WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Multiple H1 tags in an HTML5 web page

for

From: Steve Faulkner
Date: Mar 8, 2014 2:04AM

this current discussion on the public HTML mailing list is relevant:
http://lists.w3.org/Archives/Public/public-html/2014Mar/0000.html

--

Regards

SteveF
HTML 5.1 <http://www.w3.org/html/wg/drafts/html/master/>;


On 8 March 2014 09:02, Steve Faulkner < <EMAIL REMOVED> > wrote:

> Hi all,
>
> Jared is correct. There are no implementations of the outline algorithm in
> browsers or other user agents and there is unlikely to be as there is no
> requirements on user agents to implement it. There are a few 3rd party
> implementations in JavaScript (demos) and one in XSLT in ePub. (this is
> used to produce a table of contents for a book).
>
> It will probably not be dropped from the spec as features are dropped when
> they have implementation requirements but no implementations, this is not
> the case for the outline algorithm (there are no requirements).
>
> There are requirements on user agents to map heading levels in acc APIs
> based on the headings outline depth
> http://www.w3.org/html/wg/drafts/html/master/dom.html#sec-implicit-aria-semantics
>
> These have not been implemented because they rely on the outline algorithm
> being implemented (which is not a requirement and browser vendors have
> shown no intention to implement) I am in the process of talking with
> browser acc implementers to drop these requirements from the html spec.
>
> this article may be helpful
> http://blog.paciellogroup.com/2013/10/html5-document-outline/
>
> The advice in the HTML specification<http://www.w3.org/html/wg/drafts/html/master/sections.html#headings-and-sections>;is clear on this:
>
> Sections may contain headings of any rank, and authors are strongly
> encouraged to use headings of the appropriate rank for the section's
> nesting level.
>
>
> ePub also advocates use of correct heading levels
> http://www.idpf.org/accessibility/guidelines/content/xhtml/headings.php
>
>
>
> --
>
> Regards
>
> SteveF
> HTML 5.1 <http://www.w3.org/html/wg/drafts/html/master/>;
>
>
> On 8 March 2014 02:37, Jared Smith < <EMAIL REMOVED> > wrote:
>
>> Duff Johnson wrote:
>> >> Of course this assumes that everything is done correctly
>> >> and that the algorithm properly reflects the intended structure.
>> > You make it sound as if this is all very unlikely. Is that fair?
>>
>> Considering the general abuse and mis-use of HTML5 sectioning elements
>> (particularly <section>), I have seen very few instances where if the
>> HTML5 outlining algorithm were applied it would result in improved
>> accessibility.
>>
>> >> And there's a good chance this algorithm may change or be dropped from
>> >> HTML5 entirely.
>> >
>> > Fascinating. How would one track this question?
>> >
>> > Also, if it was dropped from HTML5 how would that affect EPUB3?
>>
>> These are better questions for Steve Faulkner or others. I know that
>> due to very limited implementations there has been some rumors of
>> changing or dropping it.
>>
>> > Just to be clear: you are referring to HTML 4.01, which specifies <H1>
>> - <H6>, right?
>>
>> Yes.
>>
>> > So, if the author needs to go deeper (H7, H8...) - which was actually
>> the point of Greg's question - you are saying they out of luck?
>>
>> The answer to that question would be "yes". Greg was, I believe,
>> instead referring to the fact that you could give each sectioning
>> element its own independent heading structure beginning with an <h1>
>> (or really any heading level for that fact) and the outline presented
>> to the user would be 'reset' and adjusted to reflect a proper outline
>> regardless of the heading levels actually used in markup. If fact, you
>> could use <h1> for EVERY heading and just wrap each section in
>> appropriately nested <section> elements - and this would result in a
>> proper outline for the document, assuming the outlining mechanisms
>> were actually implemented, which they presently are not.
>>
>> Here's a slide I've used that demonstrates that the heading levels
>> used wouldn't really matter that much if the outlining mechanisms
>> actually worked -
>> http://webaim.org/presentations/2013/ariahtml5/sectionarticle3
>>
>> Jared
>> >> >> >>
>
>