W3C home > Mailing lists > Public > public-html@w3.org > July 2009

Re: How to make complex data tables more accessible to screen-reader users

From: James Graham <jgraham@opera.com>
Date: Mon, 06 Jul 2009 12:38:03 +0200
Message-ID: <4A51D40B.6090505@opera.com>
To: Ian Hickson <ian@hixie.ch>
CC: Lachlan Hunt <lachlan.hunt@lachy.id.au>, public-html@w3.org
Ian Hickson wrote:
> On Mon, 6 Jul 2009, Lachlan Hunt wrote:
>> Ian Hickson wrote:
>>> I've updated the spec to encourage describing the table in general and
>>> showing a variety of doing so.
>> I think that's an improvement over what was there previously.  However, 
>> I think the spec should make it clear that it isn't an exhaustive list 
>> of techniques, and that authors may use other variations or combinations 
>> of these techniques that may be more appropriate for the design and 
>> structure of the page.
> Done.

I was going to recommend spec test that unambiguously advocated the 
<details> element in cases where the table is complex enough that the 
structure cannot be grasped by inspecting the outermost row and column 
of the table; something like

"For tables with a complex structure that cannot be understood simply by 
inspecting the outermost row and column, authors should provide an 
additional description of the table structure for the benefit of those 
who have difficulty determining it, notably the visually impaired. For 
this purpose, the <details> element should be used as a descendant of 
either the table's <caption> element or, where the table is a child of 
<figure>, the <figure>'s <legend> element. The <legend> of the <details> 
element should be used for the ordinary table caption, whilst the 
details should be used for an extended description of the table's 
logical structure."

However I am not sure what the right tradeoff between simplicity and 
flexibility is in this case; give authors too many options and they may 
suffer from the paradox of choice and decide it is easier to do nothing 
than work out the best option; give them too few options and they may 
decide that their design constraints are more important than universal 
Received on Monday, 6 July 2009 10:38:54 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:47 UTC