Path: utzoo!attcan!uunet!seismo!sundc!potomac!jtn
From: jtn@potomac.ads.com (John T. Nelson)
Newsgroups: news.admin
Subject: Re: USENET is dead! Long live USENET!
Keywords: spines local backbones mail rationalization
Message-ID: <6357@potomac.ads.com>
Date: 27 Jun 88 21:38:26 GMT
References: <316@applga.UUCP>
Reply-To: jtn@potomac.UUCP (John T. Nelson)
Organization: Advanced Decision Systems, Arlington VA
Lines: 113


> From: simmons@applga.uucp (Steve Simmons)
> Subject: Re: USENET is dead! Long live USENET!

> In southeastern Michigan we are taking some steps that a lot of other
> folks ought to emulate.  We are rationalizing news and mail.

> The sysops of the major local system got together, drew a map, and
> have created a southeastern Michigan backbone.  It covers roughly the
> southeast quarter of the lower penninsula, and will do wonders for
> local distribution when we're done.  Six sites are backbones, all
> carrying full feeds, all triply connected.  We have four "outside world"
> connections for news, more for mail.  This connection will end the
> stupidity of mail from Ann Arbor to Detroit going thru Rutgers and
> ihnp4, will reduce almost all sites long distance costs, improve mail
> and news delivery, *and reduce the load that southeastern michigan
> puts on the rest of usenet*.

Well well... at LAST someone has come up with a rational approach!
Instead of prognosticating the death of Usenet, you self-proclaimed
software wizards should be cleaning up the networking code so that it
accomoodates the loss of such a site.  And you site administrators
should fix YOUR MAILERS so that they GENERATE LEGAL MAIL ADDRESSES.

Perhaps we should move away from dependencies on backbones and design
network software with distributed intelligence and routing information
that would tolerate dynamic shifts in the networks topology.  I've
talked to a number of people with some great ideas on how the network
software should be improved.  Alas... who has the time.


From: mack@inco.UUCP (Dave Mack)
Subject: Re: The death of USENET

>>The need is to bring USENET volume back down to tolerable levels -- which
>>I'm somewhat arbitrarily building a cutoff level of a megabyte of news a
>>day. About 1/3 of current levels. 
>>
>>This is going to be painful. For me, personally, it's especially painful
>>because if you look closely, I've targetted just about every USENET group
>>that means anything to me. But these are not times to be selfish. These are
>>times of survival. Which I hope sinks in around the net. But I doubt it.

>This strikes me as a panic reaction. Why don't we wait and see what impact
>the disappearance of ihnp4 actually has?

Chuq makes a good point by claiming that these are times for survival
and not selfishness.  Sometimes you have to amputate to save a life.
However, I think that removing soc.all, talk.all, etc.all is a much
too reactionary measure.  Rather, site administrators should begin to
take more local responsibility and the network software should be
improved to accomodate the network's growing needs.

Much of the noise in the rec, binary and talk groups can be eliminated
by expiring these groups quickly or rejecting them from your site
entirely in your sys files.  My site does this because I don't
tolerate flame-fests.  And since there are too many viruses rampant in
the world, we don't keep a lot of the binary groups around either.

Given this... one can restrict entire groups as Chuq suggests without
hacking the net code and explicitly tearing out the rec and soc
groups.

The prediliction towards deleting the comp, soc and rec groups comes
from the fact that most of us sponge off of our employers or
Universities for access and what employer wants to pay for these noisy
groups?  With only technical groups though, Usenet would no longer
appeal to as wide a range of users.  It would no longer be a Use-net.

>Chuq's approach to the survival of the net represents an attitude which
>I think is far too prevalent. "The net is strained: we have to reduce
>the amount of inessential information." Why not try to *improve* the
>system instead of accepting its deficiencies? C news is a step in this
>direction: let's get a beta version of it out. (Henry? Geoff? Help!)

I agree.  Instead of tearing down the net to accomodate the software,
we should start thinking about the design of that software and why it
can't accomodate the complexity of the network.   Remember that there
are both social and technical complexities involved.

>Maybe it's time for people who care about the net, the *whole* net
>and not just the technical groups, to get together and try to solve
>some of these problems.

I'm afraid people would rather flame about JJ.


From: russ@wpg.UUCP (Russell Lawrence)
Subject: Re: The death of USENET -- Appropriate Medicine

> Some of these arguments may well be valid, but nevertheless, we'll need 
>  to remind ourselves of Chuq's warning that the usenet can't be all 
> things to all people.  Usenet is not the Universe.  It is, however, a 
> wonderful place to exchange ideas about unix.  

Only Unix?  I don't think so.

> Perhaps the sites where net connections are in jeopardy could be
> persuaded to drop non-comp groups as an alternative to dropping out 
> altogether.

Right.  Don't force all of us to adopt the all or nothing alternative.


-- 


John T. Nelson			UUCP: sun!sundc!potomac!jtn
Advanced Decision Systems	Internet:  jtn@potomac.ads.com
1500 Wilson Blvd #512; Arlington, VA 22209-2401		(703) 243-1611

"Hi... My name is Hobbes.  I'm the product of a malicious 5-year old's
twisted and destructive imagination.  Would YOU like to be my friend?"