Archive index

A11y Slackers Gitter Channel Archive 26th of October 2016

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    @zakim-robot
    Oct 26 00:19
    [estelle] The accessibility color contrast ratio requirements: is there a study that led to the recommended numbers or were those numbers pulled out of a hat? I am looking for the study, but can't find one
  2. [caesar] I'm pretty sure I remember reading an explanation on WCAG 2.0 for Understanding 1.4.3 about the contrast being based on some scientific basis.
  3. [caesar] See https://www.w3.org/TR/UNDERSTANDING-WCAG20/visual-audio-contrast-contrast.html
  4. [caesar] Under the heading "Rationale for the Ratios Chosen"
  5. [caesar] Several cited references there
  6. zakim-robot
    @zakim-robot
    Oct 26 00:24
    [estelle] thanks
  7. [estelle] I still don't see a study that involved people though. They explain the problems of different variations in sight that can impact an assumed need for bigger contrasts, but i don't see that those numbers and colors were actually tested on people with visual differences
  8. zakim-robot
    @zakim-robot
    Oct 26 00:30
    [karlgroves] You’d probably have to go to the sources cited directly to see if that was the case
  9. [estelle] http://www.visibilitymetrics.com/sites/visibilitymetrics.com/files/downloads/Effective%20Color%20Contrast%20&%20Low%20Vision%201996.pdf
  10. [estelle] unfortunately that's a non-alt text PDF
  11. [karlgroves] #irony
  12. [mattmay] The numbers were based on research Gregg Vanderheiden did at TRACE, afaicr. I'm thinking somewhere in the 2001-2002 timeframe.
  13. [estelle] "Under most practical situations, color contrast was found not to enhance reading performance and in some low vision observers, color, as opposed to luminance or brightness, contrast interfered with reading performance. "
  14. [estelle] http://li129-107.members.linode.com/research/archived-studies/investigations/
  15. [estelle] I thought the ratios came out before 2001.
  16. [caesar] There's also a reference to both an ISO and an ANSI standard. Unfortunately those are behind paywalls, but would I be too optimistic in thinking that they would be based on a sound scientific basis?
  17. zakim-robot
    @zakim-robot
    Oct 26 00:35
    [cordelia] There was a WebAIM forum thread about this topic back in February. Might have some extra context: http://webaim.org/discussion/mail_thread?thread=7325
  18. [mattmay] TRACE refers to "Quantitative Color Vision and Sensitive Contrast Guideline" as one of its "success stories."
  19. [estelle] http://juicystudio.com/article/luminositycontrastratioalgorithm.php - Luminosity Contrast Ratio Algorithm - shows the algorithm, but doesn't show scientific study of actual humans
  20. [estelle] oh, found that same article before looking at the webaim thread.
  21. [estelle] "To help get some measurable data on the actual user impacts (with a
    focus on users with low vision and color blindness) of color
    combinations (and font sizes, faces, etc.), WebAIM will be conducting
    some research on this topic in the near future to see, in part, how
    well the WCAG contrast formula thresholds relate to actual
    accessibility on modern devices." - that was in February
  22. [cordelia] Finally, a mention of humans!
  23. [estelle] "From my reading, no particular display technology was actually
    considered for the ratios, but rather on some form of idealised model of
    color rendition."
  24. [cordelia] Also from that same post: “We regularly conduct an informal poll as part of our trainings by showing various colors on screen and ask participants to guess which pass and fail WCAG. Almost universally folks choose several colors that fail as being the ones that pass.”
  25. [estelle] i kind of think these were "hat" driven. The algorithm is a sound algorithm in terms of algorithms, but it doesn't seem to have been applied in a study.
  26. zakim-robot
    @zakim-robot
    Oct 26 00:41
    [mattmay] The research appears to come from ophthalmology research, not related to display technology.
  27. [estelle] Well, I guess I'll ask @jared_w_smith if he ever did the study
  28. [estelle] I've been thinking the numbers are 72% BS for several years now
  29. [mattmay] Well, hmm. Maybe I can take that back. There's a 1994 paper by Arditi/Knoblauch cited, and it comes from a Society for Information Display conference. They're display (monitor) tech nerds.
  30. [mattmay] That said, in 1994 they were probably testing on 14" 1024x768 CRTs sooooooo
  31. zakim-robot
    @zakim-robot
    Oct 26 00:46
    [estelle] with 256 colors, if any at all ;)
  32. [estelle] in 1994 my computer was grey on grey
  33. [cordelia] a lot of UIs are still grey on grey :\
  34. [mattmay] I will say that there was a great deal of deference paid to ratios/algorithms proposed by Trace in the WCAG WG. In no small part because the participants were largely practitioners and stakeholders, while the chairs were researchers.
  35. zakim-robot
    @zakim-robot
    Oct 26 00:56
    [estelle] We've been using this AA, AAA, 3:1 / 7: 1 / 4.5:1 for years, but that doesn't seem to have ever been tested. How can we get this revisited.
  36. [estelle] I agree with ensuring that information conveyed by color differences is also available in text and the other rules, because they make obvious sense. Even having contrast makes sense. But this ratio seems like BS. An intelligent guess. but still BS.
  37. zakim-robot
    @zakim-robot
    Oct 26 01:02
    [estelle] When I try to use two different colors in text, and both have to meet the 7:1 ratio, it seems to me that my two colors have enough contrast to meet AA against the white background, but the contrast between color 1 and color 2 is probably not noticeable. And, yes, you should use something other than color to differentiate that something is a link, let's say, if you have placeholder text, and want to change the color to something other than the default grey (since the default grey doesn't meet the contrast ratio 3:1 much less 7:1, if you change it to a 7:1 color, my assumption is some people won't be able to differentiate between the, say, dark red of my new placeholder and the text of the label
  38. [caesar] I assume somebody would have to fund a public study, which the WCAG WG would then use to reassess the standards in a future update?
  39. [estelle] and while the placeholder is in an input, the grey of the input border is not enough of a contrast for it to be seen
  40. [estelle] if we came up with numbers without actually doing a study though, the numbers are "pulled out of a hat" by definition.
  41. [caesar] True, we could come up with something that sounds eminently plausible, using pseudoscience :)
  42. [caesar] And then submit that to the WG for consideration :)

  43. [karlgroves] > if we came up with numbers without actually doing a study though, the numbers are "pulled out of a hat" by definition.

    Actually…

    That’s pretty much the case. I mean, the argument in support of WCAG WG’s decisions is that there’s loads of empirical information driving the standard, but the actual way the sausage has been made is much more icky

  44. [estelle] now i am hungry for sausage
  45. [karlgroves] #nastyWoman ;)
  46. zakim-robot
    @zakim-robot
    Oct 26 01:09
    [karlgroves] There are a lot of valid criticisms of WCAG from persons who’d like to see more coverage of low-vision concerns and cognitive impairment concerns. Some strong voices tend to drive those conversations in ways that don’t seem to be based on anything more than personal preference. The hope is that the “consensus based” approach of W3C process helps being balance, but… I’m not sure sometimes
  47. [caesar] By the time the Standards get updated we'll have bionic eyes
  48. [caesar] haha
  49. [estelle] so, i am writing an article on the "simple accessible input masking" script I created a while ago, and the mask has to have enough contrast from the text and enough contrast from the background to be differentiated.
  50. [estelle] and I am going to publicly call the AA/AAA ratio BS. Maybe that will get this thing some airtime and the needed research. Or, maybe it will get me ejected from the community. Either way.... i assume it will get a reaction
  51. zakim-robot
    @zakim-robot
    Oct 26 01:14
    [caesar] In my experience they got AA and AAA right
  52. [caesar] Assuming you predict the conformance levels as the push-pull between the extremes of freedom of development vs catering for extreme disability
  53. zakim-robot
    @zakim-robot
    Oct 26 01:22
    [estelle] They didn't test anyone with low vision, extreme disability, mild or average. The idea seems right. The numbers seem like they made sense to a mathematician, but not a specific condition. It seems like in trying to cater to ALL conditions, they went extreme, but each individual condition, even an extreme one, likely doesn't need the contrast required by AAA.
  54. powrsurg
    @powrsurg
    Oct 26 03:35
    http://arstechnica.com/information-technology/2016/10/microsoft-releases-open-source-toolkit-used-to-build-human-level-speech-recognition/
  55. Peter Krautzberger
    @pkra
    Oct 26 07:46
    Morning slackers.
  56. :tea:
  57. Peter Krautzberger
    @pkra
    Oct 26 07:52
    Has anyone ever come across funky WebKit/Blink scrolling behavior when a relative link pointed to an element whose first child was a visually hidden (screenreader) hint? That is, both Chrome and Safari scroll to utterly wrong positions (though focus is right), dev tools "scroll this node into view" fails etc.
  58. zakim-robot
    @zakim-robot
    Oct 26 11:03
    [jpdevries] While taking a train back home from an a11y meetup in Amsterdam I noticed the trains note that not all disabilities are visual or immediately evident and they ask you to give up your seat to anyone who asks. thumbsup emoji
  59. Job van Achterberg
    @jkva
    Oct 26 11:20
    @jpdevries a11y meetup in Amsterdam? Interesting, which one?
  60. JP DeVries
    @jpdevries
    Oct 26 11:22
    @jkva http://www.meetup.com/Inclusive-Design-Accessibility/events/232941793/
  61. There's also a Fronteers meetup in Amsterdam on Nov8
  62. Job van Achterberg
    @jkva
    Oct 26 11:23
    @jpdevries You were at idea11y? Damn, I didn't know. Would've liked to say hi!
  63. I was hoping there was another a11y meetup in Amsterdam that I could also go to :)
  64. zakim-robot
    @zakim-robot
    Oct 26 11:23
    [karlgroves] The guy presenting at the Fronteers meetup in November is a real jerk
  65. Job van Achterberg
    @jkva
    Oct 26 11:23
    lol!
  66. Yeah, they should really pick better speakers.
  67. JP DeVries
    @jpdevries
    Oct 26 11:24
    @jkva ditto! Maybe next time.
  68. zakim-robot
    @zakim-robot
    Oct 26 11:24
    [karlgroves] Doesn’t even speak Dutch
  69. Job van Achterberg
    @jkva
    Oct 26 11:24
    @jpdevries Next one is second half of November, if I can get all speakers to cooperate. We already have a venue. Scoop!
  70. zakim-robot
    @zakim-robot
    Oct 26 11:26
    [jpdevries] @karlgroves: at least you aren't a de Vries living and working in the Netherlands that doesn't speak Dutch. I'm a disgrace!
  71. [karlgroves] lol
  72. [jpdevries] Living in a country you don't speak the primary language has taught me a lot about accessibility though ;)
  73. Job van Achterberg
    @jkva
    Oct 26 11:27
    @jpdevries: oefenen!
  74. zakim-robot
    @zakim-robot
    Oct 26 11:29
    [karlgroves] The wife and I have talked a bit about moving out of the US at some point. NL is on my top 3 preferences (would be #1 if the weather was warmer). I know one Dutch word, because @michiel taught it to me: Godver
  75. Job van Achterberg
    @jkva
    Oct 26 11:29
    That one's useful.
  76. @karlgroves: I hope the Venco licorice was well received :)
  77. JP DeVries
    @jpdevries
    Oct 26 11:30
    "zalm omelet met frietjes" is all you really need to survive
  78. Job van Achterberg
    @jkva
    Oct 26 11:31
    zalm...omelet? shudders
  79. zakim-robot
    @zakim-robot
    Oct 26 11:32

    [karlgroves] > I hope the Venco licorice was well received :)

    Um…
    Yes. I’m sure the cleaning person in the Brighton hotel loved them.
    :(

  80. Job van Achterberg
    @jkva
    Oct 26 11:35
    I'm not sure how well Brits like licorice. Paging @jnurthen.
  81. zakim-robot
    @zakim-robot
    Oct 26 11:54
    [dean] If a De Vries can’t speak Dutch then the only solution is to hold all Accessibility meetups in English.
  82. [jpdevries] @dean LOL. I’m slowly learning. Hoping to take a class soon. It is tough because they speak both Dutch and Friesian at my work
  83. [dean] Delete it before Job reads it.
  84. [dean] Mijn Nederlands sucks.
  85. [dean] (and I feel less alone)
  86. [dean] Also, sorry I didn’t see you there either.
  87. Job van Achterberg
    @jkva
    Oct 26 11:58
    Deleting won't save you! IRCCloud remembers all.
  88. zakim-robot
    @zakim-robot
    Oct 26 11:59
    [dean] Dat is niet goed
  89. Job van Achterberg
    @jkva
    Oct 26 11:59
    Niet echt, nee. Je bent geschroefd.
  90. zakim-robot
    @zakim-robot
    Oct 26 12:00
    [jpdevries] The North Remembers!
  91. James Nurthen
    @jnurthen
    Oct 26 15:10
    @jkva don't ask me. I don't even drink tea.
  92. zakim-robot
    @zakim-robot
    Oct 26 17:09
    [estelle] Is #Slack accessible? Are we missing a large chunk of the target we should be reaching due to any inaccessibilities in this platform.
  93. [emplums] @estelle last I heard it's got some accessibility flaws. There's actually a channel in here just to talk about that #slack-a11y-feedback
  94. [emplums] I've got a screen reader question about NVDA... we just got this email from a user:
  95. [emplums] "The accessibility of the Firefox extension is a little interesting. the only way to activate it is by clicking on it from the toolbar in the browser, there are no ways to get at it with the context menu or from the main menu bar. I am running Windows 7 with Firefox 49 and the NVDA screen reader free from http://www.nvdaproject.org"
  96. [emplums] I'm a bit confused by it as I don't have access to a windows machine to run NVDA on
  97. [emplums] Does anyone know what this means? Can this screen reader not access links in the toolbar?
  98. zakim-robot
    @zakim-robot
    Oct 26 17:15
    [alice] @estelle @emplums We have the gitter bridge in part because some people have found that platform works better with IRC integration.
  99. [estelle] thanks!
  100. James Nurthen
    @jnurthen
    Oct 26 17:20
    @emplums I'm not sure you can keyboard to the toolbars in FF. I always add a way to get to my extension content in both the Tools menu and (often) in the right click (shift+F10) menu
  101. zakim-robot
    @zakim-robot
    Oct 26 17:20
    [emplums] ahhh maybe the context menu he was referring to was the right click menu?
  102. James Nurthen
    @jnurthen
    Oct 26 17:23
    This would also allow someone to access your extension content in a popup window which doesn't have the toolbar
  103. zakim-robot
    @zakim-robot
    Oct 26 17:25
    [emplums] @jnurthen do you know if NVDA can access content in the right click menu? I just checked and our extension tool does show up in the right click menu
  104. James Nurthen
    @jnurthen
    Oct 26 17:27
    works ok for my toolbar
  105. zakim-robot
    @zakim-robot
    Oct 26 17:36
    [emplums] hmm i need a windows machine
  106. James Nurthen
    @jnurthen
    Oct 26 17:38
    you could use a VM
  107. zakim-robot
    @zakim-robot
    Oct 26 17:40
    [marcysutton] Our very own @george_zamfir is working on accessibility stuff at Slack. if you have feedback they would love to know! That is partly why they funded this channel.
  108. James Nurthen
    @jnurthen
    Oct 26 17:45
    @emplums - sorry just re-read.... Shift+F10 should work for an NVDA user just like it works for a keyboard only user. Just tested my toolbar and there were no issues.
  109. zakim-robot
    @zakim-robot
    Oct 26 17:46
    [emplums] @jnurthen ahh awesome! thank you!!
  110. zakim-robot
    @zakim-robot
    Oct 26 18:17

    [george_zamfir] Yes indeed :sparkles:

    We’re in a bit of a catch up mode but to accelerate the pace of improvements we’ve created a dedicated accessibility team. Some of the recent work we’ve done includes dynamic text size support on mobile, improving screen reader support across platforms, toggling animations (https://web-a11y.slack.com/archives/slack-a11y-feedback/p1477415686000030).

    We absolutely recognize that we have a lot of work ahead of us but we’re very excited to do it. Building a truly inclusive experience is a big priority for all of us at Slack! :slack:

  111. zakim-robot
    @zakim-robot
    Oct 26 18:58
    [emplums] so awesome @george_zamfir! :tada: :heart_eyes_cat:
  112. [emplums] @jnurthen sorry for all the questions! I was able to get the context menu up in a VM on firefox with NVDA... but it doesn't read the menu items for me. I can visually see which item I'm focused on but I wonder if there's something I need to enable to get NVDA to actually read these items?
  113. [emplums] @george_zamfir we're in a similar place at Buffer right now. Lots and lots of work to be done to make Buffer fully accessible. I'd love to read more or talk to someone who's working on accessibility at Slack and to learn how you're handling tackling it.
  114. zakim-robot
    @zakim-robot
    Oct 26 19:03
    [emplums] Speaking of, I'm publishing a "Why your startup should care about accessibility" post on the Buffer Open Blog tomorrow if anyone is interested in giving it a read through before it goes live
  115. sophieschoice
    @sophieschoice
    Oct 26 19:04
    Shared it on Twitter, but thought more people would like to read it; https://t.co/EUqwGqklDb
  116. How the web became unreadable by Kevin Marks is that article
  117. zakim-robot
    @zakim-robot
    Oct 26 19:05
    [emplums] ahh love that article
  118. sophieschoice
    @sophieschoice
    Oct 26 19:05
    really good examples
  119. James Nurthen
    @jnurthen
    Oct 26 19:07
    @emplums when I open the popup menu NVDA reads "menu". I can then arrow through the menu items just like in any menu
  120. this is FF45 ESR
  121. zakim-robot
    @zakim-robot
    Oct 26 19:55
    [jiatyan] Does anyone have experience in developing accessible mobile apps? iOS or Android. I'm wondering what the most common gotcha's are.
  122. zakim-robot
    @zakim-robot
    Oct 26 22:58

    [karlgroves] @jiatyan the single most important thing to keep in mind with native mobile development is to use the standard UI kits for the platform. If, for some reason, you need to create a “custom” type of control, you should extend the built-in object(s) and not create one from scratch.

    Both iOS and Android platforms have pretty good a11y docs that should be easy to find with Google

  123. zakim-robot
    @zakim-robot
    Oct 26 23:06

    [karlgroves] @emplums

    "Why your startup should care about accessibility”

    Given the current litigation atmosphere, I’m really surprised more VCs aren’t paying better attention to a11y. For instance, the negative attention that Uber’s frequently gotten for its drivers not picking up riders with assistance animals has actually created market opportunities for accessible ride sharing.

    There are loads of segments ripe for disruption because of a lack of accessible options. Data Visualization is another example. After doing some research, none of the major players in that space offer accessible platforms and I know of one company starting significant efforts to make their Data Vis product accessible which has garnered a massive amount of interest from the Federal Government space

  124. [emplums] Oh hey @karlgroves I referenced one if your articles in my article :$
  125. zakim-robot
    @zakim-robot
    Oct 26 23:14
    [karlgroves] Thanks!
  126. zakim-robot
    @zakim-robot
    Oct 26 23:46
    [caesar] Does anybody have a good example of an accessible breadcrumb navigation?
  127. [jiatyan] Thanks, @karlgroves. I figure that it was the similar to "use standard HTML first", but was are there any additional things, like even if the develop uses a standard menu in iOS, does the developer have to do extra stuff to label this to the OS?
  128. [jiatyan] @caesar https://www.w3.org/WAI/tutorials/menus/multiple-ways/
  129. [caesar] @jiatyan Native API will usually start by trying to read a native element's defined text Label. You can set an additional accessibility Text (both iOS and Android support this). Focus order can be controlled in Android, but not iOS
  130. zakim-robot
    @zakim-robot
    Oct 26 23:52
    [jiatyan] good to know, thanks @caesar
  131. [caesar] At least last I checked that was the case (focus order in iOS). Can't stay if they have added it more recently (as of iOS 10)
  132. [caesar] btw there's also a #native-mobile channel