Archive index

A11y Slackers Gitter Channel Archive 8th of December 2015

What fresh hell is THIS now? - Patrick Lauke
  1. MichielBijl
    06:21
    Godmorn Slackers
  2. garcialo
    06:21
    howdy
  3. garcialo
    06:21
    ...I should sleep =p
  4. MichielBijl
    06:25
    Haha, it's what, 00:25 over there?
  5. garcialo
    06:25
    yeah, but I have an 8 am client meeting
  6. MichielBijl
    06:25
    Ai
  7. MichielBijl
    06:26
    First meeting is in two hours. So have some time before I have to get up.
  8. MichielBijl
    06:30
    Well, for those 7 hours behind: godnat ;)
  9. garcialo
    06:30
    night
  10. MarcoZehe
    10:00
    The day aria-hidden goes the way of all earthly things will be the most joyous day of my life, and I will dance on its grave!
  11. MichielBijl
    10:15
    Why the joyful hatred @MarcoZehe?
  12. MarcoZehe
    10:22
    @MichielBijl: It is poorly specified, creates more problems than it solves, and those that it is supposed to solve could in most cases be solved with other proper coding. It was pushed into the spec by some without full consideration of the consequences.
  13. StommePoes
    10:24
    Someone asked the WebAIM list why their buttons weren't being read out in JAWS
  14. StommePoes
    10:24
    turns out they had <button aria-hidden="true">blah</button>
  15. StommePoes
    10:24
    derp
  16. StommePoes
    10:25
    Joanie found an example where a modal dialog was doing all the right things except... when people clicked on the button to open it, aria-hidden wasn't being turned to "false" so it remained hidden
  17. MichielBijl
    10:29
    That was on an a11y website too right?
  18. MichielBijl
    10:29
    @MarcoZehe fair enough.
  19. MarcoZehe
    10:31
    @StommePoes: And there with that example, you have one of the problems: What happens when there is something focusable in the tree branch that aria-hidden is hiding? There is no answer to that in the specification for browsers.
  20. MarcoZehe
    10:32
    The specifications ignore assistive technologies anyway, so whatever they do is guess work. Not just for aria-hidden, for everything. But even for browsers: Should we keep the accessible objects available somewhere so we can fire proper focus events once the user tabs into that hidden tree onto the focusable item?
  21. MichielBijl
    10:33
    @MarcoZehe the only browser involved is Safari…
  22. MichielBijl
    10:33
    And even those guys don't show up that often.
  23. MarcoZehe
    10:35
    and his users.
  24. MarcoZehe
    10:35
    @MichielBijl: Yes, I believe it was Apple's idea, but Google didn't change it when they forked Blink, and Microsoft also adopted a rather pruning approach I believe. But none answered the questions. And we at Mozilla were criticized for not following their lead. Well we fortunately had Jamie from NVDA who pointed out all the problems this would cause for him
  25. MarcoZehe
    10:35
    And those are the people we serve, right? Users, who actually have to use the shit we put out there.
  26. MichielBijl
    10:36
    Yep
  27. MichielBijl
    10:36
    There are some SR users in the PFWG (including you :P)
  28. MichielBijl
    10:36
    Or the ARIA WG (still not accepted :()
  29. MarcoZehe
    10:36
    Yes, but you know what? I have real work to do. ;-)
  30. MichielBijl
    10:37
    Hehe, yeah everybody is busy.
  31. MichielBijl
    10:37
    But it's hard to write a single document for everyone.
  32. MarcoZehe
    10:37
    Moreover, those meetings are all at times when I already have an 8 or 9 hour work day under my belt. And I do have something abstract as a private life, too. ;-)
  33. MichielBijl
    10:38
    Implementors complain it's not specific enough, developers complain there are no examples, users complain it's too tech-y.
  34. MichielBijl
    10:38
    We want to move the APG meeting to around 8-9 'o clock.
  35. MichielBijl
    10:38
    Not sure what the ARIA call will do.
  36. MarcoZehe
    10:39
    8-9 which timezone?
  37. MichielBijl
    10:39
    GMT+1
  38. MarcoZehe
    10:39
    P.M. I suppose?
  39. MichielBijl
    10:39
    Yes :)
  40. MarcoZehe
    10:39
    That's even later than now.
  41. MichielBijl
    10:40
    Doesn't really matter for the America's and gives us Europa's a bit of a rest.
  42. MarcoZehe
    10:40
    Depends. My work day usually goes until 6 P.M., when my colleagues in Toronto go to lunch.
  43. MichielBijl
    10:41
    It's either us or them that have to wake up really early or have to do it in the evening.
  44. MichielBijl
    10:42
    I was discussing this with Léonie and she preferred later too.
  45. MichielBijl
    10:42
    I don't get time for these meetings from my employer, so for me it's a moot point.
  46. MarcoZehe
    10:42
    Yeah I am also not comfortable with these phone meetings at all. I usually can explain stuff better in writing.
  47. MarcoZehe
    10:42
    Moreover, there is also stuff going on in an IRC channel while the call goes on, and I cannot, for the life of me, have two sources of input like that.
  48. MichielBijl
    10:43
    I know what you mean, I do too, unless I've thought about it for a long time.
  49. MarcoZehe
    10:43
    I don't know how other blind people do it, I cannot.
  50. MichielBijl
    10:43
    Hmm. I think most ignore IRC except when someone says something like “I pasted a link in IRC”.
  51. MarcoZehe
    10:44
    Yeah but when you look at minutes of such meetings, there are often IRC messages that are typed in response to something said on the phone, and are thus relevant to the discussion.
  52. MichielBijl
    10:45
    Unfortunately, that is true. I don't see it happening all that often, but it does happen.
  53. MarcoZehe
    10:47
    As for what you said earlier about developers wanting examples: I definitely see that, too. Developers want clear, concise examples they can follow and don't want to read lengthy and mostly theoretical documentation. The ARIA tips with examples are by far the most popular posts on my blog by visitors, for example.
  54. MichielBijl
    10:47
    Good thing we're working on examples for ARIA. But needs more simple examples in the document itself.
  55. MichielBijl
    10:48
    Takes me a lot of time to create such things, because I don't understand most the first time I read it.
  56. MarcoZehe
    10:51
    And that's a problem right there. If even those of us who are involved with this stuff have trouble understanding the documentation, how are outsiders supposed to?
  57. MichielBijl
    10:53
    That's why I joined :P
  58. MichielBijl
    10:53
    One of the reasons anyway.
  59. pkra
    11:22
    @StommePoes "Joanie found an example where a modal dialog was doing all the right things except... when people clicked on the button to open it, aria-hidden wasn't being turned to "false" so it remained hidden" I think that was on our site at mathjax.org (or at least she caught that on our site as well. And yes, it included a modal about a11y -- sigh). More a bug than a lapse but that hardly matters. Calling it out helped ;-)
  60. StommePoes
    13:03
    pkra you guys aren't the only ones to have done something like that, was more the point
  61. StommePoes
    13:04
    because aria-hidden doesn't jump out at developers, they may only notice it's sitting there hiding stuff when they test with an SR, while I think we need code linters to somehow catch this
  62. StommePoes
    13:05
    on the other hand, I don't want aria-hidden gone because more and more SRs are starting to voice more and more BS like icon-fonts and unicode silliness.
  63. StommePoes
    13:05
    And we as developers need a way to not let people wonder what the hell that crap is
  64. StommePoes
    13:14
    @michielBijl I just updated NVDA on another machine so now I know the version I tested your stuffs on
  65. StommePoes
    13:14
    2015.4
  66. StommePoes
    13:14
    @marcoZehe I would say, focusables inside aria-hidden should not be.
  67. StommePoes
    13:15
    But that's my 2 cents. Unless it's a goofy stupid emoticon being used with a separate, more sensible text alternative, aria-hidden things ought to generally be hidden from everyone in my view. So focusables inside should, from the browser, have the equiv of negative tabindex
  68. MichielBijl
    13:19
    @StommePoes coolie joolie!
  69. StommePoes
    13:28
    ug windows auto-updated my browser
  70. StommePoes
    13:29
    now I can't read anything in Zoomtext 10
  71. StommePoes
    13:29
    hope I can upgrade to ZT 10.1
  72. zakim-robot
    13:30
    [karlgroves] That was nice of them
  73. StommePoes
    13:32
    "ZoomText was not found on the system!" WTF
  74. MarcoZehe
    13:35
    @StommePoes: Yes, that's what you would say, and I, probably, too. But the spec doesn't say it. It doesn't say anything about this case, and that leads to very "interesting" behavior all over the place.
  75. StommePoes
    13:36
    I know, someone should just throw that in there (specs) until someone comes up with a good reason not to have it. It seems that was done for some of the keyboarding for some of the widgets after all
  76. StommePoes
    15:50
    @MichielBijl @MarcoZehe since Pearson wants easy-to-grab-n-use widgets and since we're going to be building a bunch and they'll all be on github etc, I'm thinking they could be examples for the ARIA specs
  77. MarcoZehe
    15:51
    That sounds awesome!
  78. StommePoes
    15:52
    It does, but they also move like molasses
  79. StommePoes
    15:53
    And the drag and drop we were looking at, and likely lots of other things, don't work with SRs that well yet due to still-bad support.
  80. zakim-robot
    16:20
    [marcysutton] We need a true inert attribute to disable interactivity on visible regions….aria-hidden and tabindex all wrapped up into one. The MS Edge team was interested :simple_smile:
  81. MichielBijl
    16:35
    @StommePoes cool!
  82. jnurthen
    17:40
    @marcysutton I agree - inert would be so useful but got dropped from the HTML5 spec IIRC
  83. zakim-robot
    17:41
    [marcysutton] that sucks…… maybe that idea can be repackaged as something else
  84. jnurthen
    17:42
    Here is the info - https://bugzilla.mozilla.org/show_bug.cgi?id=921504
  85. jnurthen
    17:42
    Yeah - I wish. It would solve all sorts of issues.
  86. jnurthen
    17:51
    If you can get edge to implement it then that would be a good way to get it pushed back into the spec
  87. zakim-robot
    17:51
    [alice] Curious: is there a use case for inert beyond "not the active modal control"?
  88. jnurthen
    17:53
    Are you implying that dialog can solve it in the reverse case?
  89. zakim-robot
    17:53
    [alice] I was in a discussion a little while back (which I neglected to get back to - sounds like now is as good a time as any) about possibly inverting the sense
  90. zakim-robot
    17:53
    [alice] Lower-level than dialog, but essentially yeah
  91. zakim-robot
    17:54
    [alice] with inert you still run the risk of having the same horrifying tree walk that Marcy did for the Angular modal dialog
  92. jnurthen
    17:54
    I think that could work fine although I can conceive of a UI with 2 dialog type things active (although I think it would be ugly and those devs would deserve all the pain they would get from that)
  93. jnurthen
    17:55
    thinking about it the inverted sense makes more sense I think
  94. jnurthen
    17:55
    Let me ping our fwk devs to discuss
  95. garcialo @garcialo wished he understood all the words flying around.
  96. zakim-robot
    18:00
    [marcysutton] "If we implement this concept it'll be as part of <dialog>. There's no need for this bug as far as I can tell.” except we have a need for inert subtrees outside of <dialog>
  97. zakim-robot
    18:04
    [alice] @marcysutton: Ooh, what's the use case you have in mind?
  98. zakim-robot
    18:04
    [marcysutton] A custom select or dropdown menus, which aren’t dialogs...
  99. zakim-robot
    18:04
    [marcysutton] it is still something covering part of the UI but it should block interaction while remaining visible
  100. zakim-robot
    18:05
    [alice] I'm re-reading through https://groups.google.com/forum/#!msg/browser-accessibility-dev/QinGGM_OM7Y/ki1aXuYH5TwJ
  101. zakim-robot
    18:05
    [marcysutton] OOH thanks for relinking! I think about this thread every so often and I had forgotten where to find it
  102. zakim-robot
    18:06
    [alice] Freaking love this reply from @marcysutton https://groups.google.com/forum/#!msg/browser-accessibility-dev/QinGGM_OM7Y/rS1Nu4D1BQAJ
  103. jnurthen
    18:10
    Sounds like this is another mailing list I need to be on!
  104. MichielBijl @MichielBijl is sick of trying to install Ubuntu. I give up.
  105. jnurthen
    19:34
    Isn't ubuntu the easiest linux dist to install?
  106. MichielBijl
    19:35
    Should be. But the ISO you download is not compatible with OS X Disc Utility.
  107. MichielBijl
    19:35
    Great work…
  108. jnurthen
    19:35
    ah OSX....
  109. garcialo
    19:36
    do you need Disc Utility to mount it on the VM?
  110. MichielBijl
    19:36
    Maybe I should try to burn the disc via a VM with Windows :P
  111. MichielBijl
    19:36
    No, I want to install it to replace OS X on my raging iMac.
  112. MichielBijl
    19:36
    aging*
  113. garcialo
    19:36
    ah
  114. garcialo
    19:36
    I was going to guess you were having problems because you were trying to install it on a toaster
  115. garcialo
    19:36
    I guess that's close enough
  116. MichielBijl
    19:37
    Haha
  117. MichielBijl
    19:37
    Well, I used hdiutil to convert to iso to an img, but that doesn't help. Even thought that is what they say on there help pages.
  118. zakim-robot
    19:37
    [blavalla] @MichielBijl Are you trying to get it onto a DVD or a USB drive?
  119. MichielBijl
    19:37
    USB.
  120. MichielBijl
    19:38
    But DU says that there is no mountable file system.
  121. zakim-robot
    19:38
    [blavalla] I think you need to convert the ISO to a .img file for Disk Utility to be able to do anything with it.
  122. MichielBijl
    19:38
    Did that :)
  123. zakim-robot
    19:38
    [blavalla] ah, damn
  124. zakim-robot
    19:38
    [blavalla] You using El Capitan?
  125. MichielBijl
    19:38
    Oh wait. I think it did something.
  126. MichielBijl
    19:38
    10.5 Leopard #oldskool
  127. zakim-robot
    19:39
    [blavalla] Snow Leopard was a classic, Leopard not so much :simple_smile:
  128. MichielBijl
    19:40
    Yeah, it also runs Yosemite, but that is slow as hell on this monster from 2008.
  129. MichielBijl
    19:41
    Running old OS X' didn't seem like an ideal solution, so maybe Ubuntu will run smoothly.
  130. zakim-robot
    19:41
    [blavalla] You might have better luck in Yosemite, since the Leopard version of Disk Utility is pretty ancient at this point. I doubt anyone is testing their images with it to amke sure they work.
  131. MichielBijl
    19:41
    I'm now in Ubuntu boot :D
  132. MichielBijl
    19:41
    The terminal command hadn't finished for copying all the stuff to the disc -.-
  133. MichielBijl
    19:42
    Not having a progress thing I assumed it to have stalled.
  134. MichielBijl
    19:42
    Thanks for the help!
  135. zakim-robot
    19:43
    [blavalla] Hah! Good luck with Ubuntu!
  136. MichielBijl
    19:43
    Oh it'll suck, but I'll try.
  137. MichielBijl
    19:45
    FML, doesn't support apple wireless peripherals.
  138. MichielBijl @MichielBijl is gonna watch some a11y talks.