Re: [URN] Re: FYI, AAP DEVELOPS DIGITAL I.D. SYSTEM FOR PUBLICATIONS

Sam X. Sun (ssun@CNRI.Reston.VA.US)
Sat, 28 Feb 1998 02:01:53 -0500


Message-ID: <001501bd4416$c4226680$d7019784@ssun2.CNRI.Reston.Va.US>
From: "Sam X. Sun" <ssun@CNRI.Reston.VA.US>
To: "Archie" <warnock@home.com>, "John C. Mallery" <jcma@ai.mit.edu>
Cc: <urn-ietf@Bunyip.Com>, <uri@Bunyip.Com>
Date: Sat, 28 Feb 1998 02:01:53 -0500
Subject: Re: [URN] Re: FYI, AAP DEVELOPS DIGITAL I.D. SYSTEM FOR PUBLICATIONS

Archie wrote:
>...
>So, what happens to name resolution times when we get something like 10**N
>(pick your favorite stressful value of N) documents in the system?
>


Sorry I didn't quite understand your question.  And I guess that you are
asking the total number of round trips required to resolve 10**N (10 to the
power N) number of handles.

The answer to this will depend on how these handles are related. In one
extreme case, if all these handles are registered under the same naming
authority, then the total number of round trips could be 10**N + 1. Another
extreme case could be that all these handles are from different naming
authority, and none of those naming authorities have ever been visited by
the client, then we are talking about a total of 2 * 10**N number of round
trips.

On the other hand, if all these documents are deems to be used together
again and again, the provider could probably create a handle as a reference
to a HTML document, where the HTML document contains all the references (say
URLs) of those documents. The HTML document can be generated
_programmatically_ by resolving handles of these documents. If this is the
case, the client could potentially need only one round trip to get the
information of all those documents.

Regards,
Sam
ssun@cnri.reston.va.us