gnu@l5.uucp (John Gilmore) (09/09/85)
In article <781@vortex.UUCP>, lauren@vortex.UUCP (Lauren Weinstein) writes: > It occurs to me that we're trying to hold back an ocean with > a sponge. Gang, there ain't NO WAY we can make this work. > It's doomed. ...though we may win occasional battles, I think we've already > lost the war. The patient is terminal and cannot be prolonged > indefinitely. I've seen this happen for four years running now...every time the colleges start up and a lot of junk comes in until people learn the ropes, we start hearing bells tolling for the Usenet. I just set up my system here in San Francisco and am getting fed by ptsfa. Batched and compressed, at 1200 baud, it doesn't seem to be a big load on the system, except for disk space. Frankly, I notice the mail traffic a lot more. The volume is up, and the tools are evolving to deal with it. In the "good old days" when there were 40 sites, there was no compression. There was no "rn". There wasn't even a "vnews"! Of course, the evolution is piecemeal and fragmented -- that's a problem of volunteer projects. But in a thousand sites, it's also easier to find volunteers and split up the work (eg, newsgroup moderators, code hackers, routing designers, net mappers, group czars, satellite experimenters...). > Plus we have the increasing number of small nodes which will > eventually swamp the larger sites in terms of sheer numbers. People are also assuming that the "deluge" of little sites will degrade the quality of the net below that maintained by the existing "large" sites. This sounds like a latent prejudice against micro users; why is a Vax wizard presumed to be more courteous than an Apple wizard, or a computer salesman, or a kid hacker? I think that the new users are MORE amenable to forming good manners, if we will present them with guidelines early enough in their exposure to the net.
lauren@vortex.UUCP (Lauren Weinstein) (09/10/85)
The reason I presume the influx of small machines will degrade netnews is quite clear. They could easily outnumber all of the big machines in a very short order and represent many thousands of new users. So the total number of users on the net grows immensely. As the number of users grow, we will get more and more people (absolute quantity) who don't give one damn about following the rules and who even enjoy causing trouble. If 1% of the population falls into one of those categories (and it's bound to be more than that) that's still enough people (in a numerical sense) to flood us with garbage. If 1% of the people on the network are "jerks" or even if they don't follow the rules by accident (and there will always be new people who will have the same "accidents") it still adds up to a large of amount of nonsense traffic in an *absolute* sense. Remember, even ONE PERSON, if they set their mind to it, could bring netnews to its knees and be extremely difficult to track down if they were clever enough. As we grow, we're going to start finding people (just like hardcore phone phreaks) who ENJOY causing that sort of trouble. When we have 500K or 600K people on this net, how is ANYONE going to afford (money or time!) to wade through even the GOOD stuff? How will anyone be able to ask a question in an unmoderated environment? So there are really two different issues here: 1) The small percentage of jerks that could cause us real trouble on purpose, who will grow in absolute numbers and take advantage of our "open door free-for-all" policy. 2) The systemic problems associated with sending EVERYTHING that ANYONE posts to EVERYWHERE (in most cases) regardless of the fact that many messages simply duplicate others, consist largely of copies of other messages, etc. This latter problem won't be solved by better transmission techniques, in fact it might only make things worse by encouraging MORE traffic! Only some fundamental change in the way we view the material can help in the long run. My opinion, anyway. --Lauren--
chuqui@nsc.UUCP (Chuq Von Rospach) (09/11/85)
In article <85@l5.uucp> gnu@l5.uucp (John Gilmore) writes: >In article <781@vortex.UUCP>, lauren@vortex.UUCP (Lauren Weinstein) writes: >> It occurs to me that we're trying to hold back an ocean with >> a sponge. Gang, there ain't NO WAY we can make this work. >I've seen this happen for four years running now...every time the >colleges start up and a lot of junk comes in until people learn the >ropes, we start hearing bells tolling for the Usenet. >The volume is up, and the tools are evolving to deal with it. In the >"good old days" when there were 40 sites, there was no compression. >There was no "rn". There wasn't even a "vnews"! It may well be the John has a point. A lot of the 'old-timers' may be suffering from "Golden Age" syndrome, and bemoaning the lost times of the good old days. We HAVE gotten this far, mind you, and when I look at where we were a couple of years ago, I'm amazed. Of course, that was then, and this is now, and doomsday really IS around the corner this time... Really... Trust me. Well, trust me anyway... One thing that DOES worry me is that the problems that I see are no longer the technological ones, they are the humanistic ones. Usenet has always been able to come up with the better widget when needed (rn, compression) and network technology will always keep ahead of technological overload. What we need to worry about is overloading the human circuits. I think this will require a shift in our emphasis from newer widgets to better widgets. Rn is a great first step in this direction, but it may just be time to evolve away from the concept of the newsgroup completely into something else. I'm playing with the idea, but I don't have any firm concepts yet. I wouldn't be at all suprised to see the net survive. I WOULD be quite suprised to see USENET survive, because it simply wasn't built for this size of a network. We need to find a solution that is, and that solution is somewhere in the user interface -- making it easier for the user to find what they need. -- Chuq Von Rospach nsc!chuqui@decwrl.ARPA {decwrl,hplabs,ihnp4}!nsc!chuqui An uninformed opinion is no opinion at all. If you dont know what you're talking about, please try to do it quietly.
rcj@burl.UUCP (Curtis Jackson) (09/12/85)
In article <85@l5.uucp> gnu@l5.uucp (John Gilmore) writes: >I just set up my system here in San Francisco and am getting fed by >ptsfa. Batched and compressed, at 1200 baud, it doesn't seem to be a >big load on the system, except for disk space. Frankly, I notice the >mail traffic a lot more. > 1) Does ptsfa call you local? If so, where do they get their news? How has my article from the East Coast reached your machine? 2) What system(s) do you feed? 3) Assuming you feed anyone systems at all, do you call any of them long distance, Daddy Warbucks? The major load we are talking about is not disk space or CPU time, although they do contribute. We are talking about PHONE TIME. I have received flak before here at AT&T for using our corporate phone network in the wee hours of the morning to send news. This network is totally within the company and doesn't "cost" AT&T anything that I know of over and above normal long distance line maintenance. It has to be there anyway because we ship REAMS of data back and forth with Bell Labs in Whippany, NJ (also on our project). Now, if I am getting flak here with all that going for me, just what do you think is happening on a system like decvax that provides incredible amounts of phone $$$$ (what was it running, $100,000/month for them??) ?? Think about it, -- The MAD Programmer -- 919-228-3313 (Cornet 291) alias: Curtis Jackson ...![ ihnp4 ulysses cbosgd mgnetp ]!burl!rcj ...![ ihnp4 cbosgd akgua masscomp ]!clyde!rcj
preece@ccvaxa.UUCP (09/16/85)
> The reason I presume the influx of small machines will degrade netnews > is quite clear. They could easily outnumber all of the big machines in > a very short order and represent many thousands of new users. /* > Written 12:37 pm Sep 10, 1985 by lauren@vortex.UUCP in > ccvaxa:net.news.group */ ---------- Not to deny that the net is unlikely to continue in its present form, but I think there's a logical inconsistency here. The small machines deluge consists of small machines, with only one or a few users. It's hard to see how they're going to have much effect on the net compared to the big sites with hundreds of users each. They WILL have an effect on the nature of the net, though: it will be harder and harder to find feeds. The current distribution method, coupled with the current volume, makes feeding to small sites just as expensive as feeding to large sites. It's hard to see how even the friendliest news feed could go on supplying feeds to all who ask when there are dozens or hundreds of one-user Unix sites in the local environment. Much as I dislike the idea of mandatory oversight that has been proposed for stargate, I think some kind of mass distribution system is going to turn up eventually, operating over some kind of high bandwidth medium. Cable tv? FM radio? Whatever. Then everybody can get the bulk feed and the only phone costs will be for passing messages into the distribution. -- scott preece gould/csd - urbana ihnp4!uiucdcs!ccvaxa!preece
hsd@uvacs.UUCP (Harry S. Delugach) (09/19/85)
I am a fairly new USENET news reader (approximately one year). While I have
no solution to newsgroup proliferation, I want to make the following points:
First, I don't believe the problem would disappear if ONLY we had the *right*
software. There are lots of programs for reading news, and that won't change
in the forseeable future. New software might ease the problem for *some* sites,
in *some* situations, but isn't this avoiding the real issues of just what
should be on the net in the first place?
Secondly, adopting a more finely grained newsgroup structure may tend to
discourage me from browsing in new groups, because I must pinpoint my personal
interests down to a detailed level, to the exclusion of all others. By analogy,
when browsing through the Sunday newspaper, I sometimes read articles whose
subjects are normally outside my usual interests if something about the headline
or caption grabs me. This is often how my interests are broadened.
Lastly, with regard to all the "noise" on the net, I offer this: When I was
first given an account, our system administrator gave newcomers an
orientation. Part of his spiel was an introduction to USENET. He gave us
a short and simple admonition --
READ a newsgroup for a while before POSTING to it.
Wouldn't this simple rule go a long way in cutting down the "noise"?
--
Harry S. Delugach University of Virginia, Dept. of Computer Science
UUCP: cbosgd!uvacs!hsd CSNET: hsd@virginia