Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1+some 2/3/84; site dual.UUCP
Path: utzoo!watmath!clyde!burl!mgnetp!ihnp4!dual!fair
From: fair@dual.UUCP (Erik E. Fair)
Newsgroups: net.news.group
Subject: The Feeling Of Impending Doom and What You Can Do About It
Message-ID: <746@dual.UUCP>
Date: Wed, 8-Aug-84 22:47:17 EDT
Article-I.D.: dual.746
Posted: Wed Aug  8 22:47:17 1984
Date-Received: Sat, 11-Aug-84 00:13:34 EDT
Distribution: net
Organization: Dual Systems, Berkeley, CA
Lines: 84

A number of people have said that USENET will eventually die of traffic
overload. Too many articles for one site to transfer in a day. Too much
disk space needed to get everything. Such arguments evoke the image of
a frail structure creaking under the weight of a large and heavy
object.

However, we have alternatives. One of them is moderated groups. This is
an argument in favor of that alternative.

Most people that I have talked to, when they hear the phrase
``moderated newsgroup'' immediately think ``censorship'' which is a
word with many nasty connotations. I have a counter-example where
moderation has worked very well, over a long period of time.
It is the ARPANET Digest Mailing List called ``HUMAN-NETS.''
Here is a section from the ``Welcome to Human-Nets'' message I got,
when I joined the list in March of 1982 (Please ignore the reference
to MIT-AI, that host no longer exists on the ARPANET).

--- BEGIN HUMAN-NETS WELCOME EXCERPT ---

The simplest way to get you started with this list, is to tell you a
little bit of the history of the list since it began two years ago.

In January 1980 Al Poskanzer  wanted to start a mailing
list that would consider ways of informing the public about the
potential of telematics technology (ie. the fusion of computer and
telecommunications technology).  Unfortunately in that message he
mentioned publicizing the ARPAnet and its activities as an example.
This resulted in an avalanche of warning and protest mail on 9 Jan.
This avalanche and a flood of mail for SF-LOVERS on the same day
effectively halted two of our machines and resulted in the invention
of the digest option.  Under this option people receive a single
message from HUMAN-NETS once each day. This message is termed a digest
and contains all of the messages which people have submitted about the
current discussions since the last issue. This option was originally
intended only as a stopgap measure until something could be done.
However, over the last eleven months these discussion lists have grown
into a valid research project in their own right.

You submit mail for the digest by addressing it to HUMAN-NETS@MIT-AI.
For convenience and prompt handling administrative requests / problem
reports / etc. should be sent to HUMAN-NETS-REQUEST@MIT-AI.

--- END HUMAN-NETS WELCOME EXCERPT ---

The idea of digests or moderated mailing lists has grown to include
someone actually editing the incoming articles and collecting into a
coherent whole, rather than just batching the day's incoming articles
out to the list once a day.

One of the problems with USENET is that when someone asks an obvious
question, many people reply immediately, and a newsgroup is flooded
with a dozen or more replies containing answers to this question. This
doesn't happen on HUMAN-NETS. The moderator collects all the responses,
throws out those which are inaccurate, and posts the most complete
reply or replies to the question, usually with a notation to the effect
that replies were received from a number of people (``Thanks also to JQ
Public, and Jane Doe for their replies'').

Another of our problems is people posting inappropriate messages to the
wrong news groups. Again, moderated groups solve this, in the same
fashion as the last problem. On a few occasions, I have submitted a
message to HUMAN-NETS and gotten a letter back from the moderator
saying that it was inappropriate for HUMAN-NETS and suggesting another
mailing list to post it in.

Getting back to the idea that moderated newsgroups mean censorship,
they do; but not in the negative sense of preventing communication. It
is more in the sense of a newspaper editor or book editor working
toward enhanced communication. The one problem that is as yet unsolved
on USENET that the ARPANET got over a long time ago is one of trust.
How can you trust the moderator? Well, remember that if a moderator
steps out of line, you can always use the standard `net' groups where
there is no censorship and everything gets to everybody. The trust
comes from having a tradition of good moderation, which we don't have
as yet. I think it's time to establish this tradition.

	after all, isn't this network all neighbors trusting each other?

	Erik E. Fair	ucbvax!fair	fair@ucb-arpa.ARPA

	dual!fair@BERKELEY.ARPA
	{ihnp4,ucbvax,hplabs,decwrl,cbosgd,sun,nsc,apple,pyramid}!dual!fair
	Dual Systems Corporation, Berkeley, California