Gitter Channel Archive 9th of December 2015
Gitter Channel Archive 9th of December 2015
What fresh hell is THIS now? - Patrick Lauke
<input type="number">) is exposed as four individual elements in the form controls list; once with the calculated name, as simply “stepper” and twice as “button” without any name. Is this expected? How do other AT handle this?
<label for="name">Name</label> <input type="text" name="name" id="name" autocomplete="off" aria-required="true">
<label for="name" id="name-label">Name</label> <input type="text" name="name" id="name" autocomplete="off" aria-required="true" aria-labelledby="name-label">
for="idName"
<label for=“foo”> where foo matches an id of a form field.
[karlgroves] @callumacrae:
Dude.
<label class="label--bubble label--bubble label--error error-name" data-js-tracking="error--name-field--homepage--form" aria-live="polite" for="name" style="visibility: hidden; display: none;"></label>
[karlgroves] New post by @stevef
"ARIA allows developers to re-invent and extend native HTML features in meaningful ways. But like all bolt-on technologies its features are brittle compared to its built-in counterparts."
http://html5doctor.com/html-developers-please-consider/
[karlgroves] Not sure I understand the question. HTML5 references ARIA roles, states, and properties as valid.
ARIA’s purpose is to bridge gaps between the current UI/ Widget paradigms in use on the web and the existing platform features. The example I use a lot are tree menus. There are no elements in HTML that can create a tree menu widget. However, people have been doing them for years, using divs and spans or (hopefully) properly structured HTML lists along with JS. Using ARIA, you can apply roles, states, and properties to these things and allow AT to communicate to the user about what the widget is and what it is doing.
button's with div's, point is you can.
[karlgroves] @powrsurg can you do me a favor? Try TPG’s color contrast analyzer on that. https://www.paciellogroup.com/resources/
I’d be interested in knowing how that works in your use case
[blavalla] @karlgroves, the first thing that came to mind was this:
1.) Add the image into a canvas and get the image data, store this in some variable.
2.) Draw the text via canvas onto that same image, and get the image data again.
3.) Compare the two and throw out any image data that is unchanged.
4.) Average the color of the remaining pixels, and compare.
It definitely won't be perfect, but it'll give you a decent ballpark estimate.
