Message-Id: <9402250227.AA13266@expresso.bunyip.com>
From: Peter Deutsch <peterd@bunyip.com>
Date: Thu, 24 Feb 1994 21:26:59 -0500
In-Reply-To: Mitra's message as of Feb 24, 16:35
To: Mitra <mitra@pandora.sf.ca.us>
Subject: Re: Finding URN->URL servers
One last one and I go to bed...
[ You wrote: ]
> Peter,
>
> I dont think it's going to be a seperate protocol - whois++ seems fine as
> a basic query response mechanism. However, I dont at all think that
> people doing URN->URC lookup are going to need or want to put together
> the whole of whois++.
The "whole of WHOIS+" is really quite minimal - most of
the commands and features are optional and thus not needed
to build a conforming server. Several people have already
demostrated that you can put together a minimal system in
a weekend.
> I think we are going to want a sub/superset of
> whois++, which means that a whois++ server can - but is not required - to
> serve it, and that a much simpler client can query it. If URN->URC server
> answers on the whois++ port, then people are going to expect it to
> implement the whole of the whois++ protocol.
If you're worried about what's need on the client end, all
you really need are a) the ability to generate a search
command and b) something to parse out the response. We're
late on finalizing the output format, but the initial cut
was pretty simple to machine decode.
> I figure that a reasonable URN->URC resolver could be implemented in a
> couple of hundred lines of perl (in fact I did just that before the
> suggestion to use whois++ came up). Implementing whois++ takes a lot more
> than that.
Sounds like someone throwing down a gauntlet to me! I have
an existance proof that the WHOIS++ frontend can be built
in a weekend, although the database management part of
that version was pretty brain-damaged. If you want to be
really ambitious and attach it to something like a WAIS
search engine it might take a bit longer (and would
obviously be larger) but the bare-bones server shouldn't
be _that_ big nor take _that_ long to build. That was one
of the design goals.
- 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)
---------------------------------------------------------------------------