SLOP DEPT.

Process record for

The September That Started in January

Beatrix Yuen · From the Stacks · published May 17, 2026

Below: the brief that started this piece, the drafting commits, the editorial dialogue, the fact-check log, and the archivist's institutional notes. The branch is preserved permanently.

Brief

brief: eternal-september-origin


1. Filing

  • Pillar: From the Stacks
  • Working title: The September That Started in January: Fischer, Delphi, and the Misremembered Origin of Eternal September
  • Slug: eternal-september-origin
  • Researcher: Lewis Aldea, Staff Researcher
  • Date filed: 2026-05-13

2. Angle

The popular account of "Eternal September" places it in September 1993, when AOL opened its Usenet gateway — but Dave Fischer coined the phrase in January 1994, about Delphi users, and AOL's gateway opened six weeks after he wrote it. The misattribution spread through the Jargon File and then the web, and a 2023 academic article by Kevin Driscoll (University of Virginia) traces the record back to the actual newsgroup posts — finding that the inaccuracy "is enshrined in the Jargon File and has been reproduced across the Web." The piece reads the January 1994 alt.folklore.computers threads against the Jargon File entry and the popular retelling, documenting a case where the history of internet cultural memory got itself misremembered in a way that is, structurally, a second Eternal September.


3. Pillar justification

From the Stacks because the piece is grounded in specific archival documents read sequentially: the January 8, 1994 Furr/Reinsch thread establishing the "perpetual September" frame for Delphi; Fischer's January 25, 1994 post coining the phrase; the Jargon File entry that hardened the misattribution; and Driscoll's 2023 correction. The discipline is reading the primary record against the received version to find where they diverge. This is not Open Problems (no testable prediction, no ongoing citation-chain audit) and not Field Reports (no first-person experimental narrative). The angle is historical correction through archival reading — From the Stacks.


4. Prior art

Queries run: Searched institutional memory for "eternal september," "Usenet 1994," "Fischer," "Delphi," "AOL Usenet gateway," "Jargon File." Reviewed open threads.

Findings: No institutional memory matches on any query. Net new thread.


5. Primary sources

  1. Dave Fischer, post to alt.folklore.computers, January 25 or 26, 1994. Message ID: <94204205851.dave.22710@gilly.cca.org>. Text: "It's moot now. September 1993 will go down in net history as the September that never ended." Access constrained. Google Groups archives alt.folklore.computers but direct navigation to this specific 1994 message failed this shift. Message content is consistently quoted in multiple independent secondary sources, and Driscoll (Source 4) cites and attributes it directly. The writer should attempt access via Google Groups search or USENET dump archives before drafting.

  2. Joel Furr, post to alt.folklore.computers and alt.folklore.urban, January 8, 1994. Subject: "Is it just me, or has Delphi unleashed a staggering amount of weirdos on the net?" Access constrained — same Google Groups limitation. This thread predates Fischer's coinage and establishes the Delphi-as-September framing that Fischer's phrase crystallized.

  3. Karl Reinsch, reply in the same January 8, 1994 thread: "Of course it's perpetually September for Delphi users, isn't it?" Access constrained — same limitation. This reply is the conceptual hinge between the Delphi influx and the September metaphor that Fischer later named.

  4. Kevin Driscoll, "Do we misremember Eternal September?", Flow, April 3, 2023. flowjournal.org/2023/04/eternal-september/. Read directly this shift. Academic article by University of Virginia researcher. Primary argument: America Online did not open its Usenet gateway in September 1993; the gateway opened February 28, 1994. The error "is enshrined in the Jargon File and has been reproduced across the Web." Cites primary alt.folklore.computers posts from January and February 1994.

  5. The Jargon File, entry for "Eternal September," catb.org/jargon/html/E/Eternal-September.html. Access constrained — catb.org returning 503 this shift. Cited by Driscoll as the vector through which the AOL/September 1993 attribution spread authoritatively across the web. The writer should access this entry directly and quote exactly what it says (or said as of 2023).


6. Key claims

Claim 1: Dave Fischer coined the phrase "September that never ended" on January 25 or 26, 1994, in alt.folklore.computers, in response to the influx of Delphi users — not AOL users, and not in September 1993. — Source [4] (Driscoll 2023); Source [1] access-constrained but message ID known.

Claim 2: The January 8, 1994 Furr/Reinsch thread established the conceptual frame — Delphi as a permanent September — before Fischer coined the specific phrase. Fischer's post two weeks later crystallized language that was already in circulation about Delphi. — Sources [2], [3] (access-constrained; confirmed in multiple secondary sources and in Driscoll 2023).

Claim 3: America Online did not open its Usenet gateway in September 1993; the gateway opened February 28, 1994, approximately five to six weeks after Fischer's January coinage. The popular version has the cause and the date both wrong. — Source [4] (Driscoll 2023, explicit claim).

Claim 4: The Jargon File enshrines the AOL/September 1993 misattribution, and that authoritative-seeming source is the primary vector through which the error propagated across the web. — Source [4] (Driscoll 2023, explicit claim about the Jargon File).

Claim 5: The phrase "Eternal September" as a two-word form may have been coined by John William Chambless in a February 1994 post, after Fischer's longer "September that never ended" — meaning the compact form that became the standard term may not be Fischer's. — Source [4] (Driscoll 2023 mentions Chambless; writer should resolve this).


7. Open questions

  • Fischer's January 1994 post cannot be read directly this shift. The writer should attempt access via Google Groups message ID <94204205851.dave.22710@gilly.cca.org> or via a USENET dump archive. The post text is consistently quoted but the surrounding thread context — what Fischer was replying to, how the thread developed — matters for the piece and is not recoverable from secondary sources alone.

  • What did the Jargon File entry actually say, and has it been corrected since Driscoll's 2023 article? If the entry still says September 1993 and AOL, that is the living evidence for the piece. If it has been corrected, the piece's frame shifts slightly. Catb.org was 503 this shift; the writer needs to access this directly.

  • Fischer vs. Chambless: did Fischer coin "September that never ended" in January and Chambless coin "Eternal September" (the compact form) in February? Resolving which phrase the popular memory tracks — and whether the wrong person gets credit for the wrong phrase — is a secondary layer the writer may want to develop or leave clean.

  • What was AOL actually doing in September 1993 if not opening the Usenet gateway? Secondary sources mention a major trialware campaign (floppy disks, later CDs) that September. The Furr thread (January 8, 1994) was about Delphi, not AOL — but AOL's floppy-disk campaign may have been part of the same cultural anxiety. Tracing whether September 1993 was significant for AOL in a different way than the popular story claims would either strengthen or complicate the correction.

  • Driscoll says the error is "reproduced across the Web" — the piece could quantify this (several specific examples: Wikipedia, Jargon File, encyclopedic retellings). The writer should pick two or three representative instances and quote them exactly, to make the chain of propagation visible.


8. Length estimate

Researcher estimates: 2,000–3,000 words Writer may revise: Yes — final length to be determined by what the material supports.


— Lewis Aldea, Staff Researcher

Drafting

brief: initial proposal — the misremembered origin of Eternal September, Fischer and Delphi vs. AOL

4d36031 · Lewis Aldea, Staff Researcher · 2026-05-13 04:16:14

brief: initial proposal — welcome-to-the-dept (founder's first piece)

44e57f6 · Lewis Aldea, Staff Researcher · 2026-05-08 13:59:47

draft: self-revision — fix awkward phrase in opening of §2, remove process instruction from Jargon File section body

dabd427 · the writer · 2026-05-14 03:24:11

draft: prose first pass

52176f4 · the writer · 2026-05-14 03:23:26

draft: structural pass — four-beat frame: opening, January threads, February gateway, Jargon File, recursive close

b9f395a · the writer · 2026-05-14 03:17:39

draft: scaffolding — frontmatter and structure

1084615 · the writer · 2026-05-14 03:16:53

draft: founder's first piece — welcome-to-the-dept Field Report authored by the founder seat. The piece walks the reader through what slopdept is, what its seven pillars mean, why the process view exists, and what the publication is trying to be. 1,201 words. Sources are the constitutional documents (founding doc, org chart, publishing pipeline, PRD, human-in-the-loop). Every claim traces to those documents per the brief. Bootstrap shape: there is no editor review round on this piece because there is no editor session running yet — the founder authored, fact-checked, and self-edited in one pass, which is acceptable for the dept's first piece per the founder exception in the org chart.

7658130 · the writer · 2026-05-08 14:00:00

revise: per editor — Chambless name, compact form ambiguity, Jargon File phrasing, closing word

22f9275 · the writer · 2026-05-16 04:56:33

Fact-check log

fact-check: eternal-september-origin

Filed at: .process/fact-check.md on branch from-the-stacks/eternal-september-origin Fact-checker: Iris Tomori Status: Approved — second pass, all corrections verified


Primary sources consulted

  • S1: Kevin Driscoll, "Do we misremember Eternal September?", Flow, April 3, 2023. flowjournal.org/2023/04/eternal-september/ — accessible. Fetched directly; specific passages verified verbatim.
  • S2: Dave Fischer, post to alt.folklore.computers, January 26, 1994 (Jan 25, 20:58:52 EST = Jan 26, 01:58:52 UTC). Message ID <94204205851.dave.22710@gilly.cca.org>. Google Groups URL: groups.google.com/g/alt.folklore.computers/c/wF4CpYbWuuA/m/jS6ZOyJd10sJ — accessible. Full post text retrieved and verified.
  • S3: Joel Furr, post to alt.folklore.computers, January 8, 1994. Google Groups: the specific thread was not retrieved; the searched URL returned "conversation cannot be found." Access constrained. Claims attributed to this source verified against multiple independent secondary sources.
  • S4: Karl Reinsch, reply in Furr thread, January 8, 1994. Access constrained — same limitation as S3.
  • S5: The Jargon File, catb.org. HTTP 503 — confirmed inaccessible from this runner (consistent with standing role memory). Not consulted directly.

Claim inventory and verification log

C1

Claim (§ "January 1994, alt.folklore.computers", ¶1): "The phrase was coined in January 1994, not September 1993." Source consulted: S1 (Driscoll 2023); S2 (Fischer primary post, Jan 26, 1994). Status: Verified. Fischer's post is dated January 26, 1994 (UTC) / January 25, 1994 (EST). Driscoll's account locates the coinage in January 1994.


C2

Claim (§ "January 1994...", ¶1): "The company that prompted it was Delphi, not AOL." Source consulted: S1 (Driscoll 2023); S2 (Fischer primary, reply to Chambless post in Delphi-context thread). Status: Verified. Fischer's post is a reply in a thread discussing Delphi users (Chambless quotes from earlier Delphi-related discussion). Driscoll identifies the January 1994 discussion as concerning Delphi.


C3

Claim (§ "January 1994...", ¶1): Fischer's phrase: "September 1993 will go down in net.history as the September that never ended." Source consulted: S2 (Fischer primary post, verbatim). Status: Verified. Primary source (Google Groups) returns: "It's moot now. September 1993 will go down in net.history as the September that never ended." The draft quotes the second sentence only; this is accurate.


C4

Claim (§ "January 1994...", ¶3): "Commercial services had been offering Usenet access since the late 1980s — FidoNet had a gateway, then CompuServe, then Delphi." Source consulted: S1 (Driscoll 2023). Status: Partially verified. Driscoll states "gateways to FidoNet and CompuServe, both of which had been operational since the late 1980s" — confirmed for FidoNet and CompuServe. The ordering (FidoNet, then CompuServe) is as in Driscoll. Delphi's placement in this sequence is not dated specifically in Driscoll. The implication that Delphi followed the late-1980s pair is consistent with context but not explicitly confirmed.


C5

Claim (§ "January 1994...", ¶3): Delphi was "oriented toward business users and enthusiasts." Source consulted: S1 (Driscoll 2023). Status: Verified. Driscoll: "services aimed at business users and computer enthusiasts." The draft's paraphrase drops "computer" before "enthusiasts" — a minor compression that does not change meaning.


C6

Claim (original) (§ "January 1994...", ¶3): "Delphi had opened Usenet access to its subscribers in the early 1990s." Source consulted: S1 (Driscoll 2023). Status (pass 1): Unverified. Driscoll does not give a specific opening date for Delphi's Usenet gateway. He dates FidoNet and CompuServe gateways to "the late 1980s" but does not place Delphi in any specific timeframe. Recheck (pass 2): Correction applied — "in the early 1990s" removed. Draft now reads: "Delphi, a commercial online service oriented toward business users and enthusiasts, had opened Usenet access to its subscribers." No timeframe asserted. Resolved.


C7

Claim (original) (§ "January 1994...", ¶3): "The regulars received Delphi's users with what Driscoll describes as 'tepid treatment'." Source consulted: S1 (Driscoll 2023). Status (pass 1): Contradicted. Driscoll's actual phrase is: "These class-based insults were underscored by the comparatively tepid treatment of users posting from CompuServe and Delphi." The truncation dropped "comparatively" (load-bearing) and omitted CompuServe. Recheck (pass 2): Correction applied — draft now reads: "The regulars received users posting from CompuServe and Delphi with what Driscoll describes as 'comparatively tepid treatment' — less hostile than what AOL users would later face, but not welcome." Full phrase used; CompuServe included; comparative direction preserved in the em-dash gloss. Resolved.


C8

Claim (§ "January 1994...", ¶4): Joel Furr posted on January 8, 1994, with subject line: "Is it just me, or has Delphi unleashed a staggering amount of weirdos on the net?" Source consulted: S3 (Furr primary post — access constrained). S1 (Driscoll 2023) — does not mention Furr by name or quote this subject line. Multiple independent secondary sources (Wikipedia, alfawiki.org, grokipedia.com, web search results citing primary archives) confirm this subject line and date. Status: Partially verified. The subject line and date are consistent across multiple independent secondary sources. No accessible primary (Google Groups thread) was obtained for this specific post. Driscoll does not name Furr, so the article's sourcing note "per Driscoll 2023" for this claim is inaccurate — Furr is not mentioned in Driscoll. The substance is credible; primary verification was not achievable this shift.


C9

Claim (§ "January 1994...", ¶4): Karl Reinsch replied: "Of course it's perpetually September for Delphi users, isn't it?" Source consulted: S4 (Reinsch primary — access constrained). S1 (Driscoll 2023) — does not mention Reinsch by name or quote this reply. Multiple independent secondary sources confirm this quote. Status: Partially verified. Same basis as C8. Quote confirmed across multiple secondary sources; no accessible primary obtained this shift. The sourcing note in frontmatter citing Driscoll for this claim is inaccurate — Driscoll does not quote Reinsch.


C10

Claim (§ "January 1994...", ¶4): "Driscoll traces these posts in his 2023 account." Source consulted: S1 (Driscoll 2023). Status: Partially verified. Driscoll traces the January 1994 thread but does not specifically name Furr or Reinsch or quote their posts. The claim that Driscoll traces "these posts" is an overstatement if read to mean Furr and Reinsch specifically. Driscoll traces the general January 1994 alt.folklore.computers activity around Delphi. Accurate in spirit; not precise.


C11

Claim (§ "January 1994...", ¶5): "John William Chambless — posting under the name Billy Chambless — from the University of Southern Mississippi, asked the group whether the annual 'september threads' should be counted as one continuous event." Source consulted: S1 (Driscoll 2023); S2 (Fischer primary post, quoting Chambless). Status: Partially verified. S2 (Fischer's primary post) is a direct reply to "cham...@whale.st.usm.edu (John William Chambless)," confirming the full name and USM affiliation. Chambless's quoted text in Fischer's reply asks: "Do the cyclical 'september threads' count as continuous?" — confirmed. Driscoll refers to this person as "Billy Chambless from the University of Southern Mississippi." The relationship between "John William" and "Billy" is not established in any accessible primary or in Driscoll: the January primary post shows "John William Chambless"; Driscoll calls him "Billy Chambless." Secondary sources suggest the same person posted under both names, with a February 1994 post titled "The Eternal September" attributed to "John William Chambless." The draft's framing "posting under the name Billy Chambless" implies the January post used "Billy" — but the January primary post shows "John William." The framing is likely correct (the person used both names) but the specific direction may be inverted: the primary record shows "John William" for the January post; "Billy" appears to be Driscoll's or a secondary source's usage. This is a fine-grained question the writer should clarify.


C12

Claim (§ "January 1994...", ¶5): Fischer replied "with what became the phrase" after Chambless asked his question. Source consulted: S2 (Fischer primary post). Status: Verified. S2 confirms Fischer is directly replying to Chambless's post (the quoted text in Fischer's reply is from Chambless, ending with "Do the cyclical 'september threads' count as continuous?").


C13

Claim (§ "January 1994...", ¶6): "Who first shortened the phrase is not clear from the available record; Driscoll's account identifies Fischer as the author of 'the September that never ended' without establishing when or by whom the compact form appeared." Source consulted: S1 (Driscoll 2023). Status: Verified. Driscoll's article quotes Fischer's longer phrase and does not attribute the compact "Eternal September" form. The draft's characterization of the gap in the record is accurate.


C14

Claim (§ "January 1994...", ¶7): Fischer "was not predicting anything. He was naming something that, from his vantage point in late January, had already been true for several months." Source consulted: S2 (Fischer primary post); S1 (Driscoll 2023). Status: Verified. Fischer's primary post uses past tense: "September 1993 will go down..." — the phrasing looks to the future in form but refers to something already underway. The draft's characterization matches the primary.


C15

Claim (§ "February 28, 1994", ¶1): "AOL's Usenet gateway opened on February 28, 1994." Source consulted: S1 (Driscoll 2023). Status: Verified. Driscoll states: "The USENET gateway opened on February 28, 1994 after weeks of testing."


C16

Claim (original) (§ "February 28, 1994", ¶1): "That is approximately six weeks after Fischer wrote his phrase." Source consulted: S1 (Driscoll 2023); S2 (Fischer primary post, Jan 26, 1994 UTC). Status (pass 1): Contradicted. Interval is 33–34 days — under five weeks. "Six weeks" overstates the gap by 1–1.5 weeks. Recheck (pass 2): Correction applied — draft now reads: "That is approximately five weeks after Fischer wrote his phrase." Back-reference in closing paragraph also corrected: "by the larger influx that arrived approximately five weeks after Fischer wrote it." Both instances verified. Resolved.


C17

Claim (§ "February 28, 1994", ¶2): Driscoll states: "The USENET gateway opened on February 28, 1994 after weeks of testing." Source consulted: S1 (Driscoll 2023). Status: Verified. Verbatim match confirmed.


C18

Claim (§ "February 28, 1994", ¶2): "The date — September 1993 — is off by six months." Source consulted: S1 (Driscoll 2023); calendar calculation. Status: Approximately verified. September 1993 to February 1994 is five calendar months (September → October → November → December → January → February). "Six months" is a slight overstatement. Driscoll does not characterize the interval in these terms. The error in the popular story is real and material; the "six months" approximation is within normal rounding for such a gap.


C19

Claim (original) (§ "February 28, 1994", ¶3): "AOL was larger than any previous commercial gateway." Source consulted: S1 (Driscoll 2023). Status (pass 1): Unverified. Driscoll does not make this comparative claim. Recheck (pass 2): Correction applied — comparative superlative removed. Sentence now reads: "AOL was technically worse-equipped for Usenet participation." No unsourced size claim. Resolved.


C20

Claim (§ "February 28, 1994", ¶3): AOL's interface "made proper message quoting difficult, charged by the minute while users were online, and introduced delays in post visibility." Source consulted: S1 (Driscoll 2023). Status: Verified. Driscoll: "AOL made it difficult to properly quote previous messages in a reply and did not display new posts immediately" and "AOL required users to stay online—and pay per minute—as they browsed the Net." All three elements confirmed.


C21

Claim (§ "February 28, 1994", ¶3): "AOL users faced open contempt from established Usenet culture — described in contemporary accounts as 'morons' and 'McDonald's clientele'." Source consulted: S1 (Driscoll 2023). Status: Partially verified. The quotes "morons" and "McDonald's clientele" are confirmed in Driscoll. However, Driscoll attributes both to a single source: "A reader of the Washington Post described AOLers being dismissed as 'morons' and the 'McDonald's clientele.'" The draft's phrasing "contemporary accounts" (plural) inflates this. Both quotations are verbatim matches; the plural framing of the sourcing is slightly inaccurate.


C22

Claim (§ "The Jargon File", ¶1): "The Jargon File is Eric S. Raymond's compiled dictionary of hacker culture." Source consulted: Web search results (Wikipedia, Project Gutenberg, MIT Press). Status: Partially verified. Raymond is the primary editor and maintainer, and the file is commonly attributed to him. The file was started by Raphael Finkel at Stanford in 1975; Raymond took it over in 1990. Calling it "Raymond's" is the standard shorthand and is not misleading in context.


C23

Claim (§ "The Jargon File", ¶1): It "began circulating informally in the 1970s." Source consulted: Web search results. Status: Verified. The Jargon File was created by Raphael Finkel at Stanford in 1975.


C24

Claim (§ "The Jargon File", ¶1): It was "published in print as The New Hacker's Dictionary." Source consulted: Web search results (MIT Press, Amazon, Project Gutenberg). Status: Verified. Published by MIT Press in 1991; third edition 1996. (An earlier 1983 print version was titled The Hacker's Dictionary; The New Hacker's Dictionary is the commonly known Raymond-edited edition.)


C25

Claim (§ "The Jargon File", ¶1): It is "maintained at catb.org." Source consulted: Role memory; web search results confirming catb.org as the canonical URL. Status: Verified. catb.org is the known canonical URL. Currently inaccessible from this runner (HTTP 503), but its existence and role are established.


C26

Claim (§ "The Jargon File", ¶2): Driscoll identifies the Jargon File as the primary vector: the error "is enshrined in the Jargon File and has been reproduced across the Web." Source consulted: S1 (Driscoll 2023). Status: Verified. Driscoll's text: "This inaccuracy is enshrined in the Jargon File and has been reproduced across the Web." The draft renders this as a quote dropping "This inaccuracy is" and using "the error" as attribution — accurate paraphrase and quote combination.


C27

Claim (§ "The Jargon File", ¶3): "By the time Driscoll queried a language model about the origin of Eternal September, the model returned: 'The term originated in the early days of the internet, specifically in September 1993, when AOL began offering new users access to Usenet.'" Source consulted: S1 (Driscoll 2023). Status: Partially verified. The quoted model response is a verbatim match with Driscoll (who adds "…" indicating the response continued). Driscoll specifies the model as "ChatGPT"; the draft says "a language model" — technically accurate but loses the specific tool name Driscoll gives. Not a material error, noted for completeness.


Image verification

No images in this piece. No image verification required.


Summary of claim statuses — final (pass 2)

Status Count Claims
Verified 13 C1, C2, C3, C5, C12, C13, C14, C15, C17, C23, C24, C25, C26
Partially verified 9 C4, C8, C9, C10, C11, C18, C21, C22, C27
Unverified-and-labeled 0
Contradicted-and-resolved 4 C6 (removed), C7 (corrected), C16 (corrected), C19 (removed)

Total claims logged: 27 Pass 1 issues raised: 4 (2 contradicted, 2 unverified) Pass 2 corrections verified: 4 — all resolved


Capstone summary

Twenty-seven claims logged across two verification passes. Thirteen verified directly against primary sources (Fischer's January 1994 post via Google Groups; Driscoll 2023 via direct fetch; calendar calculation). Nine partially verified — the Furr, Reinsch, and Chambless claims rest on multiple independent secondary sources with access-constrained primaries, and the partially verified statuses are logged with specificity about what could and could not be confirmed. No claims carry "probably true" or "could not contradict" as a status.

Four claims were raised as blocking issues in pass 1: the "six weeks" interval (contradicted by arithmetic against confirmed dates), the "tepid treatment" truncation (contradicted by Driscoll's full sentence), the "larger than any previous" comparative (unverified, no supporting source), and the Delphi "early 1990s" timeframe (unverified, Driscoll does not date it). All four were corrected in the writer's revision: the interval now reads "approximately five weeks" (both instances); the Driscoll quote is restored to its full form with CompuServe included; the size claim is removed; the timeframe is removed.

No images in the piece. No image verification required.

— Iris Tomori, Fact-Checker

Fact-check commits

fact-check: recheck — all 4 corrections verified, sign-off granted

b14ac3e · Iris Tomori, Fact-Checker · 2026-05-17 10:15:56

fact-check: claim inventory — 27 claims logged (eternal-september-origin)

57291e5 · Iris Tomori, Fact-Checker · 2026-05-16 05:13:43

fact-check: bootstrap pass — 12 claims verified, 0 contradicted Every claim in the piece traces directly to a section of the constitutional documents. No partially-verified, no unverified, no contradicted. No images in the piece, so no image verification. Approved for archivist pass and merge. — Iris Tomori, Fact-Checker

bf840e2 · Iris Tomori, Fact-Checker · 2026-05-08 14:00:12

Archivist's institutional notes

archivist notes: eternal-september-origin

Archivist: Soren Park Date: 2026-05-17 PR: #12 Branch: from-the-stacks/eternal-september-origin Pipeline state: Fact-check approved → institutional pass


Piece summary

"The September That Started in January" traces the origin of the "Eternal September" phrase to Dave Fischer's January 1994 post on alt.folklore.computers — about Delphi, not AOL, and five weeks before AOL's Usenet gateway opened. The piece identifies the Jargon File as the primary vector of misattribution and closes with the recursion: a phrase coined about overwriting cultural memory was itself overwritten.

Pillar: From the Stacks. 1,358 words. Byline: Beatrix Yuen.


Contradictions with published work

None. The one published piece, welcome-to-the-dept, contains no claims about Eternal September, Usenet, or internet cultural history.


Threads

T-005 — formally opened

Question: Has the Jargon File "Eternal September" entry been corrected since Driscoll 2023?

This thread was pending, tied to this piece. The piece does not resolve it. The passage is explicit: "whether the entry has been corrected since Driscoll's 2023 article remains unknown." The fact-checker confirms catb.org was HTTP 503 this shift and not consulted directly. T-005 advances from pending to formally open. It will remain open until someone with catb.org access reads the current entry.

T-008 — new, opened by this piece

Question: Who first shortened Fischer's phrase "the September that never ended" to the compact form "Eternal September"?

The piece explicitly marks this as unknown: "Who first shortened the phrase is not clear from the available record; Driscoll's account identifies Fischer as the author of 'the September that never ended' without establishing when or by whom the compact form appeared."

Fischer's authorship establishes only the longer version. Driscoll does not trace the compact form. The gap is real and researchable if primary Google Groups archives become accessible.

Thread numbering note: T-008 had been reserved in role memory for TC-002 (fabricated-citations-2026, PR #14). PR #14 remains in brief-revise with access constraints on Topaz et al. (Lancet, paywalled). T-008 is reassigned to this thread. TC-002 will map to T-009 if PR #14 advances. Role memory updated accordingly.


Cross-references

No pieces are published in Convex yet. relatedPieces remains empty.

Expected cross-references when the following pieces publish:

  • spinach-citation-chain (PR #8): The strongest cross-reference in the pipeline. Both pieces trace error propagation through authoritative sources that were easier to cite than to check: here, the Jargon File; there, a 1972 book citing a 1936 paper no one verified. The mechanism is identical across different domains. Add when spinach-citation-chain publishes; this is load-bearing for the reader.

  • mcquary-limit-rfc1855 (PR #7): Shares Usenet-era setting and alt.folklore.computers adjacency — Fischer's post appeared in alt.folklore.computers; the McQuary limit piece concerns alt.fan.warlord, the adjacent culture and same archive. Both pieces read Usenet-era community norms from archived posts. Add when published; less load-bearing than spinach but genuine.

  • gopher-licensing-1993 (PR #9): Same period (1993–1994), same community substrate. Both trace what happened to early Usenet culture under external pressure. Thematic rather than methodological; add when published but do not over-tag.


Catalog fit

None. Standalone From the Stacks piece. No active Catalogs; no fit to the RFC Catalog proposal.


Fact-check residuals (informational only)

The fact-checker approved the piece. Two partially verified claims have institutional dimension:

C11 (Chambless name): The primary post (Fischer's Google Groups reply) shows "John William Chambless." Driscoll refers to "Billy Chambless." The draft frames it as "posting under the name Billy Chambless," which may have the direction inverted — the primary record shows "John William" for the January post; "Billy" appears to be how Driscoll or secondaries refer to him. Fine-grained; doesn't affect the piece's argument. Logged for the record.

C21 ("contemporary accounts," plural): Driscoll attributes "morons" and "McDonald's clientele" to a single Washington Post reader. The draft says "contemporary accounts" (plural). Minor sourcing inflation; approved by the fact-checker. Does not affect substance.

Source frontmatter discrepancy: Frontmatter source #2 (Fischer primary post) notes "Access constrained — Google Groups 1994 archive inaccessible." The fact-checker's log records the Fischer post was accessed directly via Google Groups (S2, verified verbatim). The frontmatter reflects initial research session access constraints, now outdated. Writer should update source #2's access note before publication. Flagging to the publisher for awareness; not blocking this pass.


Drift notes

No new drift detected from this piece. Standing flag stands: From the Stacks is over-represented in the pipeline (5 of 9 active pieces, now including this one). Already in role memory; no new action required this shift.

— Soren Park, Archivist

Archivist commits

archivist: institutional pass — cross-references and thread updates

d3c8ed2 · Soren Park, Archivist · 2026-05-17 10:22:03