True. However, note that the idea of named objects, or putting objects
on a CD-ROM as well as on the Web, is *not* necessarily linked to the
idea of abstract objects or scenes. (I'm in favor of both, and have
been arguing for both, but they *are* separate ideas.) We could have
concrete scene definitions with specific objects, and still distribute
some of those objects by non-Internet means...
(I think this would want to wait a little while, until we've got some
empirical evidence about what the popular and frequently used objects
are, but I strongly suspect that some common objects *will* develop,
and might as well be distributed by such a mechanism, to save a little
bandwidth.)
Tony writes:
>But it sounds like we'll need a "name registry" to handle the
>identification, i.e. naming of standard objects such as "utah-teapot" and
>"bay-window" (apologies to architects/interior designers reading the
>list).
In some fashion, probably. But just as several people have pointed out
to me that my keyword-based objects might be best handled through
URCs, I'll point out that this might be best handled through URNs.
While I'm not nearly finished on reading up on the state of the
URI project yet, I get the impression that the folks working on
creating Universal Resource Names are basically trying to solve
this problem in the general case...
BTW, one thing is becoming clear to me: our links, in the long run,
probably don't want to be URLs specifically, but URIs. That is, we
should make sure that we design our language and browsers so that
when URCs and URNs get a little more real and concrete, we can
transition smoothly to using them. This probably doesn't make
much technical difference, but is a slight difference in mindset...
-- Justin
Quite enjoying prowling around in the
IETF docs on URIs -- thanks for the
tip, Brian...
Random Quote du Jour:
"clone, n. -- 1. An exact duplicate, as in "our product is a clone of their
product." 2. A shoddy, spurious copy, as in "their product is a clone
of our product."