- From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
- Date: Mon, 5 Jul 2004 17:32:58 +1000
- To: "Gregg Vanderheiden" <gv@trace.wisc.edu>
- Cc: "'Ineke van der Maat'" <inekemaa@xs4all.nl>, <w3c-wai-gl@w3.org>
Gregg Vanderheiden writes: > > Hi Ineke, > > Actually, James' recommendation does conform to the W3C tip if I read it > correctly. > > At level 3 he says recommend full text (no verb phrase) > > AT level 1 he says 'must at least have more than verb phrase available > programmatically' > So instead of writing <a href="article.html">read more </a> I can write read <a href="article.html"> more </a> and become a good Web citizen satisfying this proposal at level 3! Perhaps the proposal should instead require "a noun phrase identifying the destination of the link". Even so, I can still write read <a href="article.html"> the article </a> so perhaps we need to resort to stipulating that the text can be understood out of context. I think most informed evaluators would agree that the above examples all fail this requirement. The requirement for programmatic identification could be at level 1 or level 2. I think Gregg's argument that the user can always read the context of the link, and hence that this issue is really a level 2 concern, is a strong reason for not placing it at level 1. The counter-argument (which I don't personally agree with) would be that someone with both a vision and cognitive disability is thereby disadvantaged/ We also need to clarify that by "link" we mean a single arc having one source and one destination, where the user can typically choose whether or not to follow the link. That is, each arc is a separate link for purposes of WCAG, and we are not discussing links that are automatically activated. These stipulations arise from consideration of the XLink specification.
Received on Monday, 5 July 2004 03:33:28 UTC