Does the surviving original Microsoft/Altair BASIC source listing confirm Davidoff's credit comment and his omission from the copyright notice?

Routed from question-verify-davidoff-altair-basic-source-header-primary, which flagged that claim-davidoff-wrote-third-of-altair-basic-uncredited rested only on Harry Lewis's 2011 secondary blog analysis (Tier 2), not on the surviving BASIC source listing itself.

The listing does survive and is now directly readable: Bill Gates released a scan of the original 1975 fan-fold printout on his own site (gatesnotes.com) on April 2, 2025, as part of Microsoft's 50th-anniversary observance. A community mirror of that release, OCR'd to text, is hosted at the Internet Archive (item original-microsoft-source-code). This capture fetched that OCR'd text directly via extract_pdf (sha256 recorded above) and read it — this is the primary document itself, not a retelling of it.

Claim: The surviving listing's own header comment names Davidoff and credits him with the math package

Page 1 of the listing (headed "BASIC MCS 8080 GATES/ALLEN/DAVIDOFF") carries a block comment opening "COMMENT +" that reads, per the OCR'd primary text:

"ORIGINALLY WRITTEN ON THE PoPe1¢ FROM / FEBRUARY 9 TO APRIL 9 / BILL GATES WROTE THE RUNTIME STUFF, / PAUL ALLEN WROTE THE NON®RUNTIME STUFF, / MONTE DAVIDOFF WROTE THE MATH PACKAGE,"

(OCR renders "PDP-10" as "PoPe1¢" and "NON-RUNTIME" as "NON®RUNTIME" — scan-recognition noise on punctuation, not a substantive ambiguity; the words themselves are unambiguous.) This confirms, from the primary artifact rather than through Lewis's retelling, that Davidoff was credited by name in the source comments as author of the math package — the first half of the question this capture was routed to answer. Historical/documentary claim; Tier 1 (the primary listing itself).

Claim: The same header's copyright line names only Gates and Allen, omitting Davidoff

Immediately above the "originally written" block, the same comment reads (OCR'd):

"COPYRIGHT 1979 6Y BILL GATES AND PAUL ALLEN"

The "1979" is an OCR misread of "1975" — the scan's print-date stamp two pages earlier reads "PRINTED: 10-SEP-75," the document's own internal date range for its writing is "FEBRUARY 9 TO APRIL 9" [1975], and every independent secondary account (Harry Lewis's 2011 post; The Register's 2025 coverage) gives the copyright year as 1975 — the OCR engine appears to have confused the "5" glyph with a "9" on this particular scan line. The substantive point survives untouched by that OCR artifact: the copyright line names exactly two people, Gates and Allen, and does not name Davidoff — even though the very next lines of the same comment credit him by name for the math package. This is the second half of the question this capture was routed to answer, and it now rests on the primary listing rather than solely on Lewis's paraphrase. Historical/documentary claim; Tier 1.

Claim: This is the actual surviving original artifact, not a later reconstruction

The released document is a 157-page scan of assembly-language fan-fold paper, internally titled (per its own TITLE pseudo-op) "BASIC MCS 8080 GATES/ALLEN/DAVIDOFF" and internally date-stamped "PRINTED: 10-SEP-75" — consistent with a listing pulled from the PDP-10 account Gates used at Harvard in 1975, not a modern reconstruction. Bill Gates released it himself, on his own site, framing it as "the coolest code I've ever written," to mark Microsoft's 50th anniversary (The Register, 2025-04-03; Tier 3 for the release event itself). Because Gates is releasing his own historical artifact on his own venue, this clears the floor as a Tier 1 primary document for the two claims above; the release-event framing (date, occasion, Gates's quote) is corroborated at Tier 3.

Claim: The primary listing corroborates, and upgrades the sourcing of, the existing vault claim

claim-davidoff-wrote-third-of-altair-basic-uncredited previously rested entirely on Harry Lewis's Tier-2 blog post, which stated Davidoff was "acknowledged by them but not included on the copyright notice." The two claims above confirm that statement directly against the primary artifact: the header names Davidoff for the math package, and the copyright line does not. Of the three sub-questions question-verify-davidoff-altair-basic-source-header-primary raised, this capture resolves the first two (verbatim header wording; copyright-notice omission) from the primary source. The third sub-question — Lewis's claim that Davidoff's file ran "about 31 pages" of a roughly 65-page program — is not addressed here; confirming it would require tallying page boundaries of the math-package section within the 157-page listing, which this capture did not do. That figure remains [unverified-quant — needs primary] pending that tally.

Further leads

• Lewis's "about 31 of 65 pages" page-share figure for Davidoff's contribution is still untallied against this primary listing — a page-by-page pass to find where the math-package (F4I) routines begin and end would settle it. [unverified-quant — needs primary]
• Secondary accounts disagree on how and when Harry Lewis found/preserved the listing at Harvard (one summary said "behind furniture" in the late 1970s, another said "behind a filing cabinet" in 1999) — not resolved in this session; needs a direct read of Gates's own gatesnotes.com post or a Harvard Crimson archival piece, both of which returned 403 or were only available via AI-summarized fetch this session. [unverified — needs direct read]
• The document's own internal writing-date range, "FEBRUARY 9 TO APRIL 9" [1975], lines up with the well-attested history of Paul Allen's April 9, 1975 flight to MITS in Albuquerque for the first Altair BASIC demo — worth its own dating claim-note.
• Microsoft separately released GW-BASIC source (a later, MS-DOS-era descendant product) on GitHub in 2020, which per one secondary summary carries a related but more informal Davidoff attribution ("MONTE DAVIDOFF WROTE THE MATH PACKAGE (F4I.MAC)"); this is a different artifact from the 1975 Altair BASIC listing and was only seen via AI-summarized fetch here, not verified directly. [unverified-mechanism — needs primary]
• The Internet Archive item also hosts CHOCR/HOCR/DjVu-text derivatives of the same scan, plus a smaller "_text.pdf" (the one fetched here) — useful cached entry points for any future page-tally work.
• GitHub mirror pronoiac/altair-basic-source-code carries its own independent OCR pass (ocrmypdf.txt) of the same underlying scan — a second OCR rendering that could cross-check garbled numerals like the "1979/1975" case above.

Entity candidates

• Harry Lewis — person — Gates's Harvard tutor/professor, discovered and preserved the original listing at Harvard, and authored the 2011 secondary analysis this capture was routed to verify against the primary; the foundational figure this whole capture measures its claims against. Flag first.
• Monte Davidoff — person — the contributor whose credit-vs-copyright discrepancy is the subject of this capture; already has vault notes.
• Bill Gates — person — co-author of the listing and, fifty years later, the one who released the scan publicly himself.
• Paul Allen — person — co-author, named in both the header comment and the copyright line.
• Altair 8800 / MITS — concept — the hardware platform the BASIC interpreter targeted; context for the whole episode.
• PDP-10 — concept — the Harvard machine on which the interpreter was originally written.
• Internet Archive item "original-microsoft-source-code" — artifact/term — the specific mirror used as this capture's primary source; worth a pointer note for future citation.