- From: Daniel Dardailler <danield@w3.org>
- Date: Fri, 23 Jan 1998 11:38:26 +0100
- To: "Pawson, David" <DPawson@rnib.org.uk>
- cc: w3c-wai-ig@w3.org
> 3. How to ensure aligned interpretation. Context provides the clue. > Missing context could > put a wrong light on the image description. [I drew a 'dog in a > manger', you described a > puppy] Yes, context is important (Len Kasday already commented to me privately about that) but it's also hard to achieve, as the page URL itself is often too broad a reference (when the image is buried down the bottom of a long page for instance). > 4. How to resource. > A misguided volunteer using graffiti as the alt text is a real > 'threat' to the service. Yes, although I don't think this is likely to happen, one could imagine a subscription with access passwd for describers, just requiring a valid email. Also possible is heuristic code that looks for dirty words (but vandals are likely to just say "a dog" when a picture of "a car" is shown...) > Linguistically appropriate descriptions? The source associated > text could be in any > language, the description should be in the same one. Good point, the language of the including page should be indicated, but we also need to take for a given that english is the esperanto of the 21st century :-) > Hope I haven't sounded too negative Not at all. Thanks.
Received on Friday, 23 January 1998 05:39:24 UTC