Re: Reading order issue

On Wed, 15 Oct 1997, Al Gilman wrote:
> The scenario by which I would expect that CSS does deal with the
> reading-order issue is one where the total text content is in the
> HTML in logical sequence with sub-ranges marked with SPAN or DIV.
> The stylsheet then places the individual sub-ranges of text in
> layout in a graphical presentation.  In this case the stylsheet
> is all presentation and the HTML communicates the logical units
> of text.
This is indeed the kind of scenario which I had principally in mind,
except that it should not be restricted to the use of SPAN and DIV
elements, but must extend to any container which can be selected on the
basis of a CLASS or ID value. CSS properties can then be defined that will
enable such units of text to be moved into other locations within the
document's structure.

 > 
> A second scenario would be one where the HTML contains a table,
> and fragments of one logical text are arranged in various cells
> in the table.  An audio stylesheet contains a directive which
> calls for the reading of several table cells in a particular
> order.  In this case, the sense of the text is not salvageable
> for dumping to line-mode interaction via TTD without the "audio"
> stylesheet which contains critical content information.  I think
> that this scenario puts information in the stylesheet that should
> have been captured in the base HTML document.
I agree. However, with the increasing availability of CSS, authors are
likely to turn toward style sheets and away from abuse of table constructs
as their principal formatting mechanism. Thus, one can expect this problem
to diminish in importance as style sheets achieve greater acceptance. We
should be careful not to entrench the abuse of tables by acknowledging it
in markup development, thereby running counter to the aims of the entire
HTML standardisation process.

Received on Wednesday, 15 October 1997 19:00:12 UTC