- From: Gez Lemon <gez.lemon@gmail.com>
- Date: Wed, 28 Jun 2006 18:40:01 +0100
- To: "Gregg Vanderheiden" <gv@trace.wisc.edu>
- Cc: "Katie Haritos-Shea" <kharitos-shea@cri-solutions.com>, ryladog@earthlink.net, public-wcag-teamb@w3.org
Hi Gregg, > My suggestion was only a SUFFICIENT technique. It would mean that if you > did this it would be sufficient. > > A sufficient technique doesn't not mean the only way. Just one acceptable > way. I thought that I might have misunderstood what you were suggesting. I understand the concept of sufficient, but disagree that this technique could be regarded as sufficient, as there is no method of discovery short of physically changing the last part of the URI. > I only mentioned this technique so that we had one way we could count on for > alternate forms that were completely inaccessible. I was concerned that > there was not method without it. But other methods could exist -- like > content negotiation with accessible version as default that we discussed > before. Or in the future – all browsers could support an "alternate > accessible" button or something. We have techniques for alternate forms that are completely inaccessible: * Address the accessibility issues. * Ensure visitors see the accessible version first, with a link to the inaccessible version. * Provide a link to the accessible version (for technologies that support accessible links). * Style sheet switching (for technologies that support style sheet switching). * User preferences (temporary measure until content negotiation is possible) - provide an area for people to set their preferences on the website, and then deliver the content according to those preferences. I'm not sure how much mileage there is content negotiation. Negotiation is usually performed by passing data in the HTTP headers, but it's difficult to imagine what someone would pass to receive the inaccessible version. The only way I could ever see that technique working would be if everyone was to keep a personal RDF file that indicated their preferences, and websites delivered content according to those preferences. For it to work, the data would need to be standardised around the world, and there's the question of privacy - I'm not sure people would volunteer to maintain a document that they pass to anyone that detailed their ability, or even if that approach is ethical. I understand this technique is part of the semantic web, but I think it overlooks the privacy issues. Best regards, Gez -- _____________________________ Supplement your vitamins http://juicystudio.com
Received on Wednesday, 28 June 2006 17:40:06 UTC