Re: Mozilla - like JAWS or like Hal?

[Luke Yelavich <themuso themuso com>]

At 11:47 PM 5/05/2004, Janina Sajka wrote:
> Tom, you are very correct about tables. Just to be clear, I think we need 
> to know more than what row and column. We really need the label for 
> each--the column header and the row label. I suppose a smart interface 
> would announce this once and then announce only the one that changes as 
> one navigates the table. I do think, by the way, that we'd prefer to 
> navigate around in tables. So, if it's left and right movement across a 
> single row, then the new column header and cell contents should be read as 
> each new cell gains focus.
>
> Problem is, though, how do you distinguish real tables from the rather 
> ubiquitous use of table tags to force layout? And, which cells contain the 
> labels in those cases when the HTML is improperly tagged? Are these the 
> limits of a smart tables approach?

This is true, as tables are often used to lay out websites in columns. 
However, One way of getting around this is looking for the border 
statements in the TD tags, at least I think thats what they are called. 
When these are set to 0, there is visually no border round the table on the 
screen. This is commonly used for website layout. So maybe reading the 
column title as well as row titles could happen if the border is set.

But then there still might be a problem elsewhere.

Just a thought.