User Agent Accessibility Guidelines Working Group Teleconference -- 27 Jun 2013

<trackbot> Date: 27 June 2013

<allanj> scribe: Kim

Face to Face Reminder

Kelly: inquired about rates again
... doing research on other hotels
... will have the information today or tomorrow

Proposed Speech Synthesizer Language Success Criteria

<allanj> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0097.html

Jim: have we finished this?
... suggestion that synthesizers should switch voices automatically

Jeanne: we talked about it and Jan suggested to include in another SC, not sure if we did

<Jan> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0098.html

Jan: does it clip automatically only in response to a Lang tag?

Kelly: most screen readers today are responsive to the Lang tag

Jeanne: proposal is user has control of it so one step command that will allow you to switch languages

Kelly: today either the user has no control or they can say turn off that language switching

Greg: heuristics often fall down when page contains pieces of multiple languages. Doesn't recognize passages of a foreign language embedded in something else. That's the main case I can see were having manual control would be useful. I don't have enough knowledge to say whether it would be an appropriate AA setting or not

Jim: proposal on the table to add a new SC 1.6.5, Yan said we might squeeze it into 1.6.1, but then we have whether it's the user or the system itself that switches it.
... does someone want to take it up or do we want to push it to next

Greg: 161 is A, I don't see it as that critical

Kelly: UAAG next – are we deficient without this

Jim: I could go with next

Jan: we don't put a ton of requirements on them, maybe we can just keep it, maybe we should remove the requirement about making it easy

<Jan> JR: Suggests this: 1.6.5 Synthesized Language: If synthesized speech is produced, the user can change the language (AA).

Greg: I defer to the group – I wish we had some users who actually are more familiar with this thing – specifically in multilingual situations which I assume is where it would be most useful.

Eric: W3C specs – try to be judicious about how much repair work they do – this is a manual repair issue – at least by following the specs involved it doesn't appear to be presented in the right language so it's a way of manually repairing the content, or its rendering. That would argue for not having it as a high priority and just being careful about including stuff that is sort of...
... repair work

Greg: let's say I speak Swahili, no native support, but I can say use this module. If that breaks, given the manual override can I access that page that would be otherwise inaccessible
... my attempts to come up with an example where it might be justifiable that it would completely block access of the feature was not there

<Jan> JR: Again...in the style of 1.6.1: 1.6.5 Synthesized Language: If synthesized speech is produced, the user can change the language, when more than one language is available (AA).

Jeanne: I thought about that too which is why I didn't propose it as a level a I think it's more for people with multi-languages and also for writing as well. If you're composing an email – the ability to command French for the next sentence is pretty necessary in any multilingual world

<Jan> JR: Again...in the style of 1.6.1: 1.6.5 Synthesized Language: If synthesized speech is produced and more than one language is available, then the user can change the language. (AA)

<Jan> JR: 1.6.5 Synthesized Language: If synthesized speech is produced and more than one language is available, then the user can change the language. (AA)

<jeanne> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0097.html

Jim: I'm fine with adding it with a change that Jan proposed
... any objections with leaving intent and examples as they are and using the Jans new wording

no objections

Jim: are we okay at AA?

<Jan> JR: OK with AA

Jim: not hearing any objections to AA

<jeanne> ACTION: Jeanne to add 1.6.5 with Jan's wording: 1.6.5 If synthesized speech is produced and more than one language is available, then the user can change the language. (AA) and http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0097.html [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action01]

<trackbot> Created ACTION-840 - Add 1.6.5 with Jan's wording: 1.6.5 If synthesized speech is produced and more than one language is available, then the user can change the language. (AA) and http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0097.html [on Jeanne F Spellman - due 2013-07-04].

resolved: accept 1.6.5 as AA

Proposed "programmatically available" definition

<allanj> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0116.html

Greg: sent email about that just before the meeting

<jeanne> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0121.html

Greg: we need to be more explicit about not excluding – not limiting it to published
... almost comes off as examples as how it can be done rather than requiring that

<allanj> proposed: programmatically available

<allanj> Information that is encoded in a way that allows different software, including assistive technologies, to extract and present the information in different modalities. This means making use of platform accessibility services, APIs, and, in some cases, document object models (DOM). For web-based user interfaces, this means ensuring that the user agent can pass on the information (e.g., through...

<allanj> ...the use of ARIA).

Greg: I didn't think that it came across as this is explicit as opposed to examples

<jeanne> Information that is encoded in a way that allows different software, including assistive technologies, to extract and present the information in different modalities using platform accessibility services, APIs, and, in some cases, document object models (DOM). For web-based user interfaces, this means ensuring that the user agent can pass on the information (e.g., through...

<jeanne> <allanj> ...the use of ARIA).

Jim: this is similar to the ATAG definition

<Jan> From WCAG2: programmatically determined (programmatically determinable)

<Jan> determined by software from author-supplied data provided in a way that different user agents, including assistive technologies, can extract and present this information to users in different modalities

<Jan> Example 1: Determined in a markup language from elements and attributes that are accessed directly by commonly available assistive technology.

<Jan> Example 2: Determined from technology-specific data structures in a non-markup language and exposed to assistive technology via an accessibility API that is supported by commonly available assistive technology.

<allanj> ATAG def - http://www.w3.org/WAI/AU/ATAG20/#def-Programmatically-Determined

Jan: seems okay, Jeannes change

<allanj> +1 jeanne's change

Eric: use – is the use of the term present or extract and use. Is it about content that is going to be presented to users or if it's programmatically determinable, isn't it about any kind of information regardless of whether it's for presentation?

Jeanne: do we even need that – is this a just about extracting information

Eric: in a way that can be used by other software

Jan: in a way others can make use of it

Eric: not sure it has to be that restrictive or not – maybe it does

Jeanne: are there other ways to be programmatically determined?

<jeanne> Information that is encoded in a way that allows different software, including assistive technologies, to extract and use the information in different modalities relying on platform accessibility services, APIs, or the document object models (DOM). For web-based user interfaces, this means ensuring that the user agent can pass on the information (e.g., through the use of ARIA).

Eric: do we need programmatically available if we already have something like programmatically determinable? Is what were trying to get across the same idea as what WCAG calls programmatically determinable?
... on revision – do we need the phrase in different modalities – makes sense when we had the word present, but right now I don't know that we need in different modalities, it's just they are able to extract and use it

<jeanne> Information that is encoded in a way that allows different software, including assistive technologies, to extract and use the information relying on platform accessibility services, APIs, or the document object models (DOM). For web-based user interfaces, this means ensuring that the user agent can pass on the information (e.g., through the use of ARIA).

<Greg> I will defer to the group, but a developer could still argue that they comply because, for example, a screen reader can use published API to get the text and font sizes, and from that can infer the locations of headings. I'd prefer to use phrasing that incorporates "published, supported mechanisms", "explicit" and "unambiguous".

<allanj> Information that is encoded in a way that allows different software, including assistive technologies, to extract and use the information relying on published, supported mechanisms, such as, platform accessibility services, APIs, or the document object models (DOM). For web-based user interfaces, this means ensuring that the user agent can pass on the information (e.g., through the use of ARIA)

Eric: no problem adding more of the kind of language that Greg mentions.

<Greg> The "published, supported" phrase was based on experience where developers would let assistive technology vendors use "undocumented, unsupported" API as a stop-gap measure, with the caveat that those mechanisms could be removed at any time, and customers inquiring could not get technical support because unsupported mechanisms were being used.

Jim: any objections?

Jeanne: do we want to change it to programmatically determined instead of programmatically available

Jan: keep available

Jim: I like available

Greg: the only thing missing from the latest definition was the unambiguous and/or not relying on heuristics, that kind of thing. So it would still allow one to pass if you exposed all the font information, therefore headers are implied but not explicit

Jeanne: I think we have it covered – we don't have infer anywhere in there

Greg: we are required to expose the headers, but if we don't expose the headers in any way other than – screen reader would have to recognize that larger font is headers – does that pass

Jeanne: do you have an idea of what we could use to close that down?
... we wouldn't be able to pass that as an implementation
... I can add a sentence to the intent

Greg: if you put something based on that paragraph for my email into the intent

<jeanne> ACTION: jeanne to add a definition of programmatically available: Information that is encoded in a way that allows different software, including assistive technologies, to extract and use the information relying on published, supported mechanisms, such as, platform accessibility services, APIs, or the document object models (DOM). For web-based user interfaces, this means ensuring that the user [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action02]

<jeanne> agent can pass on the information (e.g., through the use of ARIA)

<trackbot> Created ACTION-841 - Add a definition of programmatically available: Information that is encoded in a way that allows different software, including assistive technologies, to extract and use the information relying on published, supported mechanisms, such as, platform accessibility services, APIs, or the document object models (DOM). For web-based user interfaces, this means ensuring that the user [on Jeanne F Spellman -

<trackbot> ... due 2013-07-04].

<jeanne> ACTION: jeanne to add to the definition "... something is truly programmatically determinable only if the entity presenting the information does so in a way that is explicit and unambiguous, in a way that can be understood without reverse-engineering or complex (and thus potentially fallible) heuristics, and only relying on methods that are published, and officially supported by the developers of [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action03]

<jeanne> the software being evaluated."

<trackbot> Created ACTION-842 - Add to the definition "... something is truly programmatically determinable only if the entity presenting the information does so in a way that is explicit and unambiguous, in a way that can be understood without reverse-engineering or complex (and thus potentially fallible) heuristics, and only relying on methods that are published, and officially supported by the developers of [on Jeanne F Spellman

<trackbot> ... - due 2013-07-04].

Eric: one quick comment – maybe take out the word truly

Jeanne: I agree

Crowd source implementation - moving forward

Jeanne: temporarily on hold will I get the intellectual property issues – if people are not on a group and are contributing the rights are ambiguous. Have to make sure people are contributing everything royalty-free – just need to figure out the procedure for people to contribute their going to have to sign something that says I'm not going to try to copyright this work. I'm giving my work...
... to the W3C. So we can move forward on that one once I get that straightened out.
... people liked it

Jim: once we get that straightened out we will forward and get people to contribute implementations if they find them

We have Comments!! Dealing with them

<allanj> http://w3.org/WAI/UA/2013/commentsWD.html

Jim: last count there were 114 – more I haven't put in yet
... explaining comment syntax

Greg: one JR: that's missing its number

Kelly: this is very useful and great

Greg: 1.5.1 that's missing its number

Jan: when it's missing its number just give it the next number

Greg: only needed if we need to file a response and extract information into its own format
... a bunch of on the don't have the :

Jan: in the responses column, generally paste in there is some kind of highlighting when there is a proposal and then when it's approved by the group

@@approved with green highlighting or something just to make it easier to scan the whole list

Jim: start processing these the week after the Fourth of July or wait for you and Kim to thrash through the editorial ones?

Jeanne: we should definitely get started

Jim: meeting for next week canceled because it's on the Fourth of July

<Eric_Hansen_> Eric needs to sign off

Jeanne: I can do a survey for some of them that are not editorial for next week, that will get us ahead for the following week. I think a lot of these we can do with just voting the surveys maybe without a lot of discussion

Jim: any other comments, thoughts, suggestions on comments?

writing tests - while doing comments - write test as we address comments for specific SCs

Jim: we need to continue writing tests

Jan: anything useful we can borrow for keeping track from ATAG since you are a step ahead of us?
... I use a lot of @@'s, red, green, yellow
... @@approved etc.

Jim: Jan you had written a test on 2.11.2, Greg wrote back

<Jan> http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0115.html

Jan: answering Greg questions – yes doesn't have to be W3 formats, can be web formats – flash, PDF. So it comes down to what has the evaluator set out to do. When they set out to do an evaluation that actually specify the included technology. The thing might also render plaintext, Word files. Whatever's in scope
... answering what does contained within the area mean – really are just areas of the screen, but it's not the most watertight definition

Greg: would they were or would they not be – I couldn't figure it out from the test

Jan: would what?

Greg: would a script which modifies only a region – the HTML is not contained to that area, but its effect is contained so therefore does it count as executable content which would normally be contained to that area

Jan: this is pointing to a hole in the SC really
... I'm good to start with the types of elements that are limits to those areas, of course those may be generated on-the-fly by a script though

Jim: I'm concerned about this whole, is this something we need to patch in the SC or is it an edge case hole, or what

Greg: I think it is something which would need to be not ambiguous

Kelly, Jan agree

Jim: this is talking about Java scripting

Greg: it seems like it would need to be clarified in something normative, either in the SC or a note, not just implementing

<allanj> 2.11.2 Execution Placeholder:

<allanj> The user can render a placeholder instead of executable content that would normally be contained within an on-screen area (e.g. Applet, Flash), until explicit user request to execute. (Level A)

<allanj> 2.11.3 Execution Toggle:

<allanj> The user can turn on/off the execution of executable content that would not normally be contained within a particular area (e.g. Javascript). (Level A)

Greg: better phrase for contained within a particular area?

<Greg> Intent of Success Criterion 2.11.2:

<Greg> Documents that do things automatically when loaded can delay, distract, or interfere with user's ability to continue with a task. Replacing executable content like embedded objects, applets and media with a placeholder tells the user what has been blocked and provides a mechanism (e.g. a play button) for unblocking when the user is ready.

<Greg> Note: A placeholder should take up the same space as the object it is replacing, so that the presentation doesn't need to be reflowed when the execution is started. However, people using mobile devices or screen enlargers, or those who have difficulty with scroll commands may benefit from having the option of a smaller placholder.

Greg: you might have an applet which traps focus like flash or it might fear that it would display content that would trigger an epileptic seizure or start to make sound or whatever. So you want to be able to postpone the execution of it until you decide what do you really want to do it or not.
... the other one is you can have a placeholder in the document

Jan: why not change the SC to be before launching put a placeholder there and have the user have to

Jim: instead of executable content we would say a plug-in, separate from JavaScript

Jan: may be more easily testable way to break the two up

Greg: sounds promising. In some cases a video is rendered by a nested user agent, in some cases rendered natively. Unless the placeholder says the user wouldn't know the difference. It wouldn't apply to the latter case – is that handled by some other SC?

Jan: yes, there's a video SC.

<Eric_Hansen_> Eric has returned

Kim: this is a case where you would one test for several SC's?

Jeanne: problems with that

Jan: later on you could write another layers that combines them

Jim: summary says it nicely

<Jan> Propose rewrite of 2.11.2 Execution Placeholder: and 2.11.3 Execution Toggle that focus more on the fact that plug-in user agents are/are not activated. Also take into account comments: http://www.w3.org/WAI/UA/2013/commentsWD.html

<Jan> ACTION: JR to Propose rewrite of 2.11.2 Execution Placeholder: and 2.11.3 Execution Toggle that focus more on the fact that plug-in user agents are/are not activated. Also take into account comments: http://www.w3.org/WAI/UA/2013/commentsWD.html [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action04]

<trackbot> Created ACTION-843 - Propose rewrite of 2.11.2 Execution Placeholder: and 2.11.3 Execution Toggle that focus more on the fact that plug-in user agents are/are not activated. Also take into account comments: http://www.w3.org/WAI/UA/2013/commentsWD.html [on Jan Richards - due 2013-07-04].

<jeanne> ACTION: Change summary on 2.11 to make it accurate to existing SC and check numbering. [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action05]

<trackbot> Error finding 'Change'. You can review and register nicknames at <http://www.w3.org/WAI/UA/tracker/users>.

<jeanne> ACTION: jeanne to Change summary on 2.11 to make it accurate to existing SC and check numbering. [recorded in http://www.w3.org/2013/06/27-ua-minutes.html#action06]

<trackbot> Created ACTION-844 - Change summary on 2.11 to make it accurate to existing SC and check numbering. [on Jeanne F Spellman - due 2013-07-04].

Greg: the no script extension for Firefox where you turn off a script on a play-by-play basis is much more useful than doing it globally
... JavaScript is what 2.11.3 is there for – Media

Jim: media is 2.11.1

Greg: the word media in 2.11.3 means they are overlapping?

<allanj> http://www.w3.org/TR/UAAG20/#gl-speech-config

Greg: looking at June 7 edition
... difference between the June 7 editor's draft and the version used for the comments document – which is the current one?
... When I built the spreadsheet use the latest published draft

Jeanne: because that's what people commented on

<Greg> This is the version from the 7 June editor's draft: 2.11.3 Execution Toggle: The user can turn on/off the execution of dynamic or executable content (e.g. Javascript, canvas, media). (Level A)

Kelly: as we are evaluating a comment we should check that against the latest editor's draft

<Greg> This is the version from the latest published draft and comments document:

<Greg> 2.11.3 Execution Toggle: The user can turn on/off the execution of executable content that would not normally be contained within a particular area (e.g. Javascript). (Level A)

Greg: the differences in the parenthetical list of examples – we took out canvas and media, which would obsolete my comment

<Greg> Actually we *added* "canvas, media" to the parenthetical list of examples since the last public draft.

- DRAFT -

User Agent Accessibility Guidelines Working Group Teleconference

27 Jun 2013

Attendees

Contents