Lists Home |
Date Index |
Yep. X3D/XML models so far depend on cross-translation
to the VRML97 syntax for a renderable model given that
there are no native XML X3D players yet. It is ugly
but does have one good side effect: it proves the
identity (one to one mapping) for ISO. There is
an open source player in the works (God bless the
volunteers for the web). Then it becomes an
issue of tool-appropriateness. You really don't
want to build real-time 3D in an XML-only editor;
you want a rendering window that enables you to
pick and manipulate objects then updates the
XML treeview automatically. VRML editors have
worked like that since V-Realm.
But using XSLT to downtranslate from a semantic
model to a renderable model is a joyful thing.
I assert that for authors who need to create
say, animations of stories, this is the way
to go and will be the way we finally unlock
the box for compelling, less boring, web content.
Paul Fishwick's group at Univ of Fla is working
on that. The orignal concepts for HumanML
were based on that idea. Anything taggable is
renderable if mappable. That won't make it
automatically comprehensible, but hey, that
is the art of XSLT design work.
From: Jonathan Perret [mailto:firstname.lastname@example.org]
len wrote :
> Urrrp. My mistake. RTF reverses the roles of the
> curly and the slash.
> I've written db scripts to output this stuff. Somewhat like
> bagging tigers. HTML was a heckuva lot easier.
I use XSLT to generate RTF. I've never tried bagging tigers, but I'm
sure it can't be much enjoyable than this.