>> Article 2954 of news.groups:
>> Path: purdue!spaf
>> From: spaf@cs.purdue.EDU (Gene Spafford)
>> Newsgroups: news.groups,news.admin
>> Subject: Re: Timely Notification (creating groups & other points)
>> Keywords: newsgroup creation, backbone
>> Date: 18 Apr 88 01:41:44 GMT
>> References: <442@white.gcm>
>> Followup-To: news.admin
>> Organization: Department of Computer Science, Purdue University
>>
>> I'm afraid a number of people have the wrong ideas about newsgroup
>> creation.
>>
>> First off, let me try to clear up some misimpressions about the list I
>> keep and about the backbone. The list I keep is no more "official"
>> than is my "position" -- neither is vested with any formal authority.
>> The list is an advisory from me to news admins and readers, and is
>> simply a reflection of groups that a significant percentage of sites on
>> the network keep or transmit. A new site coming on line should expect
>> to be able to get those groups (or most of them) from almost any feed.
>> The list is viewed as authoritative because I try to keep it up-to-date
>> and I try to verify that the groups are likely to be carried. The fact
>> that many, many sites tend to view it as "official" means the admins
>> either don't care, or else they believe it is a pretty close
>> representation of their view of the Usenet reality.
>>
>> In a like manner, the "backbone" is not an official policymaking body.
>> Rather, it is a collection of experienced system admins who run
>> machines with substantial connectivity and news capacity. They
>> represent academic, research and commercial sites. Their opinions as a
>> group are likely to represent the opinions of many other site admins,
>> and as such, something they overwhelming agree to is worth noting. The
>> group collectively has a few hundred years of Unix and Usenet
>> experience, and includes some or all of the authors of A, B and C
>> news. If for no other reason, that is why I tend to listen carefully
>> to anything the group says, and I suspect that is why others pay close
>> attention too. As individuals, the backbone admins don't agree on
>> everything (sometimes they don't agree on anything), but all are
>> concerned with the continuing existence of Usenet. As such, they work
>> together to try to reach a group consensus and then present a united
>> approach...including the issuing and honoring of "newgroup" and
>> "rmgroup" messages.
>>
>> So it is with creating newsgroups. Any news admin can create a group,
>> at any time. The main purpose of the voting procedure the "backbone"
>> has outlined is to try to put some time into the process for people to
>> consider consequences and to allow for comment and debate on the topic
>> and name. The "backbone" as a group will probably ignore newgroup
>> messages for groups that don't follow the procedure for the simple
>> reason that most of the backbone admins (and many others too) believe
>> it is critical to the future of the Usenet that there be some serious
>> consideration given to content and name-space issues before blindly
>> creating new groups.
>>
>> Furthermore, there is *NOTHING* that anyone can do to force a site to
>> carry a group. If 500 people or 50000 people send in votes for a
>> group, that doesn't mean it will be carried by a majority of sites.
>> Thus, groups like "soc.sex" and "net.rec.drugs" that could cause
>> difficulties will likely not be carried by many major sites because
>> their admins don't believe them appropriate. If a group is not going
>> to be carried by a significant percentage of sites, I don't include it
>> in my list-of-lists because that is misleading to people at sites where
>> the group is not received. The backbone is a representative group of
>> experienced system admins and what they carry and their concerns serve
>> as a good indicator of the likelihood of a group being carried. 3/4 of
>> the backbone won't ever carry anything like soc.sex, no matter what the
>> vote, so it will not be in the list-of-lists.
>>
>> Likewise with nonsense or frivolous groups like the eniac.tcp (or
>> whatever) group that Mr. Webber has been making noises about. No doubt
>> many people took the whole idea as a joke.... and either sent in votes
>> "for" because of that, or else they didn't bother to send in votes
>> against. Either way, it is an obvious attempt to create a "bizarre"
>> group in the comp distribution -- something that a number of sites have
>> already indicated they won't carry. I'm sure if someone wanted to take
>> a vote for "rec.lobotomize.webber" then we'd get thousands of approval
>> votes, but I wouldn't add that to the list of active newsgroups,
>> either (although I would the idea tempting).
>>
>> It comes down to "anarchy" I suppose. If you have news admin privs on
>> your machine you can issue a net-wide newgroup message for any group
>> with any name you care to mention. You can do it at any time, without
>> any vote or discussion. However, other news admins can feel free to
>> issue rmgroups for those groups too. Each site is free to ignore
>> either set of messages or honor both. At least within the "standard"
>> namespace, news admins tend to ignore such messages unless they match
>> the list I keep, since it makes it easier to maintain their news. If a
>> group appears in my list, they will probably want to get it, and the odds
>> are good that they will be able to find a feed for it. Groups that don't
>> fit these criteria get created under different names and distributions
>> (like the "alt" groups, for instance) and the admins make separate
>> decisions on those on a per-site basis.
>>
>>
>> So, bottom line: a vote doesn't guarantee that a group will be created
>> and carried by anyone. It does pretty much establish that there is a
>> desire for a group and that it should be considered seriously. It also
>> pretty much guarantees that admins won't immediately issue "rmgroups"
>> if the group shows up after the vote is taken. Groups with real topics
>> that are unlikely to cause problems for admins will usually get added
>> to the list-of-lists and created after a successful vote. And that's
>> about *all* you can say about it.
>>
>> --
>> Gene Spafford
>> NSF/Purdue/U of Florida Software Engineering Research Center,
>> Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-2004
>> Internet: spaf@cs.purdue.edu uucp: ...!{decwrl,gatech,ucbvax}!purdue!spaf
>>
This page last updated on: Jul 1 09:16