Message-Id: <9402281803.AA24895@expresso.bunyip.com>
From: Peter Deutsch <peterd@bunyip.com>
Date: Mon, 28 Feb 1994 13:03:51 -0500
In-Reply-To: Mitra's message as of Feb 28, 3:51
To: mitra@pandora.sf.ca.us (Mitra), <uri@bunyip.com>
Subject: Re: URN to URC scenario
Hi Mitra,
[ You wrote: ]
> I think if we are clever we will be able to fold each of the requirements
> people have expressed in the last few messages into a scheme
> without bending, or unneccessarily complicating things.
>
> For example - to incorporate Peter's desire to retrofit FTP archives
> with MD5 checksums, . . .
> Now lets say we have a URN, "urn:md5/12344567", Someone trying to resolve this,
> would try and look up "md5.urn.int" via DNS, and they'd either
> end up with a failure - i.e. this URN is not resolvable, or they might end
> up with the address (or URL) of an archie server that could resolve
> MD5 checksums.
In principle I think this looks fine, although I do think
it might be nice if it were possible to short-circuit the
DNS lookup for those schemes where it is known that a
server wont be found (eg. until someone takes on the task
of serving ISBN numbers, MD5 checksums, etc). If those
schemes are popular, they could be generating enough
traffic to make it worthwhile.
I accept that this is an implementation concern and the
list of schemes without servers will be changing over time
but we should try to keep service issues in mind as we
start. Having said that I'm fairly happy to start
experimenting now with the scheme as generally outlined.
- peterd
--
---------------------------------------------------------------------------
"The future belongs to neither the conduit or content players, but
those who control the filtering, searching and sense-making tools
we will rely on to navigate through the expanses of cyberspace."
- Paul Saffo, (_Wired_: March,1994)
---------------------------------------------------------------------------