Archive index

A11y Slackers Gitter Channel Archive 4th of December 2015

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    Dec 04 00:37
    [marcysutton] @devonpersing: did you ever find a workaround for aria-live not working in IE11 and JAWS 16? https://web-a11y.slack.com/?redir=%2Farchives%2Fgeneral%2Fp1437679538001257
  2. zakim-robot
    02:15
    [jitendra] I’m looking to change the job and I’m getting job offers but most of them are Sass companies. And the problem in Sass companies that SEO and Accessibility is not a concern. I think only website where SEO is important will care about A11y, these kind of companies are very few.
  3. zakim-robot
    02:44
    [billnordwall] @devonpersing I’d love to hear if you heard a workaround as well! @marcysutton - Devon discovered that bug while she and I were working together on the same project. Without divulging too much about existing client work, I do know for a fact that it has been submitted to Freedom Scientific. Unfortunately, the issue still persists in JAWS 17 and we haven’t been able to find a workaround.
  4. zakim-robot
    02:45
    [billnordwall] @marcysutton: if we do find a workaround or we hear back from Freedom, I will definitely pass it along here as well! Although I’m hoping @devonpersing has had better luck on workarounds than I have :world_map:
  5. garcialo
    03:07
    Open Question. What are your thoughts on when Level AAA should be used?
  6. zakim-robot
    03:13
    [marcysutton] @billnordwall @devonpersing @jelbourn found that JAWS 17 was better, but I haven't tested it myself
  7. zakim-robot
    03:44
    [fstorr] @garcialo: Level AAA’s rules on line height and length more-or-less tie into core design typography best practices. You can cross reference them with Robert Bringhurst’s classic “Elements Of Typographic Style”. That said, if I remember correctly, the AAA requirement for line length is 80–100 characters, but Bringhurst recommends 45–75 (http://webtypography.net/2.1.2). A line length of 75-80 characters is very readable. Likewise, AAA’s leading/line-height requirement is essentially 1.5em, which is usually the lazy designer’s default (I’ve seen that so many times over the years). Bringhurst recommends choosing "a basic leading that suits the typeface, text and measure” (http://webtypography.net/2.2.1) which for body copy is usually very close to that size. I will create a typographic scale for a project and then pick a leading that matches, or is very close to, the AAA requirement.
  8. zakim-robot
    04:39
    [devonpersing] yeah, what @billnordwall said, @marcysutton and @jelbourn!
  9. zakim-robot
    06:04
    [fstorr] JAWS and aria-live is a world of pain. I can’t offer more than sympathies unfortunately.
  10. StommePoes
    13:04
    @fstorr I have that typography book and he somehow manages to justify every page with no visible rivers. Despite that, I won't justify web text (I can't even see how he does it, does each identical word have a different width per line???)
  11. MichielBijl
    13:53
    text-align: justify; hyphens: auto;?
  12. MichielBijl
    13:53
    Relies on the browsers dictionary though.
  13. MichielBijl
    13:53
    So won't work in all languages.
  14. MichielBijl
    13:54
    @StommePoes
  15. fstorr
    13:57
    @StommePoes wow. I hadn't noticed that. I never justify text on the web because of rivers.
  16. MichielBijl
    14:19
    Can anyone give feedback on the clarity of the last table on this page: https://cdn.rawgit.com/MichielBijl/font-awesome/feature/results-table/tests/results.html /cc @StommePoes
  17. zakim-robot
    14:34
    [karlgroves] Hey slackers, I just put a post up in @jobs. Go if interested
  18. StommePoes
    16:02
    I like the last one except I'd be more concise with the first col: VoiceOver, Safari 9, OSX
  19. StommePoes
    16:02
    instead of with, on
  20. MichielBijl
    16:03
    Good point
  21. MichielBijl
    16:04
    And no, I'm totally not here 24/7
  22. garcialo
    16:04
    Neither am I.
  23. MichielBijl
    16:04
    Haha
  24. MichielBijl
    16:04
    Anyway, I figured a couple of extra chars didn't really matter
  25. MichielBijl
    16:06
    That said, it does make the result clearer
  26. MichielBijl
    16:06
    Pushed…
  27. MichielBijl
    16:07
    I think I'll update all the tables to this format
  28. MichielBijl
    16:09
    That rawgit thing won't update
  29. MichielBijl
    16:09
    Anyway, it's there
  30. StommePoes
    16:30
    It's frustrating watching someone try to improve a11y of something (so he can even use the software) and the developers seem to have communication problems wekan/wekan#408
  31. zakim-robot
    16:48
    [johnbhartley] regarding the <a> vs. <button> discussion in that thread, my understanding is that <a> should generally be used for events that will take the user to a different page while <button> is used for events that will either complete a form or manipulate the current page. Would be curious to hear what others thoughts are
  32. garcialo
    16:49
    That is pretty much best practice.
  33. garcialo
    16:54
    Well, beyond best practice, that is how the elements are intended to be used.
  34. garcialo
    16:55
    The best practice just comes in because it's best to use things the way they're intended to be used.
  35. MichielBijl
    16:55
    😱
  36. MichielBijl
    16:56
    Just noticed we use i instead of span's in the research thing.
  37. MichielBijl
    16:56
    Will correct.
  38. garcialo
    16:56
    =)
  39. deborahgu
    16:56
    and best practice is to do page navigation using legit links and buttons as above, and not put empty <a>s with JS that does the link magic
  40. MichielBijl
    16:56
    Just thought I'd mention it because I stated different before.
  41. MichielBijl
    16:57
    @StommePoes writing reply to that thread now.
  42. deborahgu
    16:57
    I read that discussion and I see them saying "we will probably forget to add some of the hrefs" and I continue to come back to "you know, if you actually use HTML to do the things HTML is good at, then you don't actually have to do anything to add accessibility, because it usually comes for free."
  43. zakim-robot
    16:57
    [johnbhartley] nice, thanks @garcialo and @deborahgu
  44. garcialo
    16:58
    np
  45. deborahgu
    16:59
    :)
  46. StommePoes
    17:01
    The more people who add in comments like that, the more likely the devs will decide to change their minds
  47. StommePoes
    17:01
    You all saw how long it took the bootstap guys to decide that yeah, focus styles and stuff oughtta work with keyboard etc was important
  48. MichielBijl
    17:01
    https://github.com/wekan/wekan/pull/408#issuecomment-162022112
  49. garcialo
    17:02
    The thing with the whole link/button/div/span thing is that people don't complain about the things I want them to complain about.
  50. MichielBijl
    17:02
    @deborahgu the problem is that a lot of people who write HTML don't know HTML.
  51. MichielBijl
    17:02
    But that is a problem with every language I guess.
  52. MichielBijl
    17:03
    But unlike the languages we speak, it has a bigger impact on stuff that must be parsed by machines.
  53. deborahgu
    17:03
    yeah; if you are JavaScript developer than that the language you know. I recently pointed out all the places we were doing fancy JavaScript sliders and showed how they could be simplified with (plus free accessibility built-in!) with input type=range.
  54. deborahgu
    17:04
    (I mean it doesn't help that a lot of the stuff having broad user agent support is incredibly new)
  55. deborahgu
    17:04
    but links should be easy.
  56. MichielBijl
    17:04
    If I tell Siri to “Ask Mallory where we will meet for DomCode?” Siri turns it into a text saying something like “where we will meet for DomCode?” Makes no sense.
  57. deborahgu
    17:05
    I have good success with siri but I think it's the result of 15 years of constant use of Dragon. I know how to talk to computers. ;) Although then I get cranky because the Siri UI is trying to be more human so it's less deterministic than it needs to be. ie. "What's the weather today" gives me inconsistent sets of results in terms of which bits of information it's giving me.
  58. MichielBijl
    17:07
    That is you learning how to talk machine rather than the machine learning how to talk human :)
  59. MichielBijl
    17:08
    @johnbhartley that is how I explain the difference too.
  60. deborahgu
    17:09
    :)
  61. MichielBijl
    17:10
    My employer can be happy today; did a full one hour worth of client work.
  62. garcialo
    17:10
    +1
  63. garcialo
    17:10
    ...literally
  64. MichielBijl
    17:11
    The rest was filing a11y issues and researching icon fonts :')
  65. garcialo
    17:11
    you do project stuff and get paid for it? jealous!
  66. MichielBijl
    17:12
    Nah, I'm on contract thing.
  67. MichielBijl
    17:12
    So I get paid no matter what :P
  68. MichielBijl
    17:12
    There wasn't much to do for the project I'm on. So I just fill in the gaps with what interests me.
  69. MichielBijl
    17:16
    “Well 3f7df34 is also supposed to automatically add aria-label on both <a> links and <i> icons...”
  70. MichielBijl
    17:16
    Erm
  71. MichielBijl
    17:16
    Accessibility is not an automated process :x
  72. MichielBijl
    17:17
    Had a look at the code, they grab the title attribute and pleur that into an aria-label
  73. deborahgu @deborahgu cries
  74. deborahgu
    17:32
    the idea that we can automate accessibility is why we have alt text all over wordpress: NIKON03421.png
  75. dylanb
    18:02
    a11ys please pay attention to this issue and make sure appropriate attention gets paid by your organization's reps at the W3C w3c/webcomponents#354, especially @stevefaulkner and @LjWatson
  76. MichielBijl
    18:32
    @StommePoes is there a difference between Orco on 64-bit or 32-bit?
  77. MichielBijl
    18:32
    Trying to make table headers more concise.
  78. StommePoes
    18:49
    not that I know of, but who knows. So I put it in there. Like, I can't run Vivaldi on a 32b machine
  79. StommePoes
    18:49
    er, not vivaldi, the new opera
  80. StommePoes
    18:49
    The version they released for linux was 64-bit only and they were all like "we're not doing the work for old machines" never mind old machines + linux are like love and marriage
  81. StommePoes
    18:51
    it's not that orca itself would be different but the atk stuff might be
  82. StommePoes
    18:51
    btw it's orca 3.18
  83. MichielBijl
    18:56
    Yeah that would have been the next question; version numbers for AT.
  84. MichielBijl
    18:57
    VO is tied to the OS, so that seemed pointless.
  85. StommePoes
    19:06
    Orca tho can be outdated, and 3.18's not the latest-latest
  86. deborahgu
    19:22
    Does anyone know who's driving http://a11yproject.com/ ? I don't recognize any names on a 5-second glance and the about doesn't say.
  87. zakim-robot
    19:23
    [cordelia] I’m guessing the 3 people in the “People” list of the github group: https://github.com/a11yproject
  88. deborahgu
    19:24
    ah, I assumed that would have all the contributors. Thanks.
  89. deborahgu
    19:30
    I was mostly surprised to see a site that had -- as far as I can tell on what is, again, pretty much a one minute looksee -- relatively accurate and contemporary crowd-sourced information where I don't recognize any of the names, and it has never been mentioned on webaim list, and doesn't seem to have a corporate sponsor. Obviously the "don't recognize any of the names" is very likely to be a failing in me. ;) But I am surprised I haven't seen it mentioned on list anywhere.
  90. zakim-robot
    19:51
    [gregtarnoff] I tend to mention it, but then I contributed. I’ve seen it pop up a few places. I think for the most part Dennis ran most of the merges, but I thought I recall Dave Rupert doing some as well.
  91. deborahgu
    20:02
    gregtarnoff, thanks. v. cool. :D
  92. deborahgu
    20:32
    I did noticed it hadn't been mentioned in the webaim archives; might be a good place to talk about it.
  93. zakim-robot
    20:48
    [jdan] any good way to prevent from “sr-only” text from being copied to the clipboard with its surroundings?
  94. zakim-robot
    20:49
    [jdan] we’ve got a bug in our exercises because we use a combination of aria-hidden and some special sr-only text. I’m thinking I may need to substitute it for a simple aria-label
  95. MichielBijl
    21:04
    user-select: none?
  96. zakim-robot
    21:17
    [jdan] @MichielBijl looks like that’ll still be copied as long as you select text around either end
  97. MichielBijl
    21:17
    Hmm.
  98. MichielBijl
    21:18
    user-copy: no-you-shall-not :P
  99. MichielBijl
    21:18
    What is the context of the sr-only text?
  100. zakim-robot
    21:28
    [jdan] our LaTeX rendering library, KaTeX http://khan.github.io/KaTeX/, produced HTML and some MathML. Additionally, on KhanAcademy.org we place an additional span of text that’s “human-readable” and has the “sr-only” class
  101. zakim-robot
    21:29
    [jdan] so if I copy the produced “400” with some surrounding text, I’ll see 400400400 when I paste. One from MathML land, one from the HTML, and one from the sr-only span.
  102. MichielBijl
    21:30
    Ah ok
  103. zakim-robot
    21:32
    [jdan] so, even if I solve the sr-only issue (and use an aria-label/aria-describedby) I’ll still be facing this MathML part. bummer!
  104. MichielBijl
    21:32
    Making it a label seems fine to me. aria-label is not supported everywhere when it comes to caret navigation. But I don't have all the data on that.
  105. zakim-robot
    22:23
    [stefsull] I have a silly question that I can’t quite find an answer for...
  106. zakim-robot
    22:24
    [stefsull] If I’m using a <fieldset> and <legend> to define a set of checkboxes, or whatever… is it “illegal” to wrap the legend in a span for styling purposes?
  107. garcialo
    22:25
    Shouldn't be.
  108. zakim-robot
    22:25
    [stefsull] The spec does say that it is the first child of a fieldset element… so my guess is that it’s not kosher. (Just discussing with a co-worker)
  109. garcialo
    22:25
    Well, what if you do <legend><span>?
  110. zakim-robot
    22:26
    [stefsull] <span><legend>Content is what he has
  111. zakim-robot
    22:26
    [stefsull] He says he can’t style it properly the other way
  112. garcialo
    22:26
    ah, does it break the announcing done by a screen reader?
  113. zakim-robot
    22:27
    [stefsull] Not on the Mac SR he was testing on… but we’ve not done exhaustive testing...
  114. garcialo
    22:27
    I'll try it out real quick
  115. garcialo
    22:27
    one sec
  116. zakim-robot
    22:27
    [stefsull] grazie
  117. garcialo
    22:33
    Laptop blue screened for the first time ever. It must really not like legends inside of spans
  118. zakim-robot
    22:34
    [cordelia] 0__0
  119. zakim-robot
    22:36
    [stefsull] hahaha… oh noes!
  120. garcialo
    22:38
    Okay, so NVDA with Firefox seems to lose the legend when it's inside a span. Like its treated as if it's just a paragraph inside the field set and then NVDA doesn't associate it with the grouping
  121. zakim-robot
    22:40
    [stefsull] ahhh…. thank you SO much
  122. garcialo
    22:40
    Yeah, i'm actually quite saddened it didn't work. =(
  123. zakim-robot
    22:41
    [stefsull] As Brandon will be :wink:
  124. zakim-robot
    22:41
    [stefsull] Very kind of you @garcialo
  125. garcialo
    22:41
    np
  126. zakim-robot
    22:48
    [gregtarnoff] Rack legend up there with selects, type=radio, type=checkbox as things that are pretty much unstylable.
  127. garcialo
    22:50
    To be fair, legends have some pretty nice style as it is.
  128. zakim-robot
    22:54
    [stefsull] Well, they’re replaced elements… which are always a pain
  129. zakim-robot
    22:55
    [stefsull] but we needed to place them horizontally in this specific styling need… and they were being ugly about it.
  130. zakim-robot
    22:55
    [stefsull] I found an article I wrote about them as early as 2008 or something. hahah (When content inside won’t wrap)
  131. zakim-robot
    23:13
    [jessebeach] Is anyone in this slack actually working for Slack? We need to talk about the accessibility of this product :simple_smile:
  132. garcialo
    23:14
    There is someone that comes in here sometimes from Slack, but I don't think they specialize in a11y. That said, I believe they have an open position specifically for a11y.
  133. zakim-robot
    23:22
    [jessebeach] Thanks @garcialo !!
  134. garcialo
    23:23
    np =)
  135. zakim-robot
    23:32
    [marcysutton] They do…and we have the @slack-a11y-feedback channel!
  136. zakim-robot
    23:33
    [marcysutton] Slack also funded this community, which is seriously awesome :simple_smile:
  137. zakim-robot
    23:34
    [marcysutton] I’m sure specific feedback would be useful, since that position is still open as far as I know
  138. MichielBijl
    23:58
    Also filing bugs :)
  139. MichielBijl
    23:58
    And you can always switch to Gitter :P