Archive index

A11y Slackers Gitter Channel Archive 28th of March 2017

What fresh hell is THIS now? - Patrick Lauke
  1. zakim-robot
    @zakim-robot
    Mar 28 01:17
    [elliottregan] @karlgroves @backwardok I’ve never done this before, but I’d love to brainstorm about it. My first approach would probably be to read a list of all the visible markers out to the user, then provide the address and other information if the users selects one. If that’s not possible with the map’s markup, then you might need to build some hidden, alternative DOM content.
  2. [backwardok] I also kind of wonder, since addresses are only useful if the user knows the area well, whether or not there should be some other kind of grouping that provides extra meaning
  3. [backwardok] but that requires a fair of knowledge of the data set you’re using
  4. zakim-robot
    @zakim-robot
    Mar 28 01:23
    [backwardok] trying to poke around google maps’s api to see if there’s a way to get more information about a general location (like neighborhood)
  5. [elliottregan] I’d treat that the same as with visual users. Addresses are how we’ve been doing it for decades, so I assume that is at least a good base to start from. That’s not to say there aren’t better ways of conveying location, but there is only so much information you can ascertain from a map. Crosstreets are helpful in large cities, as are neighborhoods. Proximity to landmarks and stuff like that, but I don’t think those data need to me added just for screen readers.
  6. zakim-robot
    @zakim-robot
    Mar 28 01:32
    [backwardok] i agree that addresses are also useful, but i don’t think it’s a sufficient replacement for what that marker is representing on a map - as a visual user, i can tell that there are some markers that are clustered together, or that they’re in a particular neighborhood, or that they’re near something that i recognize
  7. [backwardok] if i’m searching in an area that i’m not familiar with, the address wouldn’t help me very much
  8. [elliottregan] ah, that’s a good point
  9. zakim-robot
    @zakim-robot
    Mar 28 01:40
    [cordelia] Touch interactions/explorations would help with discerning clusters, but with keyboard input it’s much trickier
  10. [elliottregan] What helps me is to break the information down into a sentence, then figure out how to best organize that in DOM content. ex) "There are [number] [map markers: coffee shops, points of interest, etc.] within [8 blocks, 20min walk] of [point of reference: address, cross streets, neighborhood]."
  11. zakim-robot
    @zakim-robot
    Mar 28 01:45
    [elliottregan] You just need to make sure that your [point of reference] can be understood.
  12. [elliottregan] To be honest, I have no idea how a blind person would navigate an unfamiliar area.
  13. zakim-robot
    @zakim-robot
    Mar 28 03:25
    [caesar] I wonder if someone can take "what3words" and make an accessible map solution out of it... https://what3words.com/
  14. Serge Huijben
    @thehuijb
    Mar 28 07:49
    looking for best practise, should I include aria-label/aria-labelledby on my <article role="article">?
  15. zakim-robot
    @zakim-robot
    Mar 28 09:14
    [enwin] I was looking at http://www.html5accessibility.com and the support of html5 markup on IE11. Most of the HTML5 new element are marked as not mapped. Does it means that I have to keep the role attribute on those elements to be sure they are mapped or ATs do the work for me?
  16. zakim-robot
    @zakim-robot
    Mar 28 12:34
    [michiel] enwin: if HTML elements aren’t mapped you’ll have to manually add the appropriate ARIA attributes (role, aria-*).
  17. zakim-robot
    @zakim-robot
    Mar 28 15:39
    [jbuller] A query from a colleague: “What do you think about embedding documents in PowerPoint from an accessibility point of view?”
    The object doesn’t appear in Outline view. It can be given alt text which can be read by Voiceover. in editing mode. I’ve not found how to read slides in Slideshow mode.
  18. zakim-robot
    @zakim-robot
    Mar 28 16:45
    [albert] @karlgroves @backwardok no clue regarding google maps, but the beauty of open street maps is that its open source and can be made accessible. steps on soap box open source is better for everyone. especially in this case.
  19. [albert] there are also fragmentations in osm that are working towards physical a11y. wheelmaps based in germany is first one to pop into my head. have never, not once, seen anything like it in the google maps community.
  20. zakim-robot
    @zakim-robot
    Mar 28 16:55
    [karlgroves] @albert unfortunately I need to fix a google map. If this was from scratch I’d start with Leaflet, though I still haven’t seen the markers be accessible, I think it’d probably be less janky to hack if needed
  21. [albert] @karlgroves can i see said map?
  22. [karlgroves] Nope. Client map. ;)
  23. [albert] welp. which parts are inaccessible? just the markers? i know you can use custom overlays to get around that. at least for keyboards accessibility
  24. [karlgroves] Currently all interactive bits of Google maps are inaccessible. Sorry to the Googlers here, but the maps team really doesn’t appear to give a damn about a11y, because they could easily provide access to this stuff.

    From a “hacking it” perspective, it is pretty easy (though a bit janky) to fix the pan/ zoom functionality and the default controls

    Getting to the bits necessary to fix markers is another thing altogether

  25. zakim-robot
    @zakim-robot
    Mar 28 17:00
    [khutchins] I’ve been playing around with creating an accessible map using Esri’s JSAPI. At this point I’ve been working to duplicate (and extend a bit) the behavior google maps has where you can navigate the map content using a search box. This is just a quick demo app but I think the conversation around how to make a map accessible is fascinating and I think there is definitely work we can do to make it a better experience for all users. https://github.com/kellyhutchins/a11y-map
  26. zakim-robot
    @zakim-robot
    Mar 28 20:32
    [alice] @karlgroves I take it this is an embedded map?
  27. [karlgroves] yes
  28. [alice] Is maps.google.com any better?
  29. [alice] (just figuring out who to bug - not that it'll help you in the short term)
  30. [karlgroves] No. Bad & different. ;)
  31. [karlgroves] Maps.google.com: <body tabindex=”-1” …> (trollface emoji)
  32. [karlgroves] closes chrome devtools and walks away
  33. zakim-robot
    @zakim-robot
    Mar 28 20:38
    [karlgroves] FWIW, at least the zoom controls and stuff use real <button> elements and aria-label which is not the case on embedded maps
  34. [alice] maps.google.com seems at least fairly keyboard operable
  35. [karlgroves] Agreed. But on maps.google.com they still don’t have any keyboard access to markers - which seems super trivial to fix
  36. [alice] Hm - I can tab through the markers on a search result
  37. [alice] What markers do you mean?
  38. [karlgroves] Inside the map. The “pins” for the locations that matched your results
  39. [karlgroves] Which is weird because once you have a result open you can tab through it fine
  40. [alice] If you tab through the list (let's not discuss that we're tabbing through a list) it highlights each marker in turn
  41. [alice] Then I can press enter to open the full listing
  42. [karlgroves] yeah, when the panel is showing
  43. [alice] highlights the markers on the map, I mean
  44. [alice] I see
  45. zakim-robot
    @zakim-robot
    Mar 28 20:44
    [karlgroves] In my case, a customer has an embedded map. The default pan/zoom controls aren’t accessible (whereas they are on maps.google.com) and the pins/ markers aren’t accessible either.
  46. [alice] It works well when the sidebar is open at least, and has quite sophisticated keyboard controls for panning and exploring the map
  47. [alice] Yeah let me check up on the embedded map thin
  48. [karlgroves] Thanks