Subject: Rating systems (was Re: Free software and free music... )
From: "Karsten M. Self" <>
Date: Mon, 13 Sep 1999 00:19:33 +0000

Old thread that I've been meaning to return to

On July 24, 1999 "Stephen J. Turnbull" wrote:
> "kms" == Karsten M Self <> writes:
>     kms> Peer-reviewed journals _do_ offer the service of peer review,
>     kms> though in a webified world, this can be accomplished to some
>     kms> extent by ratings sites substituting for journals.  Another
>     kms> significant
> If by "ratings sites" you mean like slashdot's up down, or Amazon's
> reviews, I think not.  I read articles in a few journals based on
> title; I read articles in a few others because of the authors.  Other
> journals' articles I read only when I am following a citation.  Why
> does a journal fall into one of the former classes?  Because the
> editorial policy is strict and reliably so.
> No-name ratings and access stats are not sufficient.  Editorial
> continuity is crucial (although it may not be the editors, but rather
> simply that the best authors congregate in those venues).

I'm completely with you on this one.  I've been arguing (online and off)
against several of Rob Malda's (aka cmdrtaco) moderation policies.  Two
in particular:  anonymous moderation (I feel moderations should be
attributed) and the prohibition against moderating _and_ posting to the
same forum -- which has been enforced by IP address recently, not merely
user ID.

A better moderation concept IMO would involve tracking both posts and
voted preferences.  One model I've considered would extend Usenet by
maintaining a database of vote preferences -- including the voter's
identifying information (either encrypted key or clear).  Several issues
to be worked out, but a key architecture issue would be to hold all of
the filtering functionality on the client side.  Slashdot gives us
Malda's vision of both the information (moderation data) _and_ the
filtering mechanism.  If you disagree with either you're SOL.

Google might be a very rough cut of an alternative approach.

Karsten M. Self (
    What part of "Gestalt" don't you understand?

SAS for Linux:
Mailing list:  "subscribe sas-linux" to    
  5:10pm  up 2 days, 20:50,  0 users,  load average: 0.12, 0.16, 0.17