Archive index

A11y Slackers Gitter Channel Archive 12th of August 2015

What fresh hell is THIS now? - Patrick Lauke
  1. stevefaulkner
    Aug 12 00:53
    @estelle @sogami invalid="false" acc API mappings http://www.w3.org/TR/core-aam-1.1/#ariaInvalidFalse
  2. stevefaulkner
    Aug 12 00:58

    [Sam Ogami, a11y] I thought that aria-invalid="false" does work... If you look at the link in my previous post the first time you focus on the textfield it dose not read invalid but aria-invalid="false" attribute is in the input element

    until you enter some text into the required fields, state is aria-invalid=true

  3. zakim-robot
    01:03
    [Sam Ogami, a11y] So it does work. Just that http://html.cita.illinois.edu/nav/form/aria/index.php?example=3 has some js that adds the word "In valid:" and does not remove it when the field is valid :simple_smile: Thanks @stevefaulkner and @estelle
  4. stevefaulkner
    08:09

    A11ySlackers chat log 11th of August 2015 http://w3c.github.io/a11ySlackers/archives/110815.html - archive automation is coming... big up to @IanPouncey

  5. StommePoes
    08:12
    give the bot a cute name
  6. StommePoes
    08:12
    "ally" perhaps.
  7. rodneyrehm
    08:13
    .oO( maybe calling my library ally.js wasn’t such a good idea after all :D )
  8. StommePoes
    08:16
    gitterbot would make me think of the killer robots from Rainbow Brite and the Star Stealer.
  9. stevefaulkner
    08:39
    @StommePoes who zakim? It's cool and has history what morze u want
  10. StommePoes
    08:49
    Zakim is already doing the cross-gitter stuff
  11. StommePoes
    08:49
    he's in front of all the names of people coming in via slack, it would be confusing if there's then Zakim someone's name and then Zakim does stuff
  12. StommePoes
    08:50
    that, or when the bot alone does something (other than porting through someone's name/message) make it like The Real Zakim
  13. stevefaulkner
    09:26
    @StommePoes are you talking about the archive automation? No bot involved here crib job grabs harmless from the gitter archives
  14. StommePoes
    09:27
    oh
  15. StommePoes
    09:30
    hm
  16. StommePoes
    09:30
    still wondering what the best thing to do is when you have a row of tabs, mostly a tab-panel, but one of the tabs, the default one, shows all panels
  17. StommePoes
    09:31
    I don't think the container of the panels can have all the aria-labelledby associations etc esp since it can't have the panel role itself anyway
  18. StommePoes
    09:31
    the anchor that performs the action is inside the tabs, not really possible in this case to take it out and style to fake
  19. MichielBijl
    09:39
    A tabpanel that contains all tabpanels?
  20. MichielBijl @MichielBijl does not compute
  21. StommePoes
    09:43
    no, a tab-panel where all the panels are wrapped in a container, and the tabs above it include one tab which shows all the panels at once.
  22. StommePoes
    09:44
    I have (but can't actually do) the first tab being all aria-selected and controls=name-of-panel-wrapper, but obviously that isn't right
  23. StommePoes
    09:44
    I was given this setup where it was just all JS and I tried tab-paneling it long ago, but never knew a good solution for the first tab
  24. StommePoes
    09:45
    so it'll work and do its thing if you click it, but it won't name things correctly and stuff
  25. MichielBijl
    09:45
    Hmm, don't think we have a pattern for that
  26. StommePoes
    09:46
    it also still works with TAB and not arrow, and doesn't do any auto-focussing (there's nothign to focus on, it's just showing/hiding text), but those two I can fix at any time I have the time
  27. MichielBijl
    09:47
    No, there is no pattern for it, only tabpanel:
    @import "modules/subheading";
  28. MichielBijl
    09:47
    Oeps
  29. MichielBijl
    09:47
    https://rawgit.com/w3c/aria/master/practices/aria-practices.html#tabpanel
  30. MichielBijl
    09:47
    Is this something that you encounter more often?
  31. MichielBijl
    09:48
    Might be possible to add something like a "show all" tab
  32. MichielBijl
    09:48
    Seems weird to me though
  33. MichielBijl
    09:53
    What is the purpose of tabs if its default behaviour is to show everything?
  34. StommePoes
    09:55
    Who knows
  35. StommePoes
    09:55
    it's client-defined behaviour
  36. StommePoes
    09:55
    I'll show you an old one
  37. StommePoes
    09:56
    before I fixed the roles (I had them on the li, not the a)
  38. StommePoes
    09:59
    http://www.muldi.nl/p/muldi-hangmappenkast-2-laden-zwart-textured/26853/ under the product, except most shops have up to 6-7 tabs
  39. StommePoes
    10:00
    in other words, it seems someone wanted to show all by default, but let people "focus" in on the particular. But that's only a guess of mine-- i suspect it was asked for by a single client back when my company only had 3, and it's become default since then
  40. MichielBijl
    10:08
    That doesn't make sense :/
  41. stevefaulkner
    10:33
    interesting w3c history/post by doug schepers - opening standards http://schepers.cc/opening-standards
  42. StommePoes
    10:40
    @Michiel and for one shop as I was building it, I tried the arrow navigation but they actually complained
  43. StommePoes
    10:40
    they must've been like the only client of ours with a keyboard user
  44. StommePoes
    13:43
    @LjWatson sent your article re SR-defaults for disabled,read-only inputs to developers of a GtK application, and saw them reading it (turns out they had manually removed focus from read-only fields to speed up work)
  45. stevefaulkner
    14:05
    FYI
    ARIA role semantics cancel native #HTML5 element role semantics
    HTML5 attribute semantics cancel aria-* state & property attribute semantics
  46. garcialo
    14:22
    @stevefaulkner Why the HTML5 > aria-* attribute semantics?
  47. zakim-robot
    14:24
    [jitendra, a11y] Nice article https://www.activestandards.com/research-resources/accessibility-now-and-future
  48. stevefaulkner
    14:25
    @garcialo http://www.w3.org/TR/wai-aria-1.1/#host_general_conflict
  49. garcialo
    14:29
    Ah, that makes sense.
  50. zakim-robot
    14:49
    [jitendra, a11y] Is HTML and CSS both combined in the word “Markup” or Markup = HTML only?
  51. stevefaulkner
    14:51
    @garcialo demo http://s.codepen.io/stevef/debug/LVvNxW?
  52. garcialo
    14:54
    @stevefaulkner I'm not getting the disabled/readonly.
  53. stevefaulkner
    14:54
    where?
  54. stevefaulkner
    14:55
    on Mac?
  55. stevefaulkner
    14:55
    if so thats a webkit bug (bugs)
  56. garcialo
    14:56
    Oh wait, I can getting the disabled
  57. garcialo
    14:57
    What should be read for "readonly?"
  58. stevefaulkner
    15:19
    on mac? cant remember
  59. stevefaulkner
    15:19
    @garcialo btw have changed demo to include a textbox
  60. garcialo
    15:20
    Win10 + Firefox + NVDA
  61. garcialo
    15:20
    I can also do Win 8.1/7 or Mac OS if needed =p
  62. garcialo
    15:22
    Yeah, he textbox works better =)
  63. garcialo
    15:22
    What does a readonly checkbox even mean?
  64. stevefaulkner
    15:25
    @garcialo right, readonly doesn't work on checkbox :smile:
  65. garcialo
    15:26
    This is quite nice. https://18f.github.io/accessibility/
  66. MichielBijl
    15:28
    @jitendravyas Markup = HTML
  67. MichielBijl
    15:28
    HyperText Markup Language
  68. MichielBijl
    15:28
    CSS isn't markup
  69. MichielBijl
    15:29
    You might call it markup for style information…
  70. MichielBijl
    15:29
    But that's a different discussion me thinks
  71. garcialo
    15:29
    I don't think it's technically a "markup" language.
  72. garcialo @garcialo googles
  73. MichielBijl
    15:30
    Technically, no
  74. MichielBijl
    15:31
    @stevefaulkner readonly isn't announced in VoiceOver
  75. MichielBijl
    15:31
    disabled is
  76. garcialo
    15:31
    https://en.wikipedia.org/wiki/Markup_language
  77. MichielBijl
    15:31
    “A markup language is a system for annotating a document in a way that is syntactically distinguishable from the text.”
  78. MichielBijl
    15:32
    No text in css besides comments, but you wouldn't call .some-class { … } markup
  79. MichielBijl
    15:36
    @stevefaulkner, wait, it's announced as “textbox1, text”; not “textbox1, edit text”.
  80. MichielBijl
    15:36
    So all is fine
  81. MichielBijl
    15:40
    @garcialo that 18f site looks promising
  82. zakim-robot
    17:35
    [Marcy Sutton, a11y] @rodneyrehm there are definitely multiple a11y/ally/a11y.js modules, Addy Osmani’s a11y is already pretty well known
  83. powrsurg
    20:10
    So I just saw that someone posted http://w3c.github.io/aria-in-html/#checklist earlier. This made me learn more about High Contrast mode
  84. powrsurg
    20:11
    it seems to me that to better make sure things work in high contrast mode you want to make sure you specify both a background and a color in CSS (and not just one). Is this the general recommendation?
  85. powrsurg
    20:14
    no wait, it was @stevefaulkner on twitter
  86. powrsurg
    21:02
    And should high contrast mode remove all background colors/images and override them?
  87. zakim-robot
    21:09
    [Marcy Sutton, a11y] When I added high contrast support in Angular Material, most of the work was simply adding important background-colors back in where they dropped out, and borders to visually separate things. Here’s one component where I added a media query (mdChips), which simply needed borders to be turned on and off: https://github.com/angular/material/blob/master/src/components/chips/chips.scss#L168
  88. zakim-robot
    21:10
    [Marcy Sutton, a11y] But there were little to no graphic background images, so it was limited to simple CSS changes.
  89. powrsurg
    21:16
    It looks to me like Firefox wasn't actually doing high contrast stuff as it only changed my submit buttons and radio buttons. But I just restarted Firefox and it had no backgrounds anywhere
  90. zakim-robot
    21:17
    [Marcy Sutton, a11y] oh yes, admittedly I was only thinking of Windows…definitely more to consider.
  91. powrsurg
    21:23
    I never considered that there would be a media query for high contrast mode. Wow
  92. stevefaulkner
    21:24
    @powrsurg this may be helpful https://youtu.be/a-GnfF55sdA