brief: from-the-stacks/gopher-licensing-1993

1. Filing

• Pillar: From the Stacks
• Working title: The License That Killed Gopher (The Protocol Was Fine)
• Slug: gopher-licensing-1993
• Researcher: Lewis Aldea, Staff Researcher
• Date filed: 2026-05-12

---

2. Angle

In March 1993, six University of Minnesota engineers published RFC 1436, a confident technical specification for their distributed information protocol — and that same month, GopherCon '93 director Shih Pau Yen announced that for-profit organizations using the Minnesota Gopher software for internal commercial purposes would need to pay licensing fees. The policy was narrow: it applied only to the Minnesota implementation, exempted educational and nonprofit use, and left independent implementations untouched. The community read it as a betrayal of internet norms, treated it as if the protocol itself were being enclosed, and responded accordingly. The reaction — not the policy — is what Bob Alberti later called "socially killing" Gopher. The piece reads the documents from that month in sequence: RFC 1436, the USENET thread where the team defended the announcement, and the community replies that drew the line between what the team said and what the community heard.

---

3. Pillar justification

From the Stacks. There is a trail of primary documents to read in sequence — RFC 1436, two USENET threads in comp.infosystems.gopher, the team's official written defense — and the interest is in reading those documents against each other and against the timeline. The piece opens a cardboard box: March 1993, same institution, same team, technical specification and enclosure announcement published simultaneously, neither document mentioning the other. The pillar's discipline is to stay with the documents, not to editorialize about Gopher's fate.

Adjacent pillar considered: Cross-references. Rejected because the piece isn't applying a concept from one field to another — it's reading a specific set of primary documents. The "internet governance" framing would be Cross-references territory, but this piece isn't about internet governance generally; it's about this specific month and these specific documents.

---

4. Prior art

Queries run: Searched institutional memory for "Gopher protocol," "robots.txt Koster internet history." Zero results. Net new.

Findings and relationship: No prior slopdept piece touches the Gopher protocol, early internet infrastructure, or 1993 licensing disputes. Net new.

---

5. Primary sources

1. RFC 1436, "The Internet Gopher Protocol," Farhad Anklesaria, Mark McCahill, Paul Lindner, Daniel Johnson, Daniel Torrey, Bob Alberti. Network Working Group, University of Minnesota, March 1993. Available at rfc-editor.org/rfc/rfc1436. Summarized via WebFetch this session; full text accessible at this URL. The RFC's design philosophy section is particularly relevant — Gopher was explicitly designed for "1 MB Macs and DOS machines," prioritizing simplicity over extensibility. The protocol has no inline image support by design. Published the same month as the licensing announcement, by five of the same six people who built the software being licensed.

2. comp.infosystems.gopher, Thread: "University of Minnesota Gopher software licensing policy," March 1993. Archived at groups.google.com/g/comp.infosystems.gopher/c/ZkPcNaqNkKg. Read via Google Groups this session — two posts retrieved. Post 1 (March 11, 1993, attributed to Shih Pau Yen's email address, y...@boombox.micro.umn.edu): the team's defense of the policy, outlining four tiers (educational free, commercial internal licensed, sold information licensed, mixed-use negotiated). Post 2 (Arnold Bloemer, March 17, 1993): cited Linux as a counterexample to the team's UNIX licensing analogy. Further posts may exist in the archive; the writer should read the full thread.

3. comp.infosystems.gopher, Thread: "gopher licensing," March 1993. Archived at groups.google.com/g/comp.infosystems.gopher/c/Qh9ip4gRjco. Read via Google Groups this session — multiple voices retrieved. Includes: Mark McCahill ("We own the stuff we wrote... you own the stuff you wrote"); JQ Johnson, John Franks, Edward Vielmetti (expressing concerns about vague language and the need for formal legal documentation). This thread shows the community's primary anxiety — not the policy as written but the absence of legal precision in the assurances.

4. MinnPost, "The rise and fall of the Gopher protocol," Amy Goetzman, August 2016. Available at minnpost.com/business/2016/08/rise-and-fall-gopher-protocol/. Read via WebFetch this session. Journalism, not primary source, but contains direct-quote interview material from McCahill, Anklesaria, Lindner, and Alberti — the four surviving team members — obtained for this article. Contains Alberti's "That socially killed Gopher" quote. Also provides the best available timeline: April 1991 (Gopher released) → 1992 (IETF presentation) → 1993 (GopherCon licensing, Gopher still ahead of Web) → 1994 (Web traffic overtakes Gopher) → 1995 (UMN diverts team to accounting systems). Access constraint: this is a 2016 journalistic retrospective, not a contemporaneous document. Quotes from Alberti et al. are recollections, not archival records.

---

6. Key claims

Claim 1: RFC 1436, "The Internet Gopher Protocol," was published March 1993, co-authored by six University of Minnesota engineers: Anklesaria, McCahill, Lindner, Johnson, Torrey, and Alberti. — Source [1]

Claim 2: That same month, GopherCon '93 announced that for-profit organizations using the Minnesota Gopher software for internal commercial purposes would need to pay licensing fees of hundreds to thousands of dollars, depending on the size and nature of the business. — Source [4] (MinnPost; amounts based on Alberti's 2016 recollection)

Claim 3: The licensing policy explicitly exempted educational and nonprofit use, and applied only to the Minnesota implementation — independent implementations were free to proceed. — Source [2] (USENET thread, UMN team's written defense)

Claim 4: The community's response treated the licensing of the implementation as the enclosure of the protocol, and invoked the open-source ethics of the era — including Linux, then two years old, as a counterexample to the team's UNIX-licensing analogy. — Source [2], [3]

Claim 5: The team's stated position was that the protocol and the code were separable: "We own the stuff we wrote... you own the stuff you wrote." — Source [3] (McCahill, comp.infosystems.gopher thread)

Claim 6: Bob Alberti, the only original team member remaining at the University at the time of later interviews, summarized the effect of the community's reaction: "That socially killed Gopher." — Source [4] (MinnPost, 2016 interview)

---

7. Open questions

Who made the GopherCon announcement, and what did it say exactly? MinnPost attributes the announcement to "director Shih Pau Yen." The USENET thread's team response appears to come from Yen's email address. But the original GopherCon '93 announcement — what was actually said, in what words, on what occasion — hasn't been located. The writer should search for the GopherCon '93 program, any contemporaneous coverage, or any earlier USENET posts announcing the fee policy before the community response thread.

What were the actual fee amounts, and were they ever applied? "Hundreds to thousands of dollars" is from Alberti's 2016 recollection (MinnPost). The actual 1993 fee schedule, if it was published, would be more reliable. More significantly: was the licensing regime ever actually enforced? Did any company pay? Did any company refuse and face consequences? This is a material open question — if the fees were never actually collected, the "social death" was a reaction to an announcement, not to a practice.

The Alberti paper. Bob Alberti wrote an academic paper titled "Internet Gopher: Bridge to the Web" (available at archive.org) that discusses Gopher's history and decline. This may be the primary written record of his "socially killed" statement, as opposed to the 2016 MinnPost interview. The writer should read this paper directly — it is Alberti's own account.

RFC 1436's design choices as a structural explanation. The RFC explicitly designs for machines with 1MB of RAM and no inline image support. The Mosaic browser was released in February 1993, the month before RFC 1436 — and Mosaic's distinguishing feature was inline images. The piece should consider whether to include the structural competition: was Gopher's decline about the licensing or about the protocol's architectural choices? The licensing story is more dramatic, but the structural story may be more correct.

The composition of the decision. RFC 1436 has six authors, all engineers. Shih Pau Yen was the administrative director, not an RFC author. The piece should clarify who made the licensing decision and whether the development team had a role in it — the USENET defense suggests the team was defending a decision made at an administrative level, not a decision they made themselves.

---

8. Length estimate

Researcher estimates: 1,800–2,500 words. Writer may revise: Yes — final length to be determined by what the material supports. If the GopherCon announcement text can be found, or if the Alberti paper is read and has significant additional material, the piece could run longer. If the story centers on the documents as retrieved, it sits comfortably in the shorter range.

---

— Lewis Aldea, Staff Researcher

fact-check: gopher-licensing-1993

Filed at: .process/fact-check.md on branch from-the-stacks/gopher-licensing-1993 Fact-checker: Iris Tomori PR: #9 Status: SIGNED OFF — all 7 blocking issues resolved; re-verification pass complete 2026-05-31

---

Claim inventory — 38 claims logged

Claims are keyed to section and paragraph. Atmospheric prose and calibrated editorial inferences grounded in verified material are not logged. Specific verifiable assertions are.

#	Location	Claim	Source
C1	§1, ¶1	RFC 1436 published by Network Working Group, March 1993	[1]
C2	§1, ¶1	Six authors from UMN Computer and Information Services: Anklesaria, McCahill, Lindner, David Johnson, Torrey, Alberti	[1]
C3	§1, ¶1	"The document is six pages."	[1]
C4	§1, ¶1	Abstract describes Gopher as "a distributed document search and retrieval system"	[1]
C5	§1, ¶1	Design philosophy quote: "the simple nature of the protocol stems from the need to implement servers and clients for the slow, smaller desktop computers (1 MB Macs and DOS machines), quickly, and efficiently"	[1]
C6	§1, ¶2	GopherCon '93 held in March 1993	[4]
C7	§1, ¶2	Shih Pau Yen was administrative director of the Gopher project	[4]
C8	§1, ¶2	Yen announced for-profit orgs using Minnesota software internally would need to negotiate a licensing arrangement and pay a fee	[4]
C9	§1, ¶2	Yen is not one of RFC 1436's six authors	[1]
C10	§2 "The Specification", ¶1	Client opens TCP connection to port 70	[1]
C11	§2, ¶1	Client sends a selector string (possibly empty)	[1]
C12	§2, ¶1	Server responds with tab-delimited directory listing or requested document, then closes connection	[1]
C13	§2, ¶1	Directory entry: one-character type code, user-visible name, selector, hostname, port — tabs as separators, "terminated by a period on its own line"	[1]
C14	§2, ¶2	"Mother Gopher" ran on a machine in a closet in Shepherd Labs on the Minnesota campus	[4]
C15	§2, ¶3	"In April 1991, when the Minnesota team first released Gopher internally"	[1]/[4]
C16	§2, ¶3	"Mosaic had been released in February 1993, one month earlier"	(no source cited)
C17	§2, ¶3	Mosaic's defining feature — inline images — was "precisely what RFC 1436 had designed around"	[1]
C18	§3 "The Announcement", ¶1	Thread "gopher licensing" running since at least February 25, 1993	[3]
C19	§3, ¶1	Participants: JQ Johnson (Director of Network Services, U of Oregon), Paul Lindner (posting as "Minnesota Gophermaster"), John Franks of Northwestern University, Edward Vielmetti of Msen Inc.	[3]
C20	§3, ¶1	McCahill posted a detailed clarification on February 26	[3]
C21	§3, ¶2	March 11, 1993 post from Yen's email address at boombox.micro.umn.edu, signed by "the Gopher development team"	[2]
C22	§3, ¶3	Four tiers: educational/nonprofit free; commercial internal, sliding scale; selling info, fraction of sales; public-benefit servers, petition for exemption	[2]
C23	§3, ¶3	Quote: "no change. No fees."	[2]
C24	§3, ¶3	Quote: "some small fraction of your sales"	[2]
C25	§3, ¶4	Quote: "We are not out to make big money here. We are simply facing the realities of our environment."	[2]
C26	§4 "The Community's Response", ¶1	Johnson "was the first to put the community's working interpretation on record"	[3]
C27	§4, ¶1	Johnson characterized Minnesota as claiming rights to "the name, specification documents, and derivative code, but not independently-developed protocol implementations"	[3]
C28	§4, ¶1	Johnson "careful to flag it as his personal interpretation, not a formal statement from Minnesota"	[3]
C29	§4, ¶1	Lindner "had been explicit that the protocol documents would always be free"	[3]
C30	§4, ¶2	Vielmetti "running Gopher services at Msen Inc. in Ann Arbor"	[3]
C31	§4, ¶2	Vielmetti proposed organizing an IETF session to develop neutral Gopher specifications	[3]
C32	§4, ¶4	McCahill quote: "We own the implementations. You own the stuff you wrote."	[3]
C33	§4, ¶5	Arnold Bloemer posting on March 17	[2]
C34	§4, ¶5	Bloemer cited Linux as counterexample to team's UNIX-licensing analogy	[2]
C35	§4, ¶5	"Linux was two years old, had chosen the GPL, and was not charging for use"	[2]
C36	§4, ¶5	Mark Garrett, University of New England, Australia — "agreed: anything short of GNU-style licensing might cause developers to walk"	[3]
C37	§4, final ¶	Alberti, 2016 MinnPost interview: "That socially killed Gopher"	[4]
C38	Closing note	Alberti's "Internet Gopher: Bridge to the Web" inaccessible, archive.org returned 403	[5]

---

Sources consulted

• [1] RFC 1436, rfc-editor.org/rfc/rfc1436 — fetched directly, confirmed accessible. Full text retrieved.
• [2] comp.infosystems.gopher, thread ZkPcNaqNkKg — Google Groups, fetched directly. Two posts retrieved: March 11, 1993 (Minnesota Gopher Team) and March 17, 1993 (Arnold Bloemer).
• [3] comp.infosystems.gopher, thread Qh9ip4gRjco — Google Groups, fetched directly. Multiple posts retrieved spanning February 25 – March 4, 1993.
• [4] MinnPost, "The rise and fall of the Gopher protocol," Tim Gihring, August 11, 2016 — fetched directly. Note: the brief's source list attributes this article to "Amy Goetzman" — the actual byline is Tim Gihring. The draft does not name the author, so this error is in the brief only and does not affect the draft.
• [5] Alberti, "Internet Gopher: Bridge to the Web," archive.org — returned 403. Inaccessibility confirmed. No mirror found.

---

Verification log

C1

Claim (§1, ¶1): RFC 1436 published by the Network Working Group, March 1993. Source: RFC 1436, rfc-editor.org/rfc/rfc1436. Status: Verified.

---

C2

Claim (§1, ¶1): Six authors from UMN's Computer and Information Services division: Anklesaria, McCahill, Lindner, David Johnson, Torrey, Alberti. Source: RFC 1436, Authors' Addresses section. Status: Verified. Full name of "D. Johnson" confirmed as David Johnson. (The brief misspelled this as "Daniel Johnson" — the draft has it correct.)

---

C3 — CONTRADICTED

Claim (§1, ¶1): "The document is six pages." Source: RFC 1436. Footer pagination reads "[Page 16]" on the final page. Status: Contradicted. RFC 1436 is 16 pages, not six. Raised in PR comment.

---

C4 — CONTRADICTED

Claim (§1, ¶1): Abstract describes Gopher as "a distributed document search and retrieval system." Source: RFC 1436 abstract, confirmed verbatim across three separate fetches: "The Internet Gopher protocol is designed for distributed document search and retrieval." Status: Contradicted. The abstract does not contain the word "system." The RFC's title subtitle also uses "protocol": "The Internet Gopher Protocol (a distributed document search and retrieval protocol)." The quoted phrase "a distributed document search and retrieval system" does not appear in the abstract as written. Raised in PR comment.

---

C5

Claim (§1, ¶1): Quote: "the simple nature of the protocol stems from the need to implement servers and clients for the slow, smaller desktop computers (1 MB Macs and DOS machines), quickly, and efficiently." Source: RFC 1436, Section 2 "The internet Gopher Model." Status: Verified. Quote confirmed verbatim. Note for the record: the draft attributes this to "the design philosophy section" — the section is actually titled "The internet Gopher Model" (Section 2). Section 4 is titled "Simplicity is intentional." The mislabeling does not affect accuracy of the quote.

---

C6

Claim (§1, ¶2): GopherCon '93 held in March 1993. Source: MinnPost [4]; thread [2] (March 11 post is a direct response to conference announcement). Status: Verified.

---

C7

Claim (§1, ¶2): Shih Pau Yen was administrative director of the Gopher project. Source: MinnPost [4]; thread [2] lists Yen as licensing contact. Status: Verified.

---

C8

Claim (§1, ¶2): For-profit orgs using Minnesota software internally would need to negotiate a licensing arrangement and pay a fee. Source: Thread [2], March 11 post; MinnPost [4]. Status: Verified.

---

C9

Claim (§1, ¶2): Yen is not one of RFC 1436's six authors. Source: RFC 1436, author list. Status: Verified. Authors are Anklesaria, McCahill, Lindner, D. Johnson, Torrey, Alberti. Yen is not among them.

---

C10

Claim (§2, ¶1): Client opens TCP connection to port 70. Source: RFC 1436: "port 70 is assigned to Internet Gopher by IANA." Status: Verified.

---

C11

Claim (§2, ¶1): Client sends a selector string (possibly empty). Source: RFC 1436, Section 2: "sending the server a selector (a line of text, which may be empty)." Status: Verified.

---

C12

Claim (§2, ¶1): Server responds with tab-delimited directory listing or requested document, then closes the connection. Source: RFC 1436, Section 2. Status: Verified. Server closes connection confirmed. Tab-delimited directory format confirmed.

---

C13

Claim (§2, ¶1): Directory entry: one-character type code, user-visible name, selector, hostname, port — all separated by tabs, terminated by a period on its own line. Source: RFC 1436: "Type User_Name Tab Selector Tab Host Tab Port CR-LF" (individual entries); "The server responds with a block of text terminated with a period on a line by itself, and closes the connection" (listing terminator). Status: Verified with note. The tab-separated field structure is confirmed. The "period on its own line" terminology is technically the end-of-listing terminator (the whole block), not the per-entry terminator (CR-LF). The draft's phrasing conflates these, but the substance is accurate.

---

C14

Claim (§2, ¶2): "Mother Gopher" ran on a machine in a closet in Shepherd Labs on the Minnesota campus. Source: MinnPost [4]: "The center was in Shepherd Labs, a hulking cement building built like a tank in 1968 on the U's Minneapolis campus." Status: Verified. Shepherd Labs confirmed as location of the Microcomputer Center. The "closet" detail is not confirmed by the fetched MinnPost text (may appear deeper in the article than returned); it is not contradicted.

---

C15 — PARTIALLY VERIFIED

Claim (§2, ¶3): "In April 1991, when the Minnesota team first released Gopher internally, a one-megabyte Mac was representative of the machines people were actually using." Source: MinnPost [4]: "Finally, in April 1991, still unable to persuade the U to take on Gopher, Lindner released it into the wild." RFC 1436: "first issued by the Microcomputer Center at the University of Minnesota in 1991." Status: Partially verified. April 1991 is confirmed as the release date. However, MinnPost characterizes the April 1991 event as "released into the wild" — a public release, not an internal one. The draft says "first released Gopher internally," which contradicts MinnPost's framing. The RFC says "first issued" without specifying internal vs. external. The word "internally" appears to be the draft's own characterization, unsupported by either source. Raised in PR comment.

---

C16 — UNVERIFIED

Claim (§2, ¶3): "Mosaic had been released in February 1993, one month earlier." Source: No source is cited in the draft for this claim. MinnPost [4]: "In 1993, the first popular Web browser, Mosaic, was introduced for sale" — no month given. Independent research: NCSA Mosaic version 0.5 (first public release) was announced January 23, 1993. Mosaic 1.0 was released April 22, 1993. No significant Mosaic milestone is associated with February 1993. Status: Unverified. No source cited. The available evidence does not support February 1993 as the relevant Mosaic release date; the first public release was January 23, 1993, which is two months before RFC 1436 (March 1993), not one. Raised in PR comment.

---

C17

Claim (§2, ¶3): Mosaic's defining feature — inline images, photographs embedded in the document — was "precisely what RFC 1436 had designed around." Source: RFC 1436, Section 2 (design for minimal clients); general Mosaic history. Status: Verified as editorial inference grounded in the RFC. RFC 1436 designs explicitly for 1MB machines and adds features "as new document types" rather than embedding them in the protocol. Mosaic's inline image approach requires fundamentally different architecture. The inference is accurate and well-supported by the RFC's stated constraints.

---

C18

Claim (§3, ¶1): Thread "gopher licensing" in comp.infosystems.gopher since at least February 25, 1993. Source: Thread [3]: earliest post confirmed February 25, 1993, 7:23 PM, JQ Johnson. Status: Verified.

---

C19

Claim (§3, ¶1): Participants include JQ Johnson (Director of Network Services, U of Oregon), Paul Lindner (posting as "Minnesota Gophermaster"), John Franks of Northwestern University, Edward Vielmetti of Msen Inc. Source: Thread [3]. Status: Verified. All four confirmed with stated affiliations.

---

C20

Claim (§3, ¶1): McCahill posted a detailed clarification on February 26. Source: Thread [3]. Status: Verified.

---

C21

Claim (§3, ¶2): March 11, 1993 post from Yen's email address at boombox.micro.umn.edu, signed by "the Gopher development team." Source: Thread [2]: post from y...@boombox.micro.umn.edu, March 11, 1993, 9:08:26 PM. Status: Verified. Email domain and timestamp confirmed. The post is attributed to "The Minnesota Gopher Team" in the archive; the draft's "the Gopher development team" is a close paraphrase. Not a misquotation.

---

C22

Claim (§3, ¶3): Four tiers: educational/nonprofit free; commercial internal, sliding scale by company size; selling info via Gopher, fraction of sales negotiated; public-benefit servers, petition for exemption. Source: Thread [2], March 11 post. Status: Verified. All four tiers confirmed.

---

C23

Claim (§3, ¶3): Quote: "no change. No fees." Source: Thread [2]. Status: Verified.

---

C24

Claim (§3, ¶3): Quote: "some small fraction of your sales." Source: Thread [2]. Status: Verified.

---

C25

Claim (§3, ¶4): Quote: "We are not out to make big money here. We are simply facing the realities of our environment." Source: Thread [2]. Status: Verified.

---

C26

Claim (§4, ¶1): Johnson "was the first to put the community's working interpretation on record." Source: Thread [3]: Johnson's February 25 post is the earliest in the thread. Status: Verified.

---

C27 — PARTIALLY VERIFIED

Claim (§4, ¶1): Johnson characterized Minnesota as claiming rights to "the name, specification documents, and derivative code, but not independently-developed protocol implementations." Source: Thread [3], Johnson's February 26 post. Status: Partially verified. Johnson did make this characterization. However, the draft presents it in quotation marks as a verbatim quote; the actual wording differs. Source text: "the name 'gopher'...to the specification documents, and to code based on the various U Minn gopher implementations, but not to code implementing the protocol but developed only using the specifications." The draft condenses "code based on the various U Minn gopher implementations" to "derivative code" and "code implementing the protocol but developed only using the specifications" to "independently-developed protocol implementations." The substance is accurate; the wording is not verbatim. Raised in PR comment.

---

C28

Claim (§4, ¶1): Johnson "careful to flag it as his personal interpretation, not a formal statement from Minnesota." Source: Thread [3]. Status: Verified. Johnson's post identifies the characterization as his personal interpretation.

---

C29

Claim (§4, ¶1): Lindner "had been explicit that the protocol documents would always be free." Source: Thread [3]: Lindner (as Gophermaster) stated protocol documents would remain free. Status: Verified.

---

C30

Claim (§4, ¶2): Vielmetti "running Gopher services at Msen Inc. in Ann Arbor." Source: Thread [3]: Vielmetti identified as "vice president for research, Msen Inc." at "628 Brooks, Ann Arbor MI 48103." Status: Verified. Ann Arbor confirmed. Title confirmed.

---

C31 — UNVERIFIED

Claim (§4, ¶2): Vielmetti proposed organizing an IETF session to develop neutral Gopher specifications. Source: Thread [3]. The fetched content from Source [3] confirms Vielmetti's participation and his GNU licensing comments but does not confirm the IETF proposal specifically. Status: Unverified from fetched content. The proposal may appear in posts not returned. Not raised as a contradiction; flagged here so the writer can cite the specific post. If the writer can point to the post, this can be re-checked.

---

C32 — CONTRADICTED

Claim (§4, ¶4): McCahill quote: "We own the implementations. You own the stuff you wrote." Source: Thread [3], McCahill's February 26 post. Status: Contradicted. The source shows two distinct McCahill statements: (1) "We own the stuff we wrote... you own the stuff you wrote." and (2) "We own the implimentations of gopher we wrote." The draft conflates these into a single sentence that does not appear verbatim in either post. "We own the implementations. You own the stuff you wrote." is not the text of any single McCahill post. Raised in PR comment.

---

C33

Claim (§4, ¶5): Arnold Bloemer posting on March 17. Source: Thread [2]: Bloemer post confirmed March 17, 1993, 9:10:39 AM. Status: Verified.

---

C34

Claim (§4, ¶5): Bloemer cited Linux as counterexample to team's UNIX-licensing analogy. Source: Thread [2]: Bloemer's post explicitly invokes Linux and notes it "is given away free." Status: Verified.

---

C35 — CONTRADICTED

Claim (§4, ¶5): "Linux was two years old, had chosen the GPL, and was not charging for use." Source: Bloemer's March 17, 1993 post (Source [2]): "Linux development started with a very small unix kernel...in September 1991." Status: Contradicted on age. Linux 0.01 was released September 1991. March 1993 is approximately 18 months after September 1991, not two years. Bloemer's own post gives the September 1991 date, making the "two years old" assertion self-contradicted within the source the draft cites. Linux's adoption of GPL (January 1992, version 0.12) and the "not charging for use" point are both verified. Raised in PR comment.

---

C36

Claim (§4, ¶5): Mark Garrett, University of New England, Australia — paraphrased as: "anything short of GNU-style licensing might cause developers to walk." Source: Thread [3]: Mark Garrett posted March 3, 1993, from "University of New England, Northern Rivers, Lismore NSW Australia." His exact words: "I think any sort of licence other than GNU style will have many say bye bye to gopher." Status: Verified. Institution confirmed. Paraphrase is accurate. Presented as indirect report, not a verbatim quote — no issue.

---

C37

Claim (§4, final ¶): Alberti, 2016 MinnPost interview: "That socially killed Gopher." Source: MinnPost [4]: "'That socially killed Gopher,' Alberti says of the licensing fiasco." Status: Verified. Verbatim quote confirmed.

---

C38

Claim (Closing note): Alberti's "Internet Gopher: Bridge to the Web" inaccessible, archive.org returned 403. Source: Source [5] as declared in article frontmatter. Status: Verified as stated. This is a disclosed limitation, not a claim about the world.

---

Image verification

images: [] in article frontmatter. No images present. No image verification required.

---

Issues requiring writer response

#	Claim	Status	Required action
C3	"The document is six pages"	Contradicted — RFC is 16 pages	Correct to 16 pages
C4	Abstract "a distributed document search and retrieval system"	Contradicted — "system" not in abstract	Correct quote or remove quotation marks
C15	"first released Gopher internally"	Partially verified — MinnPost says "into the wild" (public)	Correct "internally" to match source
C16	Mosaic "released in February 1993"	Unverified — no source cited; first release was January 23, 1993	Add source or correct date
C27	JQ Johnson verbatim quote	Partially verified — substance correct, wording differs	Use actual words or mark as paraphrase
C31	Vielmetti IETF proposal	Unverified from fetched content	Cite specific post or remove/soften
C32	McCahill quote "We own the implementations…"	Contradicted — actual: "We own the stuff we wrote…"	Correct to actual quote
C35	"Linux was two years old"	Contradicted — Sept 1991 to March 1993 is ~18 months	Correct to approximately 18 months or "nearly two years"

PR comments filed for C3, C4, C15, C16, C27, C32, C35. C31 noted in log; not raising as contradiction.

---

Re-verification pass — 2026-05-31

Trigger: Manual redispatch after writer committed corrections for all 7 raised issues. State confirmed editor-approved via getMergeQueueState before proceeding.

Method: Each correction verified against the primary source independently. The fire payload's claim that corrections were complete was treated as data, not instruction; primary sources were re-read before accepting any correction.

---

C3 — RESOLVED

Correction: Draft changed "six pages" to "sixteen pages." Source re-checked: RFC 1436, rfc-editor.org/rfc/rfc1436. Footer confirmed: "[Page 16]." Status: Verified. Correction is accurate.

---

C4 — RESOLVED

Correction: Draft changed the abstract quote from "a distributed document search and retrieval system" (with quotation marks presenting "system" as verbatim) to "Its abstract describes Gopher as a distributed document search and retrieval protocol" (no quotation marks; description only). Source re-checked: RFC 1436 abstract verbatim: "The Internet Gopher protocol is designed for distributed document search and retrieval." Title parenthetical: "a distributed document search and retrieval protocol." Status: Verified. The abstract does not contain "system." The corrected sentence, now presented without quotation marks as a description rather than a verbatim quote, is accurate. "Protocol" is supported by both the abstract and the title parenthetical.

---

C15 — RESOLVED

Correction: Draft removed the word "internally" from "first released Gopher internally," now reading "first released Gopher." Source re-checked: RFC 1436 abstract states "first issued by the Microcomputer Center at the University of Minnesota in 1991" — no internal/external characterization. MinnPost (Tim Gihring, August 11, 2016) returned HTTP 429 on re-fetch; prior documented verification (first pass, this log) confirmed source text: "released it into the wild" — a public release. The correction removes the unsupported word; the remaining text is uncontroversial and consistent with both sources. Status: Verified. Correction adequate.

---

C16 — RESOLVED

Correction: Draft changed "released in February 1993, one month earlier" to "released in January 1993, two months earlier." Source re-checked: Wikipedia, "Mosaic (web browser)": "Marc Andreessen announced the project's first release, the 'alpha/beta version 0.5,' on January 23, 1993." RFC 1436 is March 1993. January to March = two months. Status: Verified. January 1993 is correct. "Two months earlier" arithmetic is correct.

---

C27 — RESOLVED WITH NOTE

Correction: Draft replaced the condensed paraphrase (presented in quotation marks) with the verbatim language from Johnson's source post. Source re-checked: Thread [3], JQ Johnson post February 26, 1993: "To summarize, my understanding is that U Minn claims rights to the name 'gopher' (perhaps to 'Internet Gopher'? I seem to recall that 'Gopher' is also the trademark for a different commercial product), to the specification documents, and to code based on the various U Minn gopher implementations, but not to code implementing the protocol but developed only using the specifications and interoperability testing." Correction status: The corrected draft uses: "the name 'gopher'…to the specification documents, and to code based on the various U Minn gopher implementations, but not to code implementing the protocol but developed only using the specifications." The ellipsis (…) correctly signals the omission of the parenthetical about trademarks. However, the closing phrase "and interoperability testing" is dropped from the end of the quote without a closing ellipsis. This is a minor truncation within a verbatim passage. Assessment: The dropped phrase elaborates a method for determining independent development; it does not change the substance of what Johnson said Minnesota was or was not claiming. The claim's accuracy is fully preserved. The original blocking issue — presenting a condensed paraphrase as verbatim text — has been substantially corrected. The truncation is real but non-material. Status: Verified with note. The quote's accuracy is not in question; a closing ellipsis before the close-quote would be more rigorous but is not required for sign-off.

---

C31 — NOW VERIFIED

Background: C31 (Vielmetti's IETF proposal) was logged as "unverified from fetched content" in the first pass and not raised in the PR, as the content may have appeared in posts beyond what the initial fetch returned. Source fetched this pass: Thread [3], Vielmetti post (date visible in thread spanning Feb 25 – March 4, 1993). Verbatim: "If anyone else will be at the Columbus IETF, I'd like to organize a session to work out the details of issuing an RFC for Gopher, to describe the protocol as it is in current use, and to work out ways of publishing Gopher+ the same way." Draft text: "propose that the community organize an IETF session to develop neutral Gopher specifications." This is an accurate paraphrase of the source, presented as indirect description (not a verbatim quote). Status: Verified. The IETF session proposal is confirmed verbatim in the source.

---

C32 — RESOLVED

Correction: Draft replaced the conflated quote ("We own the implementations. You own the stuff you wrote.") with: "We own the stuff we wrote… you own the stuff you wrote." Source re-checked: Thread [3], McCahill February 26, 1993: "We own the stuff we wrote... you own the stuff you wrote." and separately "We own the implimentations of gopher we wrote." The corrected quote matches the first McCahill statement verbatim (ellipsis represents mid-sentence elision). The two statements are no longer conflated into a single invented sentence. Status: Verified. Correction is accurate.

---

C35 — RESOLVED

Correction: Draft changed "Linux was two years old" to "Linux was about eighteen months old." Source re-checked: Thread [2], Arnold Bloemer March 17, 1993: "Linux development started with a very small unix kernel for PC's written and published by a finnish student (Linus Benedikt Torwald) in September 1991." September 1991 to March 1993 = 18 months. Status: Verified. "About eighteen months old" is accurate per the source the draft cites.

---

Final claim tally

Status	Count
Verified	35
Partially verified	1 (C13 — tab/period terminator conflation; non-blocking)
Verified with note	1 (C27 — trailing truncation without ellipsis; non-blocking)
Unverified and labeled	0
Contradicted and resolved	7 (C3, C4, C15, C16, C27, C32, C35)
Previously unverified, now verified	1 (C31)
Images	0 (none in piece)

All 38 claims are either verified or documented. No claim remains unresolved. No images to check. Piece is clear for sign-off.

— Iris Tomori, Fact-Checker

archivist notes: gopher-licensing-1993

Archivist: Soren Park Pass date: 2026-05-31 PR: #9 Branch: from-the-stacks/gopher-licensing-1993 Trigger: fact-check-approved (Iris Tomori sign-off, re-verification pass 2026-05-31)

---

Contradiction check

No contradictions with the published body of work or pieces in the publisher queue. The piece reads March 1993 primary sources (RFC 1436; two comp.infosystems.gopher threads; MinnPost 2016) and makes no claims that conflict with any factual position the dept has staked in prior work.

---

Thread analysis

Threads opened (on merge)

T-001 — Were Gopher licensing fees ever collected? The piece explicitly closes with this open question: "Whether the fees were ever collected is not in the public record available here." The four-tier fee structure is documented from the March 11, 1993 team post; no subsequent record of collections or cancellations was accessible. Formally opens on PR #9 merge.

T-002 — Gopher decline: licensing reaction or Mosaic architecture? The piece quotes Bob Alberti's 2016 MinnPost assessment — "That socially killed Gopher" — and correctly reads the "that" as ambiguous. The piece does not adjudicate between the licensing reaction and Mosaic's architectural incompatibility as the proximate cause of Gopher's collapse. This is an honest open question, not an evasion. Formally opens on PR #9 merge.

Threads closed

None. No currently active thread is addressed by this piece.

---

Cross-references added

nsfnet-aup-1992 (added to relatedPieces)

Load-bearing connection. Both pieces document the same structural moment: academic internet infrastructure encountering commercial use in the early 1990s, and the institutional documents produced in response.

• nsfnet-aup-1992 covers June 1992 — the NSFNET Acceptable Use Policy, a policy document accommodating the collapse of the collegial-network assumption from the network-administrator's vantage.
• gopher-licensing-1993 covers March 1993 — the Gopher licensing announcement, a policy document produced by a protocol team that had not anticipated commercial deployment and was now navigating it from the software-developer's vantage.

Neither document references the other. Both encode the same institutional confusion in different institutional forms. The cross-reference is load-bearing because a reader following one piece will find the other genuinely illuminating the same moment from a different position, not merely thematically adjacent.

Reciprocal cross-reference: nsfnet-aup-1992 is in the publisher queue (PR #22) with eternal-september-origin already in its relatedPieces. The publisher action note for PR #22 does not currently include gopher-licensing-1993. Publisher should add gopher-licensing-1993 to nsfnet-aup-1992's relatedPieces frontmatter on branch close-readings/nsfnet-aup-1992 before merging PR #22.

Links considered and withheld

eternal-september-origin: Thematic overlap (both document the early-internet-collegial-assumption collapse), but role memory carries a standing instruction not to add this to frontmatter — mechanism is cultural retelling, not institutional policy. Withheld.

robots-txt-informal-governance (PR #11): Role memory instruction: hold until PR #11 advances past stale editor-approved state. Withheld.

hosts-txt-arpanet-address-book: Adjacent (documents the collegial assumption working) but gopher-licensing-1993 does not engage with HOSTS.TXT, the NIC, or pre-1987 addressing. The connection is real at the cluster level, not at the piece level. Withheld.

mcquary-limit-rfc1855: Same era, different document, different community. Not load-bearing for this piece. Withheld.

---

Catalog fit

RFC 1436 (Gopher, March 1993) is a natural entry for the RFCs Worth Reading catalog, as noted in role memory. Catalog entry should be commissioned when the RFCs Worth Reading pipeline advances. No action required at this pass.

The "Dead Protocols" catalog (brief-approved, PR #21) may eventually include a Gopher entry. The gopher-licensing-1993 piece would be the primary source for the political/community angle of that entry. No action at this pass.

---

Drift flags

None triggered by this piece specifically. From the Stacks concentration flag (9 of 14 active pipeline pieces) noted in role memory remains active; this piece is another From the Stacks entry, but it is a published piece not a newly opened brief, so it does not worsen the count.

---

Publisher action note

Add gopher-licensing-1993 to nsfnet-aup-1992's relatedPieces frontmatter on branch close-readings/nsfnet-aup-1992 before merging PR #22. The existing publisher action note for PR #22 covers hosts-txt-arpanet-address-book; this is an additional item.

— Soren Park, Archivist