Archive index

A11y Slackers Gitter Channel Archive 10th of January 2017

What fresh hell is THIS now? - Patrick Lauke
  1. Takeshi Kurosawa
    @kurosawa-takeshi
    Jan 10 00:15
    I have a question regarding WCAG 2.0 Language of Parts.
  2. zakim-robot
    @zakim-robot
    Jan 10 00:15
    [karlgroves] OK
  3. Takeshi Kurosawa
    @kurosawa-takeshi
    Jan 10 00:15
    It is not possible to use lang attributes in title element or attributes.
  4. Does a title element violate Language of Parts if the title element contains phrases of multiple languages?
  5. zakim-robot
    @zakim-robot
    Jan 10 00:19
    [karlgroves] That’s really interesting.
    IMO Language of Parts can’t apply in this case.
  6. [caesar] I can't see how 3.1.2 can be satisfied in HTML presently, to be honest.
  7. Takeshi Kurosawa
    @kurosawa-takeshi
    Jan 10 00:22
    Thanks.
  8. My colleagues and I can't find solutions too.
  9. My colleague posted an email while ago: https://lists.w3.org/Archives/Public/w3c-wai-ig/2016OctDec/0163.html
  10. If there are good solutions, please let us know
  11. zakim-robot
    @zakim-robot
    Jan 10 00:28
    [caesar] Is there any particular reason why you believe you need to conform to this success criteria?
  12. Takeshi Kurosawa
    @kurosawa-takeshi
    Jan 10 00:32
    One of our clients has a site which introduce overseas news into Japan. In that site, many articles have to contain phrases of English, Chinese and so on.
  13. Some pages have titles like <title>Impact of (some non-Japanese language phrases here) </title>
  14. zakim-robot
    @zakim-robot
    Jan 10 00:50
    [caesar] Not that there's any solution, but I found a lot of interesting discussion on Drupal: https://www.drupal.org/node/1323338
  15. [caesar] See comment #12 on this page: https://www.drupal.org/node/1322906
  16. [caesar] They give some good examples of language changes in page body
  17. [caesar] Hope that helps a little...
  18. Takeshi Kurosawa
    @kurosawa-takeshi
    Jan 10 00:53
    Thanks. I'll review them!
  19. zakim-robot
    @zakim-robot
    Jan 10 04:42

    [caesar] Question: is a visible label always required for a text input field (including textarea) in order to meet 3.3.2?

    It seems more and more designs are using those placeholder style labels that disappear once the field receives input. There's definitely a very strong design trend, particularly on mobile.

  20. zakim-robot
    @zakim-robot
    Jan 10 14:53
    [melsumner] I require a label value even if it has a sr-only class on it.
  21. [melsumner] I am generally displeased with this trend of "oh we don't need a label" because if you are a distracted person is filling out the form, such as a busy mom with kids around, then you have to try to back out or try to remember what the placeholder wanted.
  22. [tink] @melsumner Not sure of the context, but it's generally a good idea for form labels to be visible - or at least for there to be a visible label.
  23. [melsumner] @tink I agree, although I've tolerated it in some instances where the placeholder is visible and it's in a fieldgroup where the legend is enough context. I just don't think it's very empathic design to leave off a field label "because prettiness"
  24. [tink] Definitely agree with that :)
  25. zakim-robot
    @zakim-robot
    Jan 10 15:35
    [melsumner] Would you put a role on an iFrame?
  26. [karlgroves] What does the iFrame do?
  27. [tink] The only allowed roles on an iframe are application, document or img.
  28. zakim-robot
    @zakim-robot
    Jan 10 15:46
    [melsumner] Ty @tink - I avoid them like the plague but then occasionally come across one and am inevitably left feeling a bit befuddled
  29. zakim-robot
    @zakim-robot
    Jan 10 15:58
    [tink] The best place to check is in the HTML spec - for each element definition, it has the roles that are allowed - though in this case I'm not convinced of using *any role on an iframe.
  30. Mallory
    @StommePoes
    Jan 10 18:25
    melsumner one compromise for stiff designers is the "floating label" where it mimics a placeholder but moves into a still-visible area whenever it's either focussed or contains a value (for type=text or textarea inputs).
  31. Mallory
    @StommePoes
    Jan 10 18:35
    @stevefaulkner it seems HTML5 is still using required on checkboxes as that crap setup where you can set it on any checkbox but then you might as well pre-check it and then prevent users from changing it because if it's required it must be in a checked state?
  32. So maybe aria-required for checkboxes was left out on purpose.
  33. What people want with required on checkboxes is the same behaviour as on radios-- demand required isn't satisfied if at-least-one-of-the-group isn't checked. You would want the same for aria-required, no?
  34. Mallory
    @StommePoes
    Jan 10 20:57
    I'm curious about clicking grids with speech rec, assuming a variety of aria (and non-aria) patterns but generally using tables
  35. with an interactive table where you're using a grid role and adding tabindices on grid cells, those being interactive means they need a name. The spec says they get their name from "contents (newline) author", does that mean adding an id on a part of content and aria-labelling with it?
  36. zakim-robot
    @zakim-robot
    Jan 10 21:15
    [melsumner] @StommePoes yes, I’ve seen and kind of like those forms where the placeholder sort of transitions up into the label space, but I guess I am maybe a little old-fashioned in that I really like a good form that doesn’t change on me while I am trying to use it, and really keeps things simple (including how many form fields they are asking me to fill out).
  37. Mallory
    @StommePoes
    Jan 10 21:25
    it's not ideal, but it's a trick I've used as a dev whose position on the totem pole was way way under the designer's.
  38. powrsurg
    @powrsurg
    Jan 10 21:36
    @StommePoes got a link to an example of that? I've seen that a few times on random sites but hadn't looked at how it's being done in practive
  39. I've been meaning to look into that and how sites like FB have a large block of text when you first type and then shrink the text once it moves onto a second line.
  40. zakim-robot
    @zakim-robot
    Jan 10 22:21
    [herin] Hi, I am testing something that does not support NVDA short cut keys. I understand that there is a role="application" that causes this issue. Has anyone come across this? Is Role="application" to be avoided is there any way NVDA users change the hotkey mode if there is a Role="application"?
  41. zakim-robot
    @zakim-robot
    Jan 10 23:10
    [melsumner] When you set role=application, you are telling NVDA that you will provide your own Keyboard shortcuts
  42. [melsumner] So only apply it to the widget/component where you actually control/define those & make sure to provide good instructions for the user
  43. [herin] Thanks @melsumner.
  44. [herin] That makes sense.
  45. [melsumner] You can nest these things too and it will switch back to browsing mode with role=document.
  46. [melsumner] Yw thumbsup emoji
  47. [herin] If Role=application is used in body tag, the role=document can be used to switch back to browse mode inside the form ?
  48. [herin] My issue here is a iframe using role="application" and the master html is fine.
  49. [melsumner] O.o that's a lot of native capabilities you are giving up
  50. zakim-robot
    @zakim-robot
    Jan 10 23:15
    [melsumner] Theoretically yes, but please think carefully about it and read what spec has to say :)
  51. [herin] when the focus moves into iframe, it's completely gone. I don't think application is needed here as there are native html elements. I will check with dev to remove this.
  52. [herin] Thanks @melsumner .
  53. [melsumner] Yw @herin