Archive index

A11y Slackers Gitter Channel Archive 2nd of November 2016

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    @zakim-robot
    Nov 02 07:15
    [jv] http://www.mediaaccess.org.au/digitalaccessibilityservices/the-accessibility-of-content-management-systems/
  2. Peter Krautzberger
    @pkra
    Nov 02 11:23
    hey, slackers.
  3. Quick question on nested figures.
    I remembered http://www.html5accessibility.com/tests/figures-nested.html
    Is role=group necessary on the inner figures here?
  4. or at all.
  5. Peter Krautzberger
    @pkra
    Nov 02 11:57
    It sure is quiet here today ;(
  6. zakim-robot
    @zakim-robot
    Nov 02 12:09
    [michiel] pkra: last modified 05/12/2013
  7. [michiel] My guess is that it is outdated ;)
  8. zakim-robot
    @zakim-robot
    Nov 02 12:24
    [dean] I’ll dust off the cobwebs and ask a switch question :)
  9. [dean] Can you configure different modes per app - or is it device wide.... so basically, point mode is probably not ideal if you’re using the Stopwatch function, but you may prefer to use point mode for all other apps - can you do an override for this one app?
  10. Peter Krautzberger
    @pkra
    Nov 02 14:28
    @michiel :wink:
  11. zakim-robot
    @zakim-robot
    Nov 02 17:01
    [simong] This is a bit out there, but is there any value in adding a Title when creating office document (in the document’s properties). And would that be more important than having a proper title in the document itself (marked up with the title style) ?
  12. [simong] I’m trying to apply WCAG 2.4.2 to office documents and PDFs. For PDFs there’s something such as a /Title entry which can be read by screenreaders and displayed in the window bar, but that doesn’t seem to apply to documents really
  13. [simong] https://www.w3.org/TR/WCAG20-TECHS/PDF18.html specifically
  14. zakim-robot
    @zakim-robot
    Nov 02 17:21
    [felix.berger] Hey there - I am just wondering if the wcag 1 checkpoint of increasing consistency https://www.w3.org/TR/WCAG10-CSS-TECHS/#consistency ... and advising to avoid inline css does still apply officially to wcag 2.0? Tipps are much appreciated :)
  15. [karlgroves] No.
  16. zakim-robot
    @zakim-robot
    Nov 02 17:37
    [karlgroves] Right. There were some things in WCAG 1 that really had little direct impact on a11y anymore that were not included in 2.0 (another example would be the explicit requirement that your markup be valid)
  17. zakim-robot
    @zakim-robot
    Nov 02 17:45
    [felix.berger] That's interesting since UAAG-2 still includes this topic, but maybe that is another perspective.
  18. powrsurg
    @powrsurg
    Nov 02 19:06
    In general, valid code does make things easier
  19. zakim-robot
    @zakim-robot
    Nov 02 19:23
    [michiel] <div role="button" onclick="return false;">Click me for free tacos</div> is valid code ;)
  20. [michiel] Not particularly useful though.
  21. Mallory
    @StommePoes
    Nov 02 19:29
    @dean I'd be interested to know as well. Have you asked @iAmMaccing?
  22. I thought you could set a mode per app, though.
  23. @dean hey I do have his book on switch control on iOS, it's probably not exactly the same as on OSX but you do want to read it? (digital copy)
  24. @powrsurg "I'm surprised they know about specific SR software" <-- most people in education who've heard anything about "screen readers" have heard of JAWS because of its wide use in education and employment
  25. Michael Fairchild
    @mfairchild365
    Nov 02 19:34
    I'm curious what folks think about this https://gist.github.com/mfairchild365/2dbffbcac3286413c0509ea8e8e7ad7c
  26. Mallory
    @StommePoes
    Nov 02 19:37
    Hm
  27. What if speech rec understood the aria-hidden=true?
  28. It's the dichotomy of "disabled people are using a screen reader and can't see the screen" and "everyone else is totally abled and not running AT and can see the screen"
  29. I think that gets a lot of devs in trouble
  30. Michael Fairchild
    @mfairchild365
    Nov 02 19:39
    Very good point
  31. Mallory
    @StommePoes
    Nov 02 19:39
    can a video with a role of image use the controls? since those controls don't belong on things called "img" (i had no idea)
  32. It can't have alt either
  33. powrsurg
    @powrsurg
    Nov 02 19:39
    @StommePoes Admittedly the last one was a decade ago, but no one in either of the two college websites I had worked on seemed to know much about particular screen readers. And the people we've been dealing with lately seem to ask if something does I feel because they need to check off some requirement. I think we have more people that need keyboard controls for things than SR. But I could be wrong.
  34. Mallory
    @StommePoes
    Nov 02 19:39
    so it's an image that, perhaps an SR confronted with that would thing "here's an image, no alt, I'll go look for a source to read that out loud" like they do
  35. putting text in a video tag is only rendered by the browser if it has no idea what a video tag is (which I rather dislike, if I have a silent video I would like to be able to add alt text)
  36. It's because of the lack of alt attrs on video elements that I don't use them but convert the videos to gifs instead. Stupid.
  37. Michael Fairchild
    @mfairchild365
    Nov 02 19:41
    According to the spec, the aria-label would become the alt text, right? https://www.w3.org/TR/wai-aria/roles#img
  38. Mallory
    @StommePoes
    Nov 02 19:41
    @powrsurg it might just be the kinds of orgs I'm dealing with
  39. Oh michael I couldn't see that because it's offscreen
  40. zakim-robot
    @zakim-robot
    Nov 02 19:41
    [michiel] I would be more inclined to add SR text after the video that has the alt text if you will.
  41. Mallory
    @StommePoes
    Nov 02 19:41
    yeah it might
  42. I'm not certain if it can have aria-label?
  43. powrsurg
    @powrsurg
    Nov 02 19:41
    I wish I was dealing with more cluefull orgs.
  44. Mallory
    @StommePoes
    Nov 02 19:41
    Like, putting aria-label on a div won't get it read out
  45. oh but you have the img role
  46. so it might
  47. I would leave the video's "stop" button available to AT in either case; if it plays longer than 5 seconds it's required by WCAG anyway.
  48. zakim-robot
    @zakim-robot
    Nov 02 19:42
    [michiel] ^ +1
  49. Michael Fairchild
    @mfairchild365
    Nov 02 19:43
    So if I understand you correctly, @StommePoes you would remove the aria-hidden from the button so that it is always available?
  50. Mallory
    @StommePoes
    Nov 02 19:43
    I think the button doesn't have to stop the video either, if it's silent: these auto-playing videos are usually behind text. I dunno what sorts of monsters who hate people think this is a good thing, but instead of stopping/pausing the video in those instances, what I would love more is simply an opaque div show up behind the text instead. Now I don't see the video AND I can actually read that text.
  51. I would, yes. But then keeping the img role and stuff would make it a button that the case of the totally-blind screen reader user would find puzzling indeed.
  52. But at least having the button available allows anyone else using an SR or Dragon or whatever able to interact with it.
  53. Michael Fairchild
    @mfairchild365
    Nov 02 19:44
    Yeah, I honestly think background videos are terrible and shouldn't be used. But I'm never going to convince everyone to NOT use them, so I have to provide some recommendations on how to do it best.
  54. zakim-robot
    @zakim-robot
    Nov 02 19:44
    [michiel] mfairchild365: yes, if a sighted SR user wants to interact with it for example.
  55. Mallory
    @StommePoes
    Nov 02 19:45
    People with dyslexia for example may be using either a screen reader or Dragon (or both) and would want to have access to a button they can see.
  56. zakim-robot
    @zakim-robot
    Nov 02 19:45
    [michiel] or a braille user…
  57. Mallory
    @StommePoes
    Nov 02 19:45
    Michael: same with animations, I'm no fan of them, so I let the designers do whatever they want, but then include some way of people turning them off
  58. zakim-robot
    @zakim-robot
    Nov 02 19:45
    [michiel] or… who knows.
  59. Mallory
    @StommePoes
    Nov 02 19:45
    And we may tie this into user profiles since a lot of our stuff is behind logins
  60. This way, people who like videos or animations or whatever, they get what the designers want. But people who can't stand that shit, they have a method of turning it off. In our case, it won't be near the animations but I think in a user setting. If you're using the localstorage thing for example
  61. if not, some control near the top of the page is also good. Esp if the videos did make noise.
  62. Like CNN, zomg that horrible site, it autoplays videos though only if you scroll down to a story. Then as you scroll further it moves to the side! Very hard to find how to turn it off with keyboard
  63. Michael Fairchild
    @mfairchild365
    Nov 02 19:49
    What I was worried about was this scenario: a background video exists with no audio. A screen reader user encounters the video but it doesn't have a text track associated with it (because it has no audio). Now the screen reader user doesn't know if they are missing something important or not. So marking it up as an image should fix that?
  64. Mallory
    @StommePoes
    Nov 02 19:49
    Try another role and the aria label maybe?
  65. Michael Fairchild
    @mfairchild365
    Nov 02 19:49
    Maybe i'm assuming too much and overthinking it to begin with
  66. Mallory
    @StommePoes
    Nov 02 19:49
    Do videos already have some role I'm unaware of?
  67. I do like the idea of having some text for a totally-blind screen reader user who realises there's a video.
  68. @michielBijl what's the native role for video?
  69. looks like it has an application role
  70. https://www.w3.org/TR/html5/embedded-content-0.html#the-video-element
  71. so possibly aria-label alone would work, you'd have to test it
  72. oh wait, it means it can have an application role, not that it does
  73. Michael Fairchild
    @mfairchild365
    Nov 02 19:54
    I still like giving it the role of an image, because really it is just an animated image at this point (no audio or anything else)
  74. Mallory
    @StommePoes
    Nov 02 19:57
    as in, it could have been a gif background really
  75. and it's got no controls attr
  76. yeah, I think you're right. Would test, but yes.
  77. Michael Fairchild
    @mfairchild365
    Nov 02 19:58
    Exactly, but video is so much smaller in terms of file size, so many people prefer them over the gif format.
  78. Mallory
    @StommePoes
    Nov 02 19:58
    yeah
  79. agreed
  80. but I meant in terms of semantics, you could have used the other for the same content effect.
  81. so making the video element in this case more like an animated gif can make sense.
  82. actually, I have some slides I'm making and I had text after the video as alt text, which felt dumb and took up room, so I could test if aria-label reads out on my SRs
  83. Michael Fairchild
    @mfairchild365
    Nov 02 19:59
    btw @StommePoes thank you for discussing this. :) Whatever you get payed is not enough.
  84. Mallory
    @StommePoes
    Nov 02 20:00
    I'm getting paid €0 to yack on gitter, lawlz
  85. Michael Fairchild
    @mfairchild365
    Nov 02 20:00
    :P
  86. I'm also thinking in terms of automated accessibility testing. How do I determine if a video actually NEEDS a caption track or not? If its role is not video, but an image, it probably doesn't need a caption track anymore.
  87. Mallory
    @StommePoes
    Nov 02 20:01
    Hm
  88. Most of the "videos" I watch via Twitter, I dunno if they even have sound, they DO have people talking and so have and/or need captions
  89. but a machine wouldn't know that
  90. Michael Fairchild
    @mfairchild365
    Nov 02 20:03
    A machine might be able to detect the existence of an audio track.
  91. http://stackoverflow.com/questions/21270048/html5-video-how-to-detect-when-there-is-no-audio-track
  92. zakim-robot
    @zakim-robot
    Nov 02 20:07
    [michiel] StommePoes: see you found it.
  93. Mallory
    @StommePoes
    Nov 02 20:47
    Michael: yeah but then at best the automated tester could give a warning. If there are lots of videos they'll turn into warnings devs learn to ignore (potentially). But, either there are automated tools that can do that today OR it's on their lists of features to add, because it's something a machine can do.
  94. zakim-robot
    @zakim-robot
    Nov 02 21:04
    [michiel] Sure Mellary.
  95. [michiel] :smiling_imp:
  96. zakim-robot
    @zakim-robot
    Nov 02 21:13
    [ted_drake] UK Government’s survey on assistive technology. https://accessibility.blog.gov.uk/2016/11/01/results-of-the-2016-gov-uk-assistive-technology-survey/ Sorry if this has already been posted.