Re: Layout tables

Vacating the chair for a moment, there are a few comments that I would
like to contribute to this discussion.

With XHTML 2.0, there is no need to retain the legacy of conventional
HTML table structures or patterns of usage. In the case of genuine
data tables, the elements and attributes are well suited to the
purpose of defining header and content cells, captions, summaries, and
the semantics of inter-cell relations. The use of tables for layout,
by contrast, is a kludge--the motivation is obvious (tables work
acceptably, in the absence of a superior layout mechanism, such as
that provided by style sheet positioning); but, as evinced by the
complex arrangements of table elements, often highly nested, which
appear on web pages today, they fall far short of offering an elegant
or convenient layout mechanism. More importantly for purposes of
accessibility, the underlying semantics of the components comprising
the page are often lost.

This legacy need not, and ought not to be, carried forward. Rather,
elements should be defined, if necessary, with which to express those
 semantic and structural relationships which authors endeavour to
present visually via layout tables. Formatting can be achieved either
by style sheet positioning (the preferred option) or by an appropriate
intermixing of XHTML and SVG markup. From the standpoint of
accessibility, the question is: what components of a typical web page
are poorly captured by the currently available repertoire of XHTML
elements? While it is possible to represent navigational constructs in
XHTML 1.x (and HTML 4.01) MAP elements, perhaps it would be better to
define structures that are better suited to this role (just an illustration;
I haven't considered this issue in detail).

If all that one has truly amounts to mere blocks of text, with only
spatial relations between them, then there is always the DIV element
or its analogue (plus style sheet positioning), or again, SVG, both of
which are markedly superior in terms of layout functionality.

What other common artifacts of contemporary web pages fail to be
adequately represented in XHTML 1.x? What further semantic
distinctions might be usefully captured?

Received on Friday, 9 March 2001 02:18:33 UTC