W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2003

Re: accessify.com's review of RNIB relaunch

From: Kynn Bartlett <kynn@idyllmtn.com>
Date: Wed, 25 Jun 2003 13:55:33 -0700
Cc: "Matt May" <mcmay@w3.org>, <tina@greytower.net>, <w3c-wai-ig@w3.org>
To: "John Foliot - WATS.ca" <foliot@wats.ca>
Message-Id: <5D40E985-A74F-11D7-9537-000393D9E692@idyllmtn.com>

On Wednesday, June 25, 2003, at 01:33 PM, John Foliot - WATS.ca wrote:
> Kynn, your reasoning is not totally flawed, but why then not start with
> <h7>?

Indeed, why not?  There is no logical reason for not doing so, save that
there is no such tag. :)

> Since we can style away the look of our <h> tags to suit the
> particular design considerations of our sites, the "look" is less 
> important
> today... very often the reasoning for starting with an <h2> was that 
> the
> "<h1> was too big".

Which is still a concern for Web developers who are following the W3C's
own guidelines about making pages usable without style sheets.

> The logic would suggest that we start with 1 and progress from there.  
> While
> intellectually (and technically per the specs) there is no real reason 
> to
> *not* number our headings H2, H4, H6, does it really make sense?

It could very well make sense, for the reason that in Applesoft BASIC we
used to number lines 10, 20, 30, etc.

It makes it easier to insert a line between the others.

In the same way, someone might want to insert a heading between other
headings.  There are good reasons for allowing this flexibility as
a design consideration, especially if it does not cause any harm to

> And while,
> as you stated, as long as all headers within a site are relative to 
> each
> other, and their cousin pages, there really is no harm caused, what 
> happens
> when the user moves on to another document/site?  A consistency between
> unrelated sites seems to me to be a laudable goal as well, and 
> everybody
> starting with <h1> seems to be a pretty simple starting point, no?

Actually, that's not helpful at all.  A simple example will illustrate
why the HTML 2.0 <h> tag is superior to relying on <h1> being the
"topmost" heading and everything proceeds down from there.

Let's say that I'm posting something on a Web site where the headings
go <h1> for the page, <h2> for the section, and <h3> for articles in
that section.  I type up an article which is divided into three parts
and naturally I use <h4> for those parts.

Now I want to republish that same article on another site.  That other
site also starts with <h1>, then <h2> for article titles.  I paste
my text in -- but now because it's an <h4>, you'd claim it's out of

With <h> this is solved (as the index numbers are automatically

The problem here is that you can't enforce a hierarchy across
divergent sites.  If you decide "<h1> is the page title" -- well, you
run into problems because you _can_ have multiple <h1> pages.

An article's title may be an <h1> on one page, an <h2> on another,
and an <h4> on a third.

What's my point?  Any attempts to claim that a strict <h1>-and-down-
from-there hierarchy is going to be consistent across sites is
simply provably wrong.

For each site, we rely upon the Web developer to make the decisions
as to how headings will be used -- for example, she chooses which
heading corresponds to an article title, and that may very well
vary from page to page on the site even.

Thus, we should also allow her the flexibility to decide at what
level she places (and starts) items within that hierarchy.  The
assumption that <h1> _has_ to come first is much, much less
important than the very valid requirement that structure
exists and the structure is navigable.

The requirement must be "a navigable structure"; whether it begins
at <h1> or not is really beside the point as long as the relative
hierarchy of headings is properly maintained.


Kynn Bartlett <kynn@idyllmtn.com>                     http://kynn.com
Chief Technologist, Idyll Mountain                http://idyllmtn.com
Author, CSS in 24 Hours                       http://cssin24hours.com
Inland Anti-Empire Blog                      http://blog.kynn.com/iae
Shock & Awe Blog                           http://blog.kynn.com/shock
Received on Wednesday, 25 June 2003 16:50:07 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:36:16 UTC