W3C

- DRAFT -

User Agent Accessibility Guidelines Working Group Teleconference

12 Nov 2013

See also: IRC log

Attendees

Present
jeanne, jim, greg, kim, kelly, Jan, judy, kurosawa, boryung, ijongcheol, kenny, rachel, wang, kimwoonyoung, jeonghun, stone
Regrets
eric
Chair
Jim, Kelly
Scribe
jeanne, allanj

Contents


<trackbot> Date: 12 November 2013

<kford> Warning, IRC is going to fill up for a second as GL is going to paste the minutes from the smeeting start until we started logging.

<Greg> [17:03] -->| kenny (~kenny@public.cloak) has joined #ua

<Greg> [17:05] -->| boryung (~boryung@public.cloak) has joined #ua

<Greg> [17:05] -->| rachel (~rachel@public.cloak) has joined #ua

<Greg> [17:06] jeanne Last Call Working Draft http://www.w3.org/TR/UAAG20/

<Greg> [17:07] -->| ijongcheol (~ijongcheol@public.cloak) has joined #ua

<Greg> [17:07] jeanne Implementing UAAG 2.0 http://www.w3.org/TR/2013/WD-IMPLEMENTING-UAAG20-20131107/

<Greg> [17:09] -->| jeonghun (~jeonghun@public.cloak) has joined #ua

<Greg> [17:14] -->| Judy (jbrewer@public.cloak) has joined #ua

<Greg> [17:15] -->| wang (~wang@public.cloak) has joined #ua

<Greg> [17:16] Greg scribe: Greg

<Greg> [17:16] Judy thanks Greg!

<Greg> [17:16] -->| kimwoonyoung (~kimwoonyoung@public.cloak) has joined #ua

<Greg> [17:16] Judy welcomes people to IRC

<Greg> [17:16] Greg Three accessibility guidelines for this environment:

<Greg> [17:16] Greg ATAG applies to authoring tools.

<Greg> [17:16] Greg WCAG applies to content.

<Greg> [17:17] Greg UAAG applies to user agents: browsers, media players, etc.

<Greg> [17:18] Greg These work with W3C technical specifications for HTML, CSS, etc.

<Greg> [17:18] Greg Jim: What is a user agent?

<Greg> [17:18] Greg Any software that retrieves, renders, or facilitates interaction with web content.

<Greg> [17:19] jeanne A user agent is any software that retrieves, renders, and facilitates end user interaction with Web content.

<Greg> [17:19] jeanne Web browsers (desktop, embedded...)

<Greg> [17:19] jeanne Browser plug-ins and extensions

<Greg> [17:19] jeanne Media players for Web content

<Greg> [17:19] jeanne Applications for Web content

<Greg> [17:19] jeanne E-readers

<Greg> [17:19] Greg Includes add-ins and extensions.

<Greg> [17:19] Greg The user agent (web browser)'s main task is to help the user find out what's on the page, navigate to it, and manipulate it.

<Greg> [17:19] jeanne What is here?

<Greg> [17:19] jeanne Let me move to it.

<Greg> [17:19] jeanne Let me execute it!

<Greg> [17:19] jeanne Ultimate provider of the user

<Greg> [17:19] jeanne experience

<Greg> [17:19] jeanne True test of

<Greg> [17:19] jeanne accessibility

<Greg> [17:20] Greg The browser is the ultimate provider of the user experience. The author writes content, but it is rendered by the browser.

<Greg> [17:20] Judy ("execute" here means "do")

<Greg> [17:20] jeanne Defines how browsers, media players, and other "user agents" should support accessibility for people with disabilities and work with assistive technologies.

<Greg> [17:20] Judy ("mediator" = "help do something")

<Greg> [17:22] -->| kford (~chatzilla@public.cloak) has joined #ua

<Greg> [17:22] Greg Judy explains Last Call, where readers can provide feedback before it is finalized.

<Greg> [17:23] Greg Jeanne: particularly would like companies making browsers to review this draft and provide feedback.

<Greg> [17:23] Greg The draft document contains instructions for how to return comments. We greatly appreciate the effort you put in to read and comment on drafts.

<Greg> [17:24] Greg UAAG 2.0 is organized into five principals. The first three come from WCAG: user interface should be perceivable, operable, and understandable.

<Greg> [17:24] Greg We add two more: user agents should follow standards, and work with assistive technology.

<Greg> [17:26] Greg The Perceivable principle is broken down into individual guidelines, each of which has a set of success criteria.

<Greg> [17:26] Greg s/principals/principles/

<Greg> [17:28] Greg (Note that those of us on Skype cannot hear audience questions, so cannot minute them.)

<Greg> [17:29] Greg Q&A: Discussion of ways text presentation is customized by the user, such a color, size, and font choice.

<Greg> [17:29] Judy ...people were asking about text configuration

<Greg> [17:30] Judy ...and one person was demonstrating examples of text customization in Korean

<Greg> [17:30] Greg Jeanne demonstrates how text size is increased without increasing the size of images and other content, which is useful for many people. This contrasts with Zoom, where everything is enlarged.

<Greg> [17:31] Greg Jim: the slides from this presentation will be posted on the User Agent Working Group's web page.

<Greg> [17:32] Greg Jim demonstrates reflowing text to a single column, both on screen and when printed.

<Greg> [17:34] Greg Jeanne: The Perceivable principle also deal with configuring audio, including synthesized speech.

<Greg> [17:34] Greg Users should be able to create their own style sheets which override author styles.

<Greg> [17:35] Greg They should be able to manipulate the viewport, which is the box or region in which content is displayed. That includes visual and tactile presentations.

<Greg> [17:35] Judy "point of regard" = what you are looking at

<Greg> [17:35] Greg It's important to maintain the user's point of regard, so that when you change the size of the window or fonts, the point the user is looking at should remain visible.

<Greg> [17:36] Greg That is, the user should not lose their place when changing how the page is presented to them.

<Greg> [17:38] Greg Losing their place is confusing to users, particularly for users who have limited vision or some types of cognitive difficulties.

<Greg> [17:39] -->| stone (~stone@public.cloak) has joined #ua

<Greg> [17:39] Greg Also suggest providing access to other views, such as outline view and a source view that can serve as a last-resort for users who cannot understand the content as the browser renders it.

<Greg> [17:39] kford rrsagent, make minutes

<Greg> [17:40] Greg s/outline view/outline view,/

<Greg> [17:40] Greg The principle that the user interface must be operable by the user includes making sure it is usable with the keyboard alone.

<Greg> [17:41] Greg Jeanne notes that for all requirements in UAAG only apply to content that the user agent can recognize (that is, understand).

<Greg> [17:41] Greg s/for all/all/

<Greg> [17:43] Greg For example, if a heading is indicated only by font size rather than by appropriate HTML tags, then the browser cannot understand it is a heading, and thus cannot include it in the outline view.

<Greg> [17:44] Judy "DOM" = Document Object Model

<Greg> [17:44] Greg User agents are only expected to understand information that is marked up properly according to the document format (e.g. HTML).

<Greg> [17:46] Judy "markup aware" = a software tool that recognizes HTML code

<Greg> Jim: One important section we worked a long time on is keyboard use.

<Greg> The user should be able to do anything using the keyboard alone, although today that is not fully implemented in many user agents.

<Greg> Question about how one moves the mouse using the keyboard. The answer is that rather than moving the mouse with the keyboard, there should be keyboard commands to do the same operations and access the same information.

<Greg> An example is that the user can often see title text by hovering the mouse over something, but there is often no way to to this by moving keyboard focus to the element and pressing a keyboard command.

<Greg> Jeanne notes there are exceptions, such as those that involve moving the mouse along a path (drawing).

<Greg> Jeanne notes that properly authored content should provide keyboard methods of doing anything, but for badly written content the user agent should provide fallback mechanism, such as letting the user simulate click events into the HTML element that has the focus.

<KimPatch> yes, thanks

<Greg> Jim: The other thing about keyboard use is that the browser should provide a centralized list of all keyboard commands.

<KimPatch> no

<Greg> Jeanne: Users with many types of disabilities rely on keyboard access.

<Greg> Jeanne: When a web application adds keyboard commands, it is important for the user to be able to override those commands. For example, if they type with a single hand, they want to change keyboard commands to only use the half of the keyboard they can reach easily.

<Greg> Jim: Other guidelines in this principle include navigating sequentially through all the links (and similar items). This is for example useful when assistive technology adds scanning interfaces to the content.

<Greg> Jeanne: Many features of UAAG are about allowing the user to override things that the content author has specified.

<Greg> Other guidelines include the ability to save and restore sets of user preference settings. Ideally these should be portable between systems and devices.

<Greg> Jim: Also give the user the ability to customize the display of controls, such as choosing graphical buttons, textual labels, or both.

<Greg> Also override time-out periods, and help users avoid flashing that can trigger seizures.

<Greg> Jeanne explains how some types of flashing on the screen can trigger seizures in susceptible individuals.

<Greg> She gives examples of cases where television content has triggered seizures in viewers, such as in a Japanese animated television program and in Olympic coverage.

<Greg> We want the browser to let the user choose not to have movies play automatically, so the user can decide what and when they want the video played.

<Greg> (Example is the Flash Block extension for Firefox, which prevents Flash animations from playing until the user clicks on them.)

<Greg> Jeanne notes there is only so much browsers can do. Jim gives the example that browsers recognized when they're playing animated GIFs, and allows the user to pres the ESC key to stop the animation. The browser could also stop the animation when the GIF is first displayed, until the user chooses to start it.

<Greg> Jim: There should be mechanisms to help the user save and restore their preferences, and restore them to the default configuration if they become misconfigured. Ideally a wizard would help the user set them correctly for their own needs.

<Greg> Jim notes how he destroyed a cell phone by accidentally setting the colors to black on black, and provided no way to get it back to anything usable.

<Greg> In that case, a reset function would have helped immensely.

<Greg> Jeanne shows off an early, accessible video player that was designed for people who were elderly or had learning disabilities. Buttons were large and easy to operate.

<Greg> Jim: Web Content Accessibility Guidelines require video to have captions, but if the user cannot operate the button to display those captions, they do them no good. Thus User Agent Accessibility Guidelines require that all buttons be operable using the keyboard, not just the mouse.

<Greg> Jim: The content author should not have to write a lot of Javascript to make controls (buttons, etc.) usable with the keyboard. Instead, all native controls should be keyboard accessible.

<Greg> Ideally web pages should be able to use the browser's native controls, inheriting accessibility features, and skin them to their look and feel, rather than have to implement them on their own and risk inaccessibility.

<Greg> There are a lot of success criteria about media (video and audio). These include features for easy navigation, resizing, compatibility with assistive technology, and so forth.

<Greg> We need to think about future enhanced functionality in internet video, as well.

<Greg> Soon we will see links in video text, and it's not yet clear how users will be able to interact with those in accessible ways.

<Greg> For example, press Space to pause the video, use the Tab key to navigate to links in the captions or annotations, etc. This type of content is rare today but is coming.

<Greg> This functionality will be built into HTML5, and we're trying to anticipate these features and provide guidance to the browsers when they add these new features.

<Greg> Browsers compete to offer the best user experience, and that includes features like these that are very important to people with disabilities but also benefit many other users.

<Greg> The Understandable Principle includes features to help users avoid making mistakes, providing good documentation (e.g. being able to find keyboard commands), and ways to make the browser and content behave in predictable, consistant ways.

<Greg> Jim: An important things for users of assistive technology is that the browser cooperate with assistive technology.

<Greg> For example, the browser takes information from the Document Object Model and exposes it through a platform accessibility API (e.g. MSAA).

<Greg> That should work, and efficiently, so that assistive technology can treat the browser just like any other software on the platform, rather than having to talk to the DOM directly.

<Greg> Jeanne: The final principle in UAAG is that browsers should comply with existing specifications and conventions.

<Greg> Each platform has its own accessibility conventions, and browser should comply with those.

<Greg> Jim: Some of the benefits of UAAG is that it solves many common accessibility problems of web sites. The browsers should let the user work around those problems, rather than having to fix every web site.

<Greg> There are many other benefits of compliance with UAAG.

<Greg> A key thing to remember is that "The author proposes, the user disposes", which means the user gets to have final say on how they interact with web content.

<Greg> Challenges include the fact that there are many types of user agents, such as browsers, functionality embedded in devices, on mobile devices.

<Greg> There is the problem of things that the browser cannot recognized, such as functionality implemented entirely in Javascript.

<Greg> New media may introduce new features which can be useful or present problems.

<Greg> We ask all of you to please comment on the working draft for UAAG. We will address all comments, and in fact are required to do so by W3C procedures.

<Greg> If you have plans for writing extensions or adding features to your browsers to comply with these guidelines, we would really appreciate hearing about these.

<Greg> We need to have at least two examples of real products complying with each guideline.

<Greg> Jeanne: We can to China to talk with all of you, believe that exciting work is happening in Chinese browsers, more so in many cases than elsewhere.

<Greg> We hope accessibility can help teach you how to build better user experiences.

<Greg> If you lead the way, other browser manufacturers will be forced to catch up eventually.

<Greg> You are invited to join our working group and participate, through phone calls, a public mailing list, etc. There are many ways to communicate and participate.

Q&A

<Greg> The User Agent Working Group's home page is http://www.w3.org/WAI/UA/.

<Greg> The group is taking a 15-minute break.

<Jan1> http://www.tencent.com/en-us/ps/imservice.shtml

<Jan1> http://browser.qq.com/

<AllanJ> big blue button on the right

<Greg> Kim reports that QQ for iOS is in English.

<Greg> Mac version at http://browser.qq.com/mac/.

<Greg> http://www.tencent.com/en-us/ps/imservice.shtml says "Tencent Traveler is a multi-page webpage browser.", but it has apparently been renamed to the QQ Browser.

<Greg> The Windows version is based on IE.

<KimPatch> QQ iOS browser has a nice efficient scrolling control – you go a page at a time and you can slide the control up or down on the screen. Also a really nice rotor control in the bottom corner of the screen.

<rachel> http://se.360.cn/

<stone> http://browser.qq.com

<rachel> http://www.maxthon.cn/

<stone> http://www.liebao.cn/

<Greg> The working group put out a call for Asian browsers that we can download and test.

<kenny> http://www.uc.cn/browser/index.shtml from UCweb

<stone> http://browser.taobao.com/

<kenny> UCweb is for mobile device

<stone> http://www.fsllq.com/

<stone> we should invite more developer of the ua to tpac

<Greg> Jeanne stressed how building in accessibility features makes mainstream products better. She's gratified to see that some of these browsers already implement many accessibility features, either for users with disabilities or for the general population.

Q&A continued

<Greg> Jeanne: We do expect that many browsers will comply by relying on extensions.

<Greg> Q: If the browser installs extensions, is that OK for UAAG?

<Greg> A: Yes.

<Greg> Jeanne: For example, the HeadingsMap extension for Firefox makes it easy to navigate by headings.

<Greg> Jeanne: The Developer Toolbar extension for Firefox allows disabling CSS.

<Greg> Members of the audience say some of their browsers do not allow users to write extensions, but the developer can write and distribute them.

<Greg> Jeanne: That works well, since the developer can ensure that the extensions continue to work with new versions of the browser.

<Greg> Jeanne: Companies who want to make a public statement that their product meets UAAG, can post a document ("Conformance Claim") that can include a list of extensions which let the browser comply.

<Greg> Appendix A of UAAG describes how to write a Conformance Claim document.

Resuming work on the Implementations spreadsheet

<AllanJ> extensions can be problematic in that browsers can update and break extensions, then the user must wait for the developer to update the extension. better if feature is native to browser.

1.8.7

<Greg> 1.8.7 Reflow Text: The user can specify that text content in a graphical top-level viewport reflows so the text forms a single column that fits within the width of the viewport.

<Greg> One example with HTML/CSS is turning off author stylesheets, which reduces to one column.

<Greg> IE and Firefox have built-in features to allow turning off author-styles.

<Greg> Greg: Another example is Microsoft Word, which lets the user choose Normal viewing mode which turns off multi-column display.

<Jan1> http://help.adobe.com/en_US/acrobat/X/pro/using/WS58a04a822e3e50102bd615109794195ff-7d19.w.html

<Jan1> Adobe Acrobat X

<Jan1> "You can reflow a PDF to temporarily present it as a single column that is the width of the document pane. This reflow view can make the document easier to read on a mobile device or magnified on a standard monitor, without scrolling horizontally to read the text."

<Greg> Jim: Ideally if the browser had features to override author styling that would make content inaccessible, rather than requiring the user to write a user stylesheet.

1.8.8

<Greg> 1.8.8 Customize Viewport Highlighting: When highlighting viewports as specified by 1.8.1 Highlight Viewport, the user can customize attributes of the viewport highlighting mechanism (e.g. blink rate for blinking, color and width of borders). (Level AA)

<Greg> Jim: IE would highlight the active frame with a dotted rectangle, but it was very subtle.

<Greg> Ideally that should be customizable (e.g. thicker, solid, and bright color).

<Greg> (The same is true of Firefox.)

1.8.9

<Greg> 1.8.9 Allow Viewport Resize: The user can resize viewports within restrictions imposed by the platform, overriding any values specified by the author. (Level AA)

<Greg> Greg: An example is that Firefox and Chrome allow resizing multiline text input controls.

<AllanJ> js: advertising popup, with no scrollbars

1.8.10

<Greg> 1.8.10 Provide Viewport History: For user that implement a history mechanism for top-level viewports (e.g. "back" button), the user can return to any state in the viewport history that is allowed by the content, including a restored point of regard, input focus and selection. (Level AA)

<Jan1> re: 1.8.9 I tested FF25 with open new window, with set size and no scrollbars and I was able to resize.

<Jan1> re: 1.8.10 return to point of regard works in FF25

<Greg> Jan, apparently the author can override with window.open(resizable=no), e.g.

<Greg> popupWindow = window.open(

<Greg> url,'popUpWindow','height=700,width=800,left=10,top=10,resizable=yes,scrollbars=yes,toolbar=yes,menubar=no,location=no,directories=no,status=yes')

<Greg> Greg: Do any of the Asian browsers implement their own rendering engine?

<Greg> Answer, in Japan a project is creating a new rendering engine, Ecort.

<Jan1> Greg, FF seems to have been ignoring that for awhile http://stackoverflow.com/questions/1715201/how-can-we-disable-resizing-of-new-popup-window-in-firefox

<AllanJ> ACTION: jallan to raise why 1.8.10 is AA. this is very easy. should be A. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action01]

<trackbot> Created ACTION-921 - Raise why 1.8.10 is aa. this is very easy. should be a. [on Jim Allan - due 2013-11-19].

<Jan1> AllanJ, re: AA...just because easy...doesn't mean AA...also failing has to have quite major impact on users.

<jeanne> Sogouu, QQ,

<jeanne> scribe: jeanne

<kurosawa> Ecort or escude browesr information in English http://code.google.com/p/es-operating-system/wiki/UsingEscort

<AllanJ> jr: for A it is not just because it is easy, but will make a major accessibility issue if not implemented

<AllanJ> gl: different opinion. if a feature is widely implemented, it is generally available in all new browsers, then might make it an A.

1.8.11, 12, 13

<Jan1> wording is "When I open a link in a new tab, switch to it immediately"... in FF25

<AllanJ> ACTION: greg to raise 1.8.11 wording [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action02]

<trackbot> Created ACTION-922 - Raise 1.8.11 wording [on Greg Lowney - due 2013-11-19].

<Greg> For later investigation, I’m a bit concerned that disabling opening new top-level viewports (1.8.11) could break the user experience, by forcing the user to leave a page when the author didn’t expect it, may be that on returning you have lost some author-kept state.

<AllanJ> jr: ff has built in pop-up blockers, provide information about holding control key and click to open new window.

<Jan1> FF25 has a builtin popup blocker under Options>Content https://support.mozilla.org/en-US/kb/pop-blocker-settings-exceptions-troubleshooting

<KimPatch> Many mobile browsers have pop-up blockers including Mercury and Atomic Web

<Jan1> Chrome has a popup blocker https://support.google.com/chrome/answer/95472?hl=en

<AllanJ> kenny: qq browser, need to test

<Greg> In Opera, Right click page > Edit Site Preferences > General tab > Pop-ups

1.8.13

<AllanJ> audience: allow users to customize the user interface of the browser (including toolbars)?

<Greg> By the way, in theory a user stylesheet can include "div {column-count: 1 !important; }", and a few other values such as column-width, to disable multi-column layout in CSS.

<AllanJ> jr: one way to do this is to tell browser to always open everything in a tab

<AllanJ> kf: several browsers have this feature.

<AllanJ> jr: new tab for @target=_blank will open in a new tab

<AllanJ> ... but if author uses javascript to open a new window, the browser opens an new window.

<AllanJ> gl: if they use javascript to open a new window, the browser recognizes function to open a new window.

<AllanJ> ... browser should implement only opening in a new tab NOT a window.

<AllanJ> Kim Patch is co-chair of Mobile Accessibility Taskforce

<AllanJ> we will try to have a lunch meeting about mobile accessibility...informal

<Greg> A web discussion it says "Opera 12.x can open pop-ups in a new tab, Opera 15 will open them in a window if they call for a specific size when they open."

1.8.14 webpage bookmarks

<AllanJ> important for voice input users.

<AllanJ> JAWS and WindowEyes does this, but no browsers that do.

<AllanJ> kp: webapps do this...e.g. google docs

<AllanJ> gl: this would not be difficult to implement as an extension, or as a user script installed by greasemonkey

1.9.1 and 1.9.2

<AllanJ> js: 1.9.2 is AAA because although easy to do, it is only useful for very knowledgeable users

<KimPatch> Several mobile browsers allow you to view source one example is Atomic Web

<kurosawa> An extension of Firefox for Android for S.C. 1.9.2 https://addons.mozilla.org/android/addon/view-source-mobile/

1.10.1

<AllanJ> js: screen readers do this.

<AllanJ> ... would like some mechanism in browsers to provide related elements.

JA: There is an extension that shows headers related to a specific cell

<AllanJ> tabletools2 extension for ff

GL: I think you are talking about Tabletools2

<Greg> The TableTool2 extension for Firefox is at http://www.mingyi.org/.

<kurosawa> Sony's Android Browser does something for label http://www.quirksmode.org/blog/archives/2013/10/a_brilliant_son.html

<KimPatch> Jim: one way to see the element hierarchy

<KimPatch> Jim: click here and it automatically expands the entire hierarchy – it does make it easy but most users would have no reason to go to developer tools

<KimPatch> Jim: developer tools for most browsers including Firefox and IE (press F 12) also in QQ

<KimPatch> Jeanne: this finishes principal 1 so let's stop here

<KimPatch> Jim: back an hour from now

<AllanJ> back from lunch

<AllanJ> scribe: allanj

js: talking about IndieUI. Speech input is going to present unique issues

2.1.1 Provide Full Keyboard Functionality

stone: demonstrated a mouse with keyboard modification keys, to select an area of a webpage and take a screenshot.

ja: can see a way to do that functionality with keyboard only
... using context menu "screen shot" then mark top left and bottom right corners to define region, then hit enter to capture the region

js: explaining drag and drop with keyboard
... space to highlight, arrow keys or tab to move, and space to drop

<scribe> ACTION: jallan to raise adding 'user context' profile part of indieUI as part of UA profiles and preferences [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action03]

<trackbot> Created ACTION-923 - Raise adding 'user context' profile part of indieui as part of ua profiles and preferences [on Jim Allan - due 2013-11-19].

<jeanne> scribe: jeanne

2.1.1

KF: I think IE has close to full functionality for basic functions -- tab and arrow. I can't think of anything that is jumping out that can't be done with the keyboard.

GL: I can't think of anything that can't be accessed by keyboard

Stone: the demonstration I did earlier was for the operating system, not the browser

<AllanJ> gl: classic example is the "title" attribute popping up from the keyboard

GL: The classic example is to see if you can access the title of an element (tooltip)

JA: I don't think anyone does it.

Kurosawa: I think IE 11 does it.

JA: Some people are doing it with javascript, so that may be what you are doing.

s/that may be what you are doing. /that may be what you are seeing.

<Greg> Firefox does not seem to provide this, other than using Inspect Element, which is extremely technical.

JA: It would be good if the browser took care of it, and had a preference to turn that on and off.
... they say that it is too distracting to show the tooltip with the keyboard.

JS: I can see it, because you typically don't run the mouse over all the website, but a keyboard user will have to tab across the page.

GL: There are some edge cases where tooltips would decrease accessibility

JA: If you open a tooltip, but it extends out of sight so you would need the mouse to move over to get it to appear. You could not do it from the keyboard.

GL: Do we address this in uAAG 2/-

JA: not in this detail

GL: We want to consider this for UAAG.next

<AllanJ> gl: perhaps use a context menu to specifically ask for the tooltip to be displayed

2.1.2

2.1.2 Show Keyboard Focus:

Every viewport has an active or inactive keyboard focus at all times. (Level A)

JA: I think everyone does this.

<AllanJ> ACTION: jallan to review 2.1.2 active and inactive focus and viewports. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action04]

<trackbot> Created ACTION-924 - Review 2.1.2 active and inactive focus and viewports. [on Jim Allan - due 2013-11-19].

<AllanJ> gl: if you wanted a screen reader to read an inactive windows?

<AllanJ> gl: can you use a screenreader to read inactive windows.

<AllanJ> kf: technically you can do it. screenreaders allow you to read inactive windows through advanced functions

<AllanJ> gl: example - set screenreader to check status bar of inactive window for particular information.

<AllanJ> ... in X windows, there are insertion bars in active and inactive windows.

2.1.3

<AllanJ> 2.1.3 Avoid Keyboard Traps

2.1.3 Avoid Keyboard Traps:

If keyboard focus can be moved to a component using a keyboard interface (including nested user agents), then focus can be moved away from that component using only a keyboard interface. If this requires more than unmodified arrow or Tab keys (or standard exit methods like Escape), users are advised of the method for moving focus away. (Level A)

JA: This is a legacy problem where Flash used to take control of the browser controls and you could Tab in and Shift Tab to get to the top. IE uses Alt-D to take you back to the address bar.

GL: Even if Flash has fixed this problem, it doesn't mean that all plugins have.

JA: If the author sets an accesskey for alt-D, does the browser override alt-D?

KF: The last time I was involved, IE would not let an author set the alt-D accesskey.

2.1.4

2.1.4 Separate Selection from Activation:

The user can specify that focus and selection can be moved without causing further changes in focus, selection, or the state of controls, by either the user agent or author-supplied content. (Level A)

JA: new radio buttons are not selected. If the keyboard user moves through the radio buttons, it automatically selects it. There is no way to de-select it.
... I don't know any browser that lets you go into a radio button through the keyboard and not select it.
... the state is in the DOM, so it could be possible for an extension to clear that state.

GL: BUt that would not meet the SC as stated, which is specific that the movement does not trigger the selection.

JA: Screenreaders do this, because they examine the content outside of the browser.

<AllanJ> kf: we may get pushback on this. radio button selection on focus is core behavior for most if not all browsers

JS: THis is a level A. If none of the browsers do this, we have to decide whether we can keep this at level A.

<AllanJ> ... at a level A will be tough. seem to remember that FF may allow none selection on focus.

2.1.5

2.1.5 Follow Text Keyboard Conventions:

The user agent follows keyboard conventions for the operating environment. (Level A)

KF: This is a freebie -- everyone does this.

2.1.5 Follow Text Keyboard Conventions:

The user agent follows keyboard conventions for the operating environment. (Level A)

2.1.6 efficient keyboard access

<AllanJ> kf: this is not testable, how to you measure efficiency

<AllanJ> js: count keystrokes for direct access vs sequential access

<AllanJ> kf: how do you quantify this...if I do 10 things but not 11 does my browser fail?

<AllanJ> kf: Windows menus has 2 levels of shortcuts - save page = control s and alt-f s to select save from menu item

JA: Most browsers have this feature, although some people may criticize that it cannot be tested because we do not say how many mechanisms are needed to comply.

2.2.1

2.2.1 Sequential Navigation Between Elements:

The user can move the keyboard focus backwards and forwards through all recognized enabled elements in the current viewport. (Level A)

<AllanJ> @@revisit intent for 2.1.6 seem more of an example

<kurosawa> https://bugs.webkit.org/show_bug.cgi?id=94707

<AllanJ> webkit ignores absolutely positioned elements.

<AllanJ> can click with a mouse but not get focus from keyboard

<AllanJ> gl: make clear that enabled elements include javascript actions

<jeanne3> kf: The screenreaders have done this as a way to solve this problem.

<AllanJ> kf: must think about this. if javascript events (on click) done get keyboard focus

<jeanne3> JA: the screenreader has a feature that searches for clickable events, but if you aren't using a screenreader, you won't be able to tab to it, if it is a javascript link.

<jeanne3> KF: Authors shouldn't do this, it is a WCAG violation.

<AllanJ> jaws has keys to move to clickable events. also ask user do you want dom event or mouse click

<jeanne3> JA: This is a case where WCAG and UAAG work together.

<AllanJ> UAAG is last line of defense for the user to make things work

<jeanne3> GL: An extension could search the DOM for any events that are onclick or onmouseover and wrap them with an an anchor tag with onfocus or onpress event.

<jeanne3> KF: You could add extensions easily to compensate for this problem.

2.2.2

<jeanne3> 2.2.2 Sequential Navigation Between Viewports:

<jeanne3> The user can move the keyboard focus backwards and forwards between viewports, without having to sequentially navigate all the elements in a viewport. (Level A)

<jeanne3> JA: All desktop browsers do this, I don't know about mobile.

2.2.3

<jeanne3> 2.2.3 Default Navigation Order:

<jeanne3> If the author has not specified a navigation order, the default sequential navigation order is the document order. (Level A)

<jeanne3> JA: Every browser does this.

<jeanne3> ... unless the author wrote a tabindex, everyone does this.

2.2.4

<jeanne3> 2.2.4 Options for Wrapping in Navigation:

<jeanne3> The user can request notification when sequential navigation wraps at the beginning or end of a document, and can prevent such wrapping. (Level AA)

<jeanne3> GL: I recall we did some discussion of this that isn't in the document.

<jeanne3> ... there is an extension that does part of it.

<jeanne3> GL: I may have confused this with search wrapping notification.

<jeanne3> JA: no one does it.

2.3.1

<jeanne3> 2.3.1 Allow Direct Navigation to Important Elements:

<jeanne3> The user can navigate directly to important elements in rendered content. (Level AA)

<Greg> Re 2.2.4, see the Find Toolbar With Alert Box extension (https://addons.mozilla.org/en-US/firefox/addon/find-toolbar-with-alert-box).

<jeanne3> JA: we added landmarks. There is an extension to navigate by headings. There is a extension that does search by ARIA landmarks.

2.3.2

<jeanne3> 2.3.2 Present Direct Commands from Rendered Content:

<jeanne3> The user can have any recognized direct commands in rendered content (e.g. accesskey, landmark) be presented with their associated elements (e.g. Alt+R to reply to a web email). (Level AA)

<jeanne3> JA: If there is an accesskey, there should be an indication in the user interface to indicate that. Opera and FF have extensions for this

2.3.3

<jeanne3> 2.3.3 Allow Direct Activation of Enabled Elements:

<jeanne3> The user can move directly to and activate any enabled element in rendered content. (Level A)

<Greg> Mouseless Browsing extension for Firefox.

<AllanJ> lynx browser had numbered links and form elements

<jeanne3> JA: This SC is very important for speech input users. It numbers every link and allows the user to speak the number of the link and move directly to them.

<Greg> Mouseless Browsing (MLB) enables you to browse the web entirely with the keyboard. The basic principle is to add small boxes with unique ids behind every link and/or form element. ... https://addons.mozilla.org/En-us/firefox/addon/mouseless-browsing/

<jeanne3> 'topic: 2.3.4

2.3.4

2.3.4 Present Direct Commands in User Interface:

The user can have any direct commands in the user agent user interface (e.g. keyboard shortcuts) be presented with their associated user interface controls (e.g. "Ctrl+S" displayed on the "Save" menu item and toolbar button). (Level AA)

JA: Everyone does this.

2.3.5 Allow Customized Keyboard Commands:

The user can remap any keyboard shortcut including recognized author supplied shortcuts (e.g. accesskeys) and user agent user interface controls, except for conventional bindings for the operating environment (e.g. arrow keys for navigating within menus). (Level AA)

2.3.5

JA: You should be able to remap the keyboard shortcuts or accesskeys.
... I think opera does this.

GL: Opera -> settings -> Preferences -> Advanced -> Shortcuts
... export and import would be nice.

JA: it has a Default button.

<AllanJ> rrsagent make minutes

<AllanJ> ACTION: jallan to review spreadsheet to find 'extension' and input urls [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action05]

<trackbot> Created ACTION-925 - Review spreadsheet to find 'extension' and input urls [on Jim Allan - due 2013-11-19].

<scribe> ACTION: jeanne to fix 1.8.10 to change user to user agent. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action06]

<trackbot> Created ACTION-926 - Fix 1.8.10 to change user to user agent. [on Jeanne F Spellman - due 2013-11-19].

<AllanJ> we are back

<kurosawa> ./join #svg

<AllanJ> js: security applications that lock the keyboard,

<AllanJ> ... if extension or part of the browser then UAAG applies

<AllanJ> ... if an OS level application then UAAG does not apply

<AllanJ> js: WCAg to ICT shows that WCAG can apply to software.

http://www.w3.org/TR/wcag2ict/

<Greg> I would like us to consider whether we actually introduce a new requirement into the "Limited Conformance for Extensions" section of the Conformance section, when it really should be a Success Criterion in the main body of the document.

<Greg> That is regarding the sentence "The add-in must not cause the combined user agent (hosting user agent plus installed extension or plug-in) to fail any success criteria that the hosting user agent would otherwise pass."

<AllanJ> js: this seems to apply to all SC, it is an uber SC

<AllanJ> gl: this is not very different from all features must be keyboard accessible.

<AllanJ> ACTION: greg to consider whether we actually introduce a new requirement into the "Limited Conformance for Extensions" section of the Conformance section, when it really should be a Success Criterion in the main body of the document. regarding the sentence "The add-in must not cause the combined user agent (hosting user agent plus installed extension or plug-in) to fail any success criteria... [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action07]

<trackbot> Created ACTION-927 - Consider whether we actually introduce a new requirement into the "limited conformance for extensions" section of the conformance section, when it really should be a success criterion in the main body of the document. regarding the sentence "the add-in must not cause the combined user agent (hosting user agent plus installed extension or plug-in) to fail any success criteria... [on Greg Lowney - due

<trackbot> ... 2013-11-19].

<AllanJ> ...that the hosting user agent would otherwise pass."

<AllanJ> audience: speech synthesis...web speech API, FFOS is implementing speech synthesis use web-speech API

<kurosawa> https://dvcs.w3.org/hg/speech-api/raw-file/tip/speechapi.html

http://www.w3.org/community/speech-api/

<AllanJ> jim says WOW!

<AllanJ> FFOS using PICOTTS for speech. doesn't work on windows only android OS

<AllanJ> FFOS is still work in progress, a11y is improving

<AllanJ> ACTION: jallan to review speeech-api for inclusion in UAAG in Implementing Document https://dvcs.w3.org/hg/speech-api/raw-file/tip/speechapi.html [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action08]

<trackbot> Created ACTION-928 - Review speeech-api for inclusion in uaag in implementing document https://dvcs.w3.org/hg/speech-api/raw-file/tip/speechapi.html [on Jim Allan - due 2013-11-19].

<AllanJ> gl: if there are implementations that include tone or not, does not mean we should not consider it

2.4.1

<AllanJ> Text Search: The user can perform a search within rendered content, including rendered text alternatives and rendered generated content, for any sequence of printing characters from the document character set. (Level A)

<scribe> ACTION: jeanne to check the code in 1.2.2 and find out why the first bullet item isn't recognized as a bullet by Greg's php script. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action09]

<trackbot> Created ACTION-929 - Check the code in 1.2.2 and find out why the first bullet item isn't recognized as a bullet by greg's php script. [on Jeanne F Spellman - due 2013-11-19].

JA: Blind users can search a page for alternative text with a screenreader, but users who are not using assistive technology can only search the visible text.

<AllanJ> other than screenreader don't know of anything that lets you search on alternative content (alt, title, captions)

JA: Another use case is that I put an acronym on the page, like IBM (International Business Machines). So the in page search can look for International Business Machines and see if it finds it.

2.4.2

2.4.2 Search Direction:

The user can search forward or backward in rendered content. (Level A)

JA: Pretty much every browser lets you do that.

2.4.3

2.4.3 Match Found:

When a search operation produces a match, the matched content is highlighted, the viewport is scrolled if necessary so that the matched content is within its visible area, and the user can search from the location of the match. (Level A)

JA: Everyone does it. In Chrome, it puts a mark in the scrollbar and you can click on the yellow line, and it will bring you to that location in the document.

GL: It is non standard and it is not exposed to assistive techology. They need to find a better way to give it to you.
... FF has a good example.

2.4.4

<Greg> That is, they don't have to remove the scroll bar method, but provide an alternative (e.g. a separate pane).

2.4.4 Alert on Wrap or No Match:

The user can choose to receive notification when there is no match to a search operation. The user can choose to receive notification when the search continues from the beginning or end of content. (Level A)

GL: they don't fully meet them. This is the extension I was discussing earlier, it's a little buggy, but it does address it.

<AllanJ> https://addons.mozilla.org/en-US/firefox/addon/find-toolbar-with-alert-box

2.4.5

2.4.5 Alternative Content Search:

The user can perform text searches within alternative content that is text (e.g. text alternatives for non-text content, captions) even when the alternative content is not rendered onscreen. (Level AA)

JA: the difference between 2.4.1 and 2.4.5 is that the content is rendered in 2.4.1 and in 2.4.5 it does not have to be rendered.

<Greg> I suggest we encourage user agents to comply with 2.4.5 as an option, rather than all the time.

<Greg> So that one can search a page that has a lot of non-rendered alternative content without getting hung up on all the alt text.

<AllanJ> page with test for screen reader users that is moved off screen by CSS. how to comply with 2.4.3 move the viewport to show match

JA: Youtube will allow you to search for captions

GL: IE does not find 2.4.1 with alt text displayed -- it doesn't search the alternative text even thought hte text is on the screen.

JA: Same thing with FF.

GL: Opera is the same as IE -- when it shows the alt text, it doesn't search it.

Boryong: The summary of the table should be searchable.

JA: In FF, it is difficult to turn off images.
... BUT when you use the Developer Toolbar to display alt text, the alternative text becomes searchable. It takes the alternative text and wraps it in a span, and makes it plain text on the page.

<Greg> http://images.apple.com/itunes/features/images/features_discover_hero.jpg

<scribe> meeting: UAWG Face to Face

<AllanJ> gl: this is specifically for media players

<AllanJ> ja: but apple could rewrite iTunes to only work in the browser

<AllanJ> kim: its a way of providing context for the user so they know where they are (headings)

<AllanJ> meta-information - key words, visual arrangement of information on the screen, relationship must be made available in text

<AllanJ> rdf semantic web

<AllanJ> jeanne to check with Hiroki about RDF iTunes

<Greg> This is not for general purpose web browsers, but for special-purpose user agents that are conveying information to the user purely by spatial relationships between things. If a web page did that, the browser could not know about it (it is not "recognized"), and therefore this SC does not apply. However, the iTunes media player knows the genre, artist, and album, and conveys those in...

<Greg> ...separate columns oriented from left to right, and only that layout tells the user that selecting something in one column provides information about the the entries in the columns farther to the right. Because the iTunes player understands this, it can repeat that as text as well, such as by repeating the genre, artist, and album along with the picture of the selected album.

<AllanJ> http://www.w3.org/WAI/UA/

Summary of Action Items

[NEW] ACTION: greg to consider whether we actually introduce a new requirement into the "Limited Conformance for Extensions" section of the Conformance section, when it really should be a Success Criterion in the main body of the document. regarding the sentence "The add-in must not cause the combined user agent (hosting user agent plus installed extension or plug-in) to fail any success criteria... [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action07]
[NEW] ACTION: greg to raise 1.8.11 wording [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action02]
[NEW] ACTION: jallan to raise adding 'user context' profile part of indieUI as part of UA profiles and preferences [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action03]
[NEW] ACTION: jallan to raise why 1.8.10 is AA. this is very easy. should be A. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action01]
[NEW] ACTION: jallan to review 2.1.2 active and inactive focus and viewports. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action04]
[NEW] ACTION: jallan to review speeech-api for inclusion in UAAG in Implementing Document https://dvcs.w3.org/hg/speech-api/raw-file/tip/speechapi.html [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action08]
[NEW] ACTION: jallan to review spreadsheet to find 'extension' and input urls [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action05]
[NEW] ACTION: jeanne to check the code in 1.2.2 and find out why the first bullet item isn't recognized as a bullet by Greg's php script. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action09]
[NEW] ACTION: jeanne to fix 1.8.10 to change user to user agent. [recorded in http://www.w3.org/2013/11/12-ua-minutes.html#action06]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2013-11-12 09:28:17 $

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/ESCORT/Ecort/
Succeeded: s/Sogo/Sogou/
Succeeded: s/Sogo/Sogou/
Succeeded: s/leve /level/
Succeeded: s/levelview/level view/
Succeeded: s/difficult, perhaps with /difficult to implement as an extension, or as a user script installed by /
FAILED: s/that may be what you are doing. /that may be what you are seeing./
Succeeded: s/windows/Windows menus/
Succeeded: s/onfocus/an anchor tag with onfocus/
Found Scribe: jeanne
Inferring ScribeNick: jeanne
Found Scribe: allanj
Inferring ScribeNick: AllanJ
Found Scribe: jeanne
Inferring ScribeNick: jeanne
Scribes: jeanne, allanj
ScribeNicks: jeanne, AllanJ
Present: jeanne jim greg kim kelly Jan judy kurosawa boryung ijongcheol kenny rachel wang kimwoonyoung jeonghun stone
Regrets: eric
Found Date: 12 Nov 2013
Guessing minutes URL: http://www.w3.org/2013/11/12-ua-minutes.html
People with action items: greg jallan jeanne

WARNING: Possible internal error: join/leave lines remaining: 
        <Greg> [17:03]    -->| kenny (~kenny@public.cloak) has joined #ua
        <Greg> [17:05]    -->| boryung (~boryung@public.cloak) has joined #ua
        <Greg> [17:05]    -->| rachel (~rachel@public.cloak) has joined #ua
        <Greg> [17:07]    -->| ijongcheol (~ijongcheol@public.cloak) has joined #ua
        <Greg> [17:09]    -->| jeonghun (~jeonghun@public.cloak) has joined #ua
        <Greg> [17:14]    -->| Judy (jbrewer@public.cloak) has joined #ua
        <Greg> [17:15]    -->| wang (~wang@public.cloak) has joined #ua
        <Greg> [17:16]    -->| kimwoonyoung (~kimwoonyoung@public.cloak) has joined #ua
        <Greg> [17:22]    -->| kford (~chatzilla@public.cloak) has joined #ua
        <Greg> [17:39]    -->| stone (~stone@public.cloak) has joined #ua



WARNING: Possible internal error: join/leave lines remaining: 
        <Greg> [17:03]    -->| kenny (~kenny@public.cloak) has joined #ua
        <Greg> [17:05]    -->| boryung (~boryung@public.cloak) has joined #ua
        <Greg> [17:05]    -->| rachel (~rachel@public.cloak) has joined #ua
        <Greg> [17:07]    -->| ijongcheol (~ijongcheol@public.cloak) has joined #ua
        <Greg> [17:09]    -->| jeonghun (~jeonghun@public.cloak) has joined #ua
        <Greg> [17:14]    -->| Judy (jbrewer@public.cloak) has joined #ua
        <Greg> [17:15]    -->| wang (~wang@public.cloak) has joined #ua
        <Greg> [17:16]    -->| kimwoonyoung (~kimwoonyoung@public.cloak) has joined #ua
        <Greg> [17:22]    -->| kford (~chatzilla@public.cloak) has joined #ua
        <Greg> [17:39]    -->| stone (~stone@public.cloak) has joined #ua



[End of scribe.perl diagnostic output]