Xref: utzoo news.misc:1081 news.config:385
Path: utzoo!utgpu!water!watmath!looking!brad
From: brad@looking.UUCP (Brad Templeton)
Newsgroups: news.misc,news.config
Subject: Re: Correction - Re: The USENET Backbone (Last changed: 10 December 1987)
Message-ID: <1284@looking.UUCP>
Date: 5 Jan 88 16:54:03 GMT
References: <2802@arthur.cs.purdue.edu> <14191@oddjob.UChicago.EDU> <19509@clyde.ATT.COM> <1270@looking.UUCP> <2663@killer.UUCP>
Reply-To: brad@looking.UUCP (Brad Templeton)
Organization: Looking Glass Software Ltd.
Lines: 75

In article <2663@killer.UUCP> billw@oberon.LCS.MIT.EDU (Bill Wisner) writes:
>Brad Templeton suggests that every month, the two most costly newsgroups be
>removed. He believes this would result in great savings.
>
>This is ridiculous. You may condemn comp.sys.amiga as being worthless because
>you don't own an Amiga, but (as evidenced by the high volume of that
>newsgroup) it IS useful to a hell of a lot of others. And no, I don't own
>an Amiga.
>...
>But it gets better. Templeton goes on to say that some newsgroups, like
>comp.mail.maps, should be exempted from such a policy of newsgroup genocide.
>
>I find it hard to accept such galling bigotry.

First of all, it's an idea, not a policy.  Secondly, it's quite the oposite
of bigotry.  Some groups I read would get schedule for termination under
such a scheme if they didn't mend their ways.  The Amiga group was listed
because it was up at the top of the list under my (arbitrary) criteria.
I have nothing against Amigas.  Fine machines.

The point is that high volume is NOT a reason to keep a newsgroup in my
opinion.  It's a reason to kill one.   I have long advocated that group
creation rules should require a group to demonstrate it will have
LOW volume.

The Amiga group was my sample pick because it is one of these largest
volume groups on the net, and the readership is low to medium.

The real first pick would be "talk.religion.misc".  Probably around 1500
readers, and around 1.2 megabytes of stuff per month.  Now Brian thinks
that costs $713,000 per month, and he's way off, but it's still costing
around $10,000 per month, or $120,000 per year.

The point is not that religion isn't worth talking about.  In this case
the possible lesson is that "religion isn't worth talking about on the net"
As anybody who has read groups on "religious" (in the broad sense, as in
text editors are a religion) issues knows, they are 99.9% heat and .1% light.
If it cost nothing, who would care?   But at $120,000, what's the point?

The next group on the list might be comp.mail.maps.  I suggested that would
be exempt because the "readership" figure for this group is deliberately
low.  This group is not read by people, but by programs, and it exists for
net maintenance.  Mind you, it could do with some reducing in volume.

Next is soc.culture.china.  That one gets exempt (and, as you can see, I think
we should be generous with exemptions) because it's a new group, and they're
hashing things out for the first time.

After that soc.culture.jewish.  No doubt I'll get slammed for listing this
in order, even with my Jewish ancestry.  But I can't see a lot to exempt this
one.  That's just an opinion of course.

That's followed by comp.sys.amiga.  VERY high volume (2.2 megs, over $200,000
per year) but a larger following of around 6,000 readers, I'm guessing.

etc. etc. etc.

Point would not be, as I first lightly suggested, to delete each group right
away.  A better idea would be to consider such groups on probation, with
a message delivered to the posters in these groups along the lines of:
	"Start being more careful with your posting volume, or the rest of
	the net will become less tolerant of it, and may eventually want
	your group removed"
If this were a serious threat, the hope is it would calm down the people
in the group.

Last month, talk.bizarre was only 15th on this "kill" list.  One would almost
be tempted to put it to #1 after seing a message there that read,
	"We're almost at the top of the volume stats!  Come on guys, keep
	posting, keep posting!"  (paraphrased from memory)
Is this what we desire?


-- 
Brad Templeton, Looking Glass Software Ltd. - Waterloo, Ontario 519/884-7473