Re: WIT

John C. Mallery <JCMa@reagan.ai.mit.edu>

Mail folder: WWW Talk Apr 94-present
Next message: Henrik Frystyk Nielsen: "Re: NON delivery storms!"
Previous message: MATHEWS@vms.acer.edu.au: "NON delivery storms!"
Maybe in reply to: Dave Raggett: "Re: WIT"

Errors-To: listmaster@www0.cern.ch
Date: Sat, 25 Jun 1994 14:11:50 +0200
Errors-To: listmaster@www0.cern.ch
Message-id: <19940625120913.3.JCMA@JEFFERSON.AI.MIT.EDU>
Errors-To: listmaster@www0.cern.ch
Reply-To: JCMa@reagan.ai.mit.edu
Originator: www-talk@info.cern.ch
Sender: www-talk@www0.cern.ch
Precedence: bulk
From: John C. Mallery <JCMa@reagan.ai.mit.edu>
To: Multiple recipients of list <www-talk@www0.cern.ch>
Subject: Re: WIT
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas

    Date: Fri, 24 Jun 1994 10:20 EDT
    From: Dave Raggett <dsr@hplb.hpl.hp.com>

    Phill Hallam-Baker writes

    >    What we need is a convention for typing that link as an "agree" or a 
    >    "disagree" link. It would be easy enoug to modify the mail/news system to
    >    give this capability. Instead of "reply to" or "post" we have a disagree 
    >    and agree buttons producing the lines:-

    >    Disagree <7396@cernvm.cern.ch> 
    >    Agree <7396@cernvm.cern.ch>

    I think we need other buttons to express a degree of interest, such as
	    "Awful", "Poor", "Fair", "Good", "Great"

    Seals Of APproval are a must have feature for discussion tools.

See the WIT entry I put in on peer review schemes.

Meanwhile I have implemented persistent object support for general review
schemes.

It is UNWISE to commit mechanism to a FIXED scheme because there is a large
universe of possibilities here, and no clear criteria for universally
preferring a specific one.

It would be better to provide GENERAL mechanism that can support ARBITRARY
schemes but provide some basic examples for initial use.

So what you want is:

REFERENCE: message-id|urn|url|...

REVIEW-SCHEME: <scheme-name>

REVIEW: <scheme-specific value>

So, you can consider the RESCHERIAN-scheme (agree/disagree) to have five
values, from 1 to 5, with mappings to "Awful", "Poor", "median" "Good",
"Great" You can then center on the midpoint and perform numeric operations to
weight opinions...

If anybody cares about this, we can go into more detail.