W3C home > Mailing lists > Public > public-html@w3.org > September 2007

Re: Heuristic Tests for Data Tables (Discussion)

From: Leif Halvard Silli <lhs@malform.no>
Date: Thu, 06 Sep 2007 14:35:49 +0200
Message-ID: <8656234cc0f672205837eee661d892ba@10013.local>
To: "Ben 'Cerbera' Millard" <cerbera@projectcerbera.com>
Cc: "HTMLWG" <public-html@w3.org>

2007-09-06 13:01:05 +0200 "Ben 'Cerbera' Millard" <cerbera@projectcerbera.com>:
> Leif Halvard Silli wrote:
>> But, unfortunatly, JAWS and other assistive technology users are not able 
>> to test the differentces between the algorithms.
> 
> Each cell already starts with all the text from its associated headers. 
> There's no dependancy on scripting or headers+id or anything. All the 
> information is right there in the content for each cell.
> 
> What is your recommendation based on?

It is based on how I perceive such UAs work: The user is not constantly, I believe, annoyed with information about the headers cell relations. Instead, the user can ask, in an interactive way, to be informed about who the current cell "belongs" to. So, for an AT user, the user experience is more like what sighted users see, when they use the Table Inspector - and every related cell is being highlighted.

But your question got me to think further: For AT users, and speaking about how the Table Inspector currently works, it would only be bothering to have «all the text from its associated headers […] right there». For those users, it would be better if that text was removed, and that *only* headers/id attributes be added. 

[ Ferg says ] «If your concern is not only to "code to the standard" when writing HTML, but also to work with screen readers that are not yet able to "read to the standard", then using HEADERS and IDs may be your only choice.» And this is one such use case! We do not want to test how AT works with tables in general. We want to test these spesific algorithms. And there are some spesific differences between the HTML4 and (current) HTML5 algorithms, which would be useful to test. Perhaps then Joshue or Steve could provide user testing repored based upon reading the same table with different algoritms. 

Further, if you want to *both* add the text *and* server the AT users in one and the same page, then it would be better if the text from the associated headers was added via CSS - using pseudo elements. (Perhaps I am wrong, but I believe UAs for AT does not support pseudo elements.)

PS: Could the experimental option(s) tell if  it is supposed to improve the HTML5 algorithm or the HTML4 algorithm?

[ Ferg says ] <http://www.ferg.org/section508/accessible_tables.html#contents_item_6.1>
-- 
leif halvard silli
Received on Thursday, 6 September 2007 12:36:48 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:49 UTC