Re: PHIL: more machine thoughts

Mark Waks (justin@dsd.camb.inmet.com)
Thu, 16 Jun 94 13:11:33 EDT


Kevin writes a bunch of sensible stuff about issues that we don't need to
worry about yet; I'll toss in one quibble, though:
>Likewise is
>discussion of navigation (flying/walking/etc), that is a browser issue, not a
>language issue.

Yes and no. The problem is that the *paradigm* of navigation we are using
needs to be reflected in the language. Specifically: if we go with the
Cyberspace Protocol, then we simply don't need to deal with navigation
at the language level -- all navigation is handled through the official
space-defining sites. (If I grok what Mark is saying.) But if we go for
a Web-like model (which I strongly prefer), then we need some mechanism
in the language to say, "this thing is a doorway, and goes to this URL".
We don't need to define many of the *details* at the language level, but
I believe we need to know whether we're using links or not...

(Something a lot like the CP could be implemented under a Web-like model,
with a subset of sites agreeing to conform to a standard authority. But
pure CP implies language differences from Web-based CP, I think...)

-- Justin

Random Quote du Jour:

Re: The 1990 Mass. elections
"This is our state.
This is our state on politicians.
Any Answers?"
-- Keenan Ross