Archive index

A11y Slackers Gitter Channel Archive 15th of October 2015

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    Oct 15 00:00
    [ldavis, a11y] @marcysutton: a11y test failed: The purpose of each link should be clear from the link text - https://github.com/GoogleChrome/accessibility-developer-tools/wiki/Audit-Rules#ax_text_04
  2. zakim-robot
    Oct 15 00:01
    [ldavis, a11y] link in question is "<a class="classnamehere" ></a>"
  3. zakim-robot
    Oct 15 00:01
    [Marcy Sutton, a11y] Yeah, sounds like you’re missing text in the anchors. If they are intended to be used/filled out later, they should have display: none or be hidden by other means
  4. zakim-robot
    Oct 15 00:01
    [Marcy Sutton, a11y] Otherwise, you could put a visually-hidden span with text in it or an aria-label attribute with text for the anchor.
  5. zakim-robot
    Oct 15 00:01
    [ldavis, a11y] Okay interesting. Let me try that
  6. zakim-robot
    Oct 15 00:02
    [Marcy Sutton, a11y] http://a11yproject.com/posts/how-to-hide-content/
  7. zakim-robot
    Oct 15 00:06
    [Alice Boxhall, a11y] Oh - it should look for href though
  8. zakim-robot
    Oct 15 00:06
    [ldavis, a11y] display: none didn't seem to change the outcome of the test, but I'll give that post a read through. Much appreciated!
  9. zakim-robot
    Oct 15 00:06
    [Alice Boxhall, a11y] otherwise it's not a link
  10. zakim-robot
    Oct 15 00:07
    [Alice Boxhall, a11y] (the audit rule should, I mean)
  11. zakim-robot
    Oct 15 00:07
    [Alice Boxhall, a11y] fixes
  12. zakim-robot
    Oct 15 00:08
    [ldavis, a11y] yeah I was looking into that too and it appears an anchor without a href is legal from a purely html perspective, but maybe not in the accessibility tooling? Unfortunately I don't control this content, so trying to figure out what to suggest to these folks.
  13. zakim-robot
    Oct 15 00:09
    [Alice Boxhall, a11y] I'm fixing the audit
  14. zakim-robot
    Oct 15 00:09
    [Alice Boxhall, a11y] It's a false positive
  15. zakim-robot
    Oct 15 00:09
    [ldavis, a11y] Ah okay
  16. zakim-robot
    Oct 15 00:09
    [Marcy Sutton, a11y] ahh yeah that was missing an href eh
  17. zakim-robot
    Oct 15 00:09
    [ldavis, a11y] thanks @alice
  18. zakim-robot
    Oct 15 00:10
    [ldavis, a11y] Unfortunately I have to run, but will pick this up in the AM. Thanks again for your suggestions and insights!
  19. zakim-robot
    Oct 15 00:10
    [Alice Boxhall, a11y] Will ping back here when the fix is in :simple_smile:
  20. zakim-robot
    Oct 15 00:10
    [ldavis, a11y] great!
  21. zakim-robot
    Oct 15 00:12
    [Alice Boxhall, a11y] (is this for the extension or the library? @ldavis)
  22. zakim-robot
    Oct 15 00:12
    [Alice Boxhall, a11y] it'll take a little longer to roll into the extension
  23. zakim-robot
    Oct 15 00:23
    [Alice Boxhall, a11y] GoogleChrome/accessibility-developer-tools#246 - waiting on @rickb to take a look :simple_smile:
  24. zakim-robot
    02:50
    [ldavis, a11y] @alice: Not the extention, just the library.
  25. zakim-robot
    02:51
    [ldavis, a11y] Thanks for the quick turn around! Hopefully that will get merged quickly and I'll be off to the races on that one.
  26. zakim-robot
    02:51
    [Alice Boxhall, a11y] @ldavid
  27. zakim-robot
    03:42
    [Rick Brown, a11y] @alice will do so asap this evening after work so should be waiting for you first thing in the morning your time
  28. zakim-robot
    03:43
    [Alice Boxhall, a11y] @rickb: Thanks! (It's a pretty quick one.)
  29. zakim-robot
    04:01
    [Rick Brown, a11y] Ok yeah that was small - took a coffee break and got it done. It's a good fix, glad it was raised, thanks @ldavis
  30. zakim-robot
    04:45
    [jitendra, a11y] how to make a button work without js
  31. zakim-robot
    04:46
    [jitendra, a11y] we can add event to button by adding a class
  32. zakim-robot
    04:46
    [jitendra, a11y] but what about non-js condition
  33. MichielBijl
    09:06
    @jitendra, <input type="submit"> ;)
  34. MichielBijl
    09:13
    @Claire Ryberg: I remember having a lengthy discussion about this during an PF call; just can't remember the outcome… It sound logical though, that I know.
  35. MichielBijl
    09:16
    I think it will just say something like “current item in set”
  36. MichielBijl
    09:16
    Since there is no other context; not much else to say I guess
  37. zakim-robot
    09:17
    [jitendra, a11y] input type submit is for form only? no?
  38. MichielBijl
    09:18
    Yes it is
  39. MichielBijl
    09:18
    Not sure what the action is you're looking for
  40. MichielBijl
    09:20
    What are you trying to achieve Jitendra?
  41. zakim-robot
    09:40
    [jitendra, a11y] @michiel: there is a button
  42. zakim-robot
    09:40
    [jitendra, a11y] after user click/tap on button, it opens a list of other options
  43. MichielBijl
    09:41
    You could use :target
  44. MichielBijl
    09:43
    http://codepen.io/Michiel/pen/625171e4f5fd3126b338abb77689024b
  45. MichielBijl
    09:43
    But then you'd have to use an anchor
  46. zakim-robot
    09:43
    [jitendra, a11y] and anchor should not be used as a button right?
  47. MichielBijl
    09:44
    Right
  48. MichielBijl
    09:44
    But how does the behaviour differ between no-js and with js enabled?
  49. zakim-robot
    09:46
    [jitendra, a11y] if we add a class to button to do something with javascript
  50. zakim-robot
    09:46
    [jitendra, a11y] that solution won’t work without js
  51. MichielBijl
    09:50
    True, I'm trying to figure out what it is you want the button to do. If it's just showing a list, you can use an anchor (as it links to other content). Alternatively you could use a checkbox to toggle the list.
  52. MichielBijl
    09:50
    And if you can do it without JS in the first place, why add JS afterwards?
  53. zakim-robot
    09:50
    [jitendra, a11y] it’s dynamic data
  54. zakim-robot
    09:51

    [jitendra, a11y] each button will be associated with a unique url

    to assist with normal loading and push state

    and the data will be lazy-loaded

    because it can be a lot of data'

  55. MichielBijl
    09:52
    So how does the lazy loading work if there is no JS?
  56. MichielBijl
    09:53
    Does the page load everything at once?
  57. zakim-robot
    09:55
    [jitendra, a11y] page reload
  58. zakim-robot
    09:55
    [jitendra, a11y] by clicking on button
  59. MichielBijl
    09:57
    That sounds like a link to me
  60. MichielBijl
    09:57
    If you really want to use a button in the JS version, just make it a button with JS.
  61. MichielBijl
    09:58
    Either set role="button" or make it an actual button; I'd recommend the latter for obvious reasons.
  62. MichielBijl
    09:58
    But that is without knowing all the specifics :)
  63. zakim-robot
    10:00
    [jitendra, a11y] thanks i think <a role=button> will do
  64. MichielBijl
    10:01
    Be sure to add spacebar activation.
  65. MichielBijl
    10:01
    And add tabindex=0 if you remove the href to keep it tabable :)
  66. zakim-robot
    10:08
    [jitendra, a11y] is it ok to remove href from anchor?
  67. MichielBijl
    10:11
    An <a>some text</a> is valid html, it's not focusable, but <a tabindex="0">some text</a> is.
  68. MichielBijl
    10:12
    If you have <a tabindex="0" role="button">some text</a> it is essentially the same as <button type="button">some text</button>; expect for spacebar control, so you should add that!
  69. zakim-robot
    10:15
    [jitendra, a11y] ok coo
  70. zakim-robot
    10:15
    [jitendra, a11y] cool
  71. zakim-robot
    10:16
    [jitendra, a11y] i didn’t know this that href can be removed from a
  72. zakim-robot
    10:16
    [jitendra, a11y] I used to think that it’s is mandatory
  73. stevefaulkner
    12:11
    may be of interest: Notes on use of multiple ARIA role attribute values https://www.paciellogroup.com/blog/2015/10/notes-on-use-of-multiple-aria-role-attribute-values/
  74. StommePoes
    12:19
    @stevefaulkner reading it now, I think I like the idea of fallback roles. And then I want a validator or browser debug tools to alert developer-me that I did something stupid like role="main link"
  75. StommePoes
    12:20
    similar to if I try to assign a role that really conflicts with a native role, when I should not have assigned it
  76. StommePoes
    12:20
    <hx role="button">clicky header</hx> etc
  77. StommePoes
    12:21
    wait, "toggle button" gets translated to "switch" in supporting browsers?
  78. MichielBijl
    12:24
    I didn't even know
  79. MichielBijl @MichielBijl questions everything again
  80. MichielBijl
    12:24
    Thanks a lot Steve…
  81. MichielBijl
    12:24
    ;)
  82. MichielBijl
    12:25
    @StommePoes did you see Heydon's tweet about webdev terminology?
  83. zakim-robot
    12:29
    [callumacrae, a11y] @jitendra: href isn't mandatory, but the default user agent styling for the cursor change is applied to a[href], not just a
  84. MichielBijl
    12:29
    <div role="row row row your boat"> 8)
  85. MichielBijl
    12:30
    Never taught that would be valid html
  86. MichielBijl
    12:30
    thought*
  87. MichielBijl
    12:31
    Even <div role="stop hammer timer">!
  88. MichielBijl
    12:31
    Oh the possibilities!
  89. stevefaulkner
    12:31
    only non abstract roles are conforming
  90. MichielBijl
    12:32
    Ahw :(
  91. MichielBijl
    12:32
    WHoops
  92. MichielBijl
    12:32
    “Authors MUST NOT use abstract roles in content.”
  93. MichielBijl
    12:33
    I guess I could propose a “your” and “boat” role tonight during PF :P
  94. MichielBijl
    12:33
    Not sure what they should map to
  95. zakim-robot
    12:33
    [jitendra, a11y] @callumacrae: this is new. thanks
  96. StommePoes
    12:35
    @MichielBijl yeah I saw the naming thing, Heydon and I were bitching about "wtf is a combobox anyway"
  97. StommePoes
    12:35
    And we have that same guideline in our ARIA section in some developer-rules pages, "don't use abstract roles". I had to then go look up what an abstract role was.
  98. MichielBijl
    12:36
    @stevefaulkner which iOS version did you test on?
  99. MichielBijl
    12:36
    It says prior, but not prior to what.
  100. StommePoes
    12:36
    @stevefaulkner just posted a comment on the paciello blog, Orca ran for a hot minute (then I ctrl-C'd it off in the terminal and haven't been able to get it speaking again, prolly need a reboot :( and it understands switch role with firefox 41 :)
  101. MichielBijl
    12:36
    @StommePoes yeah I saw that.
  102. StommePoes
    12:37
    goddamn I hate smilies, I need to remember to use anti-smiley spaces : (
  103. stevefaulkner
    12:37

    @jitendra

    If the a element has no href attribute, then the element represents a placeholder for where a link might otherwise have been placed, if it had been relevant, consisting of just the element's contents.

    http://www.w3.org/TR/html51/semantics.html#the-a-element

  104. MichielBijl
    12:38
    @StommePoes yeah there should be an option to turn those off
  105. stevefaulkner
    12:38
    @StommePoes cool thanks
  106. MichielBijl
    12:38
    Especially the :) smiley; nobody smiles like that!
  107. StommePoes
    12:39
    @Jitendra Yeah if you're going to do <a role="button">do stuff</a> and then you add in all the tabindex=0 and listen for spacebar... that's too much work. Be like Larry Wall: be Lazy! <button> ftw. And yeah, for something like a "read more/load more" on a lazy-loaded page where there's no JS, I have used buttons there too, because they do an action (even if the action results in a page refresh)
  108. StommePoes
    12:39
    @Michiel psychokillers do
  109. stevefaulkner
    12:41
    @stommePoes how does orca handle the main role tests?
  110. MichielBijl
    12:41
    @StommePoes how would you do a page refresh with a button if it doesn't have an event listener?
  111. StommePoes
    12:41
    the main/landmark one? I think I have to reboot to test it, orca runs now in my processes but went silent on me :)
  112. MichielBijl
    12:41
    @stevefaulkner I tested with iOS 9.0.2 and no switch implemented yet.
  113. zakim-robot
    12:42
    [jitendra, a11y] @storm - can you show an example?
  114. stevefaulkner
    12:42
    @MichielBijl the update i have notmade yet is for ios 9.0.2
  115. zakim-robot
    12:43
    [jitendra, a11y] I didn’t get how <button> can load a url when js is disabled
  116. StommePoes
    12:43
    @michiel Well jitendra said there was a page refresh without, you could do it with an anchor pointing to the current page + extra info in the URL (for the moaR-info stuff), or you can make an HTML form with a button where the onsubmit is a GET to current page URL :P
  117. MichielBijl
    12:43
    @Michiel psychokillers do
    I do not smile like that!
  118. MichielBijl @MichielBijl still doesn't know how to properly quote in Gitter :x
  119. MichielBijl
    12:44
    Sounds plausible
  120. StommePoes
    12:44
    @jitendra <buttons> don't do anything without JS if outside a form, but could refresh the page with a form. If the users are really going to another URL then <a href="new URL"> should be good
  121. MichielBijl
    12:44
    That was my point though, it sounds like an anchor.
  122. MichielBijl
    12:44
    Just not if you have JS enabled.
  123. StommePoes
    12:44
    @jitendra you can also start with an anchor that goes somewhere without JS, and if user has JS, let JS turn the anchor into a button (still better to change it to HTML <button> than add a crapton of roles and listeners ot the anchor I think, but either way would work)
  124. MichielBijl
    12:45
    I've written about this in the past: http://www.michielbijl.nl/2014/09/25/buttons-with-a-double-function/
  125. StommePoes
    12:45
    @Michiel <form method="get" action="current page URL"><input type="submit" value="KLIKKIE"></form> try it
  126. MichielBijl
    12:45
    Oh I believe you :P
  127. StommePoes
    12:45
    but so if it's really doing anchor-duty I'd do an anchor
  128. zakim-robot
    12:46
    [jitendra, a11y] is this valid html ?
  129. StommePoes
    12:46
    and if when JS is enabled it needs to be buttony, let JS turn it into a button onload
  130. StommePoes
    12:46
    @jitendra my HTML? Yeah it should be valid in HTML5 (HTML4 needs a block child for the form around the input) but it's just a thingie that works, I wouldn't ever use it on a web page :P
  131. StommePoes
    12:46
    arg smilies
  132. StommePoes
    12:47
    I'll see if I can reboot and get Orca back up so I can test for steve
  133. StommePoes
    12:47
    bbl
  134. stevefaulkner
    12:47

    the main/landmark one? I think I have to reboot to test it, orca runs now in my processes but went silent on me :)

    yup

  135. MichielBijl
    12:48
    @MichielBijl the update i have notmade yet is for ios 9.0.2
  136. MichielBijl
    12:48
    Uhm
  137. MichielBijl
    12:48
    You have an update for the article?
  138. stevefaulkner
    12:56
    @MichielBijl have added your data - thanks!
  139. StommePoes
    12:56
    huh, Orca is weird after the update. I'm only getting speech for panels, and here in pidgin, but not in browsers
  140. StommePoes
    12:57
    I'm getting speech here in pidgin, and in terminal, but not in browsers. Arg
  141. stevefaulkner
    12:57
    anybody can check in el capitan/safari would be appreciated
  142. MichielBijl
    12:58
    @stevefaulkner can check tonight
  143. MichielBijl @MichielBijl is not allowed to install it as main OS on his work machine
  144. MichielBijl
    13:02
    So will have to check at home
  145. stevefaulkner
    13:03
    :+1:
  146. StommePoes
    13:05
    ah figured it out
  147. StommePoes
    13:05
    Orca's missing stuff that was already launched before Orca launched. this is a known issue (and prolly fixed in master)
  148. StommePoes
    13:10
    Okay on a page with two divs, one with role="main landmark" and the other with role="landmark main", when I look for landmarks using alt+shift+m, I get "no landmarks"
  149. stevefaulkner
    13:10
    ok so thats an Orca bug
  150. StommePoes
    13:11
    I don't have latest master
  151. StommePoes
    13:11
    If I make a div role="main" and turn the other into <main> and I call up landmarks, I get my list of landmarks as sections with the aria-labels as their names
  152. StommePoes
    13:12
    so a table, where "landmark" column is the names (test1, test2) and the next column called Role and calls them "section"
  153. StommePoes
    13:13
    Yet I'm pretty sure from the mailing list that Orca knows what "main" is
  154. StommePoes
    13:13
    So I'm running 3.18.1 but the Master on github is newer
  155. StommePoes
    13:13
    and I'm running trusty which is not latest Ubuntu either
  156. stevefaulkner
    13:18

    so a table, where "landmark" column is the names (test1, test2) and the next column called Role and calls them "section"

    ??

  157. StommePoes
    13:19
    yeah lemme make a screenshot
  158. StommePoes
    13:21
    http://stommepoes.nl/work/orca_table.png
  159. StommePoes
    13:22
    Where in the stuff on the left, the div role="main" aria-label="test1" is around the "Explanation" heading and chunk of text, and the <main aria-label="test2"> is around the sentence "there be some weirdness here"
  160. MichielBijl
    13:22
    But role section doesn't exist :o
  161. StommePoes
    13:22
    but it is a sectioning role
  162. StommePoes
    13:22
    where aria-label should be able to "name" it
  163. StommePoes
    13:23
    so I could be misunderstanding some setup, and I could bug joanie about it but prolly better to ask on the list
  164. stevefaulkner
    13:23
    ok so it's picking up the xml-roles: main but reporting the role as ATK/AT-SPI Role section
  165. StommePoes
    13:23
    lots of the guys on the mailing list are also devs, testers, and sometimes bug-fixers
  166. StommePoes
    13:23
    Seems so, was that a paciello or wave tool that showed ff's acc tree?
  167. StommePoes
    13:23
    I don't have it but might as well download it, if it runs on Linux
  168. StommePoes
    13:24
    otherwise, have to wait til tomorrow where I get a Windows machine (with pre-installed AT!) from work
  169. stevefaulkner
    13:24

    But role section doesn't exist :o

    it does exist in iA2/ATK/AT-SPI but is not same as html5 section

  170. stevefaulkner
    13:24
    its also an ARIA abstract role
  171. StommePoes
    13:24
    oh sections, yeah
  172. StommePoes
    13:25
    so it's what the browser? assigns to native and author roles?
  173. stevefaulkner
    13:25
    @StommePoes try the DOM inspector addon for firefox, it has an acc tree view
  174. StommePoes
    13:25
    that's why we should never assign them ourselves? They are internal to browsers or the acc tree?
  175. StommePoes
    13:26
    not the built-in one?
  176. StommePoes
    13:26
    adding...
  177. stevefaulkner
    13:26
    no, the extension this may help http://asurkov.blogspot.co.uk/2013/02/dom-inspector-for-accessibility-new.html
  178. StommePoes
    13:27
    ah ok
  179. StommePoes
    13:27 
         DOM Inspector                   2.0.16                              by SeaMonkey Council
  180. StommePoes
    13:31
    trying to find the Show Accessible Nodes in the dom inspector (f12) afte installing and restarting the browser
  181. StommePoes
    13:35
    ah ctrl+shift+i
  182. StommePoes
    13:36
    you only get the menu you need from the browser menu, not the shortcut, bleh.
  183. MichielBijl
    13:37

    But role section doesn't exist :o

    it does exist in iA2/ATK/AT-SPI but is not same as html5 section

  184. MichielBijl
    13:37
    WHY SO MANY DIFFERENT NAMES
  185. MichielBijl
    13:37
    Arg
  186. MichielBijl
    13:37
    “So yeah it's a section”
  187. MichielBijl
    13:38
    “Like what, a html section?”
  188. MichielBijl
    13:38
    “No, like the role”
  189. MichielBijl
    13:38
    “Aria role?”
  190. MichielBijl
    13:38
    “No, the iA2 role…”
  191. MichielBijl
    13:38
    FFS
  192. StommePoes
    13:39
    cause browsers do aria, atk does like whole windowing system thingies, gtk/qt/etc
  193. stevefaulkner
    13:39
    @MichielBijl welcome to webstandards
  194. StommePoes
    13:39
    what I want: whatever that new thingie is with Windows replacing MSAA (UIA?) would talk to GTK thingies
  195. StommePoes
    13:40
    in applications you have objects, sections and regions
  196. StommePoes
    13:40
    ok this dom inspector thing is cool, but the keyboard shortcut just gives me what f12 gives me.
  197. MichielBijl
    13:41
    Ah well, we all know naming things is the hardest part
  198. StommePoes
    13:41
    So it shows what we expect, aria-label is test1 on div with role of main
  199. MichielBijl
    13:41
    But as @Heydon has said before me: only if you use real words!
  200. StommePoes
    13:41
    main element is aria-label test2, local name is main
  201. StommePoes
    13:42
    the joys of screen reader debugging
  202. StommePoes
    13:42
    gi._glib.GError: Method "Get" with signature "ss" on interface "org.freedesktop.DBus.Properties" could not be processed as object /org/a11y/atspi/accessible/3574 does not exist
  203. MichielBijl
    13:43
    What an accessible error message ;)
  204. StommePoes
    13:43
    they're all pythonese
  205. StommePoes
    13:43
    and Glib and gObject and other pythony thingies
  206. StommePoes
    13:44
    They're made for windowing manager and GTK hackers
  207. stevefaulkner
    13:44
    @StommePoes top left corner of UI https://www.dropbox.com/s/grglm2blpikdtbj/Screenshot%202015-10-15%2014.43.33%282%29.png?dl=0
  208. stevefaulkner
    13:44
    do you have that in linux version?
  209. StommePoes
    13:45
    it looks different...
  210. StommePoes
    13:45
    ah cool, I do have it
  211. StommePoes
    13:47
    hm browser wraps a section around my progressbar
  212. StommePoes
    13:47
    ah cool it also updates the progressbar values as they go
  213. StommePoes
    13:47
    cool
  214. StommePoes
    13:48
    Orca actually waited until it was done to announce it was at 100%, but that's prolly a good thing
  215. StommePoes
    13:49
    okay, this along with the extension I heard about from @sundress ought to help a lot with our guideline writing and basic-accessible-grab-n-go widget library
  216. StommePoes
    13:49
    I hope to add more Linux/Orca stuff to our guidelines, it's currently all JAWS this and VO that.
  217. stevefaulkner
    13:49
    note its div elements that map to section
  218. StommePoes
    13:49
    But as my company expands into developing markets, and mobiles, they'll need to have more combos
  219. StommePoes
    13:50
    yeah, the main mapped to textcontainer
  220. stevefaulkner
    13:50
    HTML5 section maps to region :-)
  221. StommePoes
    13:50
    Orca saw the <main> as a section
  222. stevefaulkner
    13:50
    section is generic role
  223. StommePoes
    13:50
    div = section
  224. StommePoes
    13:50
    region = ...
  225. StommePoes
    13:50
    section = region
  226. StommePoes
    13:50
    lol
  227. StommePoes
    13:51
    but region's not an abstract role, right
  228. StommePoes
    13:51
    I can <section role="region"/>, <div role="region" aria-label="blah"/> if I wanted
  229. MichielBijl
    13:52
    region is default for section
  230. StommePoes
    13:52
    yeah
  231. stevefaulkner
    13:52
    http://rawgit.com/w3c/aria/master/aria/aria.html#region
  232. StommePoes
    13:52
    I lucked out on that with a site where I have <sections> with <h2> children and one without an h2 with aria-label, I didn't realise at the time that the labelling-with-headings-or-aria-label was a region-only thing
  233. zakim-robot
    13:53
    [jitendra, a11y] Alt text for image embedded in a tweet alt="Embedded image permalink”
  234. StommePoes
    13:53
    I had aria-label on a bunch of other stuff at the time
  235. StommePoes
    13:54
    jitendra is the image inside a link/anchor at all?
  236. stevefaulkner
    13:54

    region is default for section

    for HTML <section> element yes

  237. StommePoes
    13:54
    or just directly in a tweet?
  238. zakim-robot
    13:54
    [jitendra, a11y] yes image is inside anchor
  239. zakim-robot
    13:55
    [jitendra, a11y] which opens the same image in a modal windows
  240. StommePoes
    13:55
    ok that's where they figured the alt must be the "link text" stating what the link is
  241. StommePoes
    13:55
    though it would be nice if it could also say "opens larger image" (if it's larger, the point of permalink is to give a dedicated canonical URL for a resource)
  242. zakim-robot
    13:56
    [jitendra, a11y] https://twitter.com/jitendravyas/status/654656200680706048
  243. StommePoes
    13:56
    @stevefaulnker "Authors should reference a visible label with aria-labelledby if a visible label is present." <-- unless the visible label is a heading? Because I've been leaving it as <section><hx>mah heading</hx> stuf stuff stuff </section>
  244. zakim-robot
    13:57
    [jitendra, a11y] slack shows direct link of the image too when we post here
  245. stevefaulkner
    13:57
    @StommePoes <div role="main"> iA2 role = section, object attribute xml-roles:main
  246. stevefaulkner
    13:57
    https://www.dropbox.com/s/hcheyfefg3qhyws/Screenshot%202015-10-15%2014.48.11%282%29.png?dl=0
  247. StommePoes
    13:57
    if I open the image "in a new tab" I do get the perma URL https://pbs.twimg.com/media/CRXOWA9VEAQtrOw.png:large
  248. StommePoes
    13:58
    @stevefaulkner so Orca needs to know the xml roles and present those (maybe the Master version already does)
  249. stevefaulkner
    13:58
    reason for this is theres no platform (iA2 in this case) accessibility API role of main
  250. stevefaulkner
    13:59
    so generic role is used in iA2 and specific role provided as object attribute in iA2
  251. StommePoes
    13:59
    hm lemme see what the generic/specific roles for say a progressbar are
  252. stevefaulkner
    13:59
    it recognises them as landmarks now right?
  253. stevefaulkner
    14:00
    it just does not list the landmark role
  254. StommePoes
    14:00
    yeah it says Landmarks found: 2
  255. StommePoes
    14:01
    and you can cycle to next/prev landmarks on a page
  256. stevefaulkner
    14:01
    right
  257. StommePoes
    14:01
    progressbar is just the same in both, interesting
  258. stevefaulkner
    14:02
    that is because there is a platform role progressbar
  259. StommePoes
    14:02
    ah like checkbox, other native apps have these as well
  260. StommePoes
    14:03
    So I'll ask on Orca list if main is recognised as main
  261. StommePoes
    14:04
    I think when I pick up my windows box, since I'm actually useless in Windows, I'll throw a VirtualBox on it with a debian or whatever and just install the latest-bleeding-edge orca there
  262. StommePoes
    14:04
    Then I can test new fixes too, yay
  263. zakim-robot
    14:25
    [ccwilcox, a11y] @stevef: In your article's Pen, VoiceOver in Yosemite/Safari 9 say tests 1-5 as main and the last two are test 6, off, switch and test 7, off, switch
  264. MichielBijl
    14:25
    Interesting
  265. MichielBijl
    14:26
    in iOS 9 those are tick boxes
  266. MichielBijl
    14:26
    Wonder why that is different.
  267. stevefaulkner
    14:29
    @ccwilcox exellent thanks! so switch is implemented in OSX
  268. stevefaulkner
    14:30

    Wonder why that is different.

    iOS lags behind OSX

  269. MichielBijl
    14:30
    Apparently so
  270. powrsurg
    14:39
    We have an audio element that is on one point on the page and an on-page transcript that is located elsewhere. Should we be using aria-describedby on the audio? Most of the examples I'm seeing for how to do accessible transcripts talk about the track element and using an external file
  271. stevefaulkner
    14:39
    @ccwilcox have added your data
  272. cmpscabral
    14:47
    Hi, sorry for the offtopic message - feel free to ban me if this is too offtopic: has any of you already seen an implementation of the cookie law (https://www.cookiechoices.org/) warning for an iOS app? I've implemented google's solution but I'd like to see other examples and I can't find any...
  273. stevefaulkner
    14:54
    @cmpscabral no worries, but no haven't seen
  274. cmpscabral
    14:57
    @stevefaulkner tks :)
  275. MichielBijl
    15:09
    A project I'm working on only has variables for min-max combined mediaqueries
  276. MichielBijl @MichielBijl ignores variables and uses normal values
  277. MichielBijl
    15:10
    Why limit yourself to a predefined set of breakpoints…
  278. StommePoes
    16:00
    easier to use?
  279. zakim-robot
    16:32
    [Marcy Sutton, a11y] Does it seem reasonable to only provide focus management when a user is navigating with the keyboard? I am cringing at the thought of removing the feature on mouse click, but JavaScript developers complain a LOT.
  280. zakim-robot
    16:32
    [Marcy Sutton, a11y] Not to mention the focused elements they are complaining about are barely visible from a color contrast standpoint :rage:
  281. deborahgu
    16:33
    you mean, disable visible keyboard focus unless a keyboard click has been used?
  282. zakim-robot
    16:33
    [Marcy Sutton, a11y] I mean for components like dialogs, sidenavs and bottom sheets where focus is sent into the component on open and returned on close. Mouse users are complaining that “buttons are focused wrongly"
  283. zakim-robot
    16:33
    [Marcy Sutton, a11y] (and I want to flip a table)
  284. zakim-robot
    16:35
    [Alice Boxhall, a11y] Sounds like pretty sound evidence for the modality concept to me
  285. zakim-robot
    16:37
    [Marcy Sutton, a11y] yup :simple_smile:
  286. zakim-robot
    16:38
    [Marcy Sutton, a11y] Button Focus Hell still going strong…sometimes i don’t recognize it as such
  287. zakim-robot
    16:39
    [Alice Boxhall, a11y] Haha, it's a sneaky one
  288. zakim-robot
    16:41
    [Justin Stockton, a11y] @marcysutton: I’ve been using https://github.com/ten1seven/what-input to manage that on a few projects
  289. zakim-robot
    16:42
    [Alice Boxhall, a11y] @poorgeek How much do you distinguish between mouse and touch?
  290. zakim-robot
    16:42
    [Marcy Sutton, a11y] I guess my question is less about detecting the input, for which there are multiple methods, and more about “is it reasonable to disable focus management for mouse users?"
  291. zakim-robot
    16:43
    [Marcy Sutton, a11y] Probably yes, I’m just thinking this through
  292. zakim-robot
    16:43
    [Alice Boxhall, a11y] @marcysutton: I don't think it makes sense to disable focus management in that case
  293. zakim-robot
    16:43
    [Alice Boxhall, a11y] I think it makes sense not to make focus visible, because it's not interesting
  294. zakim-robot
    16:43
    [Alice Boxhall, a11y] but the user may switch to a keyboard at any point, at which point it becomes interesting
  295. zakim-robot
    16:44
    [Justin Stockton, a11y] @alice: what-input does a decent job at detecting touch events. I might be relying on it more than I should
  296. zakim-robot
    16:44
    [Marcy Sutton, a11y] The people who complain about it the most are usually trying to send focus to something else, like an element with autofocus.
  297. zakim-robot
    16:45
    [Marcy Sutton, a11y] Treading newish ground with a11y features in a major JS framework…where before there was no focus management, so devs could do whatever they wanted.
  298. zakim-robot
    16:45
    [Alice Boxhall, a11y] @poorgeek: What kind of use cases is it useful for?
  299. zakim-robot
    16:46
    [Alice Boxhall, a11y] @marcysutton: Could you give an example?
  300. zakim-robot
    16:47
    [Justin Stockton, a11y] my snarky response is that its only useful for compromising with designers who think that focus outlines destroy their designs
  301. schalkneethling
    16:47
    What are people's general thoughts about the new Google recaptcha (checkbox = not a robot) in terms of accessibility?
  302. zakim-robot
    16:47
    [Marcy Sutton, a11y] An example is the mdDialog: on open, an element inside of it is focused. Developers are jamming entire forms into dialogs, and they want to focus on something else. :neutral_face:
  303. deborahgu
    16:47
    since JS developers are all about adding undocumented keystroke commands via JS "to improve accessibility" it's weird that it doesn't occur to them that they may switch back and forth between keyboard and mouse actions
  304. zakim-robot
    16:48
    [Marcy Sutton, a11y] I can detect if there is an autofocus attribute and let that take focus instead…I am just trying to understand the scope of the issue as it pertains to mouse-vs-keyboard users/focus states/what to focus on
  305. deborahgu
    16:48
    schalkneethling: It is almost unusable by voice and keyboard, sadly. It tends to identify tab-click or voice click of the checkbox as a robot,
  306. deborahgu
    16:48
    and then the image selector is not keyboard/voice accessible.
  307. deborahgu
    16:49
    It keeps locking me out of stackoverflow, since they are also overeager in their robot detection.
  308. zakim-robot
    16:49
    [Justin Stockton, a11y] but Jeremy gave a demo of it a few months ago at our meetup and showed how it lets you adjust styles based on the current input method. so you can switch back and forth between keyboard and mouse and it will style accordingly. it also has a light-weight event api that you can access
  309. zakim-robot
    16:50
    [Brian Kardell, a11y] @marcysutton: what do you mean they want to focus on somethung else?
  310. deborahgu
    16:50
    I would argue that once you are jamming an entire form that is more complex than a simple "give us your email address /submit", for example, into a dialog, then you have bigger usability and accessibility problems. Focus is a symptom there, not the cause.
  311. schalkneethling
    16:52
    @deborahgu That is really unfortunate
  312. schalkneethling
    16:52
    thanks for the feedback
  313. zakim-robot
    16:52
    [Marcy Sutton, a11y] Yup
  314. zakim-robot
    16:53
    [Marcy Sutton, a11y] Plus the dialog in question already has an API for focusing on arbitrary elements.
  315. zakim-robot
    16:54
    [Marcy Sutton, a11y] “How can I add a tab switcher with autofocus in it into a dialog” <— real issues I see out there
  316. zakim-robot
    16:55
    [Alice Boxhall, a11y] @poorgeek: I do have some empathy for designers being frustrated at the focus ring, particularly when it appears at odd times.
  317. deborahgu
    16:57
    Alice, I agree with you -- but I would have more empathy for designers if they had more empathy for the users. The right way to deal with the focus ring is to find a way to make a highly-visible focus look great in your design, not to make it go away.
  318. zakim-robot
    16:57
    [Justin Stockton, a11y] in those cases I do too. However the push back I get is usually when managing keyboard focuses uncovers holes in their design
  319. deborahgu @deborahgu made my life much easier recently when I put a stylish stylesheet on every website which shows my focus in a big high contrast bright yellow. It's kind of a revelation of how much it improved my life.
  320. zakim-robot
    17:02
    [Alice Boxhall, a11y] @deborahgu I definitely agree that designers should be designing focus along with the rest of their design!
  321. zakim-robot
    17:02
    [Alice Boxhall, a11y] And I hate these apologetic subtle focus styles
  322. zakim-robot
    17:03
    [Alice Boxhall, a11y] Which is why I'm hoping things like what-input and/or whatever the modality/button focus hell proposal evolves into will help out with that
  323. zakim-robot
    17:03
    [Alice Boxhall, a11y] Because you can be sure focus will always be interesting when it shows up
  324. StommePoes
    17:05
    See, the only way I could sneak in any usable-focus-styles for keyboard+low vision was to hide it from mousers
  325. StommePoes
    17:05
    all designers I have ever had to work with were mouse-users (of course), so they never, ever saw my focus rings
  326. StommePoes
    17:06
    because if they ever did, another issue in the issue tracker "remove ugly focus ring on click"
  327. StommePoes
    17:08
    Even Duck Duck Go had it -- I had briefly convinced someone to remove the outline: 0 stuff, and first comment on the topic was "it looks really terrible in Chrome, big blue outline, can you please remove it"
  328. StommePoes
    17:08
    I asked on WebAIM how reliable detecting if the event was touch, and Patrick H Lauke said it's not necessarily totally reliable, depends on device/browser combo.
  329. zakim-robot
    17:09
    [Marcy Sutton, a11y] The issue is definitely not going away…
  330. zakim-robot
    17:10
    [Marcy Sutton, a11y] Maybe now I’ll have momentum to dogfood the Input Modality proposal into Angular Material now that I have my head pulled out of the sand :simple_smile: angular/material#5189
  331. StommePoes
    17:12
    I personally think focus should be near where last mouse click was (as in, I would expect my next tab to go from last place mouse clicked), but perhaps we need research on these switchy users (and do they differ strongly between two-index-finger facebookers vs developery computer users)
  332. StommePoes
    17:12
    And, does simulated mouse click (as done by some AT) get measured as a mouse click or not really? (I have no idea)
  333. zakim-robot
    17:12
    [Marcy Sutton, a11y] depends on the element, I think, but i could be wrong..
  334. StommePoes
    17:13
    Yeah, I would think almost everyone would expect clicking on a form input to keep focus there for example
  335. StommePoes
    17:14
    but these new dialoggy widgetty things... esp where we're sometimes putting focus on things like sometimes the modal boxes themselves,which is kinda new.
  336. StommePoes
    17:14
    Maybe you should tweet the github issue and see what ideas pour in?
  337. StommePoes
    17:15
    When people say "the buttons focus all wrong" do they mean more than the focus ring?
  338. zakim-robot
    17:15
    [Brian Kardell, a11y] @marcysutton: so the story is something like - user activates something and that causes a modal to be displayed - focus is currently going to the modal itself instead of, say, a tab in a tabset inside the modal?
  339. zakim-robot
    17:16
    [Marcy Sutton, a11y] I think the issue is that mouse users are seeing focus management and not understanding why it’s there, since they don’t need it.
  340. zakim-robot
    17:16
    [Marcy Sutton, a11y] @bkardell: exactly.
  341. zakim-robot
    17:17
    [Marcy Sutton, a11y] We have affordances in the API for working around this, but everyone expects custom elements to also support native things like autofocus…which work unless you stomp over them with custom focus management :neutral_face:
  342. StommePoes
    17:17
    ok, and you're wondering whether you should a) just make that focus invisible if mouse (and maybe also touch) was the event or b) not do focus management at all in those cases?
  343. zakim-robot
    17:17
    [Marcy Sutton, a11y] I think the answer will be a mix of better autofocus support and CSS Input Modality so focus is only shown for keyboard users
  344. StommePoes
    17:17
    where besides some annoying (arg arg annoying hate hate hate) inputs with autofocus, what other JS-related focusing are they doing?
  345. zakim-robot
    17:18
    [Marcy Sutton, a11y] The requests I’ve seen are all either form inputs or tabs :fearful:
  346. zakim-robot
    17:18
    [Marcy Sutton, a11y] (inside of dialogs or sidenavs)
  347. zakim-robot
    17:18
    [Brian Kardell, a11y] interesting that I've never thought about this but would a reader announce that they were in a modal at that point or would it even matter as it is essentially a new view?
  348. StommePoes
    17:18
    (like when I want to log into github, the stupid autofocus makes me need to, on page load, shift-tab a few times just to log in, each time needing to figure out why my several-tabs didn't work)
  349. zakim-robot
    17:19
    [Marcy Sutton, a11y] @bkardell depends on the AT, I feel like I normally hear a dialog announced in VO even if focus is sent to a descendant
  350. StommePoes
    17:20
    Brian what confuses me is there are two directions: 1) if you set focus to a modal body and it has some text/heading aria-labelledby-attached then a user should hear the name of the box, yet 2) we're supposed to set focus to the first "focusable thingie" (if there are inputs etc)
  351. StommePoes
    17:20
    I never know when to do one or the other, except when there is no obvious first input... and then, focus the close button as it's usually first? (what I have been doing so far)
  352. zakim-robot
    17:21
    [Brian Kardell, a11y] this seems exactly like the question of whether the default focus is to <body> does it not? modal is effectively a new 'page' in a sense
  353. zakim-robot
    17:21
    [Marcy Sutton, a11y] oh yes! Dialogs need a title to be announced. minor detail.
  354. StommePoes
    17:21
    Javascript modals (confirm, alert) get that application type of focus when they open, all users know they're in those things.
  355. MichielBijl
    17:21
    @bkardell default to body yes
  356. StommePoes
    17:21
    wanted to emulate that
  357. MichielBijl
    17:21
    Body of a dialog should have a role of document and get focus on open
  358. StommePoes
    17:22
    why add a role of document?
  359. StommePoes
    17:22
    aren't we already in a docuemtn
  360. StommePoes
    17:22
    document
  361. MichielBijl
    17:22
    Because reasons
  362. StommePoes
    17:22
    I thought you'd set document role to undo an application role
  363. StommePoes
    17:22
    I use the dialog role
  364. zakim-robot
    17:22
    [Marcy Sutton, a11y] "Body of a dialog should have a role of document and get focus on open” <— I get conflicting feedback on this behavior all the time. I guess it depends on the kind of dialog?
  365. MichielBijl
    17:22
    No, I don't know the reason, I know it's in the spec :P
  366. StommePoes
    17:22
    (so far have done no alertdialogs yet)
  367. MichielBijl
    17:23
    Just gonna throw this in here: http://s.codepen.io/Michiel/debug/0f9ce27964f35b55d3e7628d91874c2e?
  368. MichielBijl
    17:23
    @marcysutton I'm working on alertdialog
  369. MichielBijl
    17:23
    So could be different for other dialogs
  370. StommePoes
    17:24
    that link sounds so terrible in orca ug
  371. MichielBijl
    17:24
    Sorry :(
  372. MichielBijl
    17:24
    Can't make it any prettier
  373. StommePoes
    17:24
    With the codepen everything's read out, except the dismiss button, and I know I'm on reply cause it's pushbutton reply dirgeful silence
  374. StommePoes
    17:25
    if I leave it open, go back to the browser address bar, then tab forward, I can focus on the main part with hey fella and I'm i normal browse mode, read by line
  375. MichielBijl
    17:25
    Hmm
  376. StommePoes
    17:25
    but that's an alertdialog, supposed to be similar to Javascript alerts, seems fine
  377. StommePoes
    17:26
    most of my modals are type of dialog
  378. MichielBijl
    17:26
    Right, I haven't finished
  379. MichielBijl
    17:26
    Good points
  380. MichielBijl
    17:26
    You should not be able to leave the dialog with focus though
  381. StommePoes
    17:26
    You can't stop my Ctrl-L
  382. StommePoes
    17:26
    (and you shouldn't ever be able to)
  383. MichielBijl
    17:27
    Haha, I'm not trying to
  384. MichielBijl
    17:27
    That is one of the things I noted too
  385. StommePoes
    17:27
    Also remember we can trap focus, but we can access other page stuff in our secret magical SR ways muhahaha
  386. MichielBijl
    17:27
    If you capture focus; you cannot even go outside the web window
  387. StommePoes
    17:27
    I use it a lot to reposition myself
  388. zakim-robot
    17:27
    [Brian Kardell, a11y] @michiel: why is that an alertdialog exactly?
  389. StommePoes
    17:27
    yeah that would suck (and sometimes JS has captured ctrl-L! friend Darius found a website that did that, arg)
  390. MichielBijl
    17:28
    On document role
  391. MichielBijl
    17:28
    The message area of a detail message dialog is focusable and has a document role so screen reader users will have complete access to the message content, e.g., the screen reader can read it by character, word, or line.
  392. StommePoes
    17:28
    But... would we not have access without that role?
  393. MichielBijl
    17:28
    @bkardell because I needed something for the APG code examples?
  394. MichielBijl
    17:28
    Or maybe I don't fully understand the question.
  395. StommePoes
    17:28
    Then it's just a question of, does the user know that text is there? so ja, if we're not forced into application mode (where text often can't be read), we can always go back and read that
  396. MichielBijl
    17:29
    @StommePoes maybe for easier access? I'll ask on monday.
  397. StommePoes
    17:29
    Some of this stuff might border on usability: what to average joe-computer-users think, what do they end up missing, what do they think is broken?
  398. zakim-robot
    17:29
    [Brian Kardell, a11y] @michiel: I mean, alert is just weird in general - it's kind of a 'special' modal intended to convey 'this is more important than the average dialog' right?
  399. StommePoes
    17:30
    (like sighted users thinking dark placeholders are filled-in default form field values and skipping them)
  400. StommePoes
    17:30
    alertdialog is meant to replicate javascript alert dialogs. They must interupt user flow and force a OK-Cancel decision, is how I read it.
  401. MichielBijl
    17:31
    Hmm yeah
  402. MichielBijl
    17:31
    The design might not be the best to replicate just that
  403. StommePoes
    17:32
    Modals/ regular dialogs are meant to be triggered by users "opening" stuff, where ESC and going back to the page should be totally fine (like, login form in a stupid trendy modal Just Because, I can choose to go back to the page and just, not log in)
  404. MichielBijl
    17:32
    Maybe this should just be a dialog thingy
  405. MichielBijl
    17:32
    But it's an alert as in a notification
  406. MichielBijl
    17:32
    So it's a live area none the less
  407. MichielBijl
    17:33
    A dialog is a small application window that sits above the application and is designed to interrupt the current processing of an application in order to prompt the user to enter information or require a response (dialog).
  408. MichielBijl
    17:33
    But that text is subject to change
  409. MichielBijl
    17:33
    Still, that could also apply to my example
  410. StommePoes
    17:34
    yeah whereas lots of modals aren't notifications, they're ... hipster whatever things to focus user flow to this one specific task (or wizard) and removes the "background noise" of the rest of the page
  411. zakim-robot
    17:34
    [Brian Kardell, a11y] I think of alert as being something to let you know/confirm (the second bit is the importnant part I think) something important like "hey, if you submit this you'll wipe out the old data" or "by clicking this you agree to this license agreement"
  412. StommePoes
    17:34
    originally they were to get people's attention while on a page
  413. MichielBijl
    17:34
    I agree with you Brian. I'll edit it.
  414. MichielBijl
    17:34
    My example should be a dialog.
  415. StommePoes
    17:35
    @michiel make a form (login, two inputs), tell people to not fill in the inputs and hit submit, make an alertdialog (instead of JS) stating "stuff's not filled in! [OK]" or whatever
  416. MichielBijl
    17:35
    I need an alertdialog so I'll edit the design
  417. MichielBijl
    17:35
    It's more that the text in my example is not what you'd expect in an alertdialog
  418. StommePoes
    17:35
    I think it was for people who hand-crafted their own alerts because they couldn't style the JS ones
  419. MichielBijl
    17:35
    Sounds logical
  420. MichielBijl
    17:35
    Thanks
  421. StommePoes
    17:35
    Agreed, hey fella's not too alerty
  422. MichielBijl
    17:36
    Nah, this should be a dialog according to APG text
  423. zakim-robot
    17:36
    [Brian Kardell, a11y] it seems that wai actually recommends that "Authors SHOULD ensure each active dialog has a focused descendant element that has keyboard focus." so maybe @marcysutton use cases are actually 'right'
  424. MichielBijl
    17:36
    And Brian's interpretation is spot on :)
  425. MichielBijl
    17:36
    @bkardell that would be the content area
  426. StommePoes
    17:36
    Brian and if there's nothing really default focusable in it, the close button is a popular choice
  427. zakim-robot
    17:36
    [Marcy Sutton, a11y] now we’re just arguing over which element is the right one :simple_smile:
  428. StommePoes
    17:37
    so we have 3 doors... two have goats behind them, and one has a sportscar
  429. MichielBijl
    17:37
    Which one has the goats? I like goats
  430. zakim-robot
    17:37
    [Brian Kardell, a11y] I'd like to remove one door
  431. StommePoes
    17:37
    except they're heisengoats... they move around
  432. zakim-robot
    17:37
    [Marcy Sutton, a11y] do you show the focus style if the user initiated the modal by mouse? /shudder
  433. StommePoes
    17:37
    Brian: always change your choice after you get shown a goat-door!
  434. MichielBijl
    17:38
    Can I buy a vowel?
  435. StommePoes
    17:38
    If i had focussed on the close button or even the whole modal body, I always hid the focus
  436. zakim-robot
    17:38
    [Marcy Sutton, a11y] have I mentioned how much I love goats? #goatlife #coolgoat <-- my favorite hashtags
  437. MichielBijl
    17:38
    Go Marcy!
  438. MichielBijl
    17:38
    Anyway, I should really focus on the call…
  439. garcialo
    17:38
    Then this is relevant to your interests. https://twitter.com/TheSassyGoats
  440. StommePoes
    17:39
    using a modified document listener (if mousedown, add mousedetected class to body, if keydown, remove it... use css to do like .mouseDetected element:focus {outline:0;})
  441. zakim-robot
    17:39
    [Marcy Sutton, a11y] https://pbs.twimg.com/media/BLlr5sBCcAENvZk.jpg
  442. StommePoes
    17:39
    That also means the moment someone switched to keyboard, yay styles
  443. zakim-robot
    17:39
    [Marcy Sutton, a11y] Goats smashing heads on walls saying (THIS IS AWESOME)
  444. zakim-robot
    17:40
    [Brian Kardell, a11y] @marcysutton: I used to have goats. they really aren't that awesome :simple_smile:
  445. MichielBijl
    17:40
    https://vine.co/v/eQ21ThVqt50
  446. MichielBijl
    17:40
    Let it goat Marcy
  447. StommePoes
    17:40
    Brian did you have those little ones that mow your grass perfectly? Cause I was thnking of getting some if I ever get a lawn
  448. zakim-robot
    17:40
    [Marcy Sutton, a11y] haha, good to know, because i have been tempted
  449. zakim-robot
    17:40
    [Marcy Sutton, a11y] foliage-eating goats are a big business around here in Seattle
  450. StommePoes
    17:40
    pygmy goats
  451. zakim-robot
    17:41
    [Brian Kardell, a11y] maybe if we want to continue this someone should start a #goats channel :stuck_out_tongue:
  452. StommePoes
    17:41
    a11ygoatz
  453. StommePoes
    17:41
    if we fuck up the aria, we get a goat. If we put the focus on the wrong element, we get a goat. If we magically hit that thpecial combination that works everyone on every combination of everything, we get a sportscar (with complimentary goat if wanted)
  454. zakim-robot
    17:42
    [Marcy Sutton, a11y] totesm’#goats
  455. StommePoes
    17:42
    but that goat is nice, no bites, has a bow around its head
  456. aardrian
    17:42
    I'd have lots of goats.
  457. StommePoes
    17:42
    and mows the lawn
  458. StommePoes
    17:42
    I'm all goats, all the time. If it works in Orca I'm happy. Tomorrow I get a Windows machine so I can discover the joy of JAWS bugs
  459. zakim-robot
    17:42
    [Marcy Sutton, a11y] ok seriously though we do now have a goats channel. :simple_smile:
  460. StommePoes
    17:42
    you like bikes too
  461. StommePoes
    17:42
    goats-on-bikes channel
  462. zakim-robot
    17:43
    [Marcy Sutton, a11y] okay now you’re going too far.
  463. zakim-robot
    17:43
    [Marcy Sutton, a11y] :stuck_out_tongue:
  464. zakim-robot
    17:43
    [Marcy Sutton, a11y] I’m mostly just happy that there is now goats discussion in the W3C logs
  465. StommePoes
    17:44
    marcy could you start hiding mousey-focusses on current projects and then see if magically complaints stop and then you can avoid some of the should-we-not-manage-focus stuff?
  466. zakim-robot
    17:44
    [Marcy Sutton, a11y] Sure, will do! :simple_smile:
  467. deborahgu
    17:44
    StommePoes I have encountered websites that trap ctrl+w, too
  468. StommePoes
    17:44
    deborah ug wtf ppl
  469. StommePoes
    17:45
    Opera had several others that would be more common to trap
  470. StommePoes
    17:45
    I would use w and s to move around
  471. deborahgu
    17:46
    man yeah, Opera presto's one key commands are the bomb
  472. StommePoes
    17:46
    Hey @deborahgu my 3 Mouse Tech friends have said they'd love to take you up on your offer of piping a message to the hands-free group (though they didn't give me a message :P)
  473. StommePoes
    17:46
    basically an NVDA-version of jsay-ish stuff, @gonz_blinko is kinda heading the whole thing
  474. zakim-robot
    17:48
    [Brian Kardell, a11y] I'm really interested in @marcysutton's feedback on this - I hope they do use (and if necessary contribute back to) the modality thing
  475. deborahgu
    17:49
    StommePoes: I am so into this plan so much
  476. zakim-robot
    17:49
    [Brian Kardell, a11y] we had a not dissimilar issue where someone would activate something, angular would change the view and then focus is set programatically -- you have a whole series of things to unwind there to figure out whether to show the ring
  477. zakim-robot
    17:49
    [Marcy Sutton, a11y] yes! It has taken me a bit to get around to actually developing alongside the proposal. We have an internal Adobe initiative that touches on it, as well…so i’m in a good position to play around with it.
  478. zakim-robot
    17:50
    [Marcy Sutton, a11y] @bkardell: with Angular or Angular Material?
  479. zakim-robot
    17:51
    [Brian Kardell, a11y] oh that was just a vanilla angular app - the angular bit isn't really the important thing, that use case in general is common... it's not the handler of an action itself that swaps out a view or creates new dom and ultimately moves focus
  480. StommePoes
    17:51
    @deborahgu can I connect you and Gonz via mail or something? I'm not working on the project myself
  481. deborahgu
    17:53
    I would still love to see the results from user testing, especially in environments where there are a lot of JS keyboard shortcuts. You think you've clicked into, say, the gmail compose box, you start typing, you accidentally trigger a chain of keyboard shortcut madness. Though I suppose that most developers agree not to stip typability focus styling on text input boxes,
  482. StommePoes
    17:53
    zomg I have done that with twitter (web client) a lot, I think I'm typing and I'm opening weird DM windows instead
  483. deborahgu
    17:53
    though I am seeing it more and more often on search fields which don't look typable or focused because that's not minimalist enough
  484. deborahgu
    17:54
    Next thing you know all the minimalist websites are going to be wrapped in bacon with a side of the Decemberists while riding unicycles.
  485. StommePoes
    17:54
    You've succeeded in minimalist design when users don't interact at all because they don't think there's anything there. Points!
  486. deborahgu @deborahgu high fives stommepoes
  487. deborahgu
    17:55
    (also, yes, totatlly connect us in email)
  488. StommePoes
    17:55
    I definitely had trouble testing a Material Design page because the whole huge areas of the page would get focus, by showing a subtle box shadow, which was often outside my field of view due to magnification (which I was specifically testing)
  489. zakim-robot
    17:56
    [Marcy Sutton, a11y] that focus style….a “known issue"
  490. StommePoes
    18:00
    well they didn't change it on the page in question even after I listed it in my "high mag usability problems"
  491. StommePoes
    18:00
    so they were okay with it :P
  492. zakim-robot
    18:01
    [Marcy Sutton, a11y] until it is addressed in Material design proper, it probably won’t be addressed sadly
  493. zakim-robot
    18:52
    [Neil Milliken, a11y] Heads up if people are interested we'll be interviewing Lainey Feingold for #Axschat next week video & twitter chat hope people can join in the discussion
  494. zakim-robot
    19:34
    [jenna, a11y] hai!
  495. zakim-robot
    19:36
    [Myles Borins, a11y] oh hai jenna
  496. zakim-robot
    19:36
    [Myles Borins, a11y] “D
  497. zakim-robot
    19:36
    [Myles Borins, a11y] :smile:
  498. zakim-robot
    19:37
    [wbinnssmith, a11y] :wave: hi everyone!
  499. zakim-robot
    19:38
    [Marcy Sutton, a11y] hello!
  500. zakim-robot
    19:38
    [Myles Borins, a11y] hi wbinnssmith
  501. zakim-robot
    19:48
    [Mark Wilcock, a11y] Hello @wbinnssmith & @jenna
  502. zakim-robot
    19:48
    [wbinnssmith, a11y] hello!! what brings you all to a11y?
  503. zakim-robot
    19:50
    [Neil Milliken, a11y] Hi @mark_wilcock @wbinnssmith @jenna @marcysutton :grinning: I work in accessibility & inclusion
  504. zakim-robot
    19:50
    [Neil Milliken, a11y] Why?
  505. zakim-robot
    19:50
    [Neil Milliken, a11y] Damn you autocorrect
  506. zakim-robot
    19:50
    [Neil Milliken, a11y] WBU?
  507. zakim-robot
    19:52
    [wbinnssmith, a11y] wow, that’s great!
  508. zakim-robot
    19:53
    [wbinnssmith, a11y] I’ve been building web things for a while now and wish a11y weren’t just an afterthought for me :confused:
  509. zakim-robot
    19:54
    [Neil Milliken, a11y] The earlier you plan for it the easier it is
  510. zakim-robot
    19:55
    [Neil Milliken, a11y] :grinning:
  511. zakim-robot
    19:55
    [wbinnssmith, a11y] hehe
  512. zakim-robot
    19:55
    [Neil Milliken, a11y] I spend most of my time doing a11y stuff in big corporations and government departments
  513. zakim-robot
    19:56
    [Neil Milliken, a11y] Less time doing web stuff
  514. zakim-robot
    19:56
    [wbinnssmith, a11y] ah yeah
  515. zakim-robot
    19:57
    [Neil Milliken, a11y] I reckon @mark_wilcock who works in my team spends way more time on web A11y than me
  516. zakim-robot
    19:58
    [Neil Milliken, a11y] Outside of work I run Axschat.com lots of videos on access & inclusion topics beyond technical web #a11y
  517. powrsurg
    21:11
    so I have a div that has a child div and an image that get altered and the parent is given focus when the user clicks on a button. The parent div has aria-live="assertive". VoiceOver on iOS is only reading out the alt text on the image and ignoring the text in the div above it. Any ideas?
  518. stevefaulkner
    21:30
    @powrsurg working code example would be nice, i have an idea as to why but would need to test