W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > September to December 1996

Re: Handling ambiguous requests

From: <hallam@ai.mit.edu>
Date: Fri, 06 Sep 96 12:31:53 -0400
Message-Id: <9609061631.AA24074@etna.ai.mit.edu>
To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Cc: hallam@ai.mit.edu
X-Mailing-List: <http-wg@cuckoo.hpl.hp.com> archive/latest/1560

I think that its easier to understand this problem if we look at
it from the point of view of author intention rather than of the 
reader. I think that the term "user" is entirely bogus, it introduces
confusion since both the author and the reader are "users". to 
refer to the reader as a user introduces an undesirable asymmetry,
it is equally valid to pander to authors as to readers. To call 
readers "users" implies that we should always strive to pander to

The question to ask is "what confusion arises". Now say that
we have a URL which has the IP address embedded. What does
that IP address stand for? surely the logical interpretation is
that it stands for the resource which the author intended and
tested for.

A logical method of handling this problem is to make the host
interpretatiopn of a raw IP address the cannonical DNS name for
that address, ie what an inverse DNS lookup would result in.

This has the advantage of providing good consistency. Only if the
cannonical dns name for the host is initially unused and then
later pressed into service as a vanity host name does a problem 
arise. Since this would be a fairly unusual thing to happen I 
don't think that its worth worrying about.

As Wittgewnstein said "usage defines meaning". I think its more 
important that URLs resolve to a particular, defined name 
than it is to consider the multitude of possible interpretations
of the name and then allow a selection. If no selection is permitted
then no ambiguity arises.

Received on Friday, 6 September 1996 09:34:16 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:16:20 UTC