W3C home > Mailing lists > Public > www-html@w3.org > January 2005

Re: tabindex on headers

From: David Woolley <david@djwhome.demon.co.uk>
Date: Thu, 20 Jan 2005 21:15:56 +0000 (GMT)
Message-Id: <200501202115.j0KLFuA01227@djwhome.demon.co.uk>
To: www-html@w3.org

> If the visually obvious tabbing order deviates from the order of elements

Whilst I might agree that markup that obeys the HTML philosophy doesn't
need tabindex, and might even agree that XHTML 2 will mainly be used by
people obeying that philosophy, tabindex is absolutely essential for
real world commercial HTML.

> in markup (e.g. because the page has been formatted that way using table
> layout or CSS), then it really depends on the browser what is obvious to

Table layout is the reason why nearly every commercial page with non-trivial
forms really ought to use tabindex (but seldom does).  However, even without
tables, a desire to achieve a particular layout is likely to put submit
buttons in the wrong place.

CSS (if generally implemented) would offer the opportunity of really making
the document download order match the logical reading and filling in order,
but there is also a risk reliable absolute positioning would be used to allow a
paste up type design, where the order is the order in which the designer built
up the design.

> each user. And this is probably a problem that should be solved by
> removing the cause, rather than tabindex settings which may work for some

But the cause is deeply ingrained in human nature and won't go away.

> Besides, why would importance matter? A header cannot be tabbed to, so
> what would the tabindex be for?

There is nothing in the specification that forbids a user agent from 
allowing tabbing to headers, and I believe some assistive technology
provides this capability.
Received on Thursday, 20 January 2005 21:16:00 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:20:55 UTC