Michael Kay wrote:
> But if we want enhanced behaviour for links, the first thing is to put it at
> the right point in the architecture. That's the user-interface vocabulary,
> not the data representation - and that's what XLink got wrong.
Now that's a good subject for a Balisage talk, or maybe a set of talks.
I think that most of what XLink got wrong was political - there wasn't
much effort at outreach to people who use links, and I think a lot of
people assumed the benefits were obvious.
The architecture level question created a lot of issues, though, as
behavior (user interface) is critical to explaining why these things are
necessary, but the initial XLink approach seemed intent on defining only
a data model. (I'd say a partial data model at that.)
I wish I could I say there were clear lessons from the experience that
might improve the odds of improving hypertext next time around, but
well... I can't. I suspect better hypertext will have to impress people
in a walled garden before crossing over into the general web.
--
Simon St.Laurent
http://simonstl.com/
