Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10    5/3/83; site cornell.UUCP
Path: utzoo!linus!philabs!cmcl2!floyd!vax135!cornell!hal
From: hal@cornell.UUCP
Newsgroups: net.news
Subject: Re: digests, etc.
Message-ID: <4906@cornell.UUCP>
Date: Sat, 30-Jul-83 19:00:17 EDT
Article-I.D.: cornell.4906
Posted: Sat Jul 30 19:00:17 1983
Date-Received: Mon, 1-Aug-83 11:46:14 EDT
References: <76@vortex.UUCP>
Organization: Cornell Computer Science
Lines: 34

There seem to be two major points raised here: (1) How can we cut
down on the number of redundant and useless messages, and (2) how
can we cut down on the cost of forwarding the news.  I'd like to
make two modest suggestions:

(1) Many of the duplicate articles are caused by multiple (often
identical) answers to a posted question.  We might be able to
reduce the volume if we establish the following custom:  when
someone posts a question, all responses should be mailed to that
person, who is then responsible for posting one (1) note with the
answer--if it is worth posting at all.  Persons posting followups
directly to the net instead of replying to the person who asked
the question should receive gentle reminders from those on the
net who are irritated by the followup--unless there is a good
reason for a followup, such as being unable to mail a reply.  Of
course, this doesn't apply to open discussions in groups like
sf-lovers, politics, and (partially) unix-wizards.  This isn't
as effective as a digest, but then it is much easier to do and
might work in the current USENET environment.

(2) About phone charges:  Has anyone looked into using WATS lines
to distribute the news?  It might help if some of the backbone
sites had outgoing WATS lines and charged each of the sites they
feed a portion of the cost.  Would this be cheaper than the present
mechanism of using regular long-distance phone lines?  It would
certainly be cheaper than a satellite link.  If this isn't practical,
please don't throw bricks at me--it just seems like it would be worth
looking into.

A final thought: it is possible to fit lots of network addresses in
a reasonably compact signature...
                           
Hal Perkins                    UUCP: {decvax|vax135|...}!cornell!hal
Cornell Computer Science       ARPA: hal@cornell  BITNET: hal@crnlcs