Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 alpha 5/22/85; site cbosgd.UUCP
Path: utzoo!watmath!clyde!cbosgd!mark
From: mark@cbosgd.UUCP (Mark Horton)
Newsgroups: net.news.group,net.news.adm
Subject: Re: Suggestions for talk.* groups follows
Message-ID: <2157@cbosgd.UUCP>
Date: Wed, 21-May-86 23:05:57 EDT
Article-I.D.: cbosgd.2157
Posted: Wed May 21 23:05:57 1986
Date-Received: Sat, 24-May-86 17:27:22 EDT
References: <5500@alice.uUCp>
Organization: AT&T Bell Laboratories, Columbus, Oh
Lines: 53
Xref: watmath net.news.group:5645 net.news.adm:724

In article <5500@alice.uUCp> jj@alice.UUCP writes (tounge in cheek, I hope):
>I've read Mark's list of newsgroups that are going (it's not
>a proposal, it's a typical fait-accompli) to be moved to talk.*.

I'd like to emphasize that this is NOT a fait-accompli.  The input I've
seen from the net is important, and I've already changed my opinions
about what should be in the list based on what I've seen.  (I can't
speak for the backbone in that regard, however, I don't know if or
how their opinions may be changing.)  A better comparision would be
the Usenix board, or City Council, conducting an open meeting to get
public comment on a proposed change in something.  The final decision
will be made by the backbone, but the input from the public has a major
impact on that decision.  (Specifically, it seems to me that the final
list of groups to be moved to talk.all is subject to change, although
the general plan to move some set of groups is unlikely to change.)

One other comment - the previous posting was not my work, it was the
result of a lot of discussion among the backbone, and represents the
position of the backbone as a group.  I had input into the process,
but so did lots of others.  But the proposed list was not my creation,
it was primarily derived from lists generated by other backbone SA's.
I don't deserve all the credit, or all the blame.  I'm just serving
as spokesperson for a group of which I'm just one member.

I think jj seems to think that the sole critereon is whether the group
is "work related."  That is not the purpose of the current proposal.
It has been suggested that we ought to use this upheaval to complete
the reorganization, dividing the rest of net.all into a few more top
level categories (tech, sci, rec, and club have been mentioned as some
possible categories, for example.)  In this sense, perhaps we should
include any group in talk which fits that description better than any
of the others.

I think what will wind up happening is that the list of groups in talk
will be viewed as "here are the groups that a fairly large part of the
net views as nearly impossible to justify their cost."  Trying to decide
what goes into that list is difficult, and the metrics we've used are
basically high cost, low readership, and low content.  This process is
imperfect, but we're doing the best we can.  With input from the net,
we'll do better than we would otherwise.  However, the process is
unavoidably political and compromises do occur.

As a rough guideline, about half the backbone intends to carry everything,
another sixth would carry nontechnical nonsoapbox groups, and a third
would carry only technical groups.  This is with the SA's making the
decision - I think you can all imagine what the decision would be if
the decision was made by the third level manager above each SA.

One question it's clear we're going to have to wrestle with is "what do
we do with high cost, high readership, low content groups?" net.jokes
and net.singles both fit this category.  The key question (in the eyes
of those making the decision) is "can we justify the cost, given the
benefit?"  Putting it another way, if you were paying the bills, would
you be willing to pay for them?  Another possibility might be to force
them to be moderated, thereby increasing the signal/noise ratio,
cutting the volume, and hopefully keeping the useful content intact.  I
don't have the answers.

A note to Steve Dyer - if my mail isn't reaching you, call me.

	Mark Horton