W3C home > Mailing lists > Public > www-style@w3.org > January 2017

Re: How are screen readers supposed to read <hr> which has icons as generated contents?

From: Ian Yang <ian.html@gmail.com>
Date: Sun, 22 Jan 2017 10:49:10 +0800
Message-ID: <CAFhBhuOS_yPTtPeFHLVOFQZ=F0Fz+sQx0LqQyvth68qi3GF4xQ@mail.gmail.com>
To: Amelia Bellamy-Royds <amelia.bellamy.royds@gmail.com>
Cc: CSS public list <www-style@w3.org>
2017-01-22 5:18 GMT+08:00 Amelia Bellamy-Royds <
amelia.bellamy.royds@gmail.com>:

> On the question of what screen-readers *should* do (note, I don't know if
> this is what they *do* do):
>
> An <hr> element should be treated like the ARIA non-interactive (assuming
> no tabindex) "separator" role:
> http://w3c.github.io/aria/aria/aria.html#separator
>
> An element with the ARIA separator role is supposed to be treated as if
> any child content is presentational, which means the child content is
> ignored by screen-readers.  Child content includes any text or child
> elements you add by script, but it also includes CSS-generated text.
>
> You may add a name to a separator element (using aria-label,
> aria-labelledby, or even title), but that would normally only make sense if
> it is an interactive separator (that is, a slider used to adjust the size
> of adjacent regions).  For interactive sliders, additional attributes & JS
> are required to make it functional.
>
> In other words, unless you give it extra attributes, an <hr/> should
> always be read out only as a "break" or "thematic break", regardless of any
> content added inside it.
>
> FYI, I don't think the @media speech query is well supported, nor is
> aria-hidden able to be set via CSS.
>
> ~ABR
>

Thank you for the info. Is there a spec defines that CSS generated content
should be treated as child content? could you please also provide the data
of the support for @media speech? Cheers.

Yes. There is currently no such declaration as "aria-hidden: true" in CSS.
That was just an example I used to illustrate the convenience if such
declaration exists.

Ian Yang
Received on Sunday, 22 January 2017 02:49:49 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:09:06 UTC