[news.admin] Too much for the reader?

andrew@teletron.UUCP (Andrew Scott) (06/30/88)

In article <Jun.29.13.41.34.1988.18001@porthos.rutgers.edu>, webber@porthos.rutgers.edu (Bob Webber) writes:
> In article <385@teletron.UUCP>, andrew@teletron.UUCP (Andrew Scott) writes:
> >...
> > The net is just too big to handle as a *reader*.
> 
> You know, I was down at the Library Of Congress the other day and noticed
> that they had more books than I could read in a lifetime (even if I
> stopped reading the net).  They have just gotten way too big to handle
> as a *reader*.

Cute analogy, but there is a major difference between the Library Of Congress
and USENET.  When I go to a library, I use the card catalog system to find
exactly what I'm looking for.  On USENET, things aren't nearly so organized.
One could scan for subject lines or keywords, but very few people bother to
add a keywords line and many subject lines are undescriptive.

With the current news software, a reader is left with having to scan through
newsgroups, "n"-ing over articles that don't look interesting.  I wouldn't
use a library at all if I had to do the analogous operation:  walk down the
aisles looking at the title of each book.

Perhaps a solution to the net explosion is better news reading software and
a better organization of articles.  I'm not so sure that either will appear
soon, as a keyword based news reader has been proposed for years with not
much to show for it, and some people insist that it doesn't really matter
what newsgroup names we pick because users will find the discussions anyway.
-- 
Andrew Scott		andrew@teletron.uucp    - or -
			{codas, ubc-cs, watmath, ..}!alberta!teletron!andrew

webber@porthos.rutgers.edu (Bob Webber) (07/02/88)

In article <394@teletron.UUCP>, andrew@teletron.UUCP (Andrew Scott) writes:
> In article <Jun.29.13.41.34.1988.18001@porthos.rutgers.edu>, webber@porthos.rutgers.edu (Bob Webber) writes:
> > In article <385@teletron.UUCP>, andrew@teletron.UUCP (Andrew Scott) writes:
> > >...
> > > The net is just too big to handle as a *reader*.
> > 
> > You know, I was down at the Library Of Congress the other day and noticed
> > that they had more books than I could read in a lifetime (even if I
> > stopped reading the net).  They have just gotten way too big to handle
> > as a *reader*.
> 
> Cute analogy, but there is a major difference between the Library Of Congress
> and USENET.  When I go to a library, I use the card catalog system to find
> exactly what I'm looking for.  On USENET, things aren't nearly so organized.

Obviously you have never been down to the Library of Congress.  The
card catalog is about as useful as newsgroup names are.  To find what
you want in a research library is not a simple matter of ``looking it
up in the card catalog.''  USENET is beautifully organized.  Each
message is available for any sort of online analysis you want to apply
to it.

> One could scan for subject lines or keywords, but very few people bother to
> add a keywords line and many subject lines are undescriptive.
>
> With the current news software, a reader is left with having to scan through
> newsgroups, "n"-ing over articles that don't look interesting.  I wouldn't
> use a library at all if I had to do the analogous operation:  walk down the
> aisles looking at the title of each book.

Actually, walking down the isles is the way I find MOST useful in using
a library.  However, on any unix system there are plenty of automatic
tools for searching and processing text files.  You could let your cpu
do alot of sifting for you.  Ignore the headers -- process the raw text!
Ask other users on your system to flag articles they found interesting.
Summarize discussions that you found helpful and keep the summaries available.
Don't kill the flow of the net just because YOU don't have time to swim
in it.

----- BOB (webber@athos.rutgers.edu ; rutgers!athos.rutgers.edu!webber)

andrew@teletron.UUCP (Andrew Scott) (07/05/88)

In article <Jul.2.03.16.21.1988.8104@porthos.rutgers.edu>, webber@porthos.rutgers.edu.UUCP writes:
> 
> However, on any unix system there are plenty of automatic
> tools for searching and processing text files.  You could let your cpu
> do alot of sifting for you.  Ignore the headers -- process the raw text!

This is interesting.  I'm curious as to how *you* read news, Bob.  If there
is a Better Way to read news, I'd like to try it out.

This is an honest question.  I'm not trying to be belligerent, I really am
interested in reading news in a more efficient way.

> Don't kill the flow of the net just because YOU don't have time to swim
> in it.

I'm assuming that there are other people out there who feel the same way.
(In fact, some of the email I've been getting does agree with what I've said.)

There are several reasons as to why I'd like to see more moderation, archive
service for sources and binaries, and more efficient software.  Most impor-
tantly, the net *will* continue to grow unless we do something.  Some people
have said "if you don't want it, don't carry it."  This is a non-solution.
Taking to it's logical extreme, soon all I'll be able to carry is a dozen
newsgroups of 200K/day each.  Even in the short term it's not a good solution.
Why?  Because I'm *obligated* to carry those groups.  In exchange for my
newsfeed, I provide one myself.  I would be doing any downstream sites a
big disservice by cutting them off from groups I don't want.  I think that
a great many sites look to Spaf's list for guidance, rather than picking
and choosing individual groups themselves.  Why not let them get the groups
elsewhere?  It may not be practical due to local newsfeed policies in that
particular region.
-- 
Andrew Scott		andrew@teletron.uucp    - or -
			{codas, ubc-cs, watmath, ..}!alberta!teletron!andrew

fair@ucbarpa.Berkeley.EDU (Erik E. Fair) (07/06/88)

In the referenced article, andrew@teletron.UUCP (Andrew Scott) writes:

	Most importantly, the net *will* continue to grow unless
	we do something.  Some people have said "if you don't want
	it, don't carry it." This is a non-solution. [...] Because
	I'm *obligated* to carry those groups.

Whoa! Please show me the obligation that I signed that forces me
to commit the resources of my system for your benefit?

Sure, the network will grow. There's no limit on the users because,
by & large, they don't pay for it directly.

But there is a limit to disk space, CPU cycles, and communication
bandwidth. They are instanteously finite at each and every site,
although each site has different upper bounds for each resource.

The USENET as a whole has been fortunate in the past that continually
improving technology has outpaced our demands on these resources.
Cheap 1200 baud modems, the public domain Lempel-Ziv Compress
program, cheap 2400 baud modems, Internet/NNTP, and now cheap
9600/19200 baud modems are making the continued growth of the
network possible. Disk space and CPU power get cheaper every year
too.

Unfortunately, any site's limits are not likely to change all that
quickly, and sometimes new technology doesn't fall within the
purchasing range of a particular site quickly enough.  When that
happens, there is only one thing they can do: cut back on the news
they get to strictly those newsgroup hierarchies or individual
newsgroups that they're interested in.

If they happen to have downstream feeds, those systems have three
options:

	1. chip in to buy more of whatever the scarce resource at
		their feed is.

	2. get the news that *they're* interested in from elsewhere.

	3. bitch at their feed for not fulfilling their "obligation."

Of course, trying option three is likely to get you some self-centered
S.O.B. like me who'll ask you what the hell you're talking about.

What's really going on is that every site on USENET (with some
exceptions) is doing every other site on USENET a big favor, and
occasionally some of them can't afford the favor any more. If you
were trained properly by your parents, you'll thank them nicely
for the favor you got, and offer to return it some day. I see this
happen a lot more often than I see system admins screaming about
being cut off.

Where does this ultimately lead when we all can't afford the big
favor?  Balkanization of the USENET:  some sites carry "comp".
Some carry "rec".  Some carry "alt".  Some carry "gnu".  Some carry
"bionet".  und so weiter.

And you get to play diner at the smorgasboard. Be careful - take
too much and you'll get indigestion.

	good eating,

	Erik E. Fair	ucbvax!fair	fair@ucbarpa.berkeley.edu

andrew@teletron.UUCP (Andrew Scott) (07/08/88)

In article <4278@pasteur.Berkeley.Edu>, fair@ucbarpa.Berkeley.EDU (Erik E. Fair) writes:
|>In the referenced article, andrew@teletron.UUCP (Andrew Scott) writes:
|>
|> 	I'm *obligated* to carry those groups.
| 
| Whoa! Please show me the obligation that I signed that forces me
| to commit the resources of my system for your benefit?

Well, I never signed anything, but an informal "handshake" obligation _does_
exist between our site, our upstream feed, and our downstream feeds.

| What's really going on is that every site on USENET (with some
| exceptions) is doing every other site on USENET a big favor, and
| occasionally some of them can't afford the favor any more. If you
| were trained properly by your parents, you'll thank them nicely
| for the favor you got, and offer to return it some day. I see this
| happen a lot more often than I see system admins screaming about
| being cut off.

I can see this happening too.  It's unfortunate, as I believe that there
are "Better Ways" of doing things.  I'm still not convinced that a steadily
increasing volume on USENET is a Good Thing.  I'd like to carry everything.
I'd much rather see a reduction of volume in each newsgroup than a reduction
of newsgroups carried by my site.  Perhaps I'm just too selfish.
-- 
Andrew Scott		andrew@teletron.uucp    - or -
			{codas, ubc-cs, watmath, ..}!alberta!teletron!andrew