Archive index

A11y Slackers Gitter Channel Archive 9th of November 2016

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    @zakim-robot
    Nov 09 00:18
    [shawn.henning] Are labels as text on top of a text box allowed in WCAG 2.0? I have not found a good example unless it is an image that I am missing. Sometimes known as "ghosted text"
  2. zakim-robot
    @zakim-robot
    Nov 09 00:24
    [caesar] You mean like the label implemented using the <input> placeholder attribute? Which disappears as soon as you click into the field?
  3. [shawn.henning] Yes
  4. zakim-robot
    @zakim-robot
    Nov 09 00:38
    [caesar] I think the lack of persistence might be a problem with people with cognitive and memory impairments
  5. [caesar] I.e. once you click into the field, you forget what you were supposed to be typing in there
  6. zakim-robot
    @zakim-robot
    Nov 09 00:44
    [shawn.henning] I agree, was just wondering if anyone had seen this situation written up in a conforming to WCAG 2.0, level ?? Write-up.no
  7. [caesar] I can only think that it would be dinged for not using <label> or <legend>, per https://www.w3.org/TR/2016/NOTE-WCAG20-TECHS-20161007/H90
  8. zakim-robot
    @zakim-robot
    Nov 09 00:50
    [caesar] oops, sorry that's just for "required" fields
  9. [caesar] https://www.w3.org/TR/2016/NOTE-WCAG20-TECHS-20161007/H44
  10. [caesar] This one
  11. zakim-robot
    @zakim-robot
    Nov 09 01:30
    [shawn.henning] We are trying to see what priority text labels should be as it will involve some design mods on a page. This is very helpful; thank you @caesar
  12. James Nurthen
    @jnurthen
    Nov 09 02:01
    IMO if there is no visible label or instruction I fail 3.3.2. Placeholder is explicitly defined in HTML as NOT being a label. Hence if placeholder is the only visible identification I pass it.
  13. Sorry. Missed a negative in that last sentence. Hence if placeholder is the only visible identification I don't pass it.
  14. The exception I normally make is search fields which are often visually labeled by the search button and username password fields which are /sometimes/ ok based on the username being plain text and the password being obscured. They are both kind of labeled by the login button. However even then there can be confusion for users. I.e. Was it a username field or an email address field.
  15. zakim-robot
    @zakim-robot
    Nov 09 02:07
    [caesar] Yeah, logins are a funny one. They usually also get a free pass from the mandatory field marking requirement as well
  16. James Nurthen
    @jnurthen
    Nov 09 02:09
    I hate mandatory field marking in many UIs. I like uis that mark optional fields better generally.
  17. James Nurthen
    @jnurthen
    Nov 09 07:27
    ok. Where in Europe should I move?
  18. Job van Achterberg
    @jkva
    Nov 09 07:45
    Dude you can have the couch
  19. James Nurthen
    @jnurthen
    Nov 09 08:01
    what country in Europe doesn't have shitty politics at the moment
  20. UK/France/Austria are out
  21. zakim-robot
    @zakim-robot
    Nov 09 08:07
    [dean] Had lunch with Australian friends in Amsterdam a couple of weeks ago and had this very discussion - as expats, and at the minute, we’re all very happy living in The Netherlands
  22. James Nurthen
    @jnurthen
    Nov 09 08:18
    Netherlands is favorite at the moment. My wife works for Philips so she should be able to transfer.
  23. Unfortunately her division is in Eindhoven :(
  24. zakim-robot
    @zakim-robot
    Nov 09 08:54
    [dean] My wife is freelancing for them in Amsterdam at the minute. Hear lots of good things. I’ve only been to Eindhoven once and it was nice enough, but I’m a pro-Amsterdammer so I’m a bit blinkered when it comes to where I would live.
  25. James Nurthen
    @jnurthen
    Nov 09 08:57
    Yeah. Maybe she can find something in Amsterdam. Eindhoven seems kind of dull - although much cheaper.
  26. I'm worried how my kids would do in a Dutch school not knowing the language.
  27. zakim-robot
    @zakim-robot
    Nov 09 09:40
    [michiel] Please come to Amsterdam. I'm also looking.
  28. James Nurthen
    @jnurthen
    Nov 09 09:43
    @michiel you are moving back?
  29. zakim-robot
    @zakim-robot
    Nov 09 09:45
    [michiel] Well, not sure you yet. Not in a hurry right? Not like I have to move before end of this month :see_no_evil:
  30. [michiel] But NL is bit cheaper than UK.
  31. [michiel] Especially housing.
  32. [michiel] Sure you pay less taxes, but all the other stuff is crazy expensive.
  33. James Nurthen
    @jnurthen
    Nov 09 09:45
    Yeah. Although Amsterdam is kind of pricy.
  34. zakim-robot
    @zakim-robot
    Nov 09 09:49
    [michiel] Still cheaper than uk apparently
  35. [michiel] Seems I'm going to miss CSUN for the next four years.
  36. [michiel] Oh wait, TPAC is also in the us next year.
  37. [michiel] -.-
  38. James Nurthen
    @jnurthen
    Nov 09 09:51
    I may not be attending that either.
  39. zakim-robot
    @zakim-robot
    Nov 09 09:59
    [michiel] Can't we just move TPAC to someplace else…
  40. zakim-robot
    @zakim-robot
    Nov 09 11:16
    [michiel] Just a reminder: https://twitter.com/SteveALee/status/796298496492109824
  41. [michiel] Like y’all need one…
  42. zakim-robot
    @zakim-robot
    Nov 09 14:09
    [felix.berger] Hello there, maybe I will find some help in here - I am in the middle of a discussion if it is (a11y wise) safe to use headlines without fieldsets for small forms (that do not need grouping). I am quite sure that the usage of at least one fieldset / legend is not mandatory ... please anyone correct me if I might be wrong here!? :)
  43. luis garcia
    @garcialo
    Nov 09 14:14
    Use of fieldset is not mandatory for all forms. What are you form fields?
  44. And by "headline" do you mean a heading?
  45. zakim-robot
    @zakim-robot
    Nov 09 14:25
    [felix.berger] Oh, yes of course H2 or H3 depending on context ...
  46. [felix.berger] So, if fieldset is not mandatory this would imply legend not beeing mandatory since I can only use those within a fieldset ...
  47. [felix.berger] Thanks a lot :)
  48. luis garcia
    @garcialo
    Nov 09 14:40
    Sure thing.
  49. luis garcia
    @garcialo
    Nov 09 15:53
    I've previously heard (and repeated without verifying) that Shift+Tab is acceptable exit method for https://www.w3.org/TR/UNDERSTANDING-WCAG20/keyboard-operation-trapping.html
  50. Mallory
    @StommePoes
    Nov 09 15:53
    Back in HTML4 day there was a requirement somewhere about fieldsets and legends but it got quietly removed during some spec upgrade
  51. luis garcia
    @garcialo
    Nov 09 15:53
    Upon rereading myself, I would say Shift+Tab wouldn't be an unmodified Tab key. Thoughts?
  52. Mallory
    @StommePoes
    Nov 09 15:54
    "I'm worried how my kids would do in a Dutch school not knowing the language." @jnurthen they'd prolly be sent to an English-lang school while getting Dutch lessons and eventually English starts being the teaching language as you go higher in school anyway
  53. though it at least used to be level4 Dutch was required to go to uni, but a few years of Dutch and they all ought to have that, so long as they get forced daily to interact with real Dutch
  54. Luis: I feel it's not a real trap if I can back out of it
  55. That is, if I must make a focus trap (like a modal dialog) it's not considered acceptable to only trap going forwards. I must trap with backwards (shift+tab)
  56. luis garcia
    @garcialo
    Nov 09 15:56
    Yeah, that was my thinking as well; however...whether I consider it a real trap or not does not necessarily mean it's a pass under WCAG
  57. Mallory
    @StommePoes
    Nov 09 15:57
    Since Shift is part of Tabbing, I wouldn't consider Shift a modified Tab
  58. I did think other parts of WCAG specifically mention Shift when talking about "standard" navving keys?
  59. shift, tab, arrows, enter, space?
  60. luis garcia
    @garcialo
    Nov 09 15:58
    I'll take a look
  61. Remember where you saw it specifically?
  62. Mallory
    @StommePoes
    Nov 09 16:01
    No, and I'm starting to think it wasn't WCAG but the Aria practices page. If it was in WCAG it would be in the section 1 can access everything by keyboard
  63. er, 2.1.1
  64. luis garcia
    @garcialo
    Nov 09 16:02
    Yeah, I didn't notice any instances of 'shift' on the Understanding page
  65. Mallory
    @StommePoes
    Nov 09 16:02
    on Macs it's like sometimes you have to hit another key anyway, option
  66. derp https://www.w3.org/WAI/GL/UNDERSTANDING-WCAG20/keyboard-operation-trapping.html how did that break
  67. luis garcia
    @garcialo
    Nov 09 16:04
    It's the working draft; the actual one still works
  68. Mallory
    @StommePoes
    Nov 09 16:06
    Karl seems to think shift counts, lawlz http://www.karlgroves.com/2016/07/24/is-wcag-2-0-too-complicated/
  69. luis garcia
    @garcialo
    Nov 09 16:07
    I guess Shift+Tab could be included in "other standard exit methods."
  70. Shift+Tab would be like "slowly back away"
  71. Mallory
    @StommePoes
    Nov 09 16:08
    To me, Shift is a natural part of Tab, to set its direction.
  72. It's not even an exit method, it's just tabbing in the other direction.
  73. luis garcia
    @garcialo
    Nov 09 16:09
    So then if a page took it to the extreme and forced you to tab through the page backwards only, it would pass.
  74. Mallory
    @StommePoes
    Nov 09 16:09
    Though while something you can't tab through but can shift-tab out of, while maybe not a "trap", is still a barrier.
  75. It would "work with keyboard" so, yeah.
  76. It would fail usability, and this is why I can't accept the "has to work for people with disabilties because of their disabilities== accessibility" definition
  77. without usability it's useless.
  78. and forcing people to do an unexpected pattern is against usability
  79. But technically that's "nothing to do with accessibility because people with and without disabilities are equally shat on"
  80. and I can't buy that, since we know people with disabilities are using keyboard-only at a much higher percentage, therefore the patterns have to match expected patterns unless you've got a real special case.
  81. luis garcia
    @garcialo
    Nov 09 16:11
    Yeah, I mostly just put the statement out there so I could stew over it
  82. Mallory
    @StommePoes
    Nov 09 16:12
    It's good stew
  83. luis garcia
    @garcialo
    Nov 09 16:12
    But thanks for typing my thoughts out for me ;)
  84. Mallory
    @StommePoes
    Nov 09 16:12
    But I know to keep their sanity, many people doing WCAG conformance checks would try to stick so closely to the guidelines that a second professional has to go through later to mark up all the UX fails
  85. Honestly looking forward to more mobile/touch-related recommendations in there
  86. luis garcia
    @garcialo
    Nov 09 16:14
    Well, that's the issue isn't it. When you have different teams doing their own testing, you want to make sure they're all doing it consistently.
  87. Mallory
    @StommePoes
    Nov 09 16:14
    A modal dialog can be 100% keyboard accessible and that makes it pass WCAG, yet without a separate CLOSE button, if the modal fills the screen the user's kinda fucked on a lot of phones. onscreen keyboard may or may not be callable, may or may not have an ESC key (many don't), modal may prevent "clicking outside" because there's no outside... etc
  88. it's why the Comments column is Mandatory by us.
  89. luis garcia
    @garcialo
    Nov 09 16:15
    "Comments" column? Sorry, I'm not familiar with your product.
  90. Mallory
    @StommePoes
    Nov 09 16:16
    So if there were a barrier like an auto-tab that, whlie not a "trap", prevented someone from getting further along the page, we'd not call it a trap (so not mark in the trap section perhaps) but fail instead under the general keyboard requirements and then the comments column states what exactly is the problem.
  91. We have some table we can fill in
  92. first col is the WCAG number
  93. second is teh name of the guideline
  94. luis garcia
    @garcialo
    Nov 09 16:16
    oh, gotcha
  95. I was thinking you required a Comments column in a modal
  96. Mallory
    @StommePoes
    Nov 09 16:16
    next three are "applicable? y/n" "passes? y/n" and "comments (mandatory)"
  97. a checklist sheet thingie
  98. luis garcia
    @garcialo
    Nov 09 16:16
    yup
  99. Mallory
    @StommePoes
    Nov 09 16:18
    I still need to figure out how to get saved HTML pages to display nicely locally in browsers on Windows... bitching about standards is making me think about a demo I'm going to do soon.
  100. James Nurthen
    @jnurthen
    Nov 09 19:31
    Have you tried starting a basic webserver and serving the document from there? If you are trying to serve it from the file system I would suspect some sort of privilege issue... browsers are much more picky about things like that these days
  101. zakim-robot
    @zakim-robot
    Nov 09 23:02

    [car] @garcialo Just wondering, for your shift+tab stew, did you stumble on an actual case before wondering about it? The case that tends to come up is the one where a multi-line text area needs to allow inserting "tab characters" into the text. I just wanted to share what we ended up with for our "code editor" widget, which needs to let the user type tab characters into the code they are editing, and also type tab/shift+tab to indent/outdent selected lines.
    We needed another way to navigate out of the widget, so we looked to the Authoring Practices doc for guidance, and noticed that the Keyboard Interaction section for Rich Text Editor had a discussion about that very topic, starting with:

    Optionally, if the developer wishes to provide the ability to insert a tab into the document, it is recommended one of the following methods be used…
    We chose to prove the 2nd method. Our menu of keyboard shortcuts (and our accessibility doc) list the Control+M shortcut to toggle the use of Tab key between 2 modes: "Tab inserts tab characters" or "Tab navigates".
    https://www.w3.org/TR/wai-aria-practices-1.1/#richtext

  102. zakim-robot
    @zakim-robot
    Nov 09 23:08
    [car] @michiel I see that the APG task force is planning to review the Rich Text Editor pattern. Please note that we need to keep the "Control+M toggles tab mode" advice so that users can tab into our editor and then type Control+M tab to tab out… (see discussion above).
  103. [garcialo] Nah, was just reading through WCAG again like a completely normal person does and noticed that what it said on the page seemed to contradict with what I had internalized previously
  104. [car] Completely normal? :stuck_out_tongue_closed_eyes:
  105. zakim-robot
    @zakim-robot
    Nov 09 23:14
    [garcialo] assuming myself as the standard for normal*
  106. [garcialo] taller than me = tall; shorter than me = short; not occasionally reading through WCAG for fun = weird
  107. zakim-robot
    @zakim-robot
    Nov 09 23:32
    [car] I think "occasionally reading through WCAG for fun" is a good title for a blog. ;)
  108. zakim-robot
    @zakim-robot
    Nov 09 23:46
    [caesar] "Occasionally reading through WCAG for fun and profit" will be the title of my next book