Return-Path: X-Sieve: cmu-sieve 2.0 Received: from dns.bl.com (dns.bl.com [207.227.241.66]) by stanley.postilion.org (8.11.0/8.11.0) with ESMTP id g973olY07777 for ; Sun, 6 Oct 2002 22:50:47 -0500 Received: from stanley.postilion.org (cowpoet.postilion.org [192.132.13.218]) by dns.bl.com (8.12.6/8.12.6/1.35) with ESMTP id g973p0c1026088 for ; Sun, 6 Oct 2002 22:51:06 -0500 (CDT) Received: (from joe@localhost) by stanley.postilion.org (8.11.0/8.11.0) id g973ocw07754; Sun, 6 Oct 2002 22:50:38 -0500 Date: Sun, 6 Oct 2002 22:50:38 -0500 Message-Id: <200210070350.g973ocw07754@stanley.postilion.org> To: joe@sfbooks.com X-Also-Posted-To: news.groups Subject: Re: Restrictions on the UVV (was Re: Moderator Elections) References: From: Joe Bernstein Organization: None Status: RO Content-Length: 14396 Lines: 253 Parts of this post are a draft of something that may be posted again (and again, and again), so I'd really appreciate corrections. Thanks. In article , Mean Green Dancing Machine wrote: > In article , > Just Me wrote: > > > >Thank you very much for all this information. It makes the pieces fit > >together better. I had no idea that Tale was a person. That is fascinating > >and I'd love to hear the story how Tale came to be in that position in the > >first place. That sounds like usenet legend material if I ever heard tell! > > Spaf handed him the position: Um, this is actually closer to an urban legend. Gather round, kiddies, while I tell the tale. In the beginning was NET.general. At some point people decided that talking about Usenet was worth a separate space and made NET.news. There are still people on Usenet who were there then - at least Tom Truscott and arguably Steve Bellovin - but they haven't told me anything about this. The archives cover the rest of the story. Changes in news software - more or less, the writing of I think A News - led to the lowercasing of the newsgroup names over the course of May to September, 1981. (Yes, Usenet was then small enough that everyone could actually catch up to a new standard in just four months. Imagine.) So we now had net.general, net.news, and a bunch of other groups. People who wanted a new newsgroup generally started the conversation in net.general if they were polite. (If they weren't, they just started the newsgroup; as you may have heard, all it took to start a newsgroup in those days was to post to it, which resulted in many typo groups. Mark Horton started net.jokes about ten minutes after he proposed it; several proponents as late as 1982 stated in their posts starting groups that previous proposals - which the archives show no evidence of their ever having made - had led to massive acclaim likewise invisible in the archives.) net.general was the group everyone on Usenet was expected to read, and the sense developed that not everyone on Usenet needed to read about newsgroup debates, but the people in net.news didn't want them either. The compromise was to create net.news.group. Jerry Schwarz, later to become famous as the inventor of netiquette and FAQs, came up with the idea, and Mark Horton did it, January 15, 1982. (news.groups shares its birthday with Martin Luther King, Jr., folks...) Meanwhile, the idea had been gathering force that someone should keep track of all the groups that were out there. Mark Horton posted a list of them in December 1981, and a week later a very premature proposal for a great renaming. Curt Stephens started posting lists in January, 1982, eventually posting four of them by autumn. But he stopped over the summer and early fall, and Adam Buchsbaum, then in high school, picked up the slack, posting by far the most complete list of groups to date in November, 1982. Buchsbaum then continued posting these lists for nearly two years. The acclaim almost immediately went to his head, and he became the first rmgrouper known to me one month later, though not all of the groups he killed stayed dead. Unbridled growth was, however, seen as a problem also by people who were not in high school. Mark Horton had already organised a group of sites which were the "backbone" of Usenet, and I think he'd started a mailing list of their admins. At some point he turned the job of maintaining the mailing list, and, in effect, the backbone itself, over to Gene Spafford, then a grad student in Georgia. When Adam Buchsbaum went off to college in August 1984, Gene Spafford volunteered for the list-maintenance job too. Let's pause a moment here for technology. In the early days, Usenet operated by computers calling each other up in the middle of the night and trading copies of files, using the Unix-to-Unix-CoPy program, that is, uucp. The resulting network was therefore the UUCPnet. It was *separate* from the Internet (and in some places still is). The importance of the backbone mailing list was that the admins on it (except for Spafford) each controlled a computer whose phone bills were paid for, a computer which could afford to initiate calls not just to one, but to many, other computers. Hence these computers were the main channels through which Usenet traffic could pass. If a newsgroup sufficiently pissed off the backbone's admins that they agreed to stop carrying its postings, that newsgroup could still survive, but only through back-channel links that cost real money to the sites that usually could just freeload off the backbone. So the group Gene Spafford ran (though was not a member of) was powerful in Usenet; and Spafford joined that power to the most widely respected canon of what was an Official Newsgroup and what was Not. (He also became moderator of various newsgroups for new users, the posting of lists, and so forth.) Many backbone administrators, often feeling heat from their bosses over the size of the phone bills, were convinced that Usenet was out of control. They got actively involved in things like inventing moderated groups (the first one, net.announce, run by Mark Horton, opened up in June 1983). There are group removals scattered all through 1984, the year moderated newsgroups really took off, and I wouldn't be surprised if backbone admins were often involved, but starting in July 1985 such removals became routine, and in October 1985, when one of the groups removed was net.bizarre, the backbone admins' effective rulership over Usenet was fully unveiled; Gene Spafford's posted explanation for net.bizarre's removal was explicitly that the backbone admins wanted it gone, and newspapers, which had previously noticed Usenet, in the protests over "New Coke" and such, as an anarchy, so reported the event: Anarchy Tamed. For the next several years, in fact, rather few unmoderated groups were started; although I haven't researched their creations in detail, I think the usual claim that the backbone admins ran the show is probably basically correct. It's worth noting that the entire fa.* hierarchy could be renamed with basically no input from the moderators of its groups, in the fall of 1985, by one admin. In 1986, the admin who controlled the link to Europe demanded a Great Renaming so he could sort out which groups to send overseas and which not to. Between September 1986 and May 1987, this took place, resulting in the Big 7 hierarchies that are now the Big 8. (At this time, net.news.group - not the bogus newsgroup net.news.groups that bard's FAQ mentions - was renamed to news.groups.) Much anger was unleashed, many people were upset, and in May and June 1987, individual backbone admins acting on their own proceeded to start both the alt.* hierarchy and the inet distribution (an independent list of groups that use the same names as Big 7/8 groups, a separate long story). It took several months before people calmed down, but in September 1987, Gene Spafford started posting lists of alt.* groups, in November he unveiled a newsgroup creation procedure in which people could actually read about the backbone's preferred way to be supplicated (it involved taking a vote), and in January 1988 he started listing inet groups too. Things then went along fairly peacefully for about a year. But in the summer or early autumn of 1988, a proposal for a group called comp.society.women - in comp.* partly to reflect its being aimed at women in the computer-related professions, but partly to get it shipped to Europe - passed its vote, but the backbone admins vetoed it. An angry backbone admin proceeded to send the newgroup on his own. Much more anger ensued, and the admins apparently basically stopped speaking to each other. Two months later, in November or so, Gene Spafford shut the mailing list down. I need to pause for technology again here, because the obvious question is, if the backbone admins were really necessary, why is Usenet still here? - but if they weren't, why could they get away with their takeover for so long? Well, what happened was that more and more of Usenet got hooked up to the Internet, with its own always-on backbone. Suddenly you didn't *need* to pay any long-distance bills to get your netnews; you could freeload off a network so big, it didn't even notice Usenet! Mind, there still *is* a UUCPnet out there getting Usenet the old-fashioned way, but already the inet distribution itself was built on the assumption that most sites were on the Internet, and by late 1988, this was true enough that the backbone admins had lost their power. The one who appears, from what I've learned so far, to have been the most hawkish, Rick Adams, even acquired a motivation for changing course: he'd started the first (?) private ISP, which meant that instead of paying for Usenet's content, he could charge people for access to it. I'm not really clear on what exactly happened next yet, but obviously things could not go on indefinitely in a situation where the official rules for creating a newsgroup called upon the decisions of a defunct mailing list. (Well, they could; but I'm trying to keep current flamewars out of this.) Finally, Greg Woods, who had been one of the most respected of the backbone admins, wrote a new set of rules, the "guidelines" that (albeit wholly rewritten) still run the show. These were first posted in April 1989. The idea that just by voting you could get a newsgroup, without having to beg *any* administrator's special permission, caught on fast, and I gather that proposals multiplied. So in the summer of 1989, Woods proposed a new group, moderated of course, for nothing but proposals and votes and results, and, well, maybe other stuff as it came up, and although I haven't yet found the vote result or charter [1], evidently the group passed; on September 4, 1989, news.announce.newgroups opened for business, with Woods as moderator. Yes, that's what I said. Greg Woods - *not* Gene Spafford - was the first moderator of news.announce.newgroups. If you read the archives of the first few months of nan, you may get the sense - I certainly have - that Woods found the job a trial. In any event, he went on vacation January 16, 1990, for three weeks. He picked Eliot Lear, previously head of the bionet.* hierarchy, as his backup. He never came back. To my mind, Eliot Lear was the real founder of nan. Woods had struggled with what to do with proposals that didn't go through his group, the way the guidelines had said they should ever since September 1989, and had changed course more than once. Lear, I think, followed a fairly steady trend from tolerance for people posting their documents elsewhere, in the beginning, towards firm enforcement of rules by the end of his tenure. Woods had frequently added moderators' notes mentioning this or that violation of the guidelines, and had at one point written "But the poster insists on posting it anyway". Lear was calmer, and I get the sense that he had a better handle on how to tame the new anarchy. He also initiated regular posts noting what groups had been proposed or were being voted on, and what groups he'd recently sent newgroups for. People started wondering just what Gene Spafford was needed for anyway, and Spafford's voluminous posting runs of lists, FAQs for new users, and other stuff became ever less frequent. He had pneumonia and some sort of breakdown in the first half of 1991 and hardly anyone even noticed that the posts were gone, or so he later wrote. But Eliot Lear didn't want to keep doing *his* job forever; he had, after all, only signed up for a three-week stint. Finally, on February 11, 1991, he announced that he was leaving, and that his replacement would be Dave Lawrence , the man better known these days as David C. Lawrence . tale had great plans when he began, though he didn't talk much about them. He indicated almost immediately that he expected to be making major changes sometime soon, an indication that would be repeated roughly once a year for the next five years. Concretely, he definitely *did* start archiving the group, almost right away. My reading of the early months of news.announce.newgroups derives from my own attempt, after Google unveiled the longer archives it now offers, to complete tale's archive by adding earlier material. I have not yet taken the opportunity to read nearly so widely in tale's archive, though it's been available longer, and so my story is almost over. But the crucial piece remains. In April of 1993, Gene Spafford finally decided that he'd had enough. This is when he posted the "farewell" letter that Aahz quoted, as sad a Usenet testament as I have seen. In it, he gave his own account of how he came to be where he was; it conflicts in several places with my account above, but the archives are now available, so people can look into this and decide for themselves. In any event, he did truthfully report what would happen to the posts he posted and the groups he ran. Mark Moraes took over the newbie FAQs and news.announce.newusers. David Lawrence took over the lists and news.lists (since renamed to news.lists.misc). In other words, David Lawrence was now back in the position Gene Spafford had been in from at least 1985 to 1988. He controlled access to the official process for creating newsgroups, *and* he controlled the lists *of* newsgroups most widely seen as official. I don't know how widely this was remarked on at the time; we've certainly gotten awfully accustomed to it since then. But as moderator of news.announce.newgroups, he was *not* Gene Spafford's direct heir, or indeed Spafford's heir at all. And the person who *did* pick him out from the crowd, and start him to the position of prominence he still holds, is too widely forgotten. To the extent that we respect the Big 8 process, we owe some respect also to Greg Woods, and especially to the man who, I think, really made it the stable success it remained for so long, Eliot Lear. Joe Bernstein -- Joe Bernstein, writer joe@sfbooks.com