GUMBY.MIT-OZ@BRL@sri-unix (10/25/82)
From: David Vinayak Wallace <GUMBY@Mit-Mc>
From: Ben Goldfarb <goldfarb.ucf-cs@Udel-Relay>
Subject: Re: For Sale messages
To: Charlie Strom <CSTROM@Mit-Mc>
Cc: info-micro at BRL
OK, I realize Frank's position is valid, but I wanted to stress that
as atomistic as Usenet is, there will be difficulty in enforcing any
such sanctions. I, too, would feel the loss if the ARPA connection
were severed. But, remember, new people join Usenet all the time and
there is no rule book, no briefing, no single person to serve as an
information source for problems and general queries. Thus, someone is
liable to make the same blunder again.
I guess Frank has done the right thing by issuing a stern warning. At
least a few folks on net.micro have no doubt taken notice.
I hate to thrash a dead horse, but if there continue to be problems
(as you mentioned, Ben, USENET is impossible to "police"), we could
reconsider digesting before considering severing the link to
net.micro. It's a lot of work, most of it redundant, but it would
answer most of the fears anyone might have. Personally, I don't like
the idea of any one person censoring (however necessary) the list,
however if it is impossible to inform new members of the policy, then
such things must happen.
How does USENET work? do new members have to send to a
net.micro.request to get on the list? whoever handles that could
perhaps inform new members of the policies.
david
-------
-------george (10/26/82)
References: sri-unix.4021 USENET is not a mailing list. It is accessed as a public data base. One does not have to request that someone send them the articles. On many systems the default is to subscribe to all groups. More recent versions of the software (everyone does not use the same software or same versions) provides a facility for a site selected message to be printed when someone attempts to post an article, giving the person a chance to abort the submission. Possibly site administrators at all such systems could include a warning about advertising in such contexts via this mechanism. This, of course would require the efforts of a very large number of people and will not work on all versions of of USENET software. George Rosenberg idis!george
goldfarb.ucf-cs@Udel-Relay@sri-unix (10/27/82)
From: Ben Goldfarb <goldfarb.ucf-cs@Udel-Relay> Your suggestion re issuing a new subscriber a set of ground rules would be reasonable, but Usenet does not handle subscriptions the way ARPANET does. One does not send his address to net.micro.request and ask to be included in the distribution, rather his host subscribes to Usenet news in toto and the individual user creates his own subscription list which the local system then administers without human intervention. When a new site joins Usenet, it gets no old archives unless it finds an existing site that will voluntarily archive and distribute old news. Polling is done between hosts at their mutual convenience. When you get right down to it, Usenet is a large, ramshackle, anarchistic message-switching system composed primarily of Bell Labs and the nation's Universities. (no flames here, please -- use net.misc).
wagner (10/30/82)
Sorry to bore everyone on USENET who has seen this stuff before, but by the nature of the discussion, and recent comments by Mark about getting stuff into ARPA, I have no other way of getting this to the appropriate people. When Ben said that there was no central person or agency to funnel people through on ARPA, he meant it. There is no list to get onto...news flows to all sites (in theory), and each user at each site chooses to read the stuff (by newsgroup) or not. USENET will not be able to police its traffic in an absolute sense...it relies on good will of the participants. However, for the USENET people, what ever happened to the idea of submitted to each newgroup an article that never expired that listed the purpose and ground rules of each group. In a few weeks when all the other mail vanished, this article would become the first that the user read in the news group, and he could at least be a well-informed breaker-of-the-rules rather than able to plead ignorance. As an aside, this discussion probably belongs in net.news (it has next to nothing to do with micros, so the arpa discussion is breaking our informal guidelines for content of *this* newsgroup), but of course, ARPA people arent gatewayed into that group. More evidence of the difficulty of glueing together two different topologies? Michael Wagner, UTCS