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

Re: danger of null value for summary attribute [Re: fear of "invisible metadata"]

From: Robert Burns <rob@robburns.com>
Date: Sun, 24 Jun 2007 20:56:29 -0500
Message-Id: <42DD0906-F7EA-4C2A-96BD-19CA816E8ACA@robburns.com>
Cc: public-html@w3.org
To: Bill Mason <w3c@accessibleinter.net>

Is this discussion really over whether @summary should be included in  
HTML5, or how it should be used in HTML5?

I've been closely following all of the threads, but I haven't seen  
anyone arguing that there is an alternative to @summary other than  
just not providing such assitive technology support. Both of you seem  
to be supporting the inclusion of @summary too, or am I reading this  
wrong?

My sense of what Gregory was saying is that it might be useful to  
clearly distinguish between tables with no @summary attribute due to  
(similar for img@alt and img@longdesc):


1) table is not a data table but instead a layout table
2) table is not complex enough or otherwise doesn't warrant a summary
and
3) author's decision to otherwise deliberately or inadvertently leave  
off the summary

The first two could make use of reserved keywords for @summary such  
as "_layout" and "_simple". In this way other a user could instead  
assume the @summary just wasn't considered by the author. In other  
words it will take the user more work to determine anything about the  
table. Having the reserved keywords could ease the work of the user  
especially as more and more authors make legitimate use of the keywords.

Take care,
Rob


On Jun 24, 2007, at 7:31 PM, Bill Mason wrote:

>
> Gregory J. Rosmaita wrote:
>> a null value for summary defeats the entire purpose of alerting  
>> the user who cannot process the table at all or in toto, how the  
>> table is laid out, what it describes, etc.
>> rather than a null value for summary, layout tables should CLEARLY  
>> be marked summary="Layout Table Containing X", so that the user  
>> knows he or she can skip that table and move on to inspect the  
>> next table, courtesy of the summary attribute... after all, aren't  
>> MOST tables found in the wild are quote layout tables unquote?   
>> even though they are misusing the table element, they are using a  
>> table to (literally) lay out content in a particular manner,  
>> alignment, ect.
>> speaking as a screen reader
>
> Considering that your assertion that all tables (including layout  
> tables) should have a non-null summary attribute value flies in the  
> face of:
>
> * Current WCAG WG thinking [1]
> * Advice from at least one major web accessibility group [2]
> * Advice from one of the better known web accessibility guides [3],  
> currently in the top ten of Google results for "layout table summary"
>
> ...it would be useful to justify this stance before using it as an  
> argument for keeping the summary attribute in HTML5.
>
> -- 
> Bill Mason
> Accessible Internet
> w3c@accessibleinter.net
> http://accessibleinter.net/
>
> [1] http://www.w3.org/TR/2007/WD-WCAG20-TECHS-20070517/ 
> Overview.html#H73
> [2] http://www.webaim.org/techniques/tables/data.php#summaries
> [3] http://diveintoaccessibility.org/ 
> day_20_providing_a_summary_for_tables.html
>
Received on Monday, 25 June 2007 07:28:08 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:01 GMT