[7972] in www-talk@info.cern.ch
Re: Session tracking
daemon@ATHENA.MIT.EDU (Dave Hollander)
Wed Apr 19 11:36:38 1995
Date: Wed, 19 Apr 1995 11:04:21 +0500
Errors-To: procmaster@www19.w3.org
Reply-To: dmh@hpsgml.fc.hp.com
From: Dave Hollander <dmh@hpsgml.fc.hp.com>
To: Multiple recipients of list <www-talk@www10.w3.org>
I like where this string is going. I suspect that the session-id problem
can be solved in a more general way, giving Gary the broader abilities
he describes. I wonder if there is not an even more general approach
that extends from client to server documents--documents may embed
cookies and query the client about what cookies it has.
The exchangable cookie (we called it tokens/knapsack) idea was discused
many hypertext design sessions held by the davenport group. Everyone was
excited about the possiblities. Cookie semantics ranged from "have you
seen document x" to "do you prefer video" were discussed. Indeed, it seems
that this is the start of a client/server agent information exchange
protocol.
As always, I am looking for more...but I would be delighted with
session ids alone.
Dave Hollander
