Archive index

A11y Slackers Gitter Channel Archive 12th of February 2016

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    Feb 12 00:00

    [jyung] @marcysutton: Thanks for the response!

    does the containing action really need to be accessible
    So the User profile_ does not (like you said, it’s redundant), but _Stats link would need to be accessible somehow, right?

  2. jnurthen
    Feb 12 00:01
    @jyung - I don't believe we got your question on gitter - just the link to the screenshot. Can you repost it?
  3. zakim-robot
    Feb 12 00:01

    [jyung] Sure! I actually posted the question as a comment to the file, but here it is!

    :wave: Hey all! I’m building a list of profiles. Both User profile and Stats are links that will navigate to another page. In addition, when clicking anywhere (except for the _Stats link), inside each container, I want it to also navigate to the user’s profile page (same as clicking the _User profile link).

    I initially wanted the container itself to be an <a> such that right clicking anywhere in the container will show the Open link in new tab menu option, but we cannot have an <a> nested inside an <a>.

    Alternatively, I was thinking of having the list container element be role=“listbox” with aria-activedescendant. What’s the right approach here to ensure everything is accessible?

  4. jnurthen
    Feb 12 00:02
    The comment didn't come through the sameroom.io connection
  5. zakim-robot
    Feb 12 00:03
    [marcysutton] Those two links would be links, right? Those should be accessible by default. Any redundant links that are only useful for mouse users don’t really need to be made accessible IMO.
  6. jnurthen
    Feb 12 00:03
    @marcysutton I agree
  7. zakim-robot
    Feb 12 00:04
    [marcysutton] “clicking in the container” === mouse action
  8. zakim-robot
    Feb 12 00:04
    [jyung] ahh gotcha
  9. zakim-robot
    Feb 12 00:05
    [jyung] yea, that makes sense!
  10. zakim-robot
    Feb 12 00:07
    [marcysutton] If you really do need them to be stacked anchors like in your screenshot, here’s an example of two cards with buttons in them, one valid and one invalid….imagine swapping button for anchor, and you can see what I meant about CSS positioning. http://codepen.io/marcysutton/pen/VYpjQz
  11. zakim-robot
    Feb 12 00:07
    [jyung] would what I'm describing be a candidate for listbox? For voiceover, I want the ability for users to navigate through the item containers first and when they are interested in a specific item, then they can jump into the item and get into the two links. make sense? (voiceover beginner here)
  12. zakim-robot
    Feb 12 00:08
    [marcysutton] That sounds like a confusing interface with multiple actions in each listbox option…but someone might convince me otherwise :smile:
  13. jnurthen
    Feb 12 00:09
    Not unless you really complicate the keyboard navigation. Listbox can only contain option elements. You would have to create some sort of "drill-in"keyboard shortcut to enable a user to move into them
  14. jnurthen
    Feb 12 00:09
    for something this simple I would say no
  15. jnurthen
    Feb 12 00:10
    We do that kind of thing for our listviews as there is no good aria role to support listviews currently
  16. jnurthen
    Feb 12 00:10
    I'm not particularly happy with it though as it leads to a more complex solution than I would like
  17. zakim-robot
    Feb 12 00:12
    [marcysutton] ^^ bingo! complicated solutions honestly aren’t worth the effort most of the time. If there’s a way to simplify the design, you’ll thank yourself later.
  18. jnurthen
    Feb 12 00:14
    Here is our listview... a keyboard user has to know to use F2 to change into "actionable" mode within each of the options.
  19. MichielBijl
    Feb 12 00:19
    And has to know they have to press esc to get out :P
  20. jnurthen
    Feb 12 00:21
    but I wish listviews were easier....
  21. zakim-robot
    Feb 12 01:30
    [jyung] right. yea, totally with you all that a simple solution is the way to go to not complicate the interface for users.
  22. sophieschoice
    Feb 12 08:52
    A question, at the moment I am building a new responsive lay-out, based on the code and guidelines from this workgroup: https://playbook.cio.gov/designstandards/ . On my Windows PC, if I use the tab-key, I can navigate through all links and buttons on my lay-out. On the site from that workgroup as well: I can tab through everything on Windows, not on Mac. Is that normal behaviour for Mac? Is there a setting I should set somewhere?
  23. jnurthen
    Feb 12 09:24
    Safari only tabs through certain
  24. MichielBijl
    Feb 12 09:30
    It's a setting.
  25. MichielBijl
    Feb 12 09:32
    alt=One can set tab behaviour in Safari settings under Advanced https://usercontent.irccloud-cdn.com/file/9cGhpwSz/
  26. sophieschoice
    Feb 12 09:32
    Ah, I see. Thanks! :D
  27. sophieschoice
    Feb 12 09:36
    I just wanted to make sure that my new lay-out is keyboard accessible on all browsers and OS's. At the moment I use Bootstrap for my website ( bad Sophie....) but it lacks accessibility-wise and usability-wise (low contrast and such) out of the box.
  28. MichielBijl
    Feb 12 09:37
    Keyboard support/testing is a very good starting point!
  29. sophieschoice
    Feb 12 09:45
    Yeah, I thought so :). I know the theory behind usability and SEO and I am really passionate about it. I apply the rules to my own website, but to be honest: filling in ALT-tags is easily forgotten.... But yeah, testing and usability/accessbility is important to me (sadly no professional experience) and since a few years I have the idea/dream of starting my own company. And before I start officially, I want to dabble some more in testing as I need experience in that.
  30. MichielBijl
    Feb 12 09:51
    Do you have any custom widgets?
  31. sophieschoice
    Feb 12 10:00
    I use the Add This-widgets for sharing content and following
  32. StommePoes
    Feb 12 10:07
    " but I get some pushback from the UI guys on this who would prefer something more subtle.."<-- those are the same guys who expect everyone to use a mouse and have never seen people confused about how a thingie works. Did it do anything? Did I click it? I'm not sure, I'll try clicking it again. Is it even clickable? I don't know. It's because I suck and was never meant to use computers why is everything so goddamn complicated and WHY is this text 9px light grey on white??????
  33. StommePoes
    Feb 12 10:08
    I really don't value the opinions of those so-called UI guys >:( because they invariably make MY life hard and obviously that's what I care about-- can I use the damn thing. Often not, or not without wailing and teeth gnashing and getting new bald spots.
  34. StommePoes
    Feb 12 10:10
    I had to create a new font file using the icomoon "app" page. ZOMG. All these pictures and even the buttons with text, I have no freaking clue wth they do and I'm clicking them almost randomly trying to figure out how they work and what they do and feeling stupid because I'm not a 20-year-old design-school student who just magically knows all this crap and then I realise I've done a bunch of things I didn't mean to do like doubling my fonts (now two copies of each icon!) and I can't figure out how to undo and closing and re-opening the browser doesn't fix it!!!!!!!!!!!!111
  35. sophieschoice
    Feb 12 10:11
    Ha, I don't trust those people either. They only care about more visitors, more conversions, more money. Testing costs money, so they don't like that option.
  36. StommePoes
    Feb 12 10:11
    They make shit pretty, I'll give them that. And people like pretty things, I'll give them that too. But if I can't use it, fuckem.
  37. sophieschoice
    Feb 12 10:12
    Amen :D
  38. MichielBijl
    Feb 12 10:16
    Hey look, StommePoes woke up :P
  39. 9_9
  40. jpdevries
    Feb 12 13:50
    I mean, this is like a deadly sin right? Thanks ExtJS.
    .ext-webkit:focus {   outline: none !important; }
    
  41. powrsurg
    Feb 12 14:27
    it's not terrible as long as you have some type of detection for keyboard users where you then turn it back on for them
  42. powrsurg
    Feb 12 14:27
    which I doubt they have and it's most likely visual designers that have no clue about anything accessibility related
  43. powrsurg
    Feb 12 14:32
    about UI people, it's because most of the time they work by producing a JPEG or something and display the mock up to a higher up who of course loves it because it looks pretty, ignoring how it would have to be built or why it has problems X/Y/Z
  44. powrsurg
    Feb 12 14:35
    one of the most frustrating I had was with someone that wanted links to have the same color as text AND no underline ...
  45. sophieschoice
    Feb 12 14:43
    OMG, what was their reasoning?
  46. powrsurg
    Feb 12 14:56
    basically they wanted all text that color, and they think that underlines make a site look dated
  47. powrsurg
    Feb 12 14:57
    It was at that point that I stopped pushing SEO for the client ...
  48. jasonday
    Feb 12 14:57
    @StommePoes - if they are promoting 9px light grey text, they aren't real UI/UX guys...they're "Creatives"
  49. jasonday
    Feb 12 14:58
    @jpdevries - CSS resets are a PITA for that exact reason
  50. powrsurg
    Feb 12 15:05
    we had someone like that apply here as a co-op. We offered him the position as we needed someone to clean up some interfaces and didn't have the time for the main team to work on it. I was so looking forward to crushing him with all of the things he'd need to actually do since I would not let our main website get a new refresh with major problems like I would have expected
  51. powrsurg
    Feb 12 15:05
    "What do you do for browser testing?" "What's that? ... I mean, it works in Chrome?"
  52. MichielBijl
    Feb 12 15:23
    Who uses CSS resets still?
  53. MichielBijl
    Feb 12 15:23
    I would've thought people had learned proper CSS by now.
  54. MichielBijl
    Feb 12 15:23
    Oh wait.
  55. sophieschoice
    Feb 12 15:23
    does normalize.css count?
  56. MichielBijl @MichielBijl wakes up from dream land.
  57. sophieschoice
    Feb 12 15:24
    if so, I raise my hand in shame….
  58. MichielBijl
    Feb 12 15:24
    I don't know what normalize does.
  59. MichielBijl
    Feb 12 15:24
    Okay, so what is the use of that one?
  60. sophieschoice
    Feb 12 15:25
    normalize.css v3.0.1
  61. MichielBijl
    Feb 12 15:32
    That looks better than a reset, but still don't quite get why I should add that to my website.
  62. sophieschoice
    Feb 12 15:36
    “When an element has different default styles in different browsers, normalize.css aims to make those styles consistent and in line with modern standards when possible.” - I guess that is the reason why most people use it. Aside from the reason that every developer uses it :P
  63. jasonday
    Feb 12 15:38
    My thought is that browser defaults shouldn't be overridden (for the most part). I tend to not like when the expectation is that all things look the same in all browsers.
  64. sophieschoice
    Feb 12 15:44
    @MichielBijl Have you never encountered the use of normalize.css in website before, when testing? Could be a good thing: meaning it is there on a website, but doesn’t interfere much with usability or accessibility.
  65. MichielBijl
    Feb 12 15:47
    @sopieschoice yeah okay I get that reason, but I'm with jasonday: http://dowebsitesneedtolookexactlythesameineverybrowser.com/
  66. MichielBijl
    Feb 12 15:48
    And I can't remember, but I mainly work on own work, so that doesn't really count.
  67. MichielBijl
    Feb 12 15:53
    For example: Removes inner padding and search cancel button in S5, Chrome on OS X
  68. MichielBijl
    Feb 12 15:53
    I like that cancel button, don't remove it!
  69. MichielBijl
    Feb 12 15:53
    Developers know best (much like Hulk Hogan)
  70. MichielBijl
    Feb 12 15:53
    Meaning they don't
  71. MichielBijl
    Feb 12 15:53
    :P
  72. sophieschoice
    Feb 12 15:54
    ouch, I didn’t know that, that it removes that cancel button…...
  73. MichielBijl
    Feb 12 15:56
    It's not an a11y issue, just an annoyance.
  74. jasonday
    Feb 12 15:57
    and you won't like me when I'm annoyed X(
  75. sophieschoice
    Feb 12 16:00
    well you might classify it as just an annoyance, but basically it is an example of trying to think for the user. And that’s something a developer/designer shouldn’t do.
  76. sophieschoice
    Feb 12 16:02
    Then again, I don’t think it hurts that much, since the US Government also uses it their accessible responsive lay-out project….
  77. jnurthen
    Feb 12 16:20
    I would not look to the US Government for accessibility advice
  78. MichielBijl
    Feb 12 16:21
    How is the 508 refresh coming?
  79. jasonday
    Feb 12 16:22
    @jnurthen - they are getting better in practice - https://playbook.cio.gov/designstandards/
  80. sophieschoice
    Feb 12 16:25
    yeah, that project I meant. It looks good in my eyes.
  81. jnurthen
    Feb 12 16:28
    I guess that is not completely horrible!
  82. zakim-robot
    Feb 12 16:30
    [dna] hi folks, posted this in aria-patterns but there are way more people in here so I'm going to repost
  83. zakim-robot
    Feb 12 16:30
    [dna] when using aria-activedescendant to manage focus of subelements (like menuitems in a menu), should the aria-activedescendant be cleared when the element in question (the menu) is no longer in focus?
  84. zakim-robot
    Feb 12 16:30
    [toddkloots] yes
  85. zakim-robot
    Feb 12 16:31
    [dna] thanks toddkloots!
  86. zakim-robot
    Feb 12 16:31
    [toddkloots] Sure thing
  87. jnurthen
    Feb 12 16:35
    @dna - ideally yes but in reality it doesn't really make a difference. If the thing isn't in focus any more then it doesn't really matter if it has aria-activedescendent. It just makes it easier if you ever move focus back to it.
  88. zakim-robot
    Feb 12 16:36
    [dna] also, unrelated question: what's the best way to visualize and get an understanding of the accessibility tree?
  89. zakim-robot
    Feb 12 16:36
    [dna] thanks jnurthen
  90. jnurthen
    Feb 12 16:36
    in which browser?
  91. zakim-robot
    Feb 12 16:37
    [dna] hmmm... the browser that represents it in the most full/useful state
  92. zakim-robot
    Feb 12 16:37
    [dna] sorry i dont have a specific goal in mind
  93. zakim-robot
    Feb 12 16:37
    [dna] i just want to see what this tree looks like and how I might use it
  94. jnurthen
    Feb 12 16:38
    easiest is to go to a new tab in chrome and type chrome://accessibility but then you get Chrome's version of the tree which isn't always the best
  95. zakim-robot
    Feb 12 16:38
    [toddkloots] @dna you might find this article helpful: https://www.paciellogroup.com/blog/2015/01/the-browser-accessibility-tree/
  96. zakim-robot
    Feb 12 16:38
    [dna] yeah, i've seen the huge text blob but I don't know what to do with it
  97. zakim-robot
    Feb 12 16:39
    [dna] i've played around a bit with firefox's DOM inspector
  98. jnurthen
    Feb 12 16:39
    well - they are all a bit different. IMO FF has the best implementation in many ways but they do have their awful (IMO) way of doing aria-hidden
  99. zakim-robot
    Feb 12 16:40
    [dna] my motivations are I'd like to get a better understanding of how setting aria-hidden on various elements actually impacts the tree, visualized in a useful way
  100. zakim-robot
    Feb 12 16:41
    [dna] i've heard rumors of some special flags you can set on chrome to get some sort of developer tool with the tree available
  101. jnurthen
    Feb 12 16:41
    well - FF still exposes aria-hidden elements in the tree and just sets an attribute. I believe all other browsers remove hidden elements from the tree
  102. zakim-robot
    Feb 12 16:42
    [dna] toddkloots I've seen that post; is there any analog for os x?
  103. zakim-robot
    Feb 12 16:42
    [toddkloots] in OS X you can leverage the Accessibility Inspector which gives you an actual visualization of the a11y tree
  104. jnurthen
    Feb 12 16:42
    @alice would probably know about any special flags in chrome :)
  105. zakim-robot
    Feb 12 16:43
    [toddkloots] you can move the mouse around and see the tree for the specific part of the UI
  106. zakim-robot
    Feb 12 16:44
    [dna] is there a good resource you're aware of for parsing the output of the Accessibility Inspector?
  107. zakim-robot
    Feb 12 16:44
    [toddkloots] not sure
  108. zakim-robot
    Feb 12 16:45
    [dna] k, thanks for the help all!
  109. jnurthen
    Feb 12 16:46
    @dna you are welcome.
  110. stevefaulkner
    Feb 12 16:46
    @dna there is also UI browser for mac, its shareware but works for 30 days (as i remember) i ended up buying a licence as it much better than the mac dev tools http://pfiddlesoft.com/uibrowser/index.html
  111. zakim-robot
    Feb 12 16:48
    [dna] great stevefaulker I'll look into it
  112. zakim-robot
    Feb 12 17:09
    [marcysutton] Here are the instructions for enabling the new accessibility debugging features in Chrome Canary, you no longer have to press shift six times: https://groups.google.com/a/chromium.org/forum/#!topic/chromium-accessibility/zmWfzZic9fE
  113. garcialo
    Feb 12 17:12
    :+1:
  114. zakim-robot
    Feb 12 17:17
    [alice] @jnurthen Two options: chrome://accessibility will show you a fairly heinous text dump of the a11y tree, as well as allowing you to turn the a11y tree on/off for specific tabs
  115. zakim-robot
    Feb 12 17:17
    [alice] second option: what Marcy said
  116. zakim-robot
    Feb 12 17:18
    [alice] (chrome://accessibility is good when you want to run something like Accessibility Inspector or aViewer on just one tab, so you don't want to put the whole browser into a11y mode)
  117. zakim-robot
    Feb 12 17:23
    [alice] oh btw with the option Marcy mentioned - I recommend using it in Canary because it's being actively worked on and new changes are landing somewhat regularly (though not at the moment cause I'm busy with another project)
  118. zakim-robot
    Feb 12 17:24
    [dna] this is great @marcysutton @alice !
  119. MichielBijl
    Feb 12 17:24
    That Steve and Karl are acting weird again on Twitter.
  120. MichielBijl
    Feb 12 17:24
    They don't look like themselves at the moment…
  121. MichielBijl
    Feb 12 17:27
    @dna, if you have Xcode installed, the accessibility inspector is very nice/gives lots of info.
  122. MichielBijl
    Feb 12 17:27
    And it has an actual tree…
  123. zakim-robot
    Feb 12 17:29
    [alice] @michielbijl An actual tree?
  124. zakim-robot
    Feb 12 17:29
    [alice] Beyond the partial ancestor tree? How do you see it?
  125. MichielBijl
    Feb 12 17:30
    That's what I meant, sorry. Would've been cool if there was an actual tree in it.
  126. zakim-robot
    Feb 12 17:30
    [alice] Ah ok :simple_smile:
  127. zakim-robot
    Feb 12 17:31
    [alice] Professional interest: how and when would you use a full tree?
  128. jonathantneal
    Feb 12 17:32
    @stevefaulkner, our paths almost just crossed. And it would have been regarding a11y. I would have loved to say hi!
  129. stevefaulkner
    Feb 12 17:32
    hey @jonathantneal were u in uk?
  130. jonathantneal
    Feb 12 17:33
    I’m not 100% sure it was you, but I did the initial html, css, and js (which was then built up upon by other cool devs) for something you were reviewing.
  131. stevefaulkner
    Feb 12 17:34
    ah OK who for?
  132. jonathantneal
    Feb 12 17:34
    A client, so I’m being vague on purpose. But it would have been the second time in my entire life my online community touched my professional community.
  133. stevefaulkner
    Feb 12 17:35
    yeah understand, would be nice to work on something togther :smile:
  134. jonathantneal
    Feb 12 17:36
    And if it was you, I just have to applaude you and your team on a really solid review and great a11y advice. I was reading this going “this is so good” then I saw something that indicated you might have worked on it, and I asked someone if they met the person who did the review and they mentioned a british guy named Steve.
  135. MichielBijl
    Feb 12 17:36
    Hey jonathantneal, I haven't used it personally, but are there any a11y issues with normalize.css?
  136. MichielBijl
    Feb 12 17:36
    I'm guessing no, but thought I'd ask :)
  137. jonathantneal
    Feb 12 17:37
    @MichielBijl, well, actually from time to time there are. There were some opinionated styles on tables that made them unreadable if the table was not styled. They were removed recently but not in a published edition.
  138. MichielBijl
    Feb 12 17:38
    Ah, that's good to know. Probably due to the heuristics in browsers for detecting layout tables?
  139. jonathantneal
    Feb 12 17:38
    @MichielBijl, another time there was a missing focus style due to a normalization on an element. They come up rarely, but when they do, we find them pretty fast and usually alongside a solution.
  140. MichielBijl
    Feb 12 17:38
    :)
  141. MichielBijl
    Feb 12 17:38
    Well, that's good news for sophieschoice ;)
  142. jonathantneal
    Feb 12 17:38
    @MichielBijl, the table issue was that Normalize removed all border spacing just cause. I think there’s a half-decent chance somebody would not style tables, end up using Normalize.css, and then the text would have been stacked without any space.
  143. MichielBijl
    Feb 12 17:39
    Ah right, that's just an annoyance I guess.
  144. jonathantneal
    Feb 12 17:39
    But since you bring up Normalize.css, I gotta ask if you’ve seen Sanitize.css, because it definitely touches on a11y, in a manner where it’s trying to help though.
  145. jonathantneal
    Feb 12 17:40
    Well, if you had single line items touching without any space, that kind of visual disorientation constitutes an a11y issue too, IMHO.
  146. MichielBijl
    Feb 12 17:40
    Hmm, I can have a look, thanks. I don't norma11y use reset/normalise/something.css
  147. jonathantneal
    Feb 12 17:41
    Sanitize does things like:
    [aria-busy="true"] {     cursor: progress; }  [aria-controls] {     cursor: pointer; }
    
  148. MichielBijl
    Feb 12 17:44
    You can collapse lines 90 and 100 ;)
  149. MichielBijl
    Feb 12 17:44
    I'll make a PR
  150. jonathantneal
    Feb 12 17:44
    Sanitize.css is an opinionated base, more similar to reset, based on newer developer cowpaths as well as the csswg’s list of mistakes.
  151. MichielBijl
    Feb 12 17:44
    Oh no wait, overlooked the asterix
  152. MichielBijl
    Feb 12 17:47
    Yeah I don't know. I guess I can see the need for something like this, but I never quite got why people use any CSS but there own; you're going to overwrite it at some point anyway.
  153. MichielBijl
    Feb 12 17:55
    @StommePoes you had some test results on lists and list-style: none right?
  154. jonathantneal
    Feb 12 17:55
    With Normalize.css and Sanitize.css, I think there’s an honest goal to avoid overwriting what you would.
  155. powrsurg
    Feb 12 20:47
    so is there an insanize.css file that overwrites things in great ways?
  156. powrsurg
    Feb 12 20:48
    insanitize
  157. jonathantneal
    Feb 12 21:23
    Insanize.css
    * { background-color: pink; background-size: 98% 99% !important; color: #ff69b4; display: none !important; }
    
  158. powrsurg
    Feb 12 21:33
    * { usability: none; }
    
  159. StommePoes
    Feb 12 22:28
    " you no longer have to press shift six times:" @marcySutton know how to turn on sticky keys in Gnome :P
  160. StommePoes
    Feb 12 22:28
    Michiel yeah the lists had a list style, just not visual because the lists were set to display: table
  161. jnurthen
    Feb 12 22:28
    windows too :)
  162. StommePoes
    Feb 12 22:28
    so bullets just visually got removed like when you float
  163. StommePoes
    Feb 12 22:29
    Yeha I didn't know what triggered sticky keys in Windows
  164. StommePoes
    Feb 12 22:29
    but I was using Gnome
  165. StommePoes
    Feb 12 22:29
    and the line tool a lot
  166. StommePoes
    Feb 12 22:29
    where you click a place and hold shift down to make a line
  167. StommePoes
    Feb 12 22:29
    making a lot of lines kept tripping sticky keys on :(
  168. MichielBijl
    Feb 12 22:32
    so display: table on a list removes bullets but keeps it a list for AT?
  169. MichielBijl @MichielBijl needs to do testing…
  170. StommePoes
    Feb 12 22:34
    you're changing it from display: list item
  171. StommePoes
    Feb 12 22:34
    so it removes the bullets
  172. StommePoes
    Feb 12 22:34
    didn't used to in IE if I recall correctly
  173. StommePoes
    Feb 12 22:34
    back in the day
  174. StommePoes
    Feb 12 22:34
    but it would in other browsres
  175. StommePoes
    Feb 12 22:34
    and luckily changing display does not fool SRs
  176. MichielBijl
    Feb 12 22:34
    Those were the days
  177. MichielBijl
    Feb 12 22:35
    Good
  178. StommePoes
    Feb 12 22:35
    we're in those days now, except it's ARIA instead of browsers doing CSS
  179. MichielBijl
    Feb 12 22:35
    Would be better if it had display: table;?
  180. MichielBijl
    Feb 12 22:36
    instead of demolishing the list item?
  181. MichielBijl
    Feb 12 22:36
    s/item/style/
  182. jnurthen
    Feb 12 22:37
    I think removing the list style is cleaner
  183. MichielBijl
    Feb 12 22:39
    But then you don't get nav, list with 9 items
  184. MichielBijl
    Feb 12 22:41
    Or should that be considered an AT/UA bug?
  185. jnurthen
    Feb 12 22:42
    I guess now FF has fixed the display table sometimes creates a table in the accessibility API issue this is probably safe (although in my tests when the object had strong semantics before it didn't do it anyway)
  186. jnurthen
    Feb 12 22:42
    removing list style makes AT not read it as a list? really?
  187. StommePoes
    Feb 12 22:43
    it won't read out the bullets, if that's what you mean
  188. StommePoes
    Feb 12 22:43
    it shoudl still announce the length
  189. StommePoes
    Feb 12 22:43
    list of x items
  190. MichielBijl
    Feb 12 22:43
    Hmm
  191. MichielBijl
    Feb 12 22:43
    Yeah I meant the length thing.
  192. jnurthen
    Feb 12 22:44
    yeah - it should still say list of 9 items but will not say "bullet" before each list item.
  193. StommePoes
    Feb 12 22:44
    With our setup, display: table, we could use css generated content for bullets/numbers we could align using display: table, however all the listy semantics were still there
  194. StommePoes
    Feb 12 22:44
    and really only JAWS+IE was iffy
  195. StommePoes
    Feb 12 22:45
    and when I manually put in style="list-style-type:the-type" (where the type matched what we wanted to show visually, switching from ol numbers to lower-alpha) even JAWS+IE worked as well
  196. StommePoes
    Feb 12 22:45
    J+IE was the only combo who honoured the start attribute
  197. jnurthen
    Feb 12 22:45
    yeah - i was shocked the list semantics stayed even in dodgy versions of FF where the same structure using pure divs (and the same css display properties) showed as a table in the a11y apis
  198. StommePoes
    Feb 12 22:45
    since it will not reflect what's on the screen, we need to keep the css generated ones to match what the list would have generated
  199. StommePoes
    Feb 12 22:46
    Jason Kiss wrote up a page about display table issues
  200. StommePoes
    Feb 12 22:46
    like safari? or VO would only recognise a table with a caption
  201. StommePoes
    Feb 12 22:47
    some of that's probably not true anymore.
  202. StommePoes
    Feb 12 22:48
    Oh and agreed, removing the bullets with list-style none should not change the display style of the li's-- they should still be display: list-item and UA's/ATs should still be able to access the length property
  203. StommePoes
    Feb 12 22:48
    Forget what Steve called it
  204. StommePoes
    Feb 12 22:48
    so not accessing the length property and stating list of x items, I would call that a bug.
  205. StommePoes
    Feb 12 22:48
    I vaguely remember hearing about it before I think...
  206. MichielBijl
    Feb 12 22:53
    @StommePoes and @jnurthen: http://dir.rawr.eu/lists.html
  207. MichielBijl
    Feb 12 22:53
    Those lists (except for the dl) are not recognised as lists by VoiceOver in Safari.
  208. MichielBijl
    Feb 12 22:54
    Safari 9 on OS X 10.10
  209. MichielBijl
    Feb 12 22:54
    so no number of items announced, not even a grouping thingy
  210. jnurthen
    Feb 12 23:04
    I think that is a safari /VO issue
  211. jnurthen
    Feb 12 23:05
    When running in both FF and Chrome they are shown as listitems in the accessibility api.
  212. jnurthen
    Feb 12 23:05
    what does the accessibility inspector show?
  213. jnurthen
    Feb 12 23:06
    (on the mac)
  214. MichielBijl
    Feb 12 23:06
    I'll check
  215. MichielBijl
    Feb 12 23:06
    Chrome+Chromevox handles them correct.
  216. MichielBijl
    Feb 12 23:06
    Except for the DL (doesn't announce items/list)
  217. MichielBijl
    Feb 12 23:06
    But am not sure of support of default dl with ChromeVox
  218. jnurthen
    Feb 12 23:07
    i actually used a DL in the real world for the first time ever last week!
  219. MichielBijl
    Feb 12 23:07
    Yeap it's a Safari bug. Chrome+VoiceOver has correct behaviour.
  220. MichielBijl
    Feb 12 23:07
    First time huh?
  221. jnurthen
    Feb 12 23:08
    yeah - DL is not a common thing in my world
  222. MichielBijl
    Feb 12 23:08
    drunk list? Surely those are common :P
  223. jnurthen
    Feb 12 23:08
    not for a few more hours
  224. MichielBijl
    Feb 12 23:08
    I wasn't allowed to drink my own whisky on the plane…
  225. MichielBijl
    Feb 12 23:09
    That really bothered me.
  226. MichielBijl
    Feb 12 23:09
    And I don't have a glass in the bus, so can't drink any.
  227. jnurthen
    Feb 12 23:11
    you are on the move?
  228. MichielBijl
    Feb 12 23:11
    Oh, and to answer your a11y related question: AI shows AXGroup in Safari with a subrole of AXContentList
  229. MichielBijl
    Feb 12 23:11
    AXList in Chrome.
  230. MichielBijl
    Feb 12 23:12
    I should probably update my OS though…
  231. jnurthen
    Feb 12 23:12
    i can check on monday in the office with latest and greatest but don't have xcode on my home mac
  232. MichielBijl
    Feb 12 23:13
    I can upgrade this weekend and check :P
  233. MichielBijl
    Feb 12 23:13
    Xcode is free… what's keeping you from installing it?
  234. jnurthen
    Feb 12 23:27
    the machine is slow and full
  235. jnurthen
    Feb 12 23:27
    and besides - i try not to work on it
  236. jnurthen
    Feb 12 23:28
    it mostly acts as my plex server :)
  237. MichielBijl
    Feb 12 23:29
    Ah, gotcha, that's what my iMac does too ><
  238. MichielBijl
    Feb 12 23:30
    I'm considering to sell it…
  239. MichielBijl
    Feb 12 23:30
    Not much use for it when you have a 15" MBP…