W3C home > Mailing lists > Public > public-html@w3.org > November 2007

Re: HTML 5 Authoring Guidelines Proposal the use of the section element and its potential impact on screen reader users

From: Ben Boyle <benjamins.boyle@gmail.com>
Date: Wed, 28 Nov 2007 23:27:44 +1000
Message-ID: <5f37426b0711280527o77e9ff5fw4f8781f2c90582b0@mail.gmail.com>
To: joshue.oconnor@cfit.ie
Cc: "Lachlan Hunt" <lachlan.hunt@lachy.id.au>, "James Graham" <jg307@cam.ac.uk>, "Terry Morris" <lsnbluff@gmail.com>, "public-html@w3.org" <public-html@w3.org>

On Nov 28, 2007 8:48 AM, Joshue O Connor <joshue.oconnor@cfit.ie> wrote:
> I am probably missing something here but can we not already do this with
> HTML? The difference being that the cognitive load is on the author as
> they have to think about *how* they structure content and apply the mark
> up accordingly. Is this what you mean? Is it the case that a generic
> element is a convenient mechanism for those authors who don't want to
> *think* about how to structure content? If so, on one level I can
> understand this.

Put the concept of a "document" aside for a moment.

If I have a block of content, I can mark that with <section>.
If that block has a heading (as is often advised), I can use <header>.

Now, that section can be nested in many different documents in many
different places, and I don't care if the heading is "h1" or "h6" ...
I only care that it is the heading for that section. This is what I
want. I think it is a very elegant approach (however I don't think the
UA market is ready for just yet, but give that time ... )

Now, I could also use <h1> for that heading. But if I then include
that section 3 levels deep in another document, does the use of <h1>
assign it higher significance? What if there are sibling sections from
other sources that use <h2> instead? Are they of less importance? It
seems messy.

So mainly my point is about reuse of sections, with arbitrary nesting
altering the heading level as relevant. This could be applied on many
conglomerate sites, portals and blogs and other indexes ... (of
course, it's reasonably simple to parse h1-h6 from various sources
into the desired heading level anyway).

> For example users of AT can extract headings and learn about their
> importance from the use of author applied h1 - h6 elements. However, in
> many cases what is important to the user is that the heading *is* there
> and not necessarily whether they have been correctly applied. So on one
> level the use of a generic heading element could be good, as it may get
> old school semantically shy authors actually using headings, and that is
> a good thing.
> However my worry about using generic elements is exactly  that the
> inferred structure or the ability to infer structure could be lost. At
> least having h1-h6 gave the author a ready made "semantic toolkit" using
> a generic element takes that away IMO.

I agree that has a use. I'm just confused by the potential conflict
for the between the heading level explicitly defined by the h1-h6 tags
vs the inferred heading level based on their placement within nested
sections of a document. Which has precedence?

Received on Wednesday, 28 November 2007 13:28:01 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:28 UTC