Re: Support for CSO and gopher type 2

Tim Berners-Lee <timbl@www3.cern.ch>
Date: Fri, 11 Jun 93 13:07:15 +0200
From: Tim Berners-Lee <timbl@www3.cern.ch>
Message-id: <9306111107.AA11007@www3.cern.ch>
To: montulli@stat1.cc.ukans.edu (Lou Montulli)
Subject: Re: Support for CSO and gopher type 2
Cc: www-talk@nxoc01.cern.ch
Reply-To: timbl@nxoc01.cern.ch



What is the status of the CSO protocol?  Should we

	(a) 	Include it in the libWWW access code to allow all W3
		clients to use it
		
	(b) 	Make a gateway so that the existing 100 CSO
		servers are
		accessible
	
	(c) 	Get someone to write an HTTP server for the CSO
		database so that all CSO servers should be
		upgraded?

There aren'tany I know of over here, someone please fill me in on
CSO history?

	Tim
	
Begin forwarded message:

From: montulli@stat1.cc.ukans.edu (Lou Montulli)
Date: Sat, 13 Mar 93 15:18:26 CST

As many of you may know, Gopher file type 2 is a CSO nameserver 

interface.  All of the WWW clients that I have seen do not support
this file type.  Why not?

A page similar to the Gopher index page could be displayed,
with some additional instructions on how to construct a query.

For instance:  To search on all fields simply say "query STRING"
	       To search on a particular field like a name field
	       say "query name=STRING"

The client simply has to send an extra "\nquit\n" to complete the
query, or cut off the connection when the string "200:Ok" or
"501:No matches to your query." is sent by the server. 


Considering the fact that there are over 100 CSO servers available
for these queries, doesn't it seem necessary to support them?

Maybe a new url type is in order:
	"cso://cso.univ.edu:105/"

:lou