alt.support.jock-strap

Long-form reference · 6,281 words · about a 27-minute read

Everything the record holds

The administrative history of alt.support.jock-strap can be read in full in about two minutes, and reading it in full is the point of this page. The Internet Systems Consortium maintains a public archive of Usenet control messages — the small administrative articles by which newsgroups are created and destroyed — filed one compressed file per group name. The file for this group contains three messages. The first, in February 1994, asks the world’s news servers to create the group. The second, sent a little over a day later, asks them to delete it. The third, posted sixteen months afterwards from a corporate site in Minnesota, creates it again inside a single company. That is the whole of the archived record: no charter, no vote, no announcement, no farewell.

Three control messages would be a thin foundation for an article if they were only a curiosity, and this page does not pretend otherwise. What makes them worth setting out carefully is that between them they demonstrate, in a few hundred words of 1994 plain text, the property of the alt.* hierarchy that is most often asserted and least often illustrated: that creating a newsgroup and removing one were both requests, addressed to strangers who were under no obligation to answer, and that nobody ever collected the answers. A group could therefore be removed and remain alive at the same moment, depending entirely on whose news server a reader happened to be using. This one was. Three decades later it is still listed in the reference files that describe the shape of Usenet, carrying a one-line description that no surviving newgroup message ever proposed.

The rest of this article works through that record message by message, then through the two configuration files that explain the outcome, then through the form of joke the group belonged to and the branch of the hierarchy that was invented to hold such jokes. It stops where the documents stop.

2 February 1994: the newgroup message

The first message in the file carries a Date header of Wed, 2 Feb 94 17:44:27 GMT. Its From line reads [email protected] (MAC); it was posted to alt.config, the discussion group in which alt.* proposals were conventionally aired; and it carries the header Control: newgroup alt.support.jock-strap. The body opens with two sentences:

As per discussion in alt.config, alt.support.jock-strap has been created. Newsadmins: Please create this group and as always, if you have any questions feel free to contact me.

Those two sentences do most of the work. “Has been created” is the tense in which alt.* creation was habitually announced, and it was never quite true: nothing had been created and nothing would be until individual administrators chose to act, so the message is a request phrased as a report of an accomplished fact. The habit was ordinary rather than deceitful — every experienced reader of alt.config knew exactly how much a newgroup message accomplished by itself, which was nothing — but it is worth noticing, because the gap between the grammar and the mechanism is the gap this whole article lives in. The invitation that follows, to contact the sender with questions, is the other half of the etiquette: a proposer was expected to be reachable by the administrators he was addressing.

The message then supplies, under the heading For your newsgroups file:, the line an administrator would paste into the server’s newsgroups file, the flat text table that maps every carried group name to a one-line description:

alt.support.jock-strap   “Cup check....*WHACK*!”

The quotation marks are in the original and are not conventional; the format guidance ISC publishes for that file asks for a group name, a tab and bare descriptive text. This proposed description is not the one the group carries today, and the discrepancy turns out to be traceable; that is the subject of a later section.

Three headers on the message repay attention, because each is load-bearing later. It carries Approved: [email protected], an address at NASA’s Goddard Space Flight Center. RFC 1036, the December 1987 Standard for Interchange of USENET Messages by Mark Horton and Rick Adams that still governed the article format, states that a newgroup message “should be ignored unless there is an ‘Approved’ line in the same message header”, and the ISC archive to this day refuses to file any control message lacking one. The header was the nearest thing the format offered to a credential, and it was no harder to type than any other line — a point the hierarchy would spend the following decade discovering. The Organization and X-Disclaimer headers place the sender on Nyx, described in the message’s own words as “a public access Unix system run by the University of Denver for the Denver community”, with the University disclaiming that it had “neither control over nor responsibility for the opinions of users”. Public-access Unix systems of that kind were one of the main routes by which people outside universities and companies reached Usenet in the early 1990s, and they are also why a university’s name appears in the headers of postings it had nothing to do with.

3 February 1994: the rmgroup

The second message is dated Thu, 3 Feb 1994 22:15:02 GMT — twenty-eight and a half hours after the first. It was posted to alt.config and to control, it carries Control: rmgroup alt.support.jock-strap, and it is formatted as an ordinary follow-up: its Subject line is “Re: newgroup alt.support.jock-strap” and it quotes the message it answers. Its author signed it with his own name and institution: John Grohol, of the Center for Psychological Studies at Nova Southeastern University in Florida. After the quoted request it reads, in its entirety:

Don’t confuse “discussion” with “support to create.”

followed by three bulleted reasons, reproduced here exactly, misspelling included:

• Newsgroup doesn’t purport to follow the same meaning that other alt.support.* newsgroups do in this hiearchy.
• Newsgroup is a stupid idea better proposed under the new alt.silly-group.* hiearchy.
• Discussion was not for the creation of this newsgroup.

The three objections are of three different kinds, and it is worth separating them because they were the standard repertoire of alt.config argument. The first is about namespace: that a prefix carries a meaning and this proposal did not honour it. The second is about placement: not that the joke should not exist, but that it belonged in a branch built for jokes. The third is procedural: a denial that the discussion the proposer cited amounted to what he said it amounted to.

The sender’s position supplies context without settling anything. He signed with the degree M.S. and his university department, and the Wikipedia article on Psych Central, the mental-health website he founded, describes him in that period as a Florida psychology graduate student who had begun answering mental-health questions in Usenet newsgroups. That is as far as the published record goes, and it is context rather than a verdict. The archive contains his three reasons and nothing at all about how they were received.

Why both messages were only requests

RFC 1036 describes the rmgroup control message in the confident voice of a specification: “This message removes a newsgroup with the given name. Since the newsgroup is removed from every host on the network, this command should be used carefully by a responsible administrator.” The deployed network never worked that way, and by 1994 nobody expected it to. A control message is an ordinary article that happens to carry a Control header; it propagates from server to server like any other article, and at each server the news software either acts on it or files it away, according to how that server’s administrator has configured the software. Since anyone could send an rmgroup and forging one cost nothing, the ordinary defensive posture in alt.* became to ignore all of them.

That posture is not folklore. It is written into the configuration file that ships with the news software. control.ctl is the file in which a server records whose control messages it will honour for which hierarchies; the copy distributed by ISC, and kept in step with the one in the INN source tree, carries an alt.* stanza of exactly four lines:

• newgroup:*:alt.*:doit
• newgroup:[email protected]:alt.*:drop
• newgroup:tale@*uu.net:alt.*:drop
• rmgroup:*:alt.*:drop

Each line is message type, sender pattern, group pattern, action, and the last matching line wins. The first says: act on a newgroup for any alt.* group from anybody at all. The second and third drop newgroups that appear to come from the Big-8 control addresses, which as the file’s own comment notes never issue alt.* control messages, so that their appearance there can only be forgery. The fourth says: ignore every rmgroup for every alt.* group, from everyone, without exception. The default configuration of the standard Usenet server, in other words, is that alt.* groups can be created by anyone and deleted by no one.

The file’s own commentary is unusually frank about the status of all this. Its alt.* section advises small sites to drop alt.* control messages entirely and add groups only on user request, notes that “tons of alt.* newgroups are sent out regularly with the intent more to create nonsense entries in active files than to actually create a useable newsgroup”, and then concedes the whole question of authority: “Be aware that there is no official, generally accepted alt.* policy and all information about alt.* groups available is essentially someone’s opinion, including these comments.” For contrast, the same file’s entry for the Austrian at.* hierarchy drops all newgroups and rmgroups from everybody and then re-admits only those from one named address whose signature verifies. alt.* had no such address to trust, and never acquired one.

Applied to this group, the arithmetic is exact. The February 1994 newgroup was a valid, well-formed control message, so a server running the default configuration created the group. The February 1994 rmgroup was equally valid and equally well-formed, and a server running the same default configuration ignored it. Both messages went to the same servers on the same feed within thirty hours of each other, and the standard configuration answered yes to one and no to the other. Neither sender was overruled, because there was nobody to overrule them; the disagreement was simply resolved differently, and privately, on every server that received it.

Where the surviving description came from

ISC also publishes the two files most people mean when they speak of a list of Usenet newsgroups: active, which names groups and their posting status, and newsgroups, which pairs each name with a one-line description. Both are regenerated hourly. The policy document governing them, originally written by David Lawrence and now maintained by Russ Allbery, treats alt.* as an explicit special case: “In each hierarchy, all properly formatted newgroups … are automatically executed, and all rmgroups are ignored. No manual changes will be made in either hierarchy except under very special circumstances, and will require substantial justification. Similarly, the moderation status and description of the newsgroup will be set to whatever was present in the last control message for that group.”

So the group is there. Today’s newsgroups file, between the entries for jaw disorders and kidney failure, carries the line:

alt.support.jock-strap   Helping genitals cope.

and today’s active file lists alt.support.jock-strap with the posting flag y, meaning an ordinary unmoderated group open to posting. Thirty-two years after a control message asked the network to delete it, the group is a normal entry in the canonical description of the namespace.

That description, however, is not the one the newgroup message proposed, and tracing where it came from is the one piece of genuine archival detective work this small page has to offer. The rule quoted above says the description is taken from the most recent valid newgroup or checkgroups control message for the group. No newgroup message in the archive carries the wording “Helping genitals cope.” It appears instead in a checkgroups message — the third kind of control article, which carries an entire group list at once and asks servers to reconcile their own lists against it — posted on 16 December 1994 from a machine at AT&T Global Information Solutions. That message is three thousand and fifty-five lines long and covers the whole namespace as that site carried it, the alt.* names being much the largest block in it. The description appears again, unchanged, in a signed alt.* checkgroups of 7,575 lines sent to alt.config on 13 July 2000 by Jay Denebeim of Deep Thought, whose Originator header describes the sending account, with a certain irony, as the “Alt Rmgroup Account”; the archive holds a second copy of the same list dated the following day. Those three appearances are the only mentions of the group in the entire run of archived checkgroups messages, which covers 1993 to the present.

Two conclusions follow, one small and one not. The small one is that somebody rewrote the description between February and December 1994 and the record does not show who, or on what authority, or why — a reminder that the descriptions in the newsgroups file are free text supplied by whoever last sent a control message, not curated metadata. The larger one is that more than six years after the deletion request, the periodic list that told administrators across the world what alt.* contained still had this group in it, under a description of unknown authorship, and every server that reconciled against that list would have been told the group existed.

June 1995: created again, inside one company

The third and last message in the file is dated Mon, 26 Jun 1995 17:50:20 GMT. It comes from Bert Hyman at rsvl.unisys.com — Unisys, in Roseville, Minnesota — carries Approved: [email protected], and its control header reads newgroup alt.support.jock-strap y. The trailing y is the mode flag: RFC 1036 specified an optional keyword moderated, but the convention that prevailed in practice, and that the ISC archive still validates against, was a single letter, y for an unmoderated group and m for a moderated one. The body of the message is three lines, all of them signature. There is no explanation and no “For your newsgroups file:” line.

The most informative thing on it is the header Distribution: unisys. RFC 1036 defines Distribution as a way to “alter the distribution scope of the message”: for an article to be transmitted, the receiving site must both carry one of the listed newsgroups and accept one of the listed distributions. A message marked unisys was meant to travel only on Unisys machines. The reason it can be read today is that it did not stay there — the Path header on the archived copy begins at uunet, so the message left the company and was filed with the rest. Distribution, like newgroup and rmgroup, turned out to be a request rather than an instruction, which is the through-line of this entire record.

What the message documents is mundane and, for this page, decisive. In mid-1995, sixteen months after the deletion request, a news administrator at a large computer company created this group on that company’s own servers. Why is not recorded, and this page does not guess. What matters is that he could, that he needed nobody’s agreement to do it, and that his action had no effect whatever on anybody else’s server. Because his newgroup carried no description line, it could not update the public list either: the archive’s rules require a newgroup message to contain a valid description, introduced by exactly that “For your newsgroups file:” line, before it is automatically processed into the ISC files. Locally the group existed. Globally nothing changed. Both statements were the normal condition of an alt.* group, and holding them in mind at once is the only way the hierarchy makes sense.

alt.config, and what counted as discussion

Both 1994 messages were addressed to alt.config, and their disagreement is entirely about what happened there. The hierarchy’s origins, the ease of creation that defined it and the later history of spam and cancels are told on the alt.* hierarchy page; what this group’s record adds is a close view of the mechanism’s weak joint.

alt.config had no membership, no officers, no procedure and no power. It was a newsgroup in which people who cared about the shape of the namespace argued about proposed names: whether an existing group already covered the subject, whether the name fitted the hierarchy’s conventions, whether the thing would attract enough traffic to be worth an entry in several thousand active files. Nothing was ever counted, because there was nothing to count with. There was no register of participants, no quorum, no closing date and no result posting of the sort the Big-8 groups produced. The proposer’s formula — “as per discussion in alt.config” — was the standard citation, and it was unfalsifiable in both directions.

Grohol’s reply names the flaw precisely. A proposer could report truthfully that a discussion had occurred while a reader of the same discussion could report truthfully that it had gone against him, and no third party had any way to adjudicate between them. This is not a defect that could have been repaired without inventing an authority, and inventing an authority was the one thing the hierarchy existed to avoid.

What stood in for a verdict was propagation. A group aired in alt.config and generally welcomed would be added by enough administrators to be worth posting to; one created over evident objection would be added by few, would attract rmgroups from the people who had objected, and would sit in a scattering of active files as a name with no readers. The period’s own vocabulary for this was blunt: in a January 1994 control message about a different disputed group, Joel Furr wrote that it would be a Fine Thing if his opponents were “forever relegated to the hell of 10% propagation”. That is the punishment the system had, and it was the only one: not deletion, but existing badly. It is a sanction of a recognisable kind — reputational, slow, and entirely unappealable — and it held the hierarchy together for decades without ever producing a rule anyone could quote.

Even alt.config could be deleted

The control archive’s file for alt.config itself makes the point better than any argument could. On 8 March 1994, about five weeks after the exchange over this group, a message carrying Control: rmgroup alt.config went out with a Subject line of “Kibo” and a body consisting of the single word “Kibo”. Fourteen hours later a newgroup message re-creating the group arrived from the University of Tennessee, its body reading “Sorry guys, alt.config stays.” and supplying a fresh description line. Two days after that another arrived from the University of North Carolina, dismissing the removal as “a nice little forged message from our friends at indiana” — the rmgroup had indeed been injected through a machine at Indiana University — and adding, above a signature, “Go away.”

The same file holds one more entry that belongs in this article. On 11 March 1994 an rmgroup for alt.config was sent from [email protected], over the signature “MAC”: the same address, the same Nyx disclaimer and the same sign-off as the newgroup message that had asked the world to create alt.support.jock-strap five weeks earlier. Its body reads, misspellings and all, “sysadmnis: It has been decided by alt.config that alt.config is no longer needed. Please remove this group from you active file.” Three newgroup messages restoring the group followed within fifteen hours. One of them consists of the quoted rmgroup and the single word “Moron.” Another, from Penn State, answers the constitutional point directly: “Some of us would prefer alt to not to be reduced to enforced anarchy.” Whether the sender meant the removal seriously or as a demonstration is not recorded, and this page does not guess; what the file records is that the person who proposed this group also, five weeks later, asked the network to delete the forum in which such proposals were made.

Between 8 and 11 March 1994 the archive holds two rmgroups for alt.config and five newgroups restoring it. The forum in which alt.* proposals were debated was thus removed and restored twice inside four days by people using precisely the mechanism under debate, and the network absorbed the episode without visible difficulty, because on almost every server neither kind of message did anything. Anyone reading this group’s file in isolation might take a same-week rmgroup as a dramatic event, a group struck down at birth. Placed beside the rest of the archive, it was ordinary weather. What is unusual about alt.support.jock-strap is not that it was rmgrouped; it is that its file is short enough, and its three messages articulate enough, to be read whole.

alt.silly-group: a branch invented for jokes, and what became of it

Grohol’s second objection pointed the proposal somewhere specific: “better proposed under the new alt.silly-group.* hiearchy”. The branch was genuinely new. Its earliest surviving newgroup in the ISC archive is alt.silly-group.lampreys, sent on 20 January 1994 — thirteen days before the jock-strap proposal — by Joel Furr, whom Wikipedia describes as a Usenet personality of the early and mid-1990s, and posted, as it happens, from the same Nyx public-access system in Denver that the jock-strap proposer used. Furr’s message states the branch’s purpose in one sentence: “I put the group under alt.silly-group because we’ve needed a hierarchy to put all the inane new alt groups (like alt.spork) in.” He adds, with the scrupulousness the situation deserved, “Believe it or not, this new hierarchy was actually discussed in passing. Well, at least a message or two.”

Two and a half hours later an rmgroup for that group arrived from the University of Toronto, objecting that “the alt.config discussions have not come to a close”. The pattern that this page treats as its subject — newgroup, rmgroup within hours, group survives anyway — was thus already thirteen days old when it happened to alt.support.jock-strap, and it involved the same host and the same argument.

The lampreys file also shows what a contested creation looked like when neither side would stop. Over the following three days the newgroup and the rmgroup were re-sent several times each, with the reasoning carried in the bodies of the control messages themselves: an accusation that the proposer was creating groups he had no intention of using and sabotaging them with snide newgroup messages, a reply that his opponents were running “an elaborate scam”, and a counter-proposal that the discussion belonged under a plainer name. It ended not in a decision but in a withdrawal, the objector writing that he formally withdrew from the contest. Grohol appears in that argument too: a remark of his is quoted in a control message of 23 January 1994 — “You have one group. Use it. And Enjoy!” — eleven days before he asked for the removal of alt.support.jock-strap and directed its proposer to the very branch he had just told someone else to be content with.

When Grohol wrote that objection on 3 February 1994, the archive records exactly two names ever newgrouped under alt.silly-group.*: the lampreys group, and alt.silly-group.im-matt, created five days earlier. The branch he was recommending as the proper address for joke proposals was thus a fortnight old and had two inhabitants.

The idea behind alt.silly-group.* was a sorting one: give joke proposals an address of their own so that they need not contest the meaning of branches other people were using in earnest. Whether it worked is documented, and the answer is instructive. A newgroup message for alt.silly-group.persian, sent on 18 November 1994 by Tom Livingston, describes what the branch had become by then:

The silly-group top level hierarchy was mostly created to house joke newsgroups, but has since mutated into the only place a proposed alt. group can be created without a consensus in alt.config and not get rmgrouped. This appears to be the only place that such discussion could find a home.

The group in question was not a joke at all. It was an attempt to provide a home for Persian discussion after several proposals had met, in the proposer’s words, serious and valid objections, and the message asks administrators not to “misjudge this group because of it’s name”, adding that its placement “is only due to realism”. The proposal it quotes, from Joel Furr, is candid about the tactic: the name is “so absurd” that the group’s likely opponents would not follow it there.

Nine months after alt.silly-group.* was recommended as the correct address for unserious proposals, it had become the address of last resort for serious ones that could not survive alt.config. A naming convention created to segregate jokes had turned into a sanctuary from the discussion process, which is close to the opposite of the sorting function it was proposed to perform. It is a small, well-documented example of a general truth about the hierarchy: conventions could be invented freely and enforced not at all, so they drifted to whatever use the population found for them.

The branch’s later history is brief. The control archive holds files for fourteen names under alt.silly-group.*, the most recent of them first newgrouped in 1996. Today’s ISC newsgroups file lists two: alt.silly-group.im-matt and alt.silly-group.persian, the latter described, after all that, as “Rugs, cats, lambs, etc.” The lampreys group that the branch was invented to hold is not among them, and alt.support.jock-strap — the group that was told to move there — is still in the list under its own name. The recommendation was never binding on anyone, including the namespace it recommended.

The parody support group as a form

alt.support.* is one of the larger branches of alt.*: the current ISC newsgroups file lists 301 names beginning alt.support., covering illnesses, injuries, bereavements, addictions, disabilities and family circumstances of every description. The branch was heavily used, and for a period in the 1990s it was for many people one of the few places where a specific and uncommon problem could be discussed with others who had it.

The list is alphabetical, and alphabetical order does not sort by intent. In today’s file, alt.support.jock-strap falls between alt.support.jaw-disorders (“Discussion of jaw-disorders.”) and alt.support.kidney-failure (“Helping people deal with kidney problems.”), a few lines below entries for incontinence and interstitial cystitis and a few above entries for learning disabilities, leprosy, lupus and Marfan syndrome. Nothing in the file distinguishes any of them from any other. Each is a name, a tab and a line of free text supplied by whoever sent the last control message, and no authority ever existed to check a description against the group’s contents. Fifty-eight of the 301 have no description at all, only the placeholder “No description.” Elsewhere in the same branch are entries that exist only to insult a named individual, and entries whose descriptions read simply “A test.”

That is the whole of the parody support group as a form, and it explains why it needed no content. The name and the description do all the work; the joke is complete at the moment a reader scanning a group list stops on it. Nobody had to post. A parody group is a joke told in the vocabulary of Usenet’s own machinery — a newgroup message, a description line, a position in the alphabet — and it is legible to people who read group lists, which in the 1990s meant most users, because the group list was the interface. A newsreader offering several thousand alt.* names in a selector gave no other clue to intent than the name itself.

The tension this produced was real, and Grohol’s first objection is its clearest contemporary statement: the prefix carried an implicit promise that a group beneath it was for people in difficulty, and a parody trades on the promise while declining to keep it. The opposing position was equally available and equally often made: alt.* had never promised that its names meant anything, and a hierarchy founded on the proposition that anyone may create anything cannot coherently police the tone of what gets created. Both arguments ran throughout the hierarchy’s life, and neither had any forum in which it could be decided. This directory preserves the earnest article alongside the joke: alt.support.ocd and its moderated sibling alt.support.ocd.moderated have pages of their own, and what a working alt.support group actually did — the reading, the answering, the long threads posted at three in the morning — belongs there and not here.

It is worth distinguishing the parody group from two neighbours preserved in this directory, because they are easily conflated. A venting group such as alt.complainers.bitch-n-moan is not a parody: it solicits real postings of a particular temper, and its name is an accurate description of what it wants. A humour group such as alt.humor.best-of-usenet is not a parody either: it is an earnest group whose subject matter happens to be funny. The parody group’s subject is the group list itself. Its content, if any, is beside the point, and its readership is everyone who ever scrolled past it.

The object in the name

Since the group’s joke rests entirely on its name, the object behind it gets a paragraph and no more. The jockstrap, or athletic supporter, is dated by Wikipedia to 1874, when it was made by C. F. Bennett of the Chicago sporting-goods company Sharp & Smith to support the bicycle jockeys working the cobblestone streets of Boston; Bennett’s later Bike Web Company patented and began mass-producing the Bike Jockey Strap in 1897, and the word itself is reported to have been in use since at least 1891, probably as a contraction of “jockey strap”. The protective cup invoked by the newgroup message’s proposed description is a later accessory: an insert of hard plastic or steel, perforated for ventilation, carried in a lined pouch, and used in cricket, boxing, hockey, baseball and other sports where being struck is expected. That is as far as this page needs to go; a longer digression into sporting goods would tell a reader nothing about the newsgroup.

What the archive is, and what its silence means

Absence of evidence in the ISC control archive has to be read carefully, because it is not the same as absence of activity. The archive’s own policy document describes it as “an archive of all newgroup, rmgroup, and checkgroups control messages seen on Usenet received since early 1991”. Its provenance and its gaps are stated: it lived at rpi.edu until the autumn of 1992, moved to uu.net and later to isc.org, and “a month or so of messages (around August 1992) are missing”. It updates every three hours, and its maintainers state plainly that no guarantee is made that it is full and complete.

It is also selective by design. Only valid control messages are filed, and validity is defined narrowly: the article must be well formed, under 64KB for a newgroup or rmgroup, must carry an Approved header, must not duplicate the Approved, Sender or X-PGP-Sig headers, and must name a group whose components use only permitted characters and obey the length and structure rules that ISC takes from RFC 5536. Anything malformed — and a great deal of the control traffic aimed at alt.* was malformed, some of it deliberately — is discarded rather than kept.

So the three messages in this group’s file are three messages that were well-formed and that reached one particular well-connected site. The correct reading of the record’s brevity is that no further valid control message concerning this group arrived at the archiving host, not that nothing further happened. It is entirely possible that other servers saw other messages; the archive cannot say.

The same caution applies in the other direction to the files that keep the group nominally alive. The ISC active and newsgroups files are a description of the namespace, not a census of servers: their stated intention is to be complete for all publicly available hierarchies, which means they record that a name exists and has a description, not that anybody carries it. The entry proves that alt.support.jock-strap is a listed and described name in the reference list that many administrators synchronise against. It proves nothing about whether any server carries the group today, and nothing whatever about whether anyone ever posted to it.

What the record does not show

No charter survives, and there is no reason to think one ever existed. alt.* groups were not required to produce the formal statement of purpose that a Big-8 proposal had to publish, and everything this group ever offered by way of self-description is the single quoted line in its newgroup message.

No discussion survives. The proposer cited a discussion in alt.config; the objector denied that it amounted to support. Neither the thread nor any individual posting from it has been located for this page, and nothing is claimed here about its length, its participants or its tenor beyond what the two control messages themselves say.

No traffic is documented. This page makes no claim that anyone ever posted to alt.support.jock-strap, and no claim that nobody did. The article numbers in the ISC active file are placeholders — all 45,003 entries in today’s file, from the busiest group to the emptiest, carry the identical pair of values 0000000000 0000000001 — and cannot be read as counts of anything.

No archive claim is made. Whether a searchable archive holds postings under this name has not been established for this article, and the standard access box above should be treated as a set of pointers worth trying rather than as an assertion that something is waiting at the end of them.

No readership or subscription figure exists for this group, or for almost any Usenet group. The network never counted its readers; the periodic surveys that estimated readership worked from a sample of volunteering sites and could not measure small groups with any reliability.

Finally, the intent behind the proposal is undocumented. The proposer signed as “MAC” from a public-access account and left no statement of purpose beyond the description line. Whether the group was meant as a joke, as a test of the creation mechanism, or as an eccentric but sincere request is nowhere recorded, and this page does not guess.

Scope and limits

The honest material about this newsgroup runs out at a definite point, and it is worth saying where. A group with one day of documented controversy, no recovered postings, no charter and no participants who can be identified beyond the senders of a handful of control messages cannot be written about in the way a decade-long working newsgroup can. What this page has instead is a complete primary record, small enough to quote in full and unusually articulate about the system that produced it. Padding beyond that would mean inventing colour, and inventing colour is the one thing this directory will not do.

What could be documented has been. The three control messages are quoted from the primary archive with their dates and headers. The two configuration files that explain why a group asked to be deleted is still listed are quoted from their current published versions. The checkgroups trail that produced the group’s surviving description has been followed through every year of the archive from 1993 to the present, and it yields exactly three appearances. The origin, the drift and the residue of alt.silly-group.* are established from contemporaneous control messages and from today’s newsgroups file rather than from recollection. The shape of the alt.support.* namespace, and the counts given for it, are taken from the current reference files. Everything else has been left out.

Two further limits are worth stating. First, this page is about a newsgroup and an administrative record, not about the people named in that record. The messages are quoted because they are the record; the individuals who sent them are described only as far as their own signatures and their own published biographies allow, and no motive is attributed to anyone.

Second, the argument the two 1994 messages had — whether a joke belongs in a namespace other people are using in earnest — is reported here and not adjudicated. It was not settled in 1994. It could not have been settled, because the hierarchy had deliberately built itself with no place where such a thing could be settled. The clearest statement of that fact remains the one in the configuration file that news servers still ship with: there is no official, generally accepted alt.* policy, and all information about alt.* groups is essentially someone’s opinion.