W3C home > Mailing lists > Public > wai-xtech@w3.org > March 2012

Re: [css3-values] accessibility review

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 30 Mar 2012 11:13:09 -0700
Message-ID: <CAAWBYDCg=_ojmpyDeyu4ySMnCo_jvPD+JMd2gpyxq+q_DQ6sDQ@mail.gmail.com>
To: Cynthia Shelly <cyns@microsoft.com>
Cc: "www-style@w3.org" <www-style@w3.org>, "wai-xtech@w3.org" <wai-xtech@w3.org>
On Fri, Mar 30, 2012 at 10:54 AM, Cynthia Shelly <cyns@microsoft.com> wrote:
> Hello CSS working group,
>
> I’ve done a review of this spec for accessibility issues, but have not yet
> discussed it with PF.  This review does not constitute PF consensus.  I’ve
> copied the wai-xtech alias to allow PF to add any additional input.
>
> http://www.w3.org/TR/css3-values/
>
> 5.1.1
> Are ‘ex’ ‘ch’ ‘rem’ supported in any existing browsers?  These all seem fine
> for accessibility, but I’m curious why ‘ex’ and ‘ch’ are needed in addition
> to ‘em’?  don’t they do basically the same thing a using fractional ‘em’
> units?

No, the size of an 'ex' or 'ch' varies across fonts.  'ex' lets you
size things relative to the height of a lowercase letter, while 'ch'
is mainly meant for setting the widths of things that contain
monospace fonts.


> 5.1.2
> Are ‘ vw’ ‘vh’ ‘vmin’ supported in any existing browsers?
> Is there a way to specify a minimum absolute size for these?  It seems like
> this could easily get too small to read.
> It does, however, seem useful for scalable/zooming UI

The v* units are currently supported in IE, and a patch applying them
to WebKit is nearly finished.  I'm not sure of their status in other
browsers.

There is no way to set an absolute minimum size for them.  The v*
units are effectively identical to using percentages; their advantage
is that you don't need the entire ancestor chain to be "100%" in the
relevant property.


> 5.2
> Please include a note that absolute units are sometimes problematic for
> accessibility, particularly when the output environment does not support
> zooming or resizing of absolute units

This isn't a problem in any modern browser.


> 6.1 can you add an example of when one might use angle measurements in CSS?

I don't have a problem adding this.  fantasai, what do you think?
'image-orientation' would be easy, but it's not a great example.
Transforms or hsla would both be good, if they hadn't both made the
mistake of being unitless.


> 6.3 frequency
> Please add a note that frequencies near 7hz can cause seizures, and link to
> WCAG.  More detail in WCAG
> http://www.w3.org/WAI/WCAG20/quickref/20071211/Overview.php#seizure

This only applies to the frequency of things that are flashing or
changing color/brightness/etc., right?  There is no such use of
frequencies in CSS right now.  The frequency units are currently used
only in the Speech spec, where such a warning would be nonsensical.


> 8.3 attribute reference
> It would be great to have an example of using this to style elements with
> ARIA attributes on them.  Would you be willing to work with PF and WCAG on a
> set of techniques around this, in addition to a simple example in the spec.

As far as I know, there's no aria attribute where you'd ever want to
use its value as a CSS value.  You want to apply properties *based on*
the values, but that's done with Selectors.

Do you have an example to the contrary?

~Tj
Received on Friday, 30 March 2012 18:14:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 13:16:16 GMT