Archive index

A11y Slackers Gitter Channel Archive 26th of January 2016

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    02:46
    [jessebeach] Yay! A Slack rep just talked about an open a11y position at the A11y Bay Meetup!
  2. zakim-robot
    03:34
    [jamesn] What was the slack reps name. And is he in here?
  3. MichielBijl
    05:30
    @jamesn, could be a she ;)
  4. jnurthen
    05:31
    No. He stood up in the room. I just can't remember his name.
  5. MichielBijl
    05:34
    Oh, my bad :P
  6. MichielBijl
    05:34
    That I didn't know.
  7. MichielBijl
    05:35
    Sorry… scuttles away
  8. MichielBijl @MichielBijl blames the lack of coffee this early in the morning
  9. jnurthen
    06:09
    How was the science?
  10. MichielBijl
    07:07
    it was great! The Evolution talks were all very good =D I didn't enjoy the Big Science talks as much though, but that might be because the previous night (Evolution) was such a blast. And got to spend time with my sweetheart =) (and we got to work on our own BAH Fast proposal)
  11. MichielBijl
    07:08
    The Evolution show is already on YouTube (was live streamed actually): https://www.youtube.com/channel/UC9v7v79mAlvKCrjrJvj-Fww
  12. MichielBijl
    07:18
    s/Fast/Fest/
  13. MichielBijl
    09:58
    Hunting for typos etc like there is no tomorrow: w3c/aria#224
  14. jkva
    10:31
    Nice going =)
  15. MichielBijl
    10:41
    all fixed :D
  16. MichielBijl
    11:25
    a very happy birthday to @stevefaulkner!
  17. StommePoes
    14:16
    Trying to find if someone had a hack around IE not sending generated content to AT
  18. StommePoes
    14:16
    Everything I can think of would require javascript
  19. MichielBijl
    14:46
    @StommePoes: don't think any browser sends CSS generated content to AT?
  20. StommePoes
    14:59
    they all do except IE
  21. StommePoes
    15:00
    So in FF for example, if there's a pseudo element showing lower-alpha counts in a regular-number ol list, you'll get a, b, c. In IE you'll get 1, 2, 3.
  22. StommePoes
    15:00
    well, I can't say they all do. FF does.
  23. StommePoes
    15:00
    And Safari seems to.
  24. MichielBijl
    15:12
    Safari+VoiceOver doesn't support character navigation of ::before/::after though.
  25. StommePoes
    15:17
    Good to know, though not an issue with our current code.
  26. StommePoes
    15:17
    I wouldn't be surprised if you couldn't spell them in any browser.
  27. StommePoes
    15:19
    Basically the developer is replacing browser-created bullets with pseudo-elements, so he can display: table those bullets for purposes of alignment.
  28. StommePoes
    15:19
    But so he made a test where he all-out replaces all sorts of weird things, which I think we're not going to do, but I have to consider a future developer possibly doing so after looking at our code.
  29. StommePoes
    15:21
    So it's one of those rabbit-holes like devs dealing with shadow-dom only because they aren't able to style things supplied by the browser (dropdown's arrows, video controls)
  30. MichielBijl
    15:27
    aha. I frequently add comments to the code just for those sort of situations (future developers not understanding).
  31. StommePoes
    15:28
    Well, we actually have a whole how-to-use page that the UX people are writing
  32. StommePoes
    15:28
    And for example I improved the buttons one by listing the thousand-warnings for devs who choose non- buttons for buttons.
  33. StommePoes
    15:29
    (I couldn't get so far as to prevent divs/spans and a's, but I could add in the warnings)
  34. StommePoes
    15:29
    So I think that's what't this designer will do, he asked me to test this idea.
  35. StommePoes
    15:29
    and the results weren't horrible
  36. StommePoes
    15:29
    but I wouldn't want text to reference a list item by a css-generated bullet that doesn't get to all students.
  37. StommePoes
    15:30
    IE+JAWS was the only combo to honour the start attrubute as well.
  38. StommePoes
    15:30
    That was interesting.
  39. MichielBijl
    15:30
    That is interesting.
  40. StommePoes
    15:34
    I should say, the only combo to honour it when it was being hidden by psuedos
  41. StommePoes
    15:35
    visually it looked like 1. 2. 3
  42. StommePoes
    15:35
    and every combo read out 1. 2. 3, except JAWS+IE, read 9. 10. 11 (start="9")
  43. StommePoes
    15:35
    because that's what IE gives you, but so then not certain what NVDA is doing.
  44. MichielBijl
    15:37
    how important is the distinction? They convey the same information no?
  45. MichielBijl
    15:37
    Unless other material references it by roman numerals for example…
  46. StommePoes
    15:39
    The start attr was, I think, created when you have a list and something interrupts it (say, an image or a bunch of stuff), so you're forced to write 2 different lists in HTML. However, text-wise and semantically, you consider them 1 big list, and don't want the number ordering to restart at 1.
  47. StommePoes
    15:39
    Like, list columns
  48. StommePoes
    15:39
    So you can set the start to a new number.
  49. StommePoes
    15:40
    In any case, yeah, I'm not able to think of a reason why a dev would first use start and then override it with pseudos, but I can see it happen as some sort of freak accident.
  50. StommePoes
    15:40
    Then you've got a freak accident that doesn't visually stand out, possibly with text referring to a visible number while some users never encounter that number.
  51. StommePoes
    15:41
    We're trying to test our CSS follows the Design rules with PhantomJS tests, but so far I'm finding it frustratingly difficult to use, and I can't tell if it's possible to write a test that would catch something like that.
  52. StommePoes
    15:41
    I've in fact given up on trying to test our hover/focus states.
  53. StommePoes
    15:42
    There's just too much node/es6/babel/abstractions/blahblah in the way.
  54. MichielBijl
    15:42
    I don't follow. Does the generated content honour the start attr?
  55. StommePoes
    15:43
    No, the developer used it as a thing to test-- he set a start attr to a high number on a new list (9 overriding 1), but then covered it up with pseudos showing 1.
  56. MichielBijl
    15:43
    :/
  57. StommePoes
    15:43
    Using counters
  58. MichielBijl
    15:43
    Then it's a fail…
  59. MichielBijl
    15:43
    Right?
  60. StommePoes
    15:43
    it's a fail for IE+JAWS, if we consider it success that AT+UA report what's visible.
  61. StommePoes
    15:44
    However it's not a test of anything we want to do. More a test of "what could happen".
  62. StommePoes
    15:44
    And while I knew from Leonie's article that IE doesn't pass on pseudos, testing showed NVDA still somehow figured it out on IE.
  63. MichielBijl
    15:45
    well, you could make it honour the number of elements if you add a class or something to every li that belongs to that particular list.
  64. MichielBijl
    15:45
    If you attach that counter to those that is.
  65. StommePoes
    15:46
    Our situation currently is, the designers are making a design, and what ought to happen if another developer grabs our stuff is, he gets our CSS.
  66. StommePoes
    15:46
    Or, he gets Design's super-approved CSS.
  67. MichielBijl
    15:46
    But that all depends on if you're gonna use it, because that is probably more trouble than it's worth.
  68. MichielBijl
    15:46
    I don't know what that is.
  69. StommePoes
    15:46
    And our testing shows that, should a developer do something like use the start attr, or need to explicitly refer to a list item showing a different symbol than what the list actually by default has, then there could be a problem.
  70. StommePoes
    15:47
    if our testing had shown that all possible combo's just read out whatever was visible, then there'd be no issue.
  71. StommePoes
    15:48
    So I have to see what all he's thinking of doing, and if there are any hacks that make sense to even think of doing, or otherwise finding a whole other solution to aligning bullets.
  72. StommePoes
    15:48
    Hm, I wonder if we could just attr(list-item) or something
  73. StommePoes
    15:48
    to get whatever bullet the browser would normally give out.
  74. StommePoes
    15:49
    then the psuedo element just copies whatever would have been made, and now we get a display: table bullet.
  75. StommePoes
    15:49
    Then there'd never be any mismatches to worry about in the first place.
  76. StommePoes
    15:50
    thing is, bullets aren't an attr.
  77. StommePoes
    15:50
    I dunno what they really are, actually.
  78. jkva
    15:50
    hangout in the bus. Mobile generation ftw
  79. StommePoes
    15:50
    I'd have to pay like 60 euros a month to do that
  80. StommePoes
    15:50
    to tether
  81. jkva
    15:51
    Yeah I don't prefer to do it either
  82. StommePoes
    15:51
    all our software are large, silly-stupid-large pages that suck MBs
  83. jkva
    15:51
    Although you can downgrade speed outside your plan
  84. jkva
    15:51
    You get 5GB for 20E
  85. StommePoes
    15:51
    But a hangout, asusming we turn off video, needs realtime
  86. StommePoes
    15:51
    how much data is that?
  87. StommePoes
    15:51
    I'm just assuming a lot
  88. jkva
    15:51
    Not sure, but it adds up
  89. StommePoes
    15:52
    Also the team pastes stuff into hipchat while we talk.
  90. StommePoes
    15:52
    Hipchat is a huge bloated zomg slack-for-jira.
  91. jkva
    15:52
    I'm just glad I don't get sick
  92. StommePoes
    15:52
    Yeah that would be nice. I could do a hangout I think, but reading and writing code churn my stomache.
  93. stevefaulkner
    16:31
    I vaguely remember someone saying in this very room recently that aria-hidden testing results need updating - The state of hidden content support in 2016 https://www.paciellogroup.com/blog/2016/01/the-state-of-hidden-content-support-in-2016/
  94. jkva
    16:31
    Sweet! Thanks @stevefaulkner , and happy birthday!
  95. stevefaulkner
    16:31

    :point_up: January 26, 2016 11:25 AM

    :+1:

  96. stevefaulkner
    16:32
    @jkva cheers
  97. garcialo
    16:33
    @stevefaulkner It's your birthday?
  98. jkva
    16:33
    Rumor has it
  99. jkva
    16:34
    He keeps his birthday cleverly hidden =)
  100. StommePoes
    16:36
    aria-berfday
  101. stevefaulkner
    16:36
    52 sweet years
  102. garcialo
    16:36
    Oh wow, you look way older.
  103. jkva
    16:36
    =D
  104. stevefaulkner
    16:37
    it's due to the last few since being stalked by @StommePoes
  105. StommePoes
    16:45
    I didn't even know! Michiel knew
  106. stevefaulkner
    16:54
    Michiel Cares
  107. MichielBijl
    17:09
    don't think I have anything to add to that ;)
  108. jnurthen
    17:14
    you should add it to your profile
  109. jnurthen
    17:14
    "Michiel cares"
  110. MichielBijl
    17:17
    Haha, don't think I can cram anything more into my profile :P
  111. MichielBijl
    17:19
    Oh, I fixed 10 million typos in the APG today. I also added some, but those 'Muricans insist on writing certain words wrong…
  112. MichielBijl
    17:19
    Should we just change the lang to en_gb and correct those =P
  113. jnurthen
    17:20
    I'm not sure Matt (or the W3C would be too happy about that)
  114. MichielBijl
    17:21
    That could be an issue I guess.
  115. jnurthen
    17:22
    @stevefaulkner do you use american or real english in your specs?
  116. stevefaulkner
    17:41
    @jnurthen I think en-US is expected, but am sure that emglish spellings creep in to my writing
  117. jnurthen
    17:43
    yeah - that is what I expected. Just looked in the style guide
  118. jnurthen
    17:43
    https://www.w3.org/2001/06/manual/
  119. jnurthen
    17:44
    Although technical reports are written in U.S. English, examples and wording should not rely on conventions and idioms used only in the United States (e.g., "ZIP code"). Use international examples (e.g., "postal code") wherever possible.
  120. jnurthen
    17:44
    happy birthday btw.
  121. MichielBijl
    17:47
    Ah, so do editor notes etc fall under “examples and wording”
  122. powrsurg
    17:54
    happy birthday
  123. powrsurg
    17:54
    @stevefaulkner
  124. zakim-robot
    17:56
    [sylvia] Hey all! I have a WCAG 2.0 question. I have some devs who have (probably accidentally) assigned the attribute aria-hidden="true" to a bunch of visible text. This is clearly a problem, and I'll report it whether or not is a strict failure of WCAG 2.0
  125. zakim-robot
    17:57
    [sylvia] But I was wondering if there is a SC that covers this situation. It fails the spirit of Perceivable, but I can't figure out if it actually violates any of the SC
  126. MichielBijl
    18:05
    1.3.1 Info and Relationships maybe?
  127. garcialo
    18:05
    Yes, 1.3.1
  128. MichielBijl
    18:06
    https://www.w3.org/TR/UNDERSTANDING-WCAG20/content-structure-separation-programmatic.html
  129. garcialo
    18:06
    Basically 1.3.1 says stuff on the page needs to marked up as the stuff it is.
  130. garcialo
    18:06
    That visible text is marked up as invisible text.
  131. zakim-robot
    18:12
    [marcysutton] Hi @sylvia!!! :wave:
  132. zakim-robot
    18:13
    [sylvia] Hi @marcysutton!
  133. zakim-robot
    18:13
    [sylvia] Thanks @michiel and @garcialo. That makes sense.
  134. MichielBijl
    18:13
    :+1:
  135. MichielBijl
    18:53
    @stevefaulkner is there anything we can tweak in the sameroom settings regarding displaying of usernames (from Slack to here)?
  136. MichielBijl
    19:07
    never mind, seems there isn't.
  137. zakim-robot
    19:36
    [thecristen] hi everyone! does anyone here have experience creating an accessibility plan for your team/department/company/etc?
  138. zakim-robot
    19:36
    [thecristen] i’m trying to write a one-pager and wondering what i should make sure to cover
  139. zakim-robot
    19:41
    [adamscott] @thecristen: the CFPB’s front-end team has accessibility guidelines. I’m not sure if this is the sort of thing you have in mind https://github.com/cfpb/front-end/blob/master/accessibility.md
  140. zakim-robot
    19:43
    [adamscott] and our design team has put together these guidelines: https://cfpb.github.io/design-manual/guides/accessible-interfaces.html
  141. zakim-robot
    19:45
    [devonpersing] @thecristen: a lot of it will depend on how your team normally works (agile? etc.), but it would probably be one thing you’ll definitely want to include how accessibility affects different roles on your team.
  142. MichielBijl
    19:49
    @thecristen: do you want to lay down some ground rules or do you want to inform people about accessibility?
  143. zakim-robot
    19:50
    [sylvia] I'd also include what standards or tools you support
  144. zakim-robot
    19:51
    [sylvia] So for example, WCAG 2.0, or JAWS and NVDA
  145. zakim-robot
    20:09
    [cameron] @MichielBijl Michiel are you free to try our screen sharing setup again?
  146. MichielBijl
    20:11
    No, sorry, can you do tomorrow night (around this time is fine)?
  147. zakim-robot
    20:11
    [cameron] sure!
  148. MichielBijl
    20:12
    Cool, see you then.
  149. MichielBijl
    20:12
    I might have a problem talking, but other than that, I'm fine (dentist appointment) =P
  150. stevefaulkner
    20:36
    @powrsurg cheers
  151. zakim-robot
    22:20
    [cameron] I don't think Screen Sharing on OS X will handle audio by itself
  152. zakim-robot
    22:21
    [cameron] Screen Sharing with Apple Remote Desktop will allow a user to control Voice Over
  153. zakim-robot
    22:22
    [cameron] but it will not pass through audio (I think it uses the vnc protocol, which does not support audio)
  154. zakim-robot
    22:22
    [cameron] my best bet so far is to try using the Messages app on OS X, which opens Screen Sharing and can also launch a FaceTime call
  155. zakim-robot
    22:24
    [cameron] @MichielBijl Michiel I think this is the approach we should try tomorrow
  156. zakim-robot
    22:24
    [cameron] ^^
  157. zakim-robot
    22:24
    [cameron] Messages + Screen Share (w/ Remote Desktop) + FaceTime
  158. jnurthen
    22:25
    some of our teams at work use zoom. On windows you can share system audio (so we can hear JAWS over the conference) - don't know how it works on mac
  159. zakim-robot
    22:25
    [cameron] interesting
  160. jnurthen
    22:26
    looks like they have free accounts with less features (and 40 min meeting limit)
  161. jnurthen
    22:26
    https://www.zoom.us/pricing
  162. zakim-robot
    22:26
    [cameron] @jnurthen anything will do really, just looking to document a decent solution for remote usability testing with screen readers
  163. jnurthen
    22:26
    the iOS and Android screen sharing is awesome too....
  164. zakim-robot
    22:27
    [cameron] there's a requirement that a user would have remote access to the host
  165. jnurthen
    22:27
    wish i had a license but my department decided to go with webex
  166. zakim-robot
    22:27
    [cameron] @jnurthen does zoom support remote management?
  167. zakim-robot
    22:27
    [cameron] I can poke around if you're not sure
  168. jnurthen
    22:28
    no idea - i have never hosted a meeting with it only been a participant
  169. jnurthen
    22:28
    but I like it way better than webex from what I have seen.
  170. zakim-robot
    22:37
    [cameron] @marcysutton do you have a way you do remote usability studies?
  171. zakim-robot
    22:37
    [cameron] specifically for blind users?
  172. zakim-robot
    22:38
    [marcysutton] I don’t have experience with that, but I can ask around...
  173. zakim-robot
    22:39
    [cameron] Thanks! Trying on OS X, but Windows would be fine too
  174. zakim-robot
    22:40
    [cameron] I think NVDA remote access is a possibility
  175. zakim-robot
    22:40
    [cameron] the biggest pain so far has been sharing audio and control
  176. garcialo
    22:42
    Knowbility (http://www.knowbility.org) is an Austin based nonprofit that can help you do some user testing with PWD. They use http://www.loop11.com/; it's quite nice.
  177. zakim-robot
    22:54
    [cameron] Thanks @garcialo! I'm trying to document an "off the shelf", free or close to free, solution.
  178. garcialo
    22:54
    Ah, gotcha.