Archive index

A11y Slackers Gitter Channel Archive 9th of March 2016

What fresh hell is THIS now? - Patrick Lauke
  1. jonathantneal
    Mar 09 00:24
    @stevefaulkner, agreed, but is <address> not a good choice any longer? It was once recommended (I thought) on HTML5Doctor. I suppose as long as they are interpreted
  2. jonathantneal
    Mar 09 00:26
    Gutter mobile is failing me, so I'm going to let that spellcheck stay. :P
  3. jonathantneal
    Mar 09 00:26
    But I did mean to say "gitter".
  4. jonathantneal
    Mar 09 00:27
    I also meant to say that as long as an AT reads <address> or <aside> or whatever correctly for a byline, then no worries.
  5. jnurthen
    Mar 09 00:42
    address normally maps to just text. See http://www.w3.org/TR/html51/semantics.html#the-address-element
  6. zakim-robot
    Mar 09 00:43
    [fstorr] <address> is the correct way to mark up that data. The rel=author could be swapped out for microdata (http://html5doctor.com/microdata/) or schema (https://developers.google.com/schemas/reference/types/CreativeWork) or similar.
  7. zakim-robot
    Mar 09 00:44
    [fstorr] Last time I checked, VoiceOver gives <address> a role of contentinfo, so I’ve started adding an aria-label to make make the contents more explicityl defined.
  8. jnurthen
    Mar 09 01:56
    Linked to the wrong url above. Meant to link to the html a11y mapping doc but
  9. zakim-robot
    Mar 09 05:58
    [jitendra] Does performance affect accessibility?
  10. stevefaulkner
    Mar 09 06:55
    Performance affects everyone
  11. garcialo
    Mar 09 06:56
    Especially hover boards.
  12. zakim-robot
    Mar 09 08:07
    [jitendra] @stevef: But does expertise in web performance count as a skill for accessibility related role?
  13. zakim-robot
    Mar 09 08:09
    [jitendra] I guess WCAG or Section 508 not concerned about performance of the web site or app
  14. Hello everybody
  15. good morning.
  16. @jnurthen Thanks for the link re chrome dev tools a11y extension.
  17. MichielBijl
    Mar 09 09:25
    Morning pkra and jkva =)
  18. @MichielBijl I'm finally working up the courage to file an issue on GitHub; sigh, n00b.
  19. MichielBijl @MichielBijl does a dinosaur dance to celebrate
  20. @MichielBijl w3c/aria#290 blargh.
  21. MichielBijl
    Mar 09 10:08
    :+1:
  22. and now I'll better go into hiding for a month or so.
  23. MichielBijl
    Mar 09 10:16
    No need for that, that is a good write up =)
  24. MichielBijl
    Mar 09 10:16
    Compare that to mine: w3c/aria#286
  25. @MichielBijl :)
  26. stevefaulkner
    Mar 09 13:02
    @pkra don't feel nervous about filing issues yours was good
  27. powrsurg
    Mar 09 14:20
    I just looked at what @pkra posted. I'm actually curious about the generic HTML listed above and why it closes and then re-opens <i> elements each time.
  28. zakim-robot
    Mar 09 15:37
    [jdan] is there a good technique for making HUGE lists of things more keyboard-friendly? for example:
  29. zakim-robot
    Mar 09 15:39
    [jdan] our menu is currently hierarchical, but we decided to unravel and put all the things in one nav. for what it's worth - we'll have roughly 1/3rd of the links in the end result (the menu is curated)
  30. @powrsurg math layout traditionally styles single-letter identifiers (variables etc) in italic (possibly specialized italic fonts)
  31. but operators are not, so you need to switch in and out. I would say it's questionable if <i> is appropriate (under HTML5 semantics).
  32. thanks @stevefaulkner, I appreciate it.
  33. powrsurg
    Mar 09 16:49
    right, I get that it should be styled as such, but it seems like overkill to style them as such. I can see if this was done automatically, but by hand I wouldn't do it like that myself ...
  34. powrsurg
    Mar 09 16:49
    for the HTML-style approach. For the <math> markup I agree with the way it's done
  35. @powrsurg right. I wouldn't expect anyone to do that either. For what I'm using most of my day, see the next-to last example at http://codepen.io/pkra/pen/zqrwQx (generated from the MathML example).
  36. That's the kind of markup that most math is turned into on the web.
  37. (not just by MathJax but that's how the example was created)
  38. powrsurg
    Mar 09 17:26
    how well does that work cross-browser? I thought MathML support was pretty crappy
  39. powrsurg
    Mar 09 17:26
    I was trying to make that a thing when I was younger and still in college and didn't get traction. But then I was an idiot college student that didn't know how to get in touch with people and wasn't doing web work professionally at the time so that goes with the territory ...
  40. powrsurg
    Mar 09 17:29
    Also, damn --moz-focus-inner
  41. zakim-robot
    Mar 09 19:20
    [cordelia] Hey folks, I’m trying to create an element with role=“status” on the fly, but NVDA+Firefox isn’t picking it up. It’s working fine JAWS+IE, VoiceOver+Safari. Am I missing something? Any suggestions for workarounds? Here’s a codepen sample: http://codepen.io/anon/pen/wGGEOM
  42. zakim-robot
    Mar 09 19:21
    [cordelia] (^ just fixed a typo in that, but still no luck)
  43. StommePoes
    Mar 09 19:55
    Status roles don't have the same issues as alerts? (ie that they work better when the container already exists on the page before changes)?
  44. zakim-robot
    Mar 09 20:15
    [cameron] NYU is launching an automated speech recognition captioning program, I think this is misguided
  45. zakim-robot
    Mar 09 20:15

    [cameron] > Closed captioning pilot now available with NYU Stream

    Now, you can request free automatic speech recognition (ASR) transcriptions or closed captions, which you can edit and share with the media you post to NYU Stream. ASR transcriptions make your media more accessible and user-friendly, and can be ordered in a variety of languages, with more coming soon. Read more about NYU Stream and ASR, which are available to all NYU community members.

  46. @powrsurg

    how well does that work cross-browser?
    all the way down to IE8

    I thought MathML support was pretty crappy

    Sure but HTML/CSS support is pretty ok ;-)

  47. zakim-robot
    Mar 09 20:16
    [cameron] automation is inaccurate enough to be semantically incorrect
  48. sorry for the edit @powrsurg
  49. jnurthen
    Mar 09 20:38
    @cameron but depending on the clarity of the speaker it can be a great start when doing manual captioning.
  50. zakim-robot
    Mar 09 21:04
    [cameron] @jnurthen yes, agreed. I hope the NYU IT dept will take that additional step.
  51. zakim-robot
    Mar 09 21:05
    [cameron] “get off my lawn!"
  52. zakim-robot
    Mar 09 21:05
    [cameron] :)
  53. zakim-robot
    Mar 09 21:07
    [cameron] this looks interesting http://edenspiekermann.github.io/a11y-toggle/
  54. zakim-robot
    Mar 09 21:08
    [cordelia] @StommePoes Unfortunately I can’t guarantee a container element on the page (refactoring some old stuff)
  55. @cordelia try creating empty elements first, then add content.
  56. @cordelia old but might still be worthwhile http://codepen.io/pkra/pen/BoVBpX
  57. zakim-robot
    Mar 09 22:21
    [cordelia] Thanks @pkra. Yeah, I also just tried having a button that first creates the empty element, then adds content, but it looks like NVDA doesn’t like that. I need to create the empty element, then press the button. sigh. Updated pen: http://codepen.io/anon/pen/wGGEOM
  58. @cordelia I might remember incorrectly, but a small timeout might help as well.
  59. zakim-robot
    Mar 09 22:28
    [cordelia] also just checked and the “Hello button!” status in that code sample you shared doesn’t read out for me on Firefox + NVDA
  60. jnurthen
    Mar 09 22:56
    If you are considering a hack to work round an NVDA issue - have you considered logging a bug against NVDA to see if they will fix it?
  61. zakim-robot
    Mar 09 23:19
    [cordelia] Yes! I’ve been combing through their open issues list on github to see if someone else has reported it. Is the best way to create a bug directly via github or contact them via email? https://github.com/nvaccess/nvda
  62. jnurthen
    Mar 09 23:25
    github issues i think
  63. jnurthen
    Mar 09 23:27
    at least that is where my old issues seem to be now (I seem to recall I used a bugzilla instance before but I either remember wrong or they migrated everything)
  64. zakim-robot
    Mar 09 23:48
    [cordelia] cool! filed a bug
  65. jnurthen
    Mar 09 23:49
    great :) Are you at the meetup tonight?