- From: Ben Boyle <benjamins.boyle@gmail.com>
- Date: Fri, 12 Oct 2007 20:22:01 +1000
- To: "Henri Sivonen" <hsivonen@iki.fi>
- Cc: "Chris Wilson" <Chris.Wilson@microsoft.com>, "HTMLWG WG" <public-html@w3.org>
There's support for semantics extraction. W3C have a proof of concept tool: http://www.w3.org/2003/12/semantic-extractor.html Where and when this might be useful, I'm not going to debate. Look to the fields of academia, research, libraries, publications, ... the places where citations are most relevant, for real supporting evidence. I don't think we geeks[1] qualify to make judgements about the usefulness/relevance of citations and the @cite attribute. I just had a quick search of mozilla add-ons and found "OpenURL Referrer" https://addons.mozilla.org/en-US/firefox/addon/4150 I don't know if it uses @cite, but it conceivably could. I think it fits the bill for a specialist, niche, semantics extractor (or "use case" in html5 parlance). cheers Ben [1] I use the term "geek" affectionately to remind all that as a group we appear to have a bit of a bias towards our industry. Compare support for semantics extraction with code@type. I just don't think we represent the entire web user community at all comprehensively. Much of HTML is abstract enough that this not an issue. Occasional reminder can't hurt (much). HTML5 is gonna be great anyway of course. Go us! On 10/12/07, Henri Sivonen <hsivonen@iki.fi> wrote: > > On Oct 11, 2007, at 20:24, Chris Wilson wrote: > > > Henri, I think that depends on your definition of a UA. > > Yeah, but by default, it is reasonable to expect HTML 5 requirements > for UAs in general to apply to browsers. > > > My expectation (as a browser developer) has always been that an > > accessibility tool would enable the following of such a link; > > exposing the link to the AT is our responsibility in the browser. > [and later] > > Whoops! > > > > Replace "AT" with "editing system". > > Are there actual editing systems that use HTML as their file format, > store reasons for changes in external files and point to them using > the cite attribute? > > Earlier today, I was notified of markup that used the cite attribute > and had traits that suggested it was generated by Microsoft Office. > However, that markup didn't use the cite attribute to point a > resource explaining the change. Instead, the attribute had been > twisted to store the identity of the maker of the change by > prepending mailto: to what looked like a userid. > > > Was your [2] reference intended to be a separate issue, or were you > > drawing an analogy? > > I was pointing out a message that went even further and questioned > the <ins> and <del> elements themselves instead of questioning an > attribute on those elements. > > > [2] http://canvex.lazyilluminati.com/misc/cgi/issues.cgi/message/% > > 3C44F4892E.9030404%40cam.ac.uk%3E > > -- > Henri Sivonen > hsivonen@iki.fi > http://hsivonen.iki.fi/ > > > >
Received on Friday, 12 October 2007 10:22:11 UTC