W3C home > Mailing lists > Public > public-html@w3.org > November 2010

Re: hgroup and ARIA outline in ISSUE-129 [Was: SSUE-129: replace or modify the ARIA section of the HTML5 spec - chnage proposal and replacement spec text]

From: Steve Faulkner <faulkner.steve@gmail.com>
Date: Wed, 1 Dec 2010 08:48:44 +1000
Message-ID: <AANLkTin6UvOLPDS+Ci1DjqB-OxVsqFAZqn1VLx1pOOMK@mail.gmail.com>
To: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Cc: James Graham <jgraham@opera.com>, Bruce Lawson <brucel@opera.com>, Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>, HTMLWG WG <public-html@w3.org>, Maciej Stachowiak <mjs@apple.com>, Paul Cotton <Paul.Cotton@microsoft.com>, Sam Ruby <rubys@intertwingly.net>, Richard Schwerdtfeger <schwer@us.ibm.com>, Cynthia Shelly <cyns@microsoft.com>, Michael Smith <mike@w3.org>, Janina Sajka <janina@rednote.net>
Hi Leif,
>Perhaps your thinking is correct. But the conclusion, as written, is
>incorrect: the highest ranking element of the example is <h3>.

i don't know how you came to this conclusion, I based my conclusion on the
output of the H5O (http://code.google.com/p/h5o/) outlining bookmarklet
which for the following

<hgroup>
 <h3>Analysis</h3>
 <h2>The problem</h2>
 <h2>A mathematical model</h2>
</hgroup>

extracts the first h2  <h2>The problem</h2> as the highest ranking heading.
and for the purposes of an outline masks the h3 and the second h2.

>1) I would suggest that screenreaders should treat all, but the first
>appearing highest ranking heading, as <p> elements. That's clearly the
>logical thing to do.

for the purposes of defining an ARIA mapping, the logical thing to do is
define them as having no role since ARIA does not have a paragraph role.
mapping them to paragragh for APIs that have a paragraph role may be a good
idea for the HTML to accessibility API implementation guide (
http://dev.w3.org/html5/html-api-map/overview.html).


regards
stevef

On 1 December 2010 00:34, Leif Halvard Silli <
xn--mlform-iua@xn--mlform-iua.no> wrote:

> Steve Faulkner, Tue, 30 Nov 2010 07:58:26 +0800:
>
> > If we say the hgroup itself does not have a default role and heading
> > level, but has the effect of removing the semantics from any headings
> > except the highest ranking within the hgroup we end up with
> >
> > Analysis
> >  <h2>The problem</h2>
> >  A mathematical model
> >
> > which appears to be correct?
>
> Perhaps your thinking is correct. But the conclusion, as written, is
> incorrect: the highest ranking element of the example is <h3>.
>
> But how would screen readers read the hgroup if the highest ranking
> header is the second child element?
>
> <hgroup>
>  <h2>The problem</h2>
>  <h3>An analysis.</h3>
>  <h2>The mathematical model</h2>
> </hgroup>
>
> And what if there are two h3 elements?
>
> > As a consequence I suggest that the hgroup itself should have no
> > default role and that any headings inside a hgroup except the highset
> > ranking should have no default role.
> >
> > thoughts?
>
> (1) I would suggest that screenreaders should treat all, but the first
> appearing highest ranking heading, as <p> elements. That's clearly the
> logical thing to do.
>
> (2)  Another option could be to forget the entire hgroup element - or
> to change it. [1] It is usually frowned upon to use h1-h6 for stylistic
> purposes. But inside hgroup, all but the hightest ranking element, are
> only used for their stylistic effect.
>
> (3) I think a hgroup element contaning a single h1-h6 element should
> definitely not have any default role. But if it is a real, multilevel
> heading, then why not make the screenreader user aware that he/she
> reads a multilevel heading?
>
> [1] http://lists.w3.org/Archives/Public/public-html/2010Nov/0409
> --
> leif halvard silli
>



-- 
with regards

Steve Faulkner
Technical Director - TPG Europe
Director - Web Accessibility Tools Consortium

www.paciellogroup.com | www.wat-c.org
Web Accessibility Toolbar -
http://www.paciellogroup.com/resources/wat-ie-about.html
Received on Tuesday, 30 November 2010 22:49:39 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:16:07 UTC