A11y Slacker archive 23/07/15

zakim-robot

Jul 23 02:05

[dylanb, a11y] @Haley Smith: looking for a fully accessible Android application https://play.google.com/store/apps/details?id=com.dequesystems.accessibility101
zakim-robot

Jul 23 02:05

[Haley Smith, a11y] i have this installed @dylanb!
zakim-robot

Jul 23 02:06

[dylanb, a11y] awesome!!
zakim-robot

Jul 23 02:06

[dylanb, a11y] we have an iOS one too
zakim-robot

Jul 23 02:06

[dylanb, a11y] https://itunes.apple.com/us/app/accessibility-101/id973198091
zakim-robot

Jul 23 02:07

[dylanb, a11y] sorry about the huge logo :disappointed:
zakim-robot

Jul 23 02:07

[dylanb, a11y] also, stay tuned for awesome upcoming free content on Android and iOS accessibility
zakim-robot

Jul 23 02:08

[dylanb, a11y] as a little spoiler, you can look at https://github.com/dequelabs/Accessibility-101-Android
zakim-robot

Jul 23 02:08

[dylanb, a11y] and https://github.com/dequelabs/Deque-University-for-iOS
zakim-robot

Jul 23 02:10

[dylanb, a11y] Shoot, I realize one of those repos is not public yet
zakim-robot

Jul 23 02:11

[dylanb, a11y] so I suppose that counts as a real teaser
stevefaulkner

Jul 23 02:36

This message was deleted
stevefaulkner

Jul 23 02:37

aria-live: the good the bad and the ugly - presentation by Russ Weakley http://www.slideshare.net/maxdesign/arialive-the-good-the-bad-and-the-ugly and related tests http://maxdesign.com.au/jobs/sample-accessibility/10-notifications/index.html
zakim-robot

Jul 23 03:54

[Francis Storr, a11y] If nothing else, that maxdesign page shows that there’s been improvements in a11y in (some combination of) OS X 10.10, Safari, and Chrome.
zakim-robot

Jul 23 04:04

[Francis Storr, a11y] Can anyone recommend accessible data viz/graph packages? We use a lot of bar charts and line graphs sometimes in canvas and other times in SVG. I realize both are fraught with poor accessibility problems. If there’s nothing good out there what workarounds can be used to make non-table-based data viz accessible.
StommePoes

Jul 23 07:25

Would be great if I could get this to work with Pidgin...
MichielBijl @MichielBijl waves at @StommePoes :D
stevefaulkner

Jul 23 08:27

@StommePoes https://github.com/w3c/a11ySlackers/tree/gh-pages#pidgin
zakim-robot

Jul 23 09:40

[zakim robot, a11y] #a11yslackers log for 22/07/15 http://w3c.github.io/a11ySlackers/archives/220715.html
stevefaulkner

10:33

Answer: Aria Role Attributes Becoming Depreciated in HTML5? http://webmasters.stackexchange.com/a/83392/54341
MichielBijl

10:37

“the rules defined in the WHATWG spec are way out of date and not maintained” positive as always ;)
MichielBijl

10:37

Not saying it ain't true though.
stevefaulkner

10:43

@MichielBijl not meant to be negative, only factual, this is the case, for the ARIA section. If you take a look at it (for example) use of default implicit semantics is a MUST NOT as per WHATWG, an author conformance requirement that has never been implemented in conformance checkers, also the browser HTML/acc implementation requirements are defined as implemented in http://www.w3.org/TR/html-aam-1.0/ , no longer in the W3C HTML spec (which normatively references HTML Accessibility API Mappings 1.0).
MichielBijl

10:47

You got that it was meant as a joke right?
MichielBijl

10:48

My track record for jokes on here isn't that great, so, just checking
stevefaulkner

10:49

sure, just got me spec editors hat, disabusement mode on this morn :-)
MichielBijl

10:49

Haha
IanPouncey

11:19

Is WHATWG working on anything interesting at the moment? I haven't been keeping track of what's going on there recently.
stevefaulkner

11:26

@IanPouncey https://github.com/whatwg
stevefaulkner

11:53

@dylanb Axe bug me thinks "Description
Ensures that lists are structured correctly
<ul> and <ol> lists must contain only <li> "
Can also contain: "Script-supporting elements" i.e <script> and <template> see under 'content model:' http://www.w3.org/TR/html5/grouping-content.html#the-ul-element
stevefaulkner

12:31

Note: you can access this channel via IRC https://github.com/w3c/a11ySlackers#connecting-to-gitter-over-irc (as well as gitter and slack)
morena

12:36

ha! I would like to access it from slack, but I am quite new with slack, will try
StommePoes

13:12

test
zakim-robot

13:12

[Michiel Bijl, a11y] @StommePoes: it works!
StommePoes

13:13

cool, dit is via pidgin
StommePoes

13:13

woops
MichielBijl

13:13

Nice :)
StommePoes

13:13

switching in my client between english and dutch...
StommePoes

13:14

So why sometimes you're via Zakim (I thought he died) and sometimes direct?
MichielBijl

13:14

Dat is jetzt no problemo?
MichielBijl

13:14

Uhm
MichielBijl

13:14

Well, I'm connected via IRC, but since I don't own a bouncer, I'm occasionally on Slack, too.
MichielBijl

13:15

Otherwise I'd miss a lot.
MichielBijl

13:15

Everythime my macbook goes to sleep/dimmed screen it disconnects IRC.
StommePoes

13:15

Ah. On the jabber protocol I get offline messages next time I sign i
StommePoes

13:15

in
StommePoes

13:16

So Slack is doing what normally Krijn Hoetmer's server would do :)
MichielBijl

13:16

Cool
MichielBijl

13:16

That is better…
MichielBijl

13:16

Ah well.
MichielBijl

13:33

stevefaulkner: I'm still undecided as to whether I should use details or the anchor solutions for the character information.
MichielBijl

13:33

Looking at AT and backwards compatibility.
MichielBijl

13:33

Since details isn't finalised yet
stevefaulkner

13:34

@MichielBijl twas only a suggestion as a pattern (disclosure) rather than saying use details/summary
dylanb

13:34

@stevefaulkner thanks for this :point_up: July 23, 2015 7:31 AM
stevefaulkner

13:35

@dylanb no problem came up as I was using axe chrome extension :-)
MichielBijl

13:36

@stevefaulkner you mean design wise? I'll stick with the anchor solution for now. I'll look into details some other time. This project has to be done by friday…
dylanb

13:37

I wish gitter had the "reactions" that Slack has and I wish Slack had the "point ups" that Gitter has
stevefaulkner

13:37

@MichielBijl yeah, sounds like a plan
MichielBijl

13:45

@dylan: +1, :big_smile:
MichielBijl

13:45

Or something like that.
mgifford

14:12

I'd love some thoughts on how to make CKEditor controls more discoverable https://dev.ckeditor.com/ticket/10015
MichielBijl

14:14

transform: scale(2)
deborahgu

14:21

on desktop, traditional way to make keyboard controls discoverable is to mark it right there on the control, as an option. Underlined letter next to the icon, with visual text that says "Show keyboard shortcuts on toolbar."
deborahgu

14:21

Optionally, having "show keyboard shortcuts" which brings up anaccessible dialog as a plain text link in the editor.
deborahgu

14:22

I know that the ideal that designers are always going for is simple and flat and as little text as possible, but all it takes is a little usability testing to show that sometimes telling people about something is the best way for them to know that it exists.
StommePoes

14:23

teh designers I work with don't want any text for most things. They want little indecipherable icons.
StommePoes

14:23

More and more sites I'm opening web developer tools just to inspect these things in hopes that they gave the thing a class name or id or something that clues me in on what might happen to me if I click it.
deborahgu

14:25

Yeah, they tell me they want usability and accessibility, and then I send them http://www.nngroup.com/articles/icon-usability/ pointing out what should be obvious (that if you invent a new icon users have never seen before for functionality on your site, they will not find the thing they are looking for and therefore they will not give you money), and suddenly they close ranks because they really like little indecipherable webfont icons without text.
StommePoes

14:26

or maybe I'm a terrible guesser. During a driving lesson I was asked what an icon seemingly made of multiple teardrop-shaped thingies was. I had no idea. Turned out it was seat warming. Back in my day such things didn't exist. ("I used to be with it, but then they changed what it was!")
deborahgu

14:26

The other day I had to go into web developer tools to figure out how to pay my credit card bill, @StommePoes. :( The webfont icons they were using for submit didn't happen to be rendering on my browser.
zakim-robot

14:26

[Winston Hearn, a11y] from what i know, there is very strong evidence that icons alone are a terrible pattern
StommePoes

14:27

My firefox on both Linux and Windows seems to not show icons fonts, I haven't figured out if it was because I had to turn off some font rendering in the OS or in FF to get ZoomText to work or some FF default I haven't found while crawling through t̶h̶e̶ ̶u̶n̶d̶e̶r̶g̶r̶o̶u̶n̶d̶ ̶t̶u̶n̶n̶e̶l̶s̶ the about:config
MichielBijl

14:28

@StommePoes: “yeah, laugh now, It'll happen to you, and you, and, yes, even you!”
StommePoes

14:28

twitter pated
MichielBijl

14:28

:D
MichielBijl

14:28

Greatest movie ever
deborahgu

14:28

No, it's not just you. Usability testing shows that people understand magnifying glass -- but ONLY if there is a text box attached to it, not the new pattern of magnifying glass on its own which expands a search box. I think people understand gear for settings, and shopping cart. Heart and star are not used consistently, nobody outside of tech understands hamburger menus.
StommePoes

14:29

MichielBijl you once repeated someone's question on why make something focusable if it's not just repeating input/link/button-ness, being forced to only let icons be seen is one of my answers... at the very least, while it doesn't help a lot of people, at least i can find out what the damn things do.
StommePoes

14:30

I was told by Jeff whatshisname on discourse that the heart icon "was a solved problem" (there was a debate open there on discourse about maybe switching to SVGs from font-awesome icons, after reading the "dyslexics might be loading custom fonts" article
StommePoes

14:30

)
StommePoes

14:30

So I never replied, but I'm constantly re-adding a custom stylesheet in my browsers so I can remember what the icons buttons do on sitepoint forums. There's title attributes, so I can just stuff those in an :after for myself.
MichielBijl

14:32

StommePoes you mean like show a label on focus?
MichielBijl

14:32

Can't seem to find the tweet…
StommePoes

14:34

For example if I can keep up with the guy writing our content (I can't, he has the energy of a damn puppy :), he does
StommePoes

14:34

whereas I try
StommePoes

14:34

<a href=somewhere><span font-awesome junk here></span></a>
StommePoes

14:34

<a href="somewhere"><span class="ac">freakin text</span><span font-awesome junk here></span></a>
StommePoes

14:34

a:focus .ac { be all like, visible n stuff;}
StommePoes

14:34

then I could
StommePoes

14:34

it's still mystery meat: users must interact to get the meaning
StommePoes

14:34

but at least you don't have to actually follow the link to figure out what it does.
deborahgu

14:35

that's a good idea, StommePoes. I should try it on github. Every time the right hand navigation bar collapses I sit there thinking, okay, angle brackets, squiggly line with dots, heart monitor... What do these mean again? And they don't show tooltips on keyboard focus, either. >:(
deborahgu

14:36

and I've never understood the motivation for collapsing them sometimes and expanding them other times. It is a responsive design, and there is plenty of screen real estate to keep them expanded.
MichielBijl

14:36

I feel you pain StommePoes
deborahgu

14:36

I for one will be really pleased when the web designer Zeitgeist moves away from minimalism and acknowledges that some people still need words.
StommePoes

14:36

If they did add title attributes, then with pure user css you can add text to all of them, and style them and everything... which is nice but my old opera 12 seems to "forget" them every so amount of time.
StommePoes

14:37

https://meta.discourse.org/t/are-icon-fonts-like-font-awesome-necessary/30953/11 (even when I add the id at the end as a hash, it doesn't seem to jump down, but...)
deborahgu

14:37

You're the second Opera 12 holdout besides me I've run into this week. fistbump
StommePoes

14:37

it's because I can navigate, I still dunno if Vivaldi has this
StommePoes

14:37

because now I can't open DMs in twitter anymore, they've recently changed something
StommePoes

14:37

i dunno what, but even in dev tools I can't get the DM box down from the negative top margin or whatever it's got now)
StommePoes

14:38

more and more stuff is breaking in Opera 12, I'm guessing here it's flex-box. Anyway, if vivaldi can bring back the stuff I liked in Opera 12, I'll switch.
MichielBijl

14:38

You should try Firefox 3.6; it's really fast.
StommePoes

14:39

(and I'm a mouse user, I just use it less than I could... mostly when I can't figure out how to do it with kb)
StommePoes

14:39

I was a cold-dead-hands FF 3.6 user
StommePoes

14:39

every new version of ff cost me so much time to reset all my settings... eventually I quit using it regularly, for that reason.
deborahgu

14:39

I've been trying Vivaldi but it is still missing some of my single key keystrokes
StommePoes

14:40

I stopped bookmarking and plugins and all that. Now I'm developing with Chrome because $work forced GMail on me (let the spyware run on the spybrowser) and Opera12 for social junk. Except now even social junk is breaking...
StommePoes

14:40

can you for example s-w navigate by headings and shift-(letters) around columns? I didn't do a whole lot more than that, but I notice how I miss it in all other browsers.
MichielBijl

14:41

Why not use Firefox?
MichielBijl

14:42

GMail works in Firefox right?
deborahgu

14:42

as I recall, you can still navigate by headings, or maybe it was the div nav in Opera; I didn't try around columns. but single key link to link was not there yet.
StommePoes

14:42

I don't want to pollute my other browsers with googley junk :)
deborahgu

14:42

Yeah, I have to use Gmail at work, and it works great in Firefox.
MichielBijl

14:42

Haha, good call.
deborahgu

14:42

ah see that's why I use a copy of portable Firefox. :-)
StommePoes

14:42

Yeah it's not about gmail not running in other browsers, it's just keeping like with like.
StommePoes

14:43

I mean, I still have FF. I test in it, and NoScript is lovely. It just stopped being my regular-use browser due to their "let's change ALL THE THINGS" hide-and-seek UX they were doing at the time (around FF3-4 or so...)
StommePoes

14:44

It's the best browser for Orca and NVDA too (not that I do enough in either)
MichielBijl

14:44

So it's Gmail running in Chrome, running in Chrome OS, in a VirtualBox, somewhere on the servers of BrowserStack and you connect via another VirtualBox on your machine to net get infected with it?
MichielBijl

14:45

I stopped using Firefox when they switched their chrome to “Let's make this ugly as fuck and not make it fit within the UI of the OS”
StommePoes

14:49

I don't care about browser chrome beauty, but constantly resetting my settings and wasting time searching for things that I had already spent time learning where they were was... well, a waste of my time.
StommePoes

14:50

And I like a single bar to do all my stuff: search or URL. But FF removed the ability to stick DuckDuckGo in there when they deprecated the keyword:url thingie in about.config. So I installed teh DDG plugin but every plugin you add to FF to fix it makes it that much bloatier and slower.
zakim-robot

14:50

[jitendra, a11y] http://teachingaccessibility.com/
StommePoes

14:50

Hi Jitendra!
MichielBijl

14:50

That too, but I appreciate consistency across my OS
deborahgu

14:50

Man, StommePoes, reading that discourse thread is really illuminating. For one thing, if icon fonts are solved problem for Dragon users,could somebody tell me the solution? Because I would really like to know it!
zakim-robot

14:51

[jitendra, a11y] Hi
deborahgu

14:51

But it also has that whole frustrating attitude that is so common that disabled users chose their adaptive tech so it's all their problem, and after all, it's not too much of a hardship for somebody who is already dealing with adaptive tech and disability and inaccessible websites to Web search until they find a thread like this one, and then install Stylish, and then play around with CSS, right? All people using adaptive tech can easily do that, right? If they can't modify their own CSS, what right do they have to install a font that enables them to read?
MichielBijl

14:53

I thought it was a package deal, get a disability, get a free course in CSS.
StommePoes

14:53

yup, exactly. I disagree with FF's current stance that for example exposing aria-landmark roles to keyboard should be a plugin. It should be in the normal browser
deborahgu

14:54

MichielBijl: rofl
StommePoes

14:54

@deborah that response of his sent me into a spiral, I didn't reply until just now. throws-up-hands icons here
MichielBijl

14:55

@StommePoes: (╯°□°）╯︵ ┻━┻
StommePoes

14:56

nah that's throwing tables
StommePoes

14:57

that's more... aggressive than just throwing up hands. I give up vs Hatebreed's Destroy Every Thing
StommePoes

14:57

(╯°□°）╯ just this perhaps...
MichielBijl

14:57

Just (╯°□°）╯ then :p
MichielBijl

14:57

Haha
deborahgu

14:59

(ʘ‿ʘ)ノ✿ Hold my flower
zakim-robot

15:02

[Katy Moe, a11y] Question: I’ve got a form which has error messages below required fields - “This field is required” - when you click the “Done” button. I’ve marked these up with role=“alert” and they announce fine on VoiceOver. How can I get the screen reader to announce the field labels as well when this live region is updated?
zakim-robot

15:04

[Katy Moe, a11y] (The reason I want this is that the copy I’ve been given just has “This field is required” as an error message for every field - a blind user won’t be able to tell which field is the required one as the only message they hear on submitting the form is “This field is required”)
MichielBijl

15:07

@Katy Moe: VO announces the label if you pause in an input right?
StommePoes

15:08

You'd think people'd want to hear whether the form was successful or not-- and if not, then what's not good.
rodneyrehm

15:08

is setting the validity of the input not enough? to engage the SR? https://developer.mozilla.org/en-US/docs/Web/API/ValidityState
zakim-robot

15:08

[Katy Moe, a11y] @Michiel Bijl: ah yes, so I could focus the field - this is what you’re saying?
zakim-robot

15:09

[Katy Moe, a11y] @stommepoes yes, there will also be an overall banner saying whether the form was successful
zakim-robot

15:09

[Katy Moe, a11y] @stommepoes in which case, maybe the information about which fields are required could be in that banner
zakim-robot

15:09

[Katy Moe, a11y] @rodneyrehm I can set the validity of the input, but the SR user won’t know what is invalid about it
StommePoes

15:10

I would put success/fail in an alert role, with the fail having the assumption that somewhere, someone's not right. Then you can go through the form fields and, if the error is associated with the input (either via inside the label or via aria-describedby) you'd get it then too
rodneyrehm

15:10

the "validity state” is a simple text message, is that not read?
StommePoes

15:10

required is a pretty "stupid" validity check, mostly it only checks for empty string or not.
StommePoes

15:11

if the error is "x is required" because it's filled in but not filled in correctly, it's horrible error msg :) and then indeed nobody is sure what exactly is wrong.
zakim-robot

15:11

[Katy Moe, a11y] @stommepos yes, of course, we will display the right error message depending on what exactly is wrong
stevefaulkner

15:12

@kmoe this may be helpful: Example form validation fedback: http://files.paciellogroup.com/training/commonsamples/formvalidator_demo/html/FormValidatorDemo2.html
StommePoes

15:12

But one nice thing with (aria-)required attributes is you can get that when you blur the input or see/hear that it's a required field even before you get to the submit.
zakim-robot

15:12

[Katy Moe, a11y] @stevefaulkner thanks, I’ll check that out. I’ve also referred to http://pauljadam.com/demos/aria-alert-validation.html
zakim-robot

15:13

[Katy Moe, a11y] @stommepoes @rodneyrehm to clarify, we’re not using HTML validation here - it has to happen through JavaScript due to the multi-part nature of the form
zakim-robot

15:13

[Katy Moe, a11y] but we do of course add aria-required, aria-invalid where appropriate
StommePoes

15:14

That helps, in that they're quite immediate.
StommePoes

15:14

for users.
zakim-robot

15:14

[Katy Moe, a11y] my question (in a more general sense) was just: is it possible to get a live region to announce additional information (from somewhere else on the page) when it is updated?
rodneyrehm

15:14

yeah, I wasn’t particularly clear about what I meant. I actually meant using elem.setCustomValidity(“Error Message”) to inform the <input> of its state
zakim-robot

15:14

[Katy Moe, a11y] @rodneyrehm ah, I haven’t used that before - let me check it out
rodneyrehm

15:15

That allows pure JS validation, but using the HTML5 validation interface - with :invalid and stuff working, <forms> not submitting etc.
rodneyrehm

15:16

To actually visualize the error I’ve always used an addition span element, though, because the browser’s default visualization of these error bubbles kinda sucks…
zakim-robot

15:16

[Katy Moe, a11y] very interesting, thank you! I’ll have to find out whether this is appropriate for our app
rodneyrehm

15:17

that way you can also use the HTML5 validation stuff (required, pattern, …) and add your custom JS validation on top, simply hooked into the interface provided by HTML5
StommePoes

15:18

rodney does that work better than adding the novalidate to the form so you can state, for example, to only use some custom :invalid input styles on blur rather than right away?
rodneyrehm

15:18

some examples for what I just said can be found here https://developer.mozilla.org/en-US/docs/Web/Guide/HTML/HTML5/Constraint_validation
StommePoes

15:19

I once started with some html5 validation because I liked being able to enhance stuff with CSS using :valid/:invalid etc but it was telling people they messed up before they even got the chance to get through
rodneyrehm

15:19

:focus:invalid {} works
StommePoes

15:19

on a required-but-empty input it gets invalid right away
rodneyrehm

15:19

but I’ve had to hack a .unvisited class onto every input, to prevent :invalid before the user focused the element for the first time
StommePoes

15:20

ah ok. Yeah, it bugs me being told I suck before I even get the chance to suck :P
zakim-robot

15:20

[Katy Moe, a11y] yeah, our form has rather different state management than in the HTML form validation model or the Angular model (ng-dirty etc) due to using React
rodneyrehm

15:20

yep. as I said, there’s a hackaround™ for that
StommePoes

15:21

oh, interesting, react. How is it otherwise/mostly working out for you guys?
rodneyrehm

15:21

angular/react make me leave the discussion - I have no clue what they’re doing, sorry.
zakim-robot

15:22

[Katy Moe, a11y] React is actually very interesting when it comes to live regions, as it’s constantly adding and removing from the DOM as state updates
zakim-robot

15:22

[Katy Moe, a11y] we’ve found our React app great for accessibility so far (as far as I’ve tested, anyway)
StommePoes

15:23

It probably works quite well with AT who don't make buffers/DOM copies, sounds like.
zakim-robot

15:23

[Katy Moe, a11y] we may have some complex drag and drop stuff happening, though, so I’ll probably be back with questions about that!
StommePoes

15:23

But how is i with NVDA and JAWS?
zakim-robot

15:24

[Katy Moe, a11y] @stommepoes NVDA is fine, haven’t tested JAWS yet
StommePoes

15:24

cool to hear
zakim-robot

15:24

[Katy Moe, a11y] Do you have any resources for finding out about how NVDA/JAWS make buffers/DOM copies?
zakim-robot

15:26

[Katy Moe, a11y] I would love to do some more research into how React interacts with AT
zakim-robot

15:27

[Katy Moe, a11y] I get the feeling it’s not a problem many people have tried to solve yet (my question in #react on Slack has gone unanswered… :p)
StommePoes

15:28

I dunno if there is much research about React and AT, other than the whole flipboard bit (that was canvas specifically anyway), but about teh virtual buffers there ought to be some stuff out there, esp when they started making them update themselves when ajax started changing bits of pages without refresh
zakim-robot

15:30

[Katy Moe, a11y] I’ve started an #accessibility channel in the Reactiflux Slack community (the biggest React Slack community), so let’s see what happens :simple_smile:
zakim-robot

15:30

[Katy Moe, a11y] As for Flipboard, that was unfortunate…
StommePoes

15:31

I know on WebAIM people making various single-page "apps" have had trouble with alerting users, even with live regions. But it was hard to tell what were JAWS bugs and what were deeper problems.
zakim-robot

15:33

[Katy Moe, a11y] I’m also interested in how it’ll play with Windows Speech Recognition, which is also bad with non-refresh repaints
zakim-robot

15:33

[Katy Moe, a11y] In fact, I’m going to test that right now...
MichielBijl

15:38

Woohoo! I'm now officially an Invited Expert on the Protocols and Formats Working Group :D I'll work on WAI-ARIA code examples, and more.
StommePoes

15:39

awesome
rodneyrehm

15:39

grats
MichielBijl

15:39

Ahw yeah :D
MichielBijl

15:41

Guess nothing really changes, but at least now I'll have access to the actions and issues we discuss during teleons :p
deborahgu

15:42

woo MichielBijl
deborahgu

15:42

I will come whine on you when DPUB wants PF to do stuff for us.
stevefaulkner

15:42

@MichielBijl welcome!
MichielBijl

15:43

Haha, that's fine ;)
MichielBijl

15:43

Thank you stevefaulkner :)
zakim-robot

15:54

[Katy Moe, a11y] Okay, I’ve tested my React app with Windows Speech Recognition, and it seems to work fine, except that ‘Show Numbers’ seems occasionally not to get updated properly - but if you speak the label of a button, it’ll still click it
deborahgu

15:55

Katy you are a rock star for testing with speech. #happyDragonGirl
zakim-robot

15:55

[Katy Moe, a11y] @deborahgu if only I could afford Dragon!
StommePoes

15:56

KatyMoe++
MichielBijl

16:00

Blast from the Passt: “How does making a site more user friendly result in reducing accessibility to these users? I hope you are not thinking that an author is going to toss in some jokes to make his content more user friendly?”—Anne Pemberton
zakim-robot

16:09

[Francis Storr, a11y] Doesn’t Dragon have an entry-level product for $79 or thereabouts? I know the professional version is super-expensive. Is there a significant difference in the products that means that the entry-level product is no good for web a11y testing?
deborahgu

16:10

you only really need to use the professional level product if you are using it all the time
deborahgu

16:11

it gets you advanced scripting. I'm not sure if you need the professional or preferred versions to run the external scripting products such as dragonfly or Vocola; I have professional already so I never looked into it. But for testing, I would assume Home would work fine. Still, $79 is still money.
stevefaulkner

16:13

Eric Wright is the dude for Dragon info - How (Not) To Train Your Dragon https://ewaccess.wordpress.com/2013/11/24/how-not-to-train-your-dragon/
Which Version of Dragon Should You Buy? https://ewaccess.wordpress.com/2014/10/19/which-version-of-dragon-should-you-buy-2/
zakim-robot

16:13

[Katy Moe, a11y] Oh, I didn’t realise it was that cheap - somehow I got the impression that it was more than that.
zakim-robot

16:14

[Katy Moe, a11y] I might make a request then! Though I’d have to run it in a VM or on my home computer, as we use OSX exclusively here.
deborahgu

16:14

JAWS is the ludicrously priced one.
deborahgu

16:15

And JAWS plus J-say is completely out of my range. (Allowing JAWS and Dragon to play nicely together.) None of the screen readers work brilliantly with Dragon; in my copious free time I keep meaning to see if I can play with NVDA and Vocola to make a fix.
stevefaulkner

16:15

Also worth a read - Assistive Technology Experiment: Dragon NaturallySpeaking http://webaim.org/blog/at-experiment-dragon/
stevefaulkner

16:17

I also wrote a few dragon related rants http://www.paciellogroup.com/blog/category/dragon-naturallyspeaking/
deborahgu

16:17

I have issues with that webaim post,
deborahgu

16:18

much as I think a JAWS or NVDA user would have issues with a post I were to make about limitations I found when testing with them
deborahgu

16:18

for example, it leads you to believe that Dragon and Firefox don't play well together. When they play great together. Now Dragon + Chrome, that's another story.
zakim-robot

16:19

[Katy Moe, a11y] Yes, I noticed he seemed to dismiss Firefox without finding out whether other Dragon users had had the same experience - in my previous role we had Dragon testers and they often used Firefox.
stevefaulkner

16:19

@deborahgu I have issues with Dragon as it does not support web standards or make use of the accessibility APIs in browsers ;-)
deborahgu

16:22

Oh, I share those issues entirely, stevefaulkner! Nuance is pretty terrible about accessibility, and they are basically resting on the laurels of the old Dragon Systems, maybe L&H work. I continue to be appalled at their attitude about accessibility, like of course you can't install the product hands-free, who would do that? (Nevermind that screen readers have figured out how to have an installation process that doesn't assume you are sighted.) And of course advanced scripting isn't accessible, why would it be?
deborahgu

16:23

But at the same time, Dragon is an incredibly usable product. I am a low-kb/mouse, sometimes hands-free, not that long ago entirely hands-free developer, and it's fine. Maybe I think too highly of it because I remember the old DragonDictate for PC one-word-at-a-time days, but it is an incredibly useful product. If I could suddenly type again I would not give it up.
deborahgu

16:23

Nuance needs to get their crap together and start understanding accessibility and web standards, but people talk about Dragon like... Well, they talk about it as the NaturallySpeaking and DragonDictate for the Mac are the same product. As opposed to NaturallySpeaking being a product that allows incredibly powerful hands-free control, scripting, built-in programming, etc.
stevefaulkner

16:24

@deborahgu :+1:
deborahgu

16:24

And -- to come back to w3c accessibility -- as long as web developers write to web standards, it can handle almost anything on the web that doesn't require WAI-ARIA functionality.
deborahgu

16:24

:D
stevefaulkner

16:27

Right, on that, I got told recently by an influential corporate standards bod, that my stance advocating for built-in (native HTML) vs bolt-on (ARIA) is backwards
deborahgu @deborahgu cries
deborahgu

16:28

goes right along with how backwards it is to advocate for browser built-ins instead of community contributed extensions
deborahgu

16:29

I have been kvetching that a long conversation was just had about how to build a library of polyfill workarounds for a huge existing standard because browser manufacturers have made it clear they are not going to implement the standard itself.
deborahgu

16:30

(I worried that that was too identifiable as I was writing it, and then I realized it described so many conversations!)
stevefaulkner

16:34

goes right along with how backwards it is to advocate for browser built-ins instead of community contributed extensions

Thats just an abrogation of responsibility
StommePoes

17:02

het deborah, about jsay
StommePoes

17:03

if someone were to make an open-source something that basically allowed an SR-user to also use dictation... would there be a number of people who'd want that?
StommePoes

17:03

I know some folks who want to make something but would need to do a kickstarter and worry than the market for such things might be too small to succeed
deborahgu

17:04

I absolutely know that there would be a number of people who'd want that. Would it be enough to justify a kick starter, though? I wish I could say yes, but I really don't know.
deborahgu

17:05

I feel like an open source nvda -- dictation tie in might be best produced with grant money, unfortunately. As much as I would like to say otherwise.
StommePoes

17:07

well, one thought someone in our group had was, that there were a number of institutions who were stil buying jaws licenses because of the jsay issue
StommePoes

17:07

so one hope was maybe those institutions would pitch in
StommePoes

17:07

but we're not sure how to do it
StommePoes

17:08

we did a kickstarter for NVDA remote access
StommePoes

17:08

and it turned out most of the donations were from IT-professional blinks, which means they were working, had the money and the need
StommePoes

17:08

that worked out really well
deborahgu

17:08

oh, that was your group? That was so cool.
StommePoes

17:08

but we worry that an nvda-say-ish thing wouldn't get the same benefit, as folks with multi-disabilities are less likely to be working in for example IR
StommePoes

17:09

that was 3MT
StommePoes

17:09

3 mouse tech
StommePoes

17:09

mostly 2 guys in it, but yeah
deborahgu

17:09

I bet it would be worth reaching out. Still, I know a lot more blind accessibility rofessionalsthan I know Dragon accessibility professionals.
StommePoes

17:09

i'm the token sightie of the group yet we still have have nice design because I have terrible design sense :)
StommePoes

17:09

yeah...
StommePoes

17:10

so we figure, it needs a decent name to call it beforehand, and we need to see if it could be a viable kickstarter, and our group doesn't really know many dragon/dictation users
StommePoes

17:11

in fact you and eric are actually the only 2 I know
zakim-robot

17:11

[Francis Storr, a11y] I’m hopeful that the refreshed Section 508 provision that @Steve Faulkner mentions here (http://www.paciellogroup.com/blog/2014/09/when-the-refresh-comes/|www.paciellogroup.com/blog/2014/09/when-the-refresh-comes/) happens and Nuance is essentially forced to get their act together
StommePoes

17:12

well, nuance doesn't see themselves as an AT vendor
StommePoes

17:12

they see themselves as serving, like, radiologists and judges
StommePoes

17:12

even tho i'
StommePoes

17:12

s used as AT all over the place
deborahgu

17:12

Besides me and Eric, the other group I would reach out to would be the programming by voice crowd, the small but very active crowd at https://groups.yahoo.com/neo/groups/VoiceCoder/
MichielBijl

17:13

So 508 needs a better definition of what AT is?
StommePoes

17:13

it's either that, or some upstart like NVDA comes along and re-invents all that work to make a Flossy dictation thing
StommePoes

17:13

deborah thanks
deborahgu

17:13

no problem. If you want, write something up and I can email it to that group, if you want me to do that for you.
StommePoes

17:13

there's a guy we know who was kinda working on jsay and he apparently knows many blinks+dictation folks, but it sounded like that wasn't so many
StommePoes

17:13

but I wonder if I should have Gonz contact you?
deborahgu

17:14

oh, yeah, sure. I met him once at an FSF meeting, I think
StommePoes

17:14

haha you might have. he either pissed you off or made you laugh
StommePoes

17:14

back in the FSF days he was even more volatile than now
StommePoes

17:17

I'll let him and the group know, if people think it could get enough support then it might become a go, which would be cool.
StommePoes

17:18

Michiel: I thought AT vendors called themselves that... whereas, Nuance simply doesn't.
StommePoes

17:18

Michiel and for example, does Apple consider Siri AT? It's supposedly even based on Nuance tech.
MichielBijl

17:19

Hmm, good question.
deborahgu

17:20

Yeah, it is. And no, I don't think they do, although it works as assistive tech and I use it as such. But if it were AT, it would be more deterministic and less chaotic-to-seem-human.
StommePoes

17:20

hehe
StommePoes

17:20

<scottish accent>Siri, doe ye ayven oonderstand me speekin'?</scottish accent>
StommePoes

17:27

Hm, I've never used teh Windows speach rec thing either, didn't know that was built in (sounds like Narrator-to-JAWS kinda thing)...
StommePoes

17:27

is it really usable?
StommePoes

17:28

is it something we should focus on instead of Dragon if any SR+Dragon stuff required Dragon Pro?
MichielBijl

17:29

StommePoes: do you work with Git at all?
StommePoes

17:30

I used to, back when I was doing stuff for DuckDuckGo, and I still have repos, but because my current job uses Mercurial (same commands do totally different things!), my Git's damn rusty.
StommePoes

17:30

but, yeah, at a very basic level (pushing, pulling, resolving conflicts in vim)
MichielBijl

17:31

Aha, just thought you'd like to know I have had git poes mapped to git push for like forever :p
MichielBijl

17:31

Makes using Git more fun anyway
StommePoes

17:31

haha
StommePoes

17:31

sounds like a kick tho
StommePoes

17:32

if nobody knew what "poes" meant, it sure sounds like a sound you'd make as you were kicking something you hated
StommePoes

17:32

"puf"
StommePoes

17:33

mercurial seems to have some aliases built in... I often mistype hg comit and it still sees the com part...
zakim-robot

17:50

[Francis Storr, a11y] IIRC, there’s voice recognition built into OSX. I think I remember trying it out a while ago.
deborahgu

17:51

StommePoes: WSR is usable for casual use, is my understanding, but you still need dragon for heavy use. Or so I've read.
zakim-robot

17:52

[Francis Storr, a11y] I have a copy of Dragon Professional on my Windows machine but found it incredibly difficult to get to grips with. One day I’ll go back and spend some serious time with it.
deborahgu

17:52

The OSX stuff is limited; I've been trying it and it's better than nothing? Almost no command-and=control (though Mac DD is not much better; there are some OS problems there).
StommePoes

17:54

would also be cool if there was good switch-control+dictation+SR stuff somewhere, and the first place I would expect it to appear would be on the mac
StommePoes

17:54

Christopher said he can just plug his switch control in any mac and it Just Works; no driver installation, no nothing
StommePoes

17:54

that sounds cool
StommePoes

17:55

like, general AT-integration stuff, even though it's generally just a bunch of separate companies making their own thing
StommePoes

17:56

Francis- I heard it takes time to get Dragon working nicely, but funny enough I was on a bus after CSUN and met some random lady and she said she used dragon and it was surprisingly good for her out of the box, had it "trained" in under a month for her emails and stuff (not programming tho)
deborahgu

17:57

except mac is terrible for SR, StommePoes, is the problem
StommePoes

17:58

you mean SR+dictation?
StommePoes

17:58

because the blinks I know, despite teh bugs, love VO
StommePoes

17:58

in a general way
deborahgu

17:58

Heh, sorry, you meant screen reader, I read speech recognition
StommePoes

17:58

Ah I meant screen reader
StommePoes

17:58

yeah
deborahgu

17:58

:D
StommePoes

17:59

heh, do they have anything builtin besides the Siri bit? Any analogy to WSR?
deborahgu

17:59

yeah, but it's not very usable for things
StommePoes

17:59

maybe Apple has simply neglected it due to Windows still grabbing a majority of businessy users.
StommePoes

18:00

Anyway Gonz said they were looking at WSR for now and I've passed on your info to the group, see what they think.
deborahgu

18:00

nah. Tim Cook has been pretty clear he thinks accesibility means screen readers.
deborahgu

18:00

cool, ta.
StommePoes

18:00

Lots of people think that.
deborahgu

18:00

sadly
StommePoes

18:01

I've been thinking of maybe giving an a11y-ish-something talk at a meetup, but then focussing more on people like Chris Hills than blinks
StommePoes

18:01

jsut because I suspect a lot of devs around here have at least heard of screen readers, even if they're still not quite sure what those are
StommePoes

18:01

I remember when those were mysterious black boxes , when I started front-end
StommePoes

18:01

we heard teh rumourz about these things... these screen reader things...
StommePoes

18:02

But for me now it's stuff like chin cups and switch controls that are the mysterious black boxes
StommePoes

18:02

I'd like to know more without actually purchasing such a thing, esp since I can't afford a mac anyway
StommePoes

18:03

Dragon too actually
zakim-robot

18:55

[sterling, a11y] Does anyone know what events cause virtual buffer to refresh in VoiceOver QuickNav and if it's possible to emit it somehow? I'm working on a project that is a narrated slideshow where an HTML slide is updated with AJAX. If I'm reading the content with QuickNav the buffer isn't updated when the content of the DOM is updated.
MichielBijl

19:08

Do you mean the webrotor?
zakim-robot

19:22

[sterling, a11y] Here's a better description of QuickNav. It's like the Virtual PC Cursor in JAWS. https://www.apple.com/voiceover/info/guide/_1133.html
stevefaulkner

19:41

FYI VO does not have a virtual buffer AFAIK, its a Windows SR only paradigm
stevefaulkner

19:42

@sterling it may be that the browser accessibility tree is not updating when the DOM updates
zakim-robot

19:49

[sterling, a11y] Interesting. Is there a way to co-ordinate the update of the accessibility tree and the DOM?
stevefaulkner

20:06

@sterling that's internal to the browser, what browser are you using?
zakim-robot

20:06

[sterling, a11y] Safari
stevefaulkner

20:07

OK, can you provide a code example that exhibits the issue?
zakim-robot

20:10

[sterling, a11y] Let me see. I was just about to say I think they are all internal to our University partners' log in pages, but I think they recently put one up on a marketing site.
zakim-robot

20:25

[Devon Persing, a11y] for any poor soul testing with JAWS 16, it looks like there’s a regression where the most recent version doesn’t honor aria-live announcements in IE 11 or IE 10 (possibly others, but I haven’t tested back further) on Windows 7. Possibly also on Windows 8. Works fine in Firefox!
zakim-robot

21:24

[Marcy Sutton, a11y] Yikes
aardrian

21:47

Does anyone here have experience with video.js ("the open source video player"). Looks promising as it leans on HTML5 <video> and the example has captions built in: videojs.com
lacymorrow

22:02

It's been around quite awhile, I've used it in a few projects with little issues.
zakim-robot

22:23

[stevenlambert, a11y] Hello everyone. I was hoping someone here would know if button order (the primary action button on the left [a la Windows], or primary action button on the right [a la Mac]) made a difference in terms of accessibility. For users who use screen readers, do they prefer having to read all the buttons before getting to the primary action button, or do they prefer getting to the primary action button first, then other actions if they need it? The same question can be asked about users who use only the keyboard: would they prefer the primary action button first for less key strokes or does it even matter?
zakim-robot

22:32

[Marcy Sutton, a11y] Might be a good question for #a11y-ux
zakim-robot

22:33

[stevenlambert, a11y] Thanks, I’ll also ask it there.
fstorr

22:49

@aardrian I was asking about accessible video players in the Slack room a few months ago and @karlgroves said he liked it, which was good enough for me. FWIW I tried following the video.js github account for a week or so and it's noisy. Lots of work and change going on with the code.
aardrian

23:48

@lacymorrow @fstorr Thanks. Was evaluating options, figured I'd try it on a personal project first. If @karlgroves says he likes it, that may be good enough for me.
Scratch2k

00:54

Hi everyone, I have a quick question about date pickers, we are using the non accessible jQueryUI datepicker, but also provide an alternate way to enter the date which has type="date" and a placeholder, is it a WCAG AA fail is the datepicker is not accessible with keyboard if there is an alternate method to enter the date i.e. te native input/date field (which is actually easier to use with a keyboard anyway)?

Gitter Channel Archive 23rd of July 2015