Lists of newsgroups were first posted, as best I can determine, in late 1981, roughly two years after Usenet began. There is much evidence that these lists were intended to be in some sense "official" from an early date. I see little sign that the posters merely wanted to inform Usenet users of the range of groups available; rather, each of them explicitly advocated listing as a means of centralising authority over newsgroup creation and carriage. Both the informational and the authoritative roles of the lists appear to have met with approval early on - not unanimous approval, but still plenty of it - as you will find if you do searches on their authors' names in appropriate groups. This post begins a series in which I summarise such information available from these lists, and from certain allied posts, as contributes toward a chronology of Usenet newsgroups. At this time, I intend to confine my overall chronological work to the Usenet hierarchies often seen as "official", and covered by these lists - in chronological order, NET.* (1979-c. 1981); net.* (the c. 1981-1987 version); fa.* (from 1981 to 1985); mod.* (the 1984-1987 version); talk.*; misc.*, sci.*, soc.*; comp.*, news.*, rec.*; and humanities.*. (I don't yet know whether these lists also covered the trial.* hierarchy, which will determine whether I attempt to.) This doesn't mean I intend to confine this project to "official" groups, that is, groups in these lists, but the lists are in fact my starting point. As it happens, there are lists of newsgroups in some other hierarchies from later years, but at this time I don't intend to pursue this project into other hierarchies. (For three reasons. First, I doubt there's adequate archival support for this sort of research on other hierarchies - this holds particularly for non-English-language hierarchies. This is my main reason for not attempting most other hierarchies. Second, in some hierarchies, most obviously alt.*, newsgroup lists are intrinsically less trustworthy anyway. Third, generalising from the second reason, it strikes me as inherently wrong-headed, missing the point, to try to enumerate alt.* groups systematically, although the history of specific alt.* groups or sub-hierarchies does interest me.) At *this* time, in any event, I'm confining my results posting mostly to what can be learnt from the lists of newsgroups, and their two main offspring, checkgroups messages and lists of moderators. At a later date, I intend to add evidence from other sources (notably control messages, first or oldest surviving posts, and FAQs) to produce more rounded chronological pictures. But my record for follow-through on net-related projects is fairly poor, so I wanted to start presenting the first-stage results now, rather than wait for more "final" ones. The summaries I'm posting here take two forms. In this post and others, I proceed chronologically, list by list. In a separate set of posts in the same thread, I proceed hierarchically, group by group. This approach is focused on net-wide events such as the removal of entire hierarchies; the other approach is focused on group-specific issues. (In particular, if you don't know what some of the groups listed below were about, *please* find them in the appropriate hierarchy summary, rather than post to ask. I *do* include newsgroup descriptions for each group in the hierarchy summaries.) This initial series of postings covers the years 1980 to 1986, and the hierarchies NET.*, net.*, fa.*, and mod.*. My second series of chronological postings, which I may not be able to post in the same thread, will cover 1987 to 1993 or 1994 (the years of the "Big 7"). I'm not yet sure whether I'll offer incomplete hierarchy summaries for the Big 7 in the second series. I have some reason to think that the second series will work from considerably smaller data sets than this first series, but am not yet persuaded that this is true, so I have no estimate as yet for when the second series of postings will appear. My third series of chronological postings, which will definitely coincide with a set of hierarchy summaries as well, will cover the Big 8 years 1994 to 2002. However, I know of no lists of newsgroups for 1997, early 1998, late 2000, or 2001, so the third series will offer considerably worse granularity than the first two. I *may* have a paper copy of a list from around August 1997, which I used in producing the "List of Inactive Newsgroups", but do know for certain that I erased the copy I had kept on computer. The list I have for January 2nd, 2002 is a list I made myself. If anyone else has homemade lists from these dark periods, please let me know. Joe Bernstein List of Newsgroups Distributed in 1980 An A-News distribution plausibly claimed as that provided at the Delaware Usenix of the summer of 1980 is available online, and includes various files dated June 5, 1980 or thereabouts. One of those files names a set of newsgroups. This could simply be a minimal auto-subscribe list, so to speak, but I have various reasons for thinking it is actually complete. Added: NET.general, NET.v7bugs, NET.news. These groups were evidently renamed to, respectively, net.general, net.v7bugs, and net.news at some time between November 11, 1980 and May 11, 1981. For a much more detailed discussion with full evidence, please see the followup post with Subject line "Early lists of newsgroups". Summary: 3 NET.* groups. Lists of Newsgroups Made or Posted in 1981 MARK HORTON "network name change is in effect" May 12, 1981 net.general Google Message-ID: anews.Aucbvax.1340 (The header was converted from A-News style by Google.) This is the announcement of the creation of the fa.* hierarchy. Note that the groups in the hierarchy had already existed as NET.* groups; this post merely announces their renaming. I'm not entirely clear on the matter, but it's my impression from available evidence that at the time of this post Mark Horton was still involved in the gatewaying of the relevant groups from the ArpaNet; thus, I'm inclined to take this post as truly representing the creation of the hierarchy. Note that there is archived discussion (search on "fa" or "renaming") from the preceding day, concerning this (at that time not yet executed) renaming. It is not, by intent, primarily a list of newsgroups, but since it does include a list of the relevant groups, it can be used as such. It covers, however, the fa.* hierarchy only. Since this is the first list, all groups listed are here shown as "Added". I haven't found any comparable post listing the groups renamed from NET.* to net.*, unfortunately; my guess would be that this renaming occurred before the archives begin, but I'd be pleased to find otherwise. (It's sheer luck that we have this post, itself one of the oldest in the archive.) Contradicting my guess, the list of newsgroups in the old A-News Archive showed all hierarchies in capital letters (i.e., FA.apollo, NET.applic); but this strikes me as probably an anachronistic usage, since this post seems to assert that no FA.* hierarchy ever actually existed. The original headers which could settle the matter are not now available via Google, of course. Added: fa.arpa-bboard, fa.info-cpm, fa.sf-lovers, fa.human-nets, fa.info-terms, fa.info-micro, fa.arms-d, fa.energy, fa.unix-wizards, fa.teletext, fa.home-sat. Assumed still available: net.general, net.v7bugs, net.news. Summary: From 0 to 11 fa.* groups, from 3 to [3] net.* groups, from 3 to 14 in all. INFORMATION LOST Two different respected Usenet participants posted, in 1995 and 1996, a list of newsgroups existing in 1981, their copies of which they traced to Steve Summit. See please Joel Furr "Early USENET" April 28, 1995 comp.society.folklore Message-ID: 3nq619$fi5@shell1.best.com which provides the date November 29, 1981, and David Wright "Re: What's a Cabal? (was Re: Can you mailbomb email spammers?)" November 21, 1996 news.admin.net-abuse.misc,news.groups Message-ID: 572o9tINNge1@bhars12c.bnr.co.uk (Only the former is included in the relevant quarterly file at my website.) Although David Wright provides details on how the list got to him, neither he nor Joel Furr provides further details on the list itself, and no original Usenet posting of this list is archived at Google (or elsewhere, to my knowledge). I have, therefore, no evidence as to who compiled it or on what basis, nor where it was made available or with what claims to authority. It could be simply a list of groups carried at a single server (although most such lists in the early 1980s included groups outside net.* and fa.*, such as "general" or "junk", and this one doesn't). I can assert that it omits groups which certainly existed by November 29, 1981 (such as net.periphs and net.4bsd-bugs), but includes a group created November 13, 1981 (net.applic). I can also assert that it's minimally consistent with the development of the official namespace, by which I mean that groups on this list, that aren't on the next one, are groups that were in fact renamed in between. It's worth noting that in the fall of 1981 there were a number of group creations that led to namespace debate - net.chess, net.columbia, net.scuba, the sub-hierarchies net.bugs.*, net.rec.* and net.sport.*, etc.; this is a plausible time for the idea of listing newsgroups to gather steam, and not all such lists need have been (or were) in any sense complete. Added: fa.poli-sci, fa.telecom, fa.space, fa.tcp-ip, fa.unix-cpm, fa.works, net.applic, net.columbia, net.eunice, net.games, net.ham-radio, net.2bsd-bugs, net.vwrabbit. Removed: fa.home-sat. Summary: From 11 to 16 fa.* groups, from [3] to 10 net.* groups, from 14 to 26 in all. MARK HORTON The next post, which appears to be the oldest surviving Usenet post with the primary purpose of listing newsgroups, is archived oddly in Google; it's attached to another post. It's unclear to me whether the archived post actually includes this post by original intention, or Google simply screwed up, but I lean towards the latter. Anyway, here are the real data: "list of usenet newsgroups" December 16, 1981 net.general Google Message-ID: anews.Acbosgd.304 (The header was converted from A-News style by Google.) Here is how I found it in the archive, however: "infect: another funny manual page" December 16, 1981 net.jokes Google Message-ID: anews.Aeagle.169 (The header was converted from A-News style by Google.) Regardless, there are several lines of evidence that this is a genuine post by Mark Horton, not an extremely clever parody by the nameless (and not visibly clever!) poster to net.jokes. Direct evidence: a post definitely by Mark Horton a week later extensively quotes this post (see below). Indirect evidence: other posts of his at this time refer to the same degree of authority a post like this implies he thought he had. (Sigh. I'm using "authority" technically here, to indicate the ability to get people to listen to you and do what you say in the absence of force, in contrast to "power" - it's an anthropological distinction, if you must know. His own phrase strikes me as apt: "people seem to listen to me", from one or another of some rmgroup flamewars sometime in 1982.) This post does not appear to be textually ancestral to any later lists, although it's pretty much incredible to me that the later list-posters would not have read and indeed kept it. It's the only list for a long time to come in which groups have no more than one line of description (with exceptions for ARPAnet addresses for fa.* groups). But the descriptions are extremely terse - they begin halfway across the screen, much further than descriptions in later lists. This is done partly to enable a hierarchical structure - all net.* groups other than net.all are indented once, and third-level groups are indented twice. (Third-level groups were a concept Mark Horton had implemented for the first time a few days earlier, as best I can tell.) The list of groups is, again, conservative, though much less so than the previous list; among the previously existing groups omitted are quite a number which would appear on lists in the coming year, and so were presumably not renegade groups in any way. Added: all, fa.all, fa.digest-p, fa.editor-p, fa.info-vax, fa.printers, general, net.all, net.auto.all, net.auto.vw, net.aviation, net.bugs.all, net.bugs.2bsd, net.bugs.4bsd, net.bugs.u3, net.bugs.v7, net.chess, net.cycle, net.games.rogue, net.jokes, net.news.b, net.periphs, net.rec.all, net.rec.ski, net.rec.scuba, net.rumor, net.sport.all, net.sport.baseball, net.sport.football, net.sport.hockey, net.test, net.ucds. Removed: fa.teletext, net.2bsd-bugs, net.v7bugs, net.vwrabbit. Summary: From 16 to 20 fa.* groups, from 10 to 32 net.* groups, from 0 to 2 miscellaneous, from 11 to 54 in all. But note that the *.all convention is not respected in later lists; of the relevant groups, fa.all and net.all appear to be bogus (as do "all" and "general"), while net.bugs.all, net.rec.all, and net.sport.all survived as net.bugs, net.rec, and net.sport respectively. So I think there are 50 real groups in this list. "proposed USENET policies" December 23, 1981 net.news Google Message-ID: anews.Acbosgd.794 (The header was converted from A-News style by Google.) This is not, at core, a newsgroup list, but a proposal, probably the oldest archived proposal, for a Great Renaming. Interestingly, it provoked a thread in which all three major group creation systems the "official" groups have experienced were proposed: Wm Leler put forward the ad-hoc system of the early years, Mark Horton in this message put forward something similar to the Backbone Cabal centred system that followed, and Steve Bellovin proposed a system essentially similar to the way the Big 8 now operate. Anyway, since it *is* a proposal for a Great Renaming, it isn't exactly foolproof evidence of the *existing* names of groups, but it isn't far off, because the proposal was extremely simple, and because Mark Horton used his previous post as the skeleton for it. On the presumption that I've correctly identified the groups referred to, here are the changes: Added: net.jokes.q, net.movies, net.music. (It may be worth mentioning that net.jokes.q was controversial - it appears to have been meant for dirty jokes, or some such - and Mark Horton's actual description line for it was "We drop this newsgroup" ! It did not, in fact, appear on any subsequent lists. But net.music, about which he was also sceptical, did make it.) Removed: all, fa.all, general, net.all. Summary: From 20 to 19 fa.* groups, from 32 to 34 net.* groups, from 2 to 0 miscellaneous, 53 total (none of them apparently bogus). I know of no other newsgroup lists from 1981. Obviously, I'd like to.