Message-Id: <9310282013.AA04162@expresso.bunyip.com>
From: Peter Deutsch <peterd@bunyip.com>
Date: Thu, 28 Oct 1993 16:13:11 -0400
In-Reply-To: Mitra's message as of Oct 28, 2:00
To: mitra@path.net (Mitra), uri@bunyip.com
Subject: Re: URN resolver scaling
[ You wrote: ]
> Chris,
. . .
> We know that other schems (such as coding the location of the resolver as part
> of the authority id) will scale.
Hard coding in locations may scale but I think it is clear
that it wont be durable, in the sense that servers come
and go and thus location info will break fairly frequently.
Also, although we don't have an existance proof that the
centroids model wont scale, I haven't yet heard what I
consider pursuasive arguments for this belief, only
expressions of vague uneasy feelings. Although I respect
this, I think the jury is still out. In contrast, I think
it fair to claim that hardcoding has been shown in the
past to be a "Bad Idea (tm)".
- peterd
--