Archive index

A11y Slackers Gitter Channel Archive 3rd of February 2016

What fresh hell is THIS now? - Patrick Lauke
  1. stevefaulkner
    Feb 03 00:01
    Good question @stacycarston I believe it does result in differing behaviour, will do some testing on the morrow
  2. stacycarston
    Feb 03 00:13
    wow, awesome, thank you!
  3. zakim-robot
    Feb 03 00:34
    [marcysutton] @estelle: you can also to a single finger swipe left or right to go through the entire tab order, or use the rotor to navigate by headings/links/landmarks/etc.
  4. MichielBijl
    Feb 03 09:30
    Is there a way to prevent iOS from zooming in on a form control without disabling zoom altogether?
  5. MichielBijl
    Feb 03 09:48
    @LjWatson on screen reader detection: http://tink.uk/thoughts-on-screen-reader-detection/
  6. StommePoes
    Feb 03 10:09
    Steve had posted that earlier, when the discussion was in full swing
  7. StommePoes
    Feb 03 10:11
    But the answer is still the same-- the issue isn't "well sure some people could do evil with this, but that's no reason to prevent its release into the wilds!" because history shows us again, and again, and again that the majority will, indeed, use it for evil. Usually unintentionally, but still.
    And they say the definition of crazy is doing the same thing over and over again while expecting magically-different results. Humans gonna Hume.
  8. StommePoes
    Feb 03 10:11
    Steve had also questioned my species. I take... umbrage at that!
  9. StommePoes @StommePoes takes the umbrage and walks away with it
  10. MichielBijl
    Feb 03 10:16
    Oh my bad, hadn't seen it earlier.
  11. MichielBijl
    Feb 03 10:17
    @StommePoes, you're a cat, what is to question about that?
  12. MichielBijl
    Feb 03 10:18
    I've been classified as a mad bad dinosaur scientist that is also a cat.
  13. MichielBijl
    Feb 03 10:18
    Not sure how that works, but it's true!
  14. MichielBijl @MichielBijl dresses up in his dinosaur onesie.
  15. jkva @jkva hands @MichielBijl a Tokyo scale model to barge through
  16. MichielBijl
    Feb 03 10:28
    =D
  17. LjWatson @LjWatson lols @jkva
  18. stevefaulkner
    Feb 03 10:37

    Steve had also questioned my species. I take... umbrage at that!

    I was being lugubrious

  19. LjWatson @LjWatson thinks she has disappeared down the rabbit hole in there today.
  20. LjWatson
    Feb 03 10:38
    s/there/here/
  21. StommePoes
    Feb 03 12:47
    https://www.youtube.com/watch?v=_TEwesv1gug Lugubrious Whing Whang
  22. StommePoes
    Feb 03 12:48
    Sorry, Steve: I did not hear any banjos in your remark.
  23. zakim-robot
    Feb 03 14:20
    [jdan] do we have a room for mobile web a11y questions?
  24. powrsurg
    Feb 03 14:20
    @MichielBijl to disable the zooming in iOS on form fields you'll need to bump up your font size. I believe old-school iphone sizes worked well with 16px, but iphone 6's bumped it to 20px. But I haven't tested this in a while
  25. MichielBijl
    Feb 03 14:21
    Found that out, sorry I didn't share. 16px still works on iPhone 6(s)
  26. MichielBijl
    Feb 03 14:21
    Thanks anyway :)
  27. powrsurg
    Feb 03 14:21
    I think that's true really for any mobile device.
  28. powrsurg
    Feb 03 14:22
    As far as I'm aware there is no way to reset a zoom level back to what it was on blur after it does the zoom-in for the form field. That would be nice
  29. zakim-robot
    Feb 03 14:23
    [jdan] (oh I guess we're talking about ios web a11y here :D) is it possible to make the ios voiceover cursor focus an element?
  30. powrsurg
    Feb 03 14:40
    I've found that voiceover doesn't care about your focus will read what it wants to read in whatever order it wants to read it. I've seen had divs with a title and then an image and then a description where ios will read the image first and then the title
  31. powrsurg
    Feb 03 14:41
    because apple is smarter than typical web designers apparently
  32. jnurthen
    Feb 03 15:58
    I've long wished for a tag I could add to a page to turn off all the shitty broken AT heuristics.
  33. MichielBijl
    Feb 03 15:59
    <!DOCTYPE xhtml>?
  34. jnurthen
    Feb 03 15:59
    The "I know what I'm doing - don't try to fix it" tag
  35. powrsurg
    Feb 03 16:11
    role="application"?
  36. powrsurg
    Feb 03 16:14
    Does anyone have a list of recommendations that you may want to share with a designer who is about to put together stuff for a site relaunch? I already have things like a color contrast and color blindness checking ...
  37. powrsurg
    Feb 03 16:14
    they won't be writing any of the HTML or CSS
  38. zakim-robot
    Feb 03 16:21
    [adamscott] @powrsurg: these are the guides our design team follows https://cfpb.github.io/design-manual/guides/accessible-interfaces.html
  39. zakim-robot
    Feb 03 16:22
    [adamscott] there are a few references to HTML, but it is mostly content and design guidelines
  40. powrsurg
    Feb 03 16:23
    Yeah, that looks good. Though a bunch of that looks to be talking about things like alt tags and the like which gets into further detail since the designer is more going to just create a PSD that is shown off and then we implement everything else
  41. zakim-robot
    Feb 03 16:24
    [adamscott] a few of the latter sections might be more useful - it may be a matter of ripping out what is needed for just a visual designer
  42. powrsurg
    Feb 03 16:26
    yeah, I see. This stuff is good.
  43. powrsurg
    Feb 03 16:26
    Thank you :)
  44. zakim-robot
    Feb 03 16:26
    [adamscott] awesome! glad it’s useful
  45. jkva
    Feb 03 16:27
    Second that
  46. powrsurg
    Feb 03 16:28
    This is one of those "I know all of these things myself, what do I need to pass on. Someone has got to have already come up with a list."
  47. MichielBijl
    Feb 03 16:32
    I've used that list for a presentation preparation.
  48. MichielBijl @MichielBijl is giving a talk to some designers soon.
  49. jnurthen
    Feb 03 17:56
    @powrsurg not application please.... I want something to: (1) turn off all label guessing heuristics, (2) turn off all table guessing heuristics - a table is a table unless I have marked it with role=presentation, (3) turn off all focus-movement guessing (I'm looking at you VO) - if I move the focus then move VO focus there too - if I don't move the focus and i needed to then let my page be broken. (4) I'm sure there are more....
  50. jnurthen
    Feb 03 17:56
    this way we could get consistent behaviour.... it would also enable QA to find bugs using AT rather than have AT cover them up.
  51. powrsurg
    Feb 03 18:13
    oh I wasn't suggesting that application would be good ... but it would disable AT heuristics (at least my understanding of application).
  52. powrsurg
    Feb 03 18:14
    This message was deleted
  53. jnurthen
    Feb 03 18:15
    not really - if i navigate to a form field AT would still guess a label if I haven't supplied one - even with role=application specified
  54. jnurthen
    Feb 03 18:59
    what html presentation libraries are folks using these days?
  55. powrsurg
    Feb 03 19:09
    like ... jQuery UI?
  56. jnurthen
    Feb 03 19:10
    no - sorry not clear for presentations (like powerpoint but in HTML)
  57. powrsurg
    Feb 03 19:11
    it's been a long time since I needed to build one from scratch, but a decade ago Eric Meyer created S5 - http://meyerweb.com/eric/tools/s5/
  58. jnurthen
    Feb 03 19:12
    I was wondering if anyone had made anything better since :)
  59. powrsurg
    Feb 03 19:13
    not that I'm aware ... but there probably has been
  60. powrsurg
    Feb 03 19:13
    that page just made me feel old
  61. jnurthen
    Feb 03 19:13
    yeah - i get that feeling frequently these days
  62. jnurthen
    Feb 03 19:14
    the worst is when you go to the Doctor about something and their reply is that it is normal - you are just getting old :(
  63. powrsurg
    Feb 03 19:31
    that tells you to just stop going to the doctor :-p
  64. MichielBijl
    Feb 03 19:42
    or do as Jim Jeffries, just move to countries with shittier and shittier health until you're the healthiest man on the North Pole or something.
  65. powrsurg
    Feb 03 19:44
    That works too ;)
  66. powrsurg
    Feb 03 19:44
    So Microsoft just published their list of priorities for 2016. Accessibility was #2 after extensions
  67. powrsurg
    Feb 03 19:45
    https://blogs.windows.com/msedgedev/2016/02/03/2016-platform-priorities/
  68. MichielBijl
    Feb 03 19:57
    Accessibility extensions eh?
  69. powrsurg
    Feb 03 19:58
    Those might take longer :-p
  70. powrsurg
    Feb 03 19:59
    oh, they mentioned they're investigating script-based accessibility ...
  71. jnurthen
    Feb 03 20:00
    WAPA
  72. jnurthen
    Feb 03 20:01
    if you haven't seen it - https://github.com/cyns/wapa/blob/master/ScriptAccessibility.md
  73. powrsurg
    Feb 03 20:11
    So setting properties on elements, but not via the DOM?
  74. powrsurg
    Feb 03 20:12
    Also, it seems like to accomplish this would require some form of AT detection, which people here jumped on me here yesterday for not hating
  75. jnurthen
    Feb 03 20:13
    no AT detection needed - although I'm guessing you could infer pretty easily whether the Accessibility APIs are being exercised via whether the scripting got run
  76. jnurthen
    Feb 03 20:14
    but you can easily do that today via flash if you really want to. Doesn't tell you any more than the fact that the accessibility APIs are being used
  77. powrsurg
    Feb 03 20:40
    Scenario #1 talks about setting ARIA stuff for everyone regardless of whether they need it being programmatically costly and of little value to those not using AT. From that I'm getting that they would rather not provide ARIA stuff to those that aren't using AT in order to improve performance ... that would mean detection
  78. jnurthen
    Feb 03 20:41
    but the browser does that. you supply the code to respond to the AT events. the browser will only fire them when AT requests them
  79. jnurthen
    Feb 03 20:42
    browsers know this already.... they only maintain the accessibility tree if something is using it
  80. powrsurg
    Feb 03 20:50
    so executing JavaScript that only changes the AT's DOM, not the browser's DOM? (I forget what the term is for the AT's copy of the DOM) ...
  81. jnurthen
    Feb 03 20:55
    I'm not sure it is quite like that. I think we still have the "accessibility tree" but when the request comes for a property on a node in that tree, before it is returned from the accessibility tree, the scripting is invoked. You could then return the correct value from your script (i.e. provide it on request) rather than having to popuplate everything in advance.
  82. jnurthen
    Feb 03 20:58
    Actually - what I pointed to is out of date. Look at https://github.com/WICG/a11yapi for more recent discussion.
  83. jnurthen
    Feb 03 21:00
    if you want to influence join WICG and take part!
  84. powrsurg
    Feb 03 21:09
    ... I just noticed that page you first posted has five scenario #1's ...
  85. jnurthen
    Feb 03 21:09
    yeah.... they are all the most important
  86. jnurthen
    Feb 03 21:09
    :)
  87. powrsurg
    Feb 03 21:11
    "element.onariarequest is a new event that gets fired when an AT requests makes an Accessibility API request that maps to aria information" - this would result in the detection of AT (maybe not what AT was in use, but the presence of AT) which everyone jumped on me for being cool with yesterday ...
  88. jnurthen
    Feb 03 21:12
    correct it would. You know you can embed a tiny flash object on your page to do this today if you really want to
  89. powrsurg
    Feb 03 21:14
    well, that wouldn't work on mobile
  90. jnurthen
    Feb 03 21:14
    correct
  91. jnurthen
    Feb 03 21:16
    perhaps the browsers could fire these events randomly and then discard the results to stop that possibility :)
  92. jnurthen
    Feb 03 21:17
    (although I can't see them doing that - performance and everything)
  93. zakim-robot
    Feb 03 22:30
    [pallezingmark] Evening!
    I'm on the search for a definitive list of HTML elements, which also receives focus after being clicked, e.g. <button>.
    Anyone here that can point me in some direction, would be much appreciated. :simple_smile:
  94. jnurthen
    Feb 03 22:34
    I think it is browser dependent
  95. jnurthen
    Feb 03 22:34
    http://www.sitepoint.com/when-do-elements-take-the-focus/
  96. zakim-robot
    Feb 03 22:35
    [pallezingmark] Yes, you are probably right.. so a definitive list is maybe to much to ask for. :wink:
  97. zakim-robot
    Feb 03 22:37
    [pallezingmark] Read that article, unfortunately didn’t really answer my question.
  98. jnurthen
    Feb 03 22:48
    .... it makes it clear that for links the behaviour differs from one browser to another.
  99. PalleZingmark
    Feb 03 22:48
    Indeed.