This file is a compilation of the usenet history discussion files from nethist.101090 through nethist.102190. In the next file, nethist.102590, the discussion takes off in several different directions, most of which depart from this list. The Players: (BT) Brad Templeton: looking!brad@watmath.waterloo.edu (Creator of the original list) smb> Steve Bellovin: smb@ulysses.att.com je> James Ellis: ellis@laura.psc.edu mh> Mark Horton: stargate!mark@cis.ohio-state.edu kk> Karl Kleinpaste: karl_kleinpaste@cis.ohio-state.edu mcm> Mike Mitchell: mcm@rti.rti.org rs> Rick Salz: rsalz@bbn.com gs> Gene Spafford: spaf@cs.purdue.edu hs> Henry Spencer: henry@zoo.toronto.edu trt> Tom Truscott: trt@rti.rti.org cvr> Chuq Von Rospach: chuq@apple.com aw> Amanda Walker: amanda@visix.com bob> Bob Sutterfield: bob@MorningStar.com [bj] Bruce Jones: bjones@ucsd.edu [list moderator] -=- -=- -=- (BT) The human-nets digest on the arpanet >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> human-nets and sf-lovers were both equally important. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Both were equally important to Usenet, but I believe that smb> human-nets came first on the ARPANET. -=- -=- -=- (BT) The uucp program > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> The release of the uucp program with v7 UNIX provided the initial impetus. smb> So did the Bourne shell. The very first version of netnews was a 3-page smb> shell script. It supported multiple newsgroups, cross-posting, and smb> subscription lists implemented as environment variables. As best I can smb> tell, this script has not survived. (A few years ago, I did search for smb> it, to no avail.) Another motivation was some sort of local news system. smb> On V6, Duke and UNC had a local news system that came from Somewhere. But smb> articles were limited to 512 bytes, and we didn't carry it forward to v7. smb> A prime requirement was that there be a very efficient way to test for smb> the presence of news (hence the checknews program). smb> smb> The original idea for netnews came from Tom Truscott and Jim Ellis. -=- -=- -=- (BT) Notesfiles >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> Notesfiles were around before Netnews at U of I - they belong mh> after "uucp prog" > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Notesfiles existed on Plato before this, but I don't think there smb> was a UNIX version until after netnews. Even if there was an smb> earlier one, the cultural thread was netnews-driven -- notesfiles smb> was adapted to speak the netnews protocols. -=- -=- -=- (BT) The A-news system (Bellovin, Truscott) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Actually, I wrote a C language translation of my shell script; on smb> an 11/45 that didn't even have 256K of memory, the shell was just smb> too slow. This version (pre-A) is also lost. It was never released smb> past Duke and UNC. Truscott and Steve Daniel wrote the version of smb> A news that was released. Netnews was announced at Boulder Usenix smb> (Jan '80), though the code wasn't quite ready. smb> smb> Some day, I'll write up what I remember of our design criteria. smb> It's worth noting now that given the speed (or the lack thereof) smb> of the machines we had, we utterly relied on the ease of writing smb> shell scripts to experiment with protocol variants; compilation smb> would have taken much too long. -=- -=- -=- (BT) The first newsgroups "fa" groups "net." groups >From Tom Truscott Tue Sep 25 06:57 PDT 1990 trt> The very first news groups were "NET." and local groups such as trt> "dept". Later Horton et al. oversaw the lower-casing of NET. Only trt> when ucbvax joined the net did "fa" appear. Indeed I was unaware trt> of the Arpanet mailing lists such as human-nets until ucbvax trt> enlightened us. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Correct. The original concept was that most of the traffic would smb> be of the form now known as unix-wizards (or whatever it's called smb> this week). Growth was slow until Mark started feeding the mailing smb> lists in because there was nothing to offer prospective customers. smb> Given a ready source of material, people were attracted. smb> smb> It's interesting to note that (in my not very humble opinion) the smb> quality of dialog (and the moderator's efforts) on human-nets are smb> still essentially unsurpassed. Only comp.risks is a serious rival smb> (and probably a superior); I exempt otherrealms from discussion smb> because it's a different sort of beast. >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> i.e. the renaming of NET to be net -=- -=- -=- >From Tom Truscott Tue Oct 16 15:01 PDT 1990 trt> My recollection of the first sites were: trt> duke - unc trt> then trt> phs (Duke Physiology dept. via 9600 baud wire) - duke - unc trt> duke was a CS department PDP 11/70. trt> unc was a CS department VAX I think. smb> No, an 11/45 at the time. trt> phs was a PDP 11/60. trt> trt> There were uucp (non-usenet) connections for one or two other duke trt> machines, and "research" at Bell Labs. Perhaps "duke34" (a machine trt> sitting next to "duke") was on Usenet too, but it barely counts. trt> trt> I think vax135 at Bell Labs (where UNIX 32V was developed) was the trt> next Usenet member. (Might have been Reed, I dunno). vax135 caused trt> an upheaval in A news, because they would not let uucp copy files trt> to their system. trt> trt> As I recall, the very first implementation of news broadcast did trt> something like: trt> trt> uucp -l remote!/usr/spool/uucppublic/inbound trt> trt> We considered using uux but that would have been less efficient trt> than a simple copy, and would have used up precious disk space. trt> But vax135 would not permit the above uucp request, so we gave up trt> and switched (overnight) to trt> trt> uux -r remote!rnews trt> trt> Since there were so few sites compatibility was not a big issue! trt> We figured the performance hit and space wastage was tolerable trt> since there would be only a few articles per day. To avoid stupid trt> "exit 0" messages from uux we had to make a source code change to trt> uuxqt, but then one needed to make source code changes to mail and trt> uucp just to set the host name! (We distributed a "setup.n" trt> document, describing how to set up uucp.) >From Steve Bellovin Tue Oct 16 15:22 PDT 1990 smb> We also had to change uucp to add to the list of legal commands, smb> it being hard-wired in at the time. Since some sites wouldn't even smb> make that change, we added a mail encapsulation very early in the smb> game. The idea was to mail to some magic user name; periodically, smb> a shell script would extract it. The shell script would use mail smb> itself to extract the mailbox, thus avoiding the need to do compatible smb> locking. >From Mark Horton Mon Oct 15 19:49 PDT 1990 mh> The first cross-country link was from duke to research, then from mh> research to ucbvax, all on research's nickel. I remember, while mh> at Berkeley, exchanging email with the original A news developers, mh> and being amazed that I could get a reply back a few hours later, mh> even though research was polling both duke and ucbvax to pick up mh> waiting mail. -=- -=- -=- (BT) The mistake of net.general > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> net.general was a group intended to be read by everyone. This smb> worked while the net was still small. -=- -=- -=- (BT) Creating groups by typo or whim - net.joke, net.bizarre, net.flame, net.gdead net.wobegon and others > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> The first joke newsgroups were net.suicide (the original article smb> called for a meeting of some university suicide club on the roof smb> of a building) and net.dead-babies, which contained a series of AP smb> wire stories on a school bus accident. I seem to recall a specific smb> incident that prompted a particular individual to create them, but smb> I no longer remember what it was. -=- -=- -=- (BT) Net goes international (Canada) (Spencer, Templeton) >From henry@zoo.toronto.edu Sun Oct 21 22:07 PDT 1990 hs> utzoo is on the 1 June 1981 map. (Copy on hs> request.) My records say we joined the net on 13 May 1981. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> I've been wondering of late if that will ultimately be the smb> greatest contribution of the net to history -- that it provides a smb> cheap, reasonably accessible, mechanism for multi-way international smb> communication. -=- -=- -=- (BT) The B news software (Horton, Glickman) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> The important point here is that B news -- as most of the enhancements smb> to netnews software -- was load-driven. The original A news had smb> a number of design choices that made it unsuitable for a large net. smb> (We estimated a maximum size of 100 sites, and 1-2 articles a day, smb> net-wide....). Many of the deficiencies could have been patched smb> around (and indeed, many were in A+ news); the key problem was that smb> the last-read time was a global concept that applied to all smb> newsgroups; you couldn't read things out of order. smb> smb> The goal there (and in many other spots) was to have software free smb> of databases, and especially central databases. Instead, we chose smb> to let the file system do the work. This also worked to eliminate smb> most of the explicit synchronization. For example, article id's smb> were limited to 14 characters or less, and named the file in smb> /usr/spool/news that held the text. No history file was needed smb> for duplicate detection; stat() would tell you that. Etc. (BT) Pathalias software (Honeyman) >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> The first version of pathalias was written by Steve Bellovin while rs> working on his doctorate at the University of North Caroline at rs> Chapel Hill. He posted it to Usenet in around 1982, and included rs> patches to picked it up and installed it on allegra, a Bell Labs rs> machine at Murray Hill, NJ. At that time, allegra was the center rs> of the UUCP mail world (a role since assumed by sites named ihnp4, rs> seismo, and now uunet), and at the time when Usenet was starting rs> to explode all over the place. Before he left for Princeton, Peter rs> had moved from tweaking the options, to doing serious redesign of rs> most of the algorithms. Peter is still maintaining the program, rs> and it has followed him out to the University of Michigan. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Rich's chronology is about right, though I no longer remember the smb> exact dates. (Incidentally, there's apparently a line missing from smb> the above paragraph. Peter brought the code up on allegra.) I smb> also tried to gather the data, but the effort was too much for me. smb> The data arrived in very ``dirty'' form, and especially after I smb> join the Labs I had no time to reconcile it all. And yes, the code smb> is essentially all Peter's now; virtually nothing remains of my smb> original. >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> (I think pathalias came later, and was written by Bellovin.) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> My memory fails me on this, but someone had a "competing" program gs> with pathalias that appeared just before (early 82?). I believe gs> it was called rpaths.... It simply took in news path information gs> and mail paths, then saved out the shortest path between each two gs> arbitrary sites. It wasn't as sophisticated as pathalias because gs> all edges were assumed to have unit length. gs> gs> I used it at Gatech for a year or so. I remember that I shared gs> path files with Rob Kolstad to feed the beast. When it became gs> evident there were a significant number of one-way links, most gs> sites switched to pathalias. We shouldn't forget that other program gs> though (although I can't remember who wrote it -- John Quarterman?). gs> When Mark started the mapping project, I shipped him the files I gs> had. >From Mike Mitchell Fri Oct 12 08:52 PDT 1990 mcm> I wrote a program called 'mkpath' back in '82-'83 that competed mcm> with pathalias. We used it on our pdp-11/60 because pathalias mcm> wouldn't fit. My first version did everything with disk files, mcm> but it was too slow. I don't think I ever released that version mcm> to the net. Someone on the net did request that I send him that mcm> version, as the version he used ran out of memory on his list of mcm> ~2000 sites. My version was fairly simple; it looked for the mcm> shortest path. Pathalias had code to deal with the speed of the mcm> connection. I didn't have the resources to use that sort of mcm> information. When I got a VAX instead of a PDP, I stopped working mcm> on 'mkpath'. I also had a front-end to the mail program that used mcm> the mkpath database. I called the front-end 'nmail'. I don't have mcm> any copies laying around, but I could check our news archives. I mcm> was working at Ikonas Graphics at the time. Their site name was mcm> 'ikonas'. >From Gene Spafford Fri Oct 12 09:18 PDT 1990 gs> gs> That's the one! It was mkpath, because I remember "nmail". gs> gs> I was using it on our Vax, and distributed it to some other sites gs> and mkpath was in some use in various places around the net. -=- -=- -=- (BT) Decvax and Bill Shannon's contribution > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> This was moderately early; they picked up the slack when assorted smb> AT&T sites no longer could/would. >From Mark Horton Mon Oct 15 19:49 PDT 1990 mh> By the way, I've seen mention crediting Bill Shannon with the mh> decvax link. Bill was there and involved, but most of the credit mh> goes to Armando Stettner of DEC, who made it happen and got the mh> phone bill footed. At one point in the early 80's he announced at mh> Usenix that decvax's phone bill was running a quarter million mh> dollars/year, officially sanctioned by DEC as a goodwill gesture mh> to the net! >From James Ellis Tue Oct 16 08:43 PDT 1990 je> Yes - I remember Armando's announcement of $250k/yr but can't recall je> which conference it was. By that time we were becoming numb to je> the expenses people would go to just to pass news and e-mail around je> to other people. Especially since Duke never had to pay any of je> those bills! Still, this was a milestone. je> je> Actually, although research was one of the original sites officially, je> I didn't think they ran news until some time into the game. At je> the Winter 80 (January) Usenix I gave a short talk announcing the je> creation of Usenet and inviting anyone to join. We even had printed je> up little forms for folks to fill out giving their uucp info, etc. je> (Sadly, I don't think any of these have survived.) There were two je> incentives for folks to be interested. je> je> First was supposed to be the news software itself. Since we expected je> traffic on the order of a handful of items a week, one of the design je> goals was that the news system be a useful tool for strictly local je> use. (Which as already noted, led to the use of the "NET." prefix je> for non-local news.) We had to convince people to pick up the je> software for its own sake and then hope they would find it easy je> and convenient to tie into network news as well. Or so we thought. je> In retrospect it clearly was the net-wide aspect of news that je> attracted everyone's attention. je> je> Second we felt we needed to get across the idea that Usenet was je> already a fully-functioning non-local network rather than the je> patchwork of hopes and scripts that really existed. By this time je> we did have a uucp connection to research at Bell Labs thanks to je> Dennis Ritchie and Tom Truscott. But I don't recall that they ran je> news. (Steve B - just what state was the news software in at all je> at that point, anyway?) [See Steve's comments below -bj] je> We put them on the map that was displayed je> at the conference anyway as a Usenet "site". (So the confusion je> between uucp-net and Usenet dates back to here!) Dennis wouldn't je> let us name research though out of fears that BTL management wouldn't je> like the idea so I had to refer to it as an unnamed research site je> with a dot over on the East Coast. je> je> This was the first Usenix I'd ever been to so I was rather nervous je> but it seemed to go well. Made everyone laugh and applaud when I je> described Tom's homebrew autodialer! It seems like it was only a je> few days after the talk that our first site requested a connection je> - Reed college in Portland, OR of all places. They had no dialer je> either so we had to call them - they were willing to be billed for je> the charges. I don't recall if we ever billed them or if we were je> ever paid, but Duke's department Chairman at the time seemed very je> willing (to me) to foot some expenses to get Usenet off the ground. je> Then there was a long drought and I don't recall who the next site je> added was. je> je> I do recall that for a long while after Berkeley and Research were je> providing cross-country connectivity, the connections were often je> very wasteful. One of the worst examples was that Tektronix, in je> Oregon, couldn't send e-mail to some other site (Reed?) a local je> phone call away because it was against policy to set up the je> connection. But they could, and did, send mail via je> Berkeley/Research/Duke going cross-country twice to reach a je> local phone call away! >From Amanda Walker Tue Oct 16 09:11 PDT 1990 aw> Indeed. I suspect that there are any number of examples of this, aw> but the most egregious in my experience was at CWRU. The ECMP aw> department had a VAX 11/780 on Usenet ("cwruecmp"), and the campus aw> computer center had a DEC-20 in the room next door. The machines aw> were separated by a grand total of about 30 feet and a piece of aw> wallboard, but the computer center was not at all interested in aw> "catering" to "those CS types" by stringing an RS-232 line between aw> them. So, it was possible to send mail between them, but only by aw> sending via a route resembling: aw> aw> crwuecmp => decvax => ucbvax (UUCP) aw> ucbvax => columbia (CU20A, I think) (ARPANET) aw> columbia => cmu-cs-c => cwru20 (CCnet) aw> aw> Yup, that's three networks, and two coasts just to get through a aw> piece of sheetrock :-). Took about a week, too. >From Steve Bellovin Tue Oct 16 09:00 PDT 1990 smb> I think that my pre-A C-language version was running then. The smb> efficiency of the shell script was too low, and I'm fairly certain smb> that A-news wasn't quite ready yet. (In the grand tradition of smb> computer companies everywhere, we announced before we had a smb> product...) smb> smb> A year or two later, CSnet was announced. I remember that we smb> commented how this showed the true difference between grad students smb> and faculty. When we wanted to connect universities and research smb> sites, we wrote some software quickly, handed out a few simple smb> forms, and scrounged everything, from dialers to connectivity. smb> When some professors wanted to do the same thing, they got a grant smb> from NSF, set up formal support and monitoring stations, secured smb> DARPA permission to connect to the ARPANET, published a bunch of smb> papers, etc. -=- -=- -=- >From Mark Horton Wed Sep 26 06:34 PDT 1990 > The first used car ad probably belongs in here somewhere. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> The true meaning here is rather amusing. When we were arguing over smb> the design of the protocols, I argued strongly for many local -- smb> single-machine -- groups, with the sole group NET to be broadcast. smb> Jim Ellis pointed out the need for regional group for things like smb> used car ads; thus, NET.* was born. (Originally, network-wide smb> groups were identified syntactically; hence the use of all caps to smb> distinguish them from anything local.) Jim prevailed (fortunately); smb> naturally, we were all very amused when the first used car ad showed smb> up on NET.general. The original designers all sent congratulatory smb> messages to the perplexed poster. smb> smb> We discussed, but never resolved, the difference between a distribution smb> and an interest group. No solution seemed right. 11 years later, smb> the current implementation suffers from all the faults we talked smb> about way back when. -=- -=- -=- (BT) Chain letters over the net. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> This was later, I think. -=- -=- -=- (BT) AT&T's contribution (allegra, ihnp4 etc.) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Research and mhtsa were much earlier, and I think vax135. -=- -=- -=- (BT) Widespread net growth and international growth -=- -=- -=- (BT) Forged article critical of Unix (1981) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> This was not so much a pure forged article as a forgery designed smb> to conceal the unauthorized posting of a critical paper by Don Norman. -=- -=- -=- (BT) K News (proposal) >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> K News is a non-topic; it generated, what, a month or two of rs> discussion? -=- -=- -=- (BT) Domain style naming (Postel et al) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> I don't think this is particularly related to Usenet. The impact smb> of Usenet on the ARPANET was more as a (strong) catalyst to force smb> re-examination (and benign neglect) on the strict policies against smb> interconnection. Uucp mail into the ARPANET became a major force smb> long before it was legit. And it was obviously known to, and smb> ignored by, many of the Powers that Were. -=- -=- -=- (BT) The net.jokes.q creation and deletion -=- -=- -=- (BT) Formation of the "Backbone" >From: mark@stargate.COM (Mark Horton) mh> The backbone is first mentioned in the July 1983 map - it was not mh> mentioned on the Jan 1983 map. In July 1983 there were both logical mh> maps (showing all hosts) and geographic maps (showing the biggest mh> host in each city on a US map.) >From Gene Spafford Mon Sep 24 18:58 PDT 1990 mh> backbone creation, circa 1984 >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> I was the "father" of the backbone. It came about from two different gs> things. gs> gs> 1) I had been maintaining the list of newsgroups for a while, and gs> was concerned at the time about the unrestrained growth in newsgroups gs> (little did I know....). I was also concerned about the very flat gs> namespace. At the time, I put together a mailing list to talk gs> about renaming newsgroups. Principles in that group were Chuq von gs> Rospach and Mark Horton. The discussion didn't go anywhere in gs> particular at the time, but I kept the list as a group to talk gs> about matters of importance. The people listed were admins or gs> prominent sites....soon to be the backbone. gs> gs> 2) I was aggressively trying to established better uucp connections gs> to gatech. I did some hand mapping with pathalias output and awk gs> counts on news articles, and came up with a small set of "regions" gs> on the net with important machines connecting those regions to gs> others. I set up connections from gatech to those machines and gs> encouraged some other cross-connections to improve propagation gs> (anybody remember those mail messages? A total stranger telling gs> you to set up connections with someone else... :-) gs> gs> Eventually, by the time of the great renaming after the 1986 Usenix gs> conference, I formalized the backbone in a regular posting with a gs> map and a description of what consituted a backbone site --- good gs> connectivity, carrying the mainstream groups, and a commitment to gs> stable news and mail software. These were the same things I had gs> encouraged earlier on, or the reasons I had put people on the gs> mailing list. gs> >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> gs> Here is what the "backbone" looked like at the end: gs> gs> A Usenet "backbone" site is one which exchanges every (non-local) news gs> article it receives with at least two other backbone sites; or which is gs> the main newsfeed for a particular geographical area (e.g., Australia) gs> or special news gateway (e.g., inet) and exchanges news with at least gs> one other backbone site. This exchange is done (theoretically) with gs> minimum delay. Thus, any article submitted to a backbone site is gs> supposed to propagate to all other backbone sites within a very short gs> time period. (Under actual conditions, your mileage may vary.) To be gs> labelled as part of the "backbone," a site must also: gs> * be of sufficient capacity to handle the load of news; gs> * be of sufficient capacity and connectivity to handle the load gs> of mail generated by replies to news articles and as submissions gs> to mailing lists and moderated groups sent through the backbone; gs> * be running a recent version of the news software, and keep gs> up-to-date with new releases and patches; gs> * provide a stable news and mail relay service, preferably including gs> a mailer understanding domains and domain-based addressing; gs> * be staffed by experienced, responsible, capable staff who gs> will maintain news and mail, and quickly respond to problem gs> reports and requests for assistance; gs> * wish to be advertised as a backbone site, thus taking part in gs> discussions and debate as well as becoming a target for abuse gs> from the net-at-large; gs> * evidence some measure of financial and/or political stability gs> so as to be able to remain as a backbone site for the gs> indefinite future. gs> Site admins wishing their site included in this posting should gs> document the above points in mail to "backbone-request@rutgers.edu". gs> gs> Each backbone site normally feeds some number of well-connected gs> secondary sites, most of which are not "leaf" (terminal) nodes. gs> These secondary sites feed the news out to other distribution and leaf gs> nodes, and so on. gs> gs> For optimal news distribution, each site should establish an "L" type gs> link with a site closely connected to a backbone site. This will help gs> ensure that any articles submitted from that site get propagated to the gs> whole net with a minimum of delay. Sites should *not* request a news gs> feed from a backbone site unless they are willing to feed at least five gs> or six (or more) other, non-terminal sites. gs> gs> Note that some backbone links (viz., ALL--munnari,ncar--nbires, gs> uunet--mcvax) do not carry all newsgroups, usually meaning "talk" and gs> some "rec" groups, but sometimes also including "soc", "sci", "news", gs> "misc", and "comp" newsgroups. The European component of Usenet gs> receives a limited number of groups from outside Europe, as well as gs> having a number of active "eunet" newsgroups, and exchanges all those gs> articles via X.25 links. gs> gs> To send mail to the administrators of all the backbone sites, address gs> your mail to "backbone@rutgers.edu". If you wish to send mail to the gs> administrator of a particular site, consult the uucp map for the name gs> and address of the appropriate individual(s). (In Europe uucp map gs> entries are obtained through the national backbone's "netdir" gs> service). [For the map itself, see the file "backbone.map" and other maps also in this directory -bj] -=- -=- -=- (BT) Monthly postings, newuser group, commonly asked questions. >From Gene Spafford Mon Sep 24 18:58 PDT 1990 mh> monthly postings of FYI stuff about 1984 >From Gene Spafford Mon Sep 24 18:58 PDT 1990 gs> monthly postings of newsgroups -- pre 1981 >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> In 1981, Adam Buschbaum was posting an article now and then that gs> listed the newsgroups as he knew them. Adam was a HS student whose gs> father was at Bell Labs. When Adam went off to college (I think) gs> in 1982, he asked me to take over the postings because he would no gs> longer have access. I have continued to do this as the list of gs> active groups. gs> gs> Shortly thereafter, by a year or two, I started collecting a few gs> other things that appeared to be particularly useful for net readers gs> to post everything all together. I don't have the dates of when gs> I added items, but the first two to be added were the "Frequently gs> asked questions" by Jerry Schwarz and Chuq's "Netiquette" article. gs> You might try to find them and ask the dates. >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> The newusers group was the second moderated group created in gs> summer 1986. >From Chuq Von Rospach Fri Oct 12 09:04 PDT 1990 cvr> A note on the Netiquette document. It was one of the earliest (and cvr> if you ask me a very successful) experiments in groupware. It was cvr> a document by consensus with about 30 people throwing email around cvr> and working up various pieces. I'm not sure I was as much writer cvr> as coordinator, although the final draft was worked up by me. -=- -=- -=- (BT) net.women.only experiment >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> I remember this as 1984-1985 -=- -=- -=- (BT) Arpa mailing lists merge with usenet groups. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Comparatively early. -=- -=- -=- (BT) Subgroups to divert traffic: -=- -=- -=- (BT) net.startrek, net.abortion, source code discussion -=- -=- -=- (BT) Hoaxes etc. (kremvax) >From Amanda Walker Thu Oct 11 16:32 PDT 1990 aw> I am pretty certain that the (in)famous kremvax posting was done aw> 1983, since I believe that I read it before I left CWRU (which aw> was around January 1984). -=- -=- -=- (BT) The re-emergence of mailing lists >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> The mailing lists never went away, really. I used to post Rich gs> Zellich's list of mailing lists along with the newsgroup list until gs> they got too large. Then I tried extracting out the uucp only gs> lists and a few things that weren't in Zellich's list. It became gs> a regular posting, and hardly a month goes by where I don't get gs> one or two new ones added from people starting new ones or advertising gs> old ones. gs> gs> The flaps over net.women.only and net.motss created a small surge gs> in new mailing lists at that time. -=- -=- -=- (BT) net.columbia and naming debates -=- -=- -=- >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> The UUCP Project (Horton et al) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> I assume this is the UUCP Mapping Project, which was designed smb> to supply the data for pathalias. >From Rick Salz Wed Sep 26 07:10 PDT 1990 rs> Mark can probably fill in the gaps, and time-frame, for this. rs> Early Usenix proceedings (I guess anything before Portland counts rs> as early these days) have reports from the UUCP Mapping Project, rs> and I think a map, too. *This is a message forwarded by Rick: >From sob@harvisr.harvard.edu Mon Nov 27 21:28:17 1989 Date: Mon, 27 Nov 89 20:36:11 EST From: Scott Bradner To: rsalz@BBN.COM The original data was collected by Steve Bellovin but he sort of gave up at some point. I think that Peter Honeyman did some collecting back then also. At that time I was running wjh12 as a internet/uucp/bitnet gateway. I needed some better data. So I: 1/ had a programmer here rig up a scan-the-news-headers program to get an operating base 2/ sent out: a) a posting to some newsgroup asking for the info b) letters to usenet & root at all nodes that I had found by scanning headers c) then to all nodes found in the new data When I started the process I got calls from Rob Kolstad & Mark Horton Rob offered to help in the data gathering, we agreed to have him start at one end of the alphebet & me at the other. Mark was about to do the same thing, Rob & I agreed to pass on the data to Mark as we got it done. After the 1st full pass, Rob & I passed all operations over to Mark & the mapping project. >From Mark Horton Fri Sep 28 16:01 PDT 1990 mh> Maps are a very good point. The first maps from about 1981 had mh> about 15 sites on them and were drawn in ASCII. (This for Usenet.) mh> They grew until in 1983 or so they were too big for ASCII and they mh> were drawn on paper. My ex-wife Karen and I did these early maps, mh> got copies, and handed them out at Usenix conferences. mh> mh> After awhile, Bill and Karen Shannon took this over (around 1984-5) mh> and made multi-page ASCII maps of Usenet. After about 1985 the mh> net was too big for this. mh> mh> Recently Brian Reid has put out Postscript geographical maps of mh> Usenet. mh> mh> The earliest UUCP (email) maps were in people's heads - they would mh> use the Usenet map to route mail. Every so often someone would mh> post a message "I'm going to make a map, everybody send me your mh> L.sys file" and 6 months later they would emerge from under the mh> avalanche of responses and give up. If I recall correctly, both mh> Lauren Weinstein and Steve Bellovin did this at one time. mh> mh> The current map started when Scott Bradner and Rob Kolstad undertook mh> the effort and stuck it out enough to finish the map. At that time mh> they handed it over to our UUCP Project which was organized with mh> 10 or so regional coordinators to maintain it and post it regularly. mh> mh> The UUCP Project, which was founded by Karen Summers and I took mh> over in 1986, ran from 1985 to 1988. It did the UUCP Map (Mel mh> Pleasant supervised the regional coordinators), produced and mh> distributed smail 2 (Chris Seiwald and Larry Auton were the authors) mh> and did domain registrations for UUCP sites (myself and Tim Thompson). mh> mh> In 1988, UUNET offered virtually free domain registrations, and we mh> closed down the domain and software parts of the UUCP Project. mh> Mel took on the mapping portion in conjunction with UUNET. He's mh> still doing it today. -=- -=- -=- (BT) The emergence of voting for creating newsgroups >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> As I remember, this phased in from about 1987 on. As the influence gs> of the backbone began to wane, in part from workload, and in part gs> from flak, we would float calls for discussion of new groups, and gs> a sort of show of support. When the backbone "died" a more formal gs> mechanism took it's place about 6 months later. I seem to recall gs> that the anarchy in place as people started creating groups and gs> flaming each other brought that on. -=- -=- -=- (BT) RN & Kill files (Wall) > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Note again that this was a load-driven development. -=- -=- -=- (BT) News batching and compression > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Ditto. -=- -=- -=- (BT) Checkgroups messages (Spafford) >From Gene Spafford Mon Sep 24 18:58 PDT 1990 > checkgroups -- late 1986 > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> The whole issue of control messages was and is troubling. The smb> original versions of netnews had no control functions because we smb> knew we couldn't authenticate the messages. (Yes, we knew about smb> public key cryptography and digital signatures. If nothing else smb> -- and I had read the RSA paper by then -- V7 included smb> enroll/xsend/xget.) smb> smb> Some of the code to handle some messages was dangerously buggy in smb> the early days. I interviewed at the Labs very soon after B news smb> came out. When I entered the office of a Very Prominent UNIX guru, smb> the first words out of his mouth were ``Netnews B is a tool of the smb> devil''. (As an example of an early bug, the code would honor smb> requests to rmgroup ../../../../../.. -- and recall that v7 did smb> not honor setuid if root did the exec.) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> Checkgroups was added to B news by Rick Adams and myself in the gs> months following the 1986 summer Usenix. I seem to remember February gs> of 87 as the actual first posting. This was in 2.11 B news -=- -=- -=- (BT) The problems with the old releases of B news >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> Continued at least into 1989 when I would get error messages gs> when creating new moderated groups. -=- -=- -=- (BT) Line Eater Bug -=- -=- -=- (BT) Moderated newsgroups: a) mod.announce b) mod.newprod and commercial information on the net > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Here we see a return to syntactic decisions. >From Gene Spafford Mon Sep 24 18:58 PDT 1990 gs> moderated groups -- late 1986 (after summer Usenix in Atlanta) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> mod.announce was the first -- an experiment. Mark Horton created gs> it after Rick Adams had modified the news to support it, backbone gs> site admins installed it, and I started sending out lists of gs> moderator addresses. gs> gs> As I remember, Larry Auton (now lda@floyd.att.com) came up with gs> the idea originally at our "backbone" meeting at the Atlanta Usenix. gs> (Present were Horton, Pleasant, Adams, Auton, Heiby, Fair, Woods, gs> Beals, Spencer?, Jackson (Curtis), and me. There were at least gs> one or two more present, but I don't remember them. It was 3/4 of gs> the "backbone" of the time.) >From Gene Spafford Thu Oct 11 20:15 PDT 1990 gs> gs> moderated groups existed prior to 1986. They came about in 1984 when gs> Rick released B News 2.10. What came in 2.11 that Larry Auton (with gs> input from Mel Pleasant, too, as I remember) thought of was the idea gs> of the "newspaths" file to get submissions to the moderators. gs> Moderated groups weren't working well up to that point because the gs> users didn't like the idea of manually mailing things to the moderator. >From Karl Kleinpaste Fri Oct 12 18:41 PDT 1990 kk> You must include mod.ber in here. mod.ber was a moderated group kk> created by/for Brian E Redman (of HDB [HoneyDanBer] UUCP) as a kk> source of weekly(?) summaries of Usenet traffic. The idea was, I kk> think, that the Usenet was getting _SO_ big (heehee) that we needed kk> volunteers summarizing things for us so we could just sort of keep kk> track of the highlights and subscribe to groups along the way when kk> we noticed interesting topic threads in mod.ber. -=- -=- -=- (BT) netnews software implemented on VM (IBM Mainframes) -=- -=- -=- (BT) Stargate (Weinstein) -=- -=- -=- (BT) NSA Baiting in messages > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Not really significant of anything but paranoia, if you ask me. -=- -=- -=- (BT) Big Net personal fights a) some result in lawsuit threats (Mark Ethan Smith) b) some result in people being bounced temporarily from the net (Maroney) c) some result in people being removed permanently. -=- -=- -=- (BT) The great renaming (Horton) >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> The renaming was instigated and done by Rick Adams because his rs> news/sys file was getting unwieldy. >From Gene Spafford Mon Sep 24 18:58 PDT 1990 gs> grand renaming -- started July 86, ended March 87 >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> Actually, Rick Adams was the prime force behind this. >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> gs> Actually, there were two discussions of renaming prior to the one gs> in 86/87. I started one, and Chuq started the other. Rick was gs> in the mailing list for my discussion. The idea was one that gs> finally "ripened" by 1986. Rick was the one who did all the gs> postings, but it certainly wasn't all his idea. As I remember, gs> the group of us involved in the design floated a number of ideas, gs> but Rick was the one who condensed it and came up with the final gs> proposal. gs> gs> The idea for the top level hierarchies I believe came from Mel gs> Pleasant at the Usenix meeting. I remember we decided the names gs> of all the hierarchies at that time except for misc or rec, which gs> Rick added later. >From Chuq Von Rospach Fri Oct 12 09:04 PDT 1990 cvr> cvr> I think the one added later was talk. Talk was the only cvr> top-level domain added specifically to allow admins to not carry cvr> groups -- the pariah groups. This was done (if I remember correctly) cvr> because it was a lot easier than simply trying to make them go cvr> away. -=- -=- -=- (BT) The merging of moderated groups into the hierarchy. -=- -=- -=- (BT) The decline in propagation for groups outside of "comp" and "news" > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> This was part of the hidden (but widely admitted) agenda of the smb> renaming. In particular, there was a struggle to keep many groups smb> out of ``talk'', which everyone knew would be the first to go. -=- -=- -=- (BT) Creation of rec.humor.funny (first ultra-moderated non-mailing-list -- Templeton) -=- -=- -=- (BT) Battle over the expiry dates on rec.mag.otherrealms. (Von Rospach) -=- -=- -=- (BT) The attempt to form comp.protocols.tcp-ip.eniac. (Webber) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> This was in the summer of 87. -=- -=- -=- (BT) The misnaming of soc.culture.china -> 30 day voting time. -=- -=- -=- (BT) Attempt to remove talk.bizarre from the net for the "VOLUME,VOLUME, VOLUME" game. -=- -=- -=- (BT) "jj@portal" (Rob Noha) begs for money on the net. -=- -=- -=- (BT) General problems with Portal, and the issue of pay-for-access net s ites. -=- -=- -=- (BT) Attempts to form groups for drugs and sex. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> This belongs earlier. -=- -=- -=- (BT) Creation of the "alt" hierarchy. >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> John Gilmore and Brian Reid, with help from Carl at pyramid, created rs> alt. I think the reasons were that Brian didn't like mod.gourmand--> rs> rec.recipes, and that John didn't like having no unmoderated source rs> group. I would contact them at first, however. > From Steve Bellovin Wed Oct 10 19:48 PDT 1990 smb> Certainly true of Brian. -=- -=- -=- (BT) UUCP project and .US domain. (Gilmore, Horton) >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> The UUCP project was started by Horton, with Weinstein, and Mel. -=- -=- -=- (BT) Arbitron news statistics (Reid) -=- -=- -=- (BT) Telebit releases the Trailblazer. -=- -=- -=- (BT) UUNET arrives (Adams) (Seismo fades) -=- -=- -=- (BT) Internet & TCP/IP begins to pervade the net. The backbone begins to fade. >From Gene Spafford Mon Sep 24 18:58 PDT 1990 gs> backbone goes away, 1987 >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> I found a note to myself, June 15 1988, where I said that I gs> considered the backbone to be dead and I would no longer post gs> a "backbone" map. I removed the word "backbone" from all my gs> regular postings starting that month. -=- -=- -=- (BT) NNTP causes news to propagate too fast. >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> NNTP and the large-scale gatewaying into newsgroups is Erik rs> Fair's doing. -=- -=- -=- (BT) PC-Pursuit -=- -=- -=- (BT) Reliable links to Europe -=- -=- -=- (BT) Crazy creation of groups in the alt hierarchy. (Weiner) -=- -=- -=- (BT) alt.gourmand kepr rather than being renamed. (Reid) -=- -=- -=- (BT) AT&T decides not to forward mail. -=- -=- -=- (BT) "biz" and "inet" hierarchies. "Gnu" hierarcy. >From mstar!root@cis.ohio-state.edu Thu Oct 11 14:20 PDT 1990 bob> gnu.* came to me in the shower on the morning of 17 Mar 88. It bob> was established using Erik Fair's gateway code (originally written bob> for the inet groups). The join-up invitation went out to the news bob> neighbors of tut.cis.ohio-state.edu on Mon, 2 May 88 17:15:29 EDT. bob> By ten days later it was well-enough populated and propagated for bob> Spaf to add it to his "alternative news hierarchies" article. It bob> had penetrated well into the UK by 27 May 88, and via mcvax onto bob> the continent (EUnet) on 18 Jul 88. Brian Reid's USENET Readership bob> Report for June 88 reported most gnu.* groups with over 25% bob> propagation, and August's showed over 50% - possibly the fastest bob> growth of any alternative hierarchy. gnu.* underwent its first bob> Great Renaming yesterday, perhaps a sign of maturity :-) bob> bob> I asked Spaf to add gnu.announce to the "backbone"'s list of bob> moderated aliases - in effect, requesting a mild policy change in bob> the administration of the Usenet (its administrative structure bob> would formally begin carrying some of the administrative freight bob> for a few non-Usenet hierarchies, removing their ability to claim bob> to be completely separate from and independent of the mainstream bob> "Usenet Proper"). In a precedent-setting step, the request was bob> provisionally approved on 14 May 88. This opened the way for the bob> mechanized moderation of Brian Reid's alt.gourmand and others since. -=- -=- -=- (BT) comp.sys.next violates voting rules -=- -=- -=- (BT) AT&T complains about source code on Killer, shuts it off temporarily(?) -=- -=- -=- (BT) comp.binaries.ibm.pc goes to 8 megs/month -- gets moderated in first, and only, moderator election. (Dhesi) -=- -=- -=- (BT) The forming of comp.society.women (proposed as comp.women) (Roberts) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> Hmm, I seem to believe this was near the end of 88, beginning of 89. -=- -=- -=- (BT) Oct/88 Internet Worm (perhipheral to usenet) >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> Nov 2, 1988, to be exact. :-) -=- -=- -=- (BT) Nov/88 The battle over rec.humor.funny (Templeton, Richmond) -=- -=- -=- (BT) C News (Spencer, Collyer) -=- -=- -=- (BT) NN (actually started earlier) -=- -=- -=- (BT) B News 3.0 (Raymond) -=- -=- -=- (BT) alt.fusion, sci.physics.fusion -=- -=- -=- (BT) More rec.humor.funny -- Jokebook, Expansion to GEnie, Denninger's vote -=- -=- -=- (BT) Aquaria alt, rec, sci (sci.skeptic?) -=- -=- -=- (BT) Battles about naming, Australian voting etc. -=- -=- -=- (BT) shareware on usenet -=- -=- -=- (BT) flow mapping -=- -=- -=- (BT) bitnet merging into USENET, more newsreaders etc. -=- -=- -=- (BT) more moderator copyright (telecom, sci.med.aids) -=- -=- -=- (BT) June 89/ClariNet -=- -=- -=- (BT) In Moderation Network -=- -=- -=- (BT) More forgery -=- -=- -=- (BT) alt.sex, alt.sex.bondage and "Cindy's Torment" -- cutoff at U of T, etc. -=- -=- -=- (BT) AlterNet, NYSERNET/PSI, NSFNet, FARRnet >From Rick Salz Mon Sep 24 16:20 PDT 1990 rs> AlterNet, NYSERNET/PSI -- non-issues. -=- -=- -=- (BT) Confiscation of USENET sites involved with phrack. >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> At this point, it is not clear that any Usenet site was confiscated gs> because of Phrack. Some Usenet nodes were confiscated, but the ones I gs> know about were confiscated because they were used for the storage and gs> transmission of trade-secret AT&T code, not because they had anything gs> to do with Phrack. gs> gs> Neidorf's machine was confiscated to search for evidence for the gs> trial, but it was not a Usenet node. -=- -=- -=- (BT) EUNet and its policies (pay for feed, limit feed, not all groups) -=- -=- -=- (BT) rec.arts.erotica -=- -=- -=- (BT) Alt explosion more alt bannings, rise of alt.sex -=- -=- -=- (BT) trial newsgroups -=- -=- -=- (BT) USENET over 10 megs/month >From Mark Horton Wed Sep 26 06:34 PDT 1990 mh> Surely you mean over 10 megs/day. It's about 12.5 now. >From Gene Spafford Thu Oct 11 20:05 PDT 1990 gs> gs> I'm not sure about this, but the figures I compiled on Oct 1, 1988 for gs> the IETF meething showed 11,000 sites; over 1800 articles per day; gs> over 4MB of traffic per day, average; and over 450 newsgroups. gs> My figures were taken from Rick Adam's summary stats posted from gs> Seismo or uunet, and from Brian Reid's arbitron stats. gs> gs> That same presentation had these figures taken from stats by Adams, gs> Spencer, Horton, Bellovin and Reid: gs> gs> 1979 3 sites, ~ 2 articles per day gs> 1980 15 sites, ~10 articles per day gs> 1981 150 sites, ~20 articles per day gs> 1982 400 sites, ~50 gs> 1983 600 sites, 120 gs> 1984 900 sites, 225 gs> 1985 1300 sites, ~375 articles per day, +1MB per day gs> 1986 2500 sites, ~500, 2MB+ gs> 1987 5000 sites, 1000 per day, 2.5+MB gs> 1988 11000 sites, 1800, 4+MB gs> gs> Volume growth is exponential in number of sites, not in number of gs> newsgroups (as some have claimed). Plot the figures -- they make nice gs> curves. -=- -=- -=- (BT) BIFF and further use of forgery -=- -=- -=- (BT) Alt.sex becomes #1 read USENET group -=- -=- -=- (BT) comp.unix.sco -- debate over corporations and the net (yet again) -=- -=- -=- (BT) June 90/"Internet Porno Ring"/Houston Chronicle