the relationship between UAAG 2.0 and WCAG 2.0 (and ATAG 2.0) UAWG has added 2 new sections to the Guidelines document Introduction:

-Relationship to the Web Content Accessibility Guidelines (WCAG) 2.0 and

-Relationship to the Authoring Tool Accessibility Guidelines (ATAG) 2.0

For the following reasons, UAWG continues to believe that keeping web-based user agents in the scope of UAAG is reasonable:

- from the perspective of users needing accessibility features (e.g. zoom, media navigation, etc.) it is not always clear whether a user agent is web-based or native, especially as browsers and operating systems converge.

- UAAG does allow web-based user agents to depend on the base (native) user agent in order to meet UAAG success criteria, just as non-web-based user agents can rely on operating system features (e.g. to programmatically communicate with assistive technology)

- the web-based user agent may be best placed to provide UAAG conformant functionality. For example a web-based video player may only be serving small portions of a video to the base browser's video player, so the base browser would not be able to provide effective navigation of the video. Similarly, a web-based user agent can zoom its content and keep its own user interface the same size,...

scribe: saving screen real estate, while the base browser's zoom function may zoom the web-based browser's user interface, leaving less space for the content.

Instead of de-scoping web-based user agents, UAWG will identify exemptions from specific success criteria where necessary (e.g. due to technical limitations).

js: like it

<kford> I like this, was taking a minute to process all.

gl: wording is technical but makes the right statement
... would add reasonable and important.

<Greg> To "UAWG continues to believe that keeping web-based user agents in the scope of UAAG is reasonable" add "and important".

kp: +1

ja: this does task us with iding the exemptions

gl: we have been doing this

ja: change last line to " UAWG will continue to identify exemptions from specific success criteria where necessary (e.g. due to technical limitations).'

jr: this is a moving target, as UAs push more to other components/objects in content
... we will continue to update exceptions

close action-973

<trackbot> Closed action-973.

gl: in the bullet items. are they in the document

jr: perhaps change the def of 'web-based UA'

gl: or in the introduction

js: better in the introduction

eh: should it be both.

<Jan> ACTION: JR to Work this text http://lists.w3.org/Archives/Public/w3c-wai-ua/2014AprJun/0014.html into a section in the introduction. [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action01]

<trackbot> Created ACTION-974 - Work this text http://lists.w3.org/archives/public/w3c-wai-ua/2014aprjun/0014.html into a section in the introduction. [on Jan Richards - due 2014-05-08].

jr: don't think it changes the def. it has been stable for a long time

MS04 Delineate between content and browser -- 2.3.4 Present Direct Commands in User Interface: , 2.2.1 Sequential Navigation Between Elements:

ms04: There should be clear separation of the guidelines that are applicable to the browser itself (such as 3.2.2 or 2.3.4) and how it handles web content (such as 3.2.3 and 2.2.1). In theory, success criteria applicable to handling of web content should always contain the term "rendered content" to make this distinction clear. But this is, unfortunately, not done with any consistency. For...
... example, guideline 2.2 are clearly applicable to browser treatment of rendered content, but the term never appears in its four success criteria. This makes it difficult for the audience to interpret the intention of the guidelines.

<jeanne> Here is a proposal for draft text in response to MS04 from GL:

<jeanne> Where there are specific success criteria that could be clarified, please point them out, and we can make an effort to clarify them. Your example is not quite accurate, as while 2.2.1 does not explicitly say it applies to rendered content, it does say it applies to recognized enabled elements, which are defined as being a subset of rendered content. We have tried to balance out being repetitive

<jeanne> with being clear, and expected that readers would quickly pick up on frequently used terms such as "elements", just as with "content"; without those the document would not be meaningful

eh: is there an antonym for 'rendered content'

ja: machine readable

js: user interface

jr: ATAG split into 2 halves, UAWG discussed and rejected splitting long ago

eh: do we use user agent user interface vs rendered content

js: we do this

jr: UAAG 10 had something like that.

js: could add classes to help sort the document, rendering vs UAUI
... would be helpful in implementations

<Jan> JR: UAAG 1.0 used this...http://www.w3.org/TR/WAI-USERAGENT/conformance.html#content-or-ua .... but not very consistently.

ja: we already use RDC and UAUI,

gl: we say SC apply to both, unless we say otherwise.

eh: where is that located?

UAUI not mentioned until 2.1.6

eh: should it be called out in an applicability note?

jr: what is default assumption?

gl: assumption is SC apply to both, unless we say otherwise

eh: written where? or just our thoughts

js: not sure, a while back we went through the doc to flag UAUI and RC

jr: only a few actually overlap

<scribe> ACTION: jallan to review document for explicit use of UAUI or RC and report back [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action02]

<trackbot> Created ACTION-975 - Review document for explicit use of uaui or rc and report back [on Jim Allan - due 2014-05-08].

<Greg> We should improve the definition of "recognize", starting with a concise definition before getting into background information.

<Jan> JR: Just noticed "rendered content" just talks about author-supplied...I can imagine user agent injected content being relevant here as well

ja: like what?
... like the injection of the mouseless browsing content injection of the numbers for links

eh: ... may include content injected by extensions etc.
... could have rendered content that the author never touched

jr: help content in chrome are html

gl: content is defined in a w3 format and presented to the user

<scribe> ACTION: jallan to review definition of Rendered content [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action03]

<trackbot> Created ACTION-976 - Review definition of rendered content [on Jim Allan - due 2014-05-08].

ja: what about SSL EV green box in address bar, is that RC

jr: sort of, more UA functionality, not necessarily RC

MS05: Examples in the implementation document do not make distinction of content, browser, assistive technologies, and OS (partial)

<Greg> ACTION: Greg to edit glossary entry for "recognize" to start with a concise definition [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action04]

<trackbot> Created ACTION-977 - Edit glossary entry for "recognize" to start with a concise definition [on Greg Lowney - due 2014-05-08].

seen http://www.w3.org/2014/01/23-ua-minutes.html#item03

<jeanne> Your criticism of the example for 2.2.1 is well taken. However, the Implementing document is not on the same timeline as the guidelines document itself, and is not submitted for last call, so if the former needs more work it is not automatically a reason to hold up the latter.

js: it is a fair point. and would be a major rewrite of all examples.

ja: a review of UAUI and RC should help

gl: need to be more explicit. write the example to say ... could be done by OS or UA or an AT

js: examples are not complex. an engineer should know what is a UA or OS function

<Greg> We don't *need to*, but we certainly could, and doing so might make it easier to use for some UA developers and testers.

eh: UA and platform, allow us to say let the developer determine where a function should be handled
... conformance, the developer says who does what. We have stated that. Its someone else's problem for determining this
... should we be clear about including or excluding AT in the conformance claim

js: comment is only addressing implementation document - non-normative, not about the conformance claim - normative

eh: is AT part of the conformance claim.

gl: including AT is part of conformance

ja: what is our response?
... UAAG is non-prescriptive to have success on an SC

js: UAWG is trying to make it easier on developers has to how they want to meet requirements of an SC. AT is a moving target and is rapidly changing.

eh: reviewed conformance wrt AT. no AT mentioned in definition of platform, and not much mentioned in conformance
... is document explicit about AT place - inside or outside of the UA

<jeanne> UAWG did wrote the examples to be illustrative of common uses by people with disabilities. In order to avoid being excessively prescriptive to the browser, we did not make divisions between platform, OS, browser and assistive technologies, knowing that the field is evolving rapidly, and the accessibility feature that is on the browser on one device, may be on the OS on another.

ja: most users don't use AT, so UA need to do as much as possible.

+1 to jeanne proposal

kp: +1

jr: =1
... +1

<jeanne> Assistive technologies may or may not be included in the user agent, the OS or the platform, so we did not try to define where it would be located.

eh: AT is external to UA, peoples def of AT is very broad.

<Greg> If we elsewhere clearly distinguish between UA and AT, then saying "AT may or may not be included in the user agent" seems to contradict that.

ja: we are talking about what is needed to conform to the SC. AT are not part of UA but they can be included in a conformance claim

eh: conformance need, part h list AT as needed to meet SC as part of the platform?

<Greg> I have to agree with the original criticism in MS05 about the example for 1.2.2 "Maria uses a screen reader. When a table lacks marked up header rows, the user agent gives her the option to have the first row treated as the table header row." That *is* vague as to what the user agent is doing. We could say something like "When a table lacks marked up header rows, the user agent provides an...

<Greg> ...option or command that changes the td (table data) elements in the first table row to be th (table headings)." This seems like something that AT would handle, rather than recommending the UA provide this feature, but this is illustrates how the example (and others) could be made clearer.

js: be careful, navigate by heading should not be passed on to the AT should be native.

ja: jeanne will write a proposal for our response
... next week we will review what we want to do about making the examples more clearly reflective of UA abilities in meeting SCs

<jeanne> ACTION: jeanne to write a proposal for the response to MS05 [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action05]

<trackbot> Created ACTION-978 - Write a proposal for the response to ms05 [on Jeanne F Spellman - due 2014-05-08].

<scribe> new browser - Aviator https://www.whitehatsec.com/aviator/

kp: drawing program with novel undo. Concepts...keeps track of all changes, and remove specific element in the chain of changes. would be great for UA or platform configuration

Summary of Action Items

[NEW] ACTION: Greg to edit glossary entry for "recognize" to start with a concise definition [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action04]
[NEW] ACTION: jallan to review definition of Rendered content [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action03]
[NEW] ACTION: jallan to review document for explicit use of UAUI or RC and report back [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action02]
[NEW] ACTION: jeanne to write a proposal for the response to MS05 [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action05]
[NEW] ACTION: JR to Work this text http://lists.w3.org/Archives/Public/w3c-wai-ua/2014AprJun/0014.html into a section in the introduction. [recorded in http://www.w3.org/2014/05/01-ua-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014-06-05 16:11:08 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/as defined in w3 terms/is defined in a w3 format/
Succeeded: s/paritlal/partial/
Found Scribe: allanj
Inferring ScribeNick: allanj
Default Present: [Microsoft], Eric, Jeanne, Greg_Lowney, Jim_Allan, Kim_Patch, Jan, wuwei
Present: [Microsoft] Eric Jeanne Greg_Lowney Jim_Allan Kim_Patch Jan wuwei
Found Date: 01 May 2014
Guessing minutes URL: http://www.w3.org/2014/05/01-ua-minutes.html
People with action items: greg jallan jeanne jr

[End of scribe.perl diagnostic output]