[thelist] defined table width and accessibility

[Ian Anderson]

aardvark wrote:

>in general, leaving off the widths of absolute pixels is a good 
>idea... let the CSS do that for you... then you can serve CSS to 
>specific devices (like palmtops, for example, or the printed page)...
>
>coding widths into the page can only wreak havoc for those not on 
>traditional PCs...
>
Do you have any documented cases you are referring to here? I may be way 
off beam, but isn't this just repeating "accepted wisdom", or 
considering cases that are theoretical? In my opinion, there is a lot of 
hoo-hah talked about narrow screen widths by people, myself included, 
who haven't actually tested it to see how these things actually work.

It's like all the bogus information you see about how "text readers" 
behave, often spread by people that have never bothered to try a screen 
reader for real. In the last year or so, I've seen a much more 
experimental approach to screen reader issues, for example in connection 
with IR techniques. Surely we need the same approach for PDAs, web 
phones and whatever else is out now.

I'm surprised they would sell many UAs that can't convert fixed width 
layouts to their own needs. A PDA that can't access bbc.co.uk, or 
virtually any corporate web site in use today? How crap would that be? I 
know some of them scale images to fit the screen - how hard is it to 
ignore table widths?

I think this is the responsibility of the UA, not the designer. I 
personally don't have access to any hand-held web device and so have no 
personal data. Does anyone else? Is there a table of narrow-screen UAs 
and their behaviour/support for web techniques somewhere?

Cheers

Ian