Archive index

A11y Slackers Gitter Channel Archive 8th of September 2015

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    15:50
    [Brian Kardell, a11y] anyone in here spent a significant amount of time thinking about CSS visual reordering?
  2. zakim-robot
    16:02
    [dylanb, a11y] @bkardell: what do you consider "significant"?
  3. zakim-robot
    16:02
    [dylanb, a11y] what is the topic?
  4. zakim-robot
    16:02
    [Brian Kardell, a11y] well, I think a lot of developers have for a long time thought of, even used the term "document order independence"
  5. zakim-robot
    16:03
    [dylanb, a11y] yeah, developers, pfffft!!!
  6. zakim-robot
    16:03
    [dylanb, a11y] :simple_smile:
  7. zakim-robot
    16:05
    [Brian Kardell, a11y] well, ok so -- the platonic ideal with HTML/CSS is that someone can author semantics about something without a lot of regard to "look" - when this was about text documents, there was a sense of kind of intrinsic order but that's about the extent
  8. zakim-robot
    16:05
    [Brian Kardell, a11y] so "ideally" the CSS could be used to present that kind of however
  9. zakim-robot
    16:07
    [Brian Kardell, a11y] so we have things like grid which can reorder and I know a whole lot of people who have been drooling over that for years and now I'm noticing that even things like abspos can create funky questions I'm not entirely sure a lot of people have been asking/thinking about
  10. zakim-robot
    16:08
    [dylanb, a11y] http://www.w3.org/TR/UNDERSTANDING-WCAG20/navigation-mechanisms-focus-order.html
  11. zakim-robot
    16:08
    [Brian Kardell, a11y] many of the use cases friends of mine have been desiring seem like they raise way more questions
  12. zakim-robot
    16:08
    [dylanb, a11y] that is the basics
  13. zakim-robot
    16:09
    [Brian Kardell, a11y] yeah I'm familliar :simple_smile:
  14. zakim-robot
    16:12
    [Brian Kardell, a11y] I guess I'm wondering if there's in the very least a good case to be made that some of us should be out in front of these features kind of reframing the idea from "document order independence" to... I dont know, something slightly less than what many are imaginging such that it's understood that the document order will still matter
  15. LjWatson
    16:13
    @bkardell not sure I follow?
  16. zakim-robot
    16:17
    [Brian Kardell, a11y] @ljwatson: sorry, I'm trying to think of how to articulate this better in a few words
  17. zakim-robot
    16:21
    [Brian Kardell, a11y] @ljwatson if, for example, you had something which was a menu in the right rail and after some time you discovered that that was less than great, it should be a submenu or something -- today this means rewriting HTML, potentially across many apps that are all a part of the same 'site'. Many devs are thinking "grid will save me because I can just write a rule and move it all around" - but this will actually cause a lot of weirdness because that doesn't affect focus order, for example
  18. zakim-robot
    16:23
    [dylanb, a11y] @bkardell: its quite simple...the DOM order should be the same as the order in which your UX designers expect the user's eye to read the content...anything else is dangerous
  19. zakim-robot
    16:23
    [Brian Kardell, a11y] @ljwatson: this actually already happens with abspos, but the limitations of what a dev can accomplish with that make it less prevalent. things like grid will make this really easy and I'm worried about creating a new norm
  20. zakim-robot
    16:24
    [Brian Kardell, a11y] @dylanb: mobile responsive/hamburger menus already currently create a situation frequently in which something that is visually before in one case is visually after in another
  21. zakim-robot
    16:25
    [Brian Kardell, a11y] @dylanb: so this seems incomplete
  22. zakim-robot
    16:26
    [Brian Kardell, a11y] @dylanb: also, what I'm saying is that one of the main values many devs see is "independence" - the ability to change their mind about what their expectations are over time without going back to the HTML
  23. zakim-robot
    16:26
    [dylanb, a11y] @bkardell: where are the UX designers in this process?
  24. zakim-robot
    16:27
    [dylanb, a11y] accessibility is, to a large degree about change...getting people to accept that part of their role is to do things they have not considered part of their role up until that point
  25. zakim-robot
    16:28
    [dylanb, a11y] UX designers need to be explicit about expected reading order in their comps
  26. LjWatson
    16:28
    Flexbox is an interesting case in point. So far it seems that most browsers change the DOM and the acc tree based on the CSS driven layout.
  27. zakim-robot
    16:28
    [Brian Kardell, a11y] thus my comment https://web-a11y.slack.com/?redir=%2Farchives%2Fgeneral%2Fp1441728764000027
  28. zakim-robot
    16:28
    [dylanb, a11y] developers need to accept that DOM order matters and think about it as much as they think about CSS positioning/tyling
  29. zakim-robot
    16:28
    [dylanb, a11y] *styling
  30. LjWatson
    16:28
    I think Chrome might not change the DOM, only the acc tree - but don't quote me on this (I haven't investigated in detail).
  31. zakim-robot
    16:31
    [Brian Kardell, a11y] dylanb you're preaching to the choir though - that's why I suggested it might be worth a concerted effort to get out in front of that as new powers and abilities are being placed in the hands of folks
  32. zakim-robot
    16:32
    [dylanb, a11y] @bkardell: when you say "grid", what specifically are you talking about?
  33. zakim-robot
    16:32
    [Brian Kardell, a11y] http://www.w3.org/TR/2015/WD-css-grid-1-20150806/
  34. zakim-robot
    16:34
    [Brian Kardell, a11y] in grid there's no correlation between the dom tree and where items are placed (same with abspos, but again- that had so many limitations to what an author could do it wasn't used at scale like grid will be)
  35. zakim-robot
    16:37
    [dylanb, a11y] The best way to get ahead of this is to get involved in the WG. I see two potential approaches - one is to build accessibility into the spec and specifically defined what should happen to the acc tree. One would be to create techniques on how to use the technologies in an accessible fashion.
  36. zakim-robot
    16:38
    [Brian Kardell, a11y] dylanb I am in the WG
  37. zakim-robot
    16:38
    [Brian Kardell, a11y] :simple_smile:
  38. zakim-robot
    16:38
    [dylanb, a11y] woo hooo
  39. zakim-robot
    16:38
    [dylanb, a11y] :simple_smile:
  40. zakim-robot
    16:40
    [Brian Kardell, a11y] this is why I am actively seeking out: a) whether someone else has already given this quite a lot of thought or discussion so I can brush up b) whether we as a community can help with the social aspect of re-aligning the expectation/practices before bad ones take root
  41. zakim-robot
    16:44
    [dylanb, a11y] Is the WG looking for input on the accessibility? Are there proposals already for how to deal with it? What is the status?
  42. zakim-robot
    16:52
    [Brian Kardell, a11y] https://lists.w3.org/Archives/Public/www-style/2015Aug/0348.html
  43. zakim-robot
    16:52
    [Brian Kardell, a11y] the status is "don't do that, its a layering violation"
  44. zakim-robot
    16:53
    [Brian Kardell, a11y] see this thread https://lists.w3.org/Archives/Public/www-style/2015Aug/0347.html
  45. zakim-robot
    17:01
    [dylanb, a11y] I am not surprised. The precedence for CSS has always been to have almost no impact on accessibility (display properties and broken table algorithms excluded)
  46. zakim-robot
    17:03
    [dylanb, a11y] This is definitely the place for some good "best practices". One thing you could do is get the WG to change this example so that it conforms to best practice http://www.w3.org/TR/2015/WD-css-grid-1-20150806/#z-order
  47. zakim-robot
    17:03
    [dylanb, a11y] in other words, make the visual, left to right, top to bottom order match the DOM order, rather than as is the case right now where it is obviously quite broken
  48. zakim-robot
    17:05
    [dylanb, a11y] I strongly believe that all examples in the standards should only show how to do things the most accessible way (even if you can break stuff with the technology)
  49. zakim-robot
    18:16
    [Val Head, a11y] Hi all! A while back I promised to share when the results of my animations and accessibility research were published. An article about some of my findings (mostly around animation and vestibular disorders) went up on ALA today: http://alistapart.com/article/designing-safer-web-animation-for-motion-sensitivity#comment-form
  50. zakim-robot
    18:29
    [Marcy Sutton, a11y] @patrickfox: in http://spa-a11y.divshot.io/#14 you advocate for "one live region to rule them all”. How do you handle differing politeness levels?
  51. zakim-robot
    18:29
    [Marcy Sutton, a11y] cc @jelbourn
  52. zakim-robot
    18:33
    [Patrick Fox, a11y] Use a helper function to 1) update the message and 2) set the politeness value: https://github.com/patrickfox/a11y_kit/blob/master/dist/a11y_kit.jquery.js#L31-L44
  53. zakim-robot
    18:34
    [Marcy Sutton, a11y] ooh nice! :ok_hand:
  54. zakim-robot
    18:36
    [Marcy Sutton, a11y] And you found that manipulating the element doesn’t cause problems in AT? I’m circling back on angular/material#3174 to discuss whether a similar approach would work for Angular Material.
  55. zakim-robot
    18:42
    [Patrick Fox, a11y] I haven't seen any issues with changing the politeness level dynamically, but @dylanb has mentioned previously that he's had issues with this with certain browsers/AT
  56. zakim-robot
    18:42
    [dylanb, a11y] Never tried changing politeness
  57. zakim-robot
    18:43
    [dylanb, a11y] but I did get issues with a single region across browsers
  58. zakim-robot
    18:43
    [dylanb, a11y] in IE (I believe version 9), there were announcements of deletions even when it should not have been announced
  59. zakim-robot
    18:44
    [dylanb, a11y] That caused a cascade of issues that were bugs related to trying to fix the IE issue
  60. zakim-robot
    18:46
    [Marcy Sutton, a11y] I wondered if a single region for each politeness level might reduce complexity and potential bugs at the same time
  61. zakim-robot
    18:46
    [Patrick Fox, a11y] weird. The $.announce() function also has a reset which clears the announcer in a way that overcomes an NVDA issue that I've seenwhere repeat/identical messages are not spoken
  62. zakim-robot
    18:46
    [Marcy Sutton, a11y] The idea for Angular Material being, the framework has no control over how many components requiring announcements might be used in an app….so centralizing messages makes sense
  63. zakim-robot
    18:47
    [Patrick Fox, a11y] we have used this technique using a single live region in a large project, and I haven't seen the need for this
  64. zakim-robot
    18:47
    [Marcy Sutton, a11y] An app !== a framework used to build apps.
  65. zakim-robot
    18:47
    [Marcy Sutton, a11y] (less control)
  66. zakim-robot
    18:47
    [Patrick Fox, a11y] are you seeing specific issues with changing the politeness levels?
  67. zakim-robot
    18:47
    [Patrick Fox, a11y] level value...
  68. zakim-robot
    18:49
    [Marcy Sutton, a11y] We haven’t implemented a centralized message system yet, it’s an item in the queue. Right now each component that needs a live region embeds its own.
  69. zakim-robot
    18:50
    [Marcy Sutton, a11y] I was just curious if we were going to run into any problems with compatibility since the framework gets used so many places..including IE and hybrid apps.
  70. stevefaulkner
    20:09

    A11ySlackers chat log 7th September 2015 http://w3c.github.io/a11ySlackers/archives/2015/09/2015-09-07.html

  71. aardrian
    20:28
    For those who I didn't assault on Twitter, I did some longdesc analysis in the wild coupled with WebAIM's SR survey and munged it all together in a post with a general bit of advice: Use On-Page Image Descriptions http://adrianroselli.com/2015/09/use-on-page-image-descriptions.html
  72. stevefaulkner
    21:55
    @aardrian don't spread your anti-longdesc propaganda here ;-)
  73. aardrian
    21:56
    @stevefaulkner I'm spreading my aria-described-by-that-guy-over-there message actually.