Provenance Record

Well-structured News piece (671 words) with clear Overview / What We Know / What We Don't Know / Why It Matters sections. Tone is neutral and appropriately professional throughout — no editorializing, no AI self-reference. The author explicitly handles a sourced disagreement (the $75K vs ~$100K acquisition price) instead of papering over it, which is good editorial practice.

Read all 6 gzipped snapshots from disk (source-0 through source-5). • source-0.html.gz (Microsoft Open Source Blog): Confirms blog post 'Continuing the story of early DOS development', published 2026-04-28, authored by Stacey Haffner (Director, Microsoft OSPO) and Scott Hanselman (VP, Member of Technical Staff, Microsoft/GitHub). Confirms team 'led by Yufeng Gao and Rich Cini', the prior open-sourcings (MS-DOS 1.25/2.11 in 2018, MS-DOS 4.0 in 2024), and the verbatim phrase 'licensed under MIT via pull request'. • source-1.html.gz (Tom's Hardware, Mark Tyson, 2026-05-02): Confirms the verbatim quote 'is perfect and recompiles byte for byte to the original binaries' attributed to a Microsoft VP, and the 'found in a garage' detail with 'reams of old dot matrix printouts'. • source-2.html.gz (It's FOSS, Sourav Rudra, 2026-04-30): Confirms the verbatim quote 'Tim did not hand over a tidy source archive; instead, what he kept were physical assembler printouts and stacks of continuous-feed paper from 1981' and the verbatim phrase 'locate, scan, and transcribe the DOS-related portions into compilable code'. Confirms '$100,000' acquisition figure and 'August 1981' shipment to IBM as PC DOS 1.0. • source-3.html.gz (PCWorld): Confirms competing '$75,000' acquisition figure and 'PC-DOS 1.0 was ready in August 1981'. The article correctly flags the $75K vs ~$100K discrepancy as unresolved. • source-4.html.gz (GitHub repo DOS-History/Paterson-Listings): Confirms repository description 'Transcription of Tim Paterson's DOS printouts', the three top-level directories (1_transcription, 2_printed_files, 3_source_code), the verbatim contents description '86-DOS 1.00 kernel, various PC-DOS 1.00 pre-release kernels and utilities, and the Microsoft BASIC-86 Compiler runtime library', and 'Seattle Computer Products' ASM assembler' targeting. README states 'MIT license'. • source-5.html.gz (Open Source For You, Apurba Sen, 2026-05-04): Confirms the verbatim phrase 'support study, preservation, and exploration of computing history' and the included materials list (86-DOS 1.00 kernel, PC-DOS 1.00 development snapshots, CHKDSK, assembler listings). Every inline link in the body matches a source that contains the specific claim attached to it. No misattribution detected.

All factual claims verified against snapshots. The article's careful handling of the $75K vs $100K discrepancy — explicitly stating 'the two figures cannot both be exact, and neither outlet documents an underlying contract reference' — is exactly the kind of bounded reporting that strengthens credibility. No hallucinated quotes; the 4+ direct quotations all appear verbatim in the cited snapshots.

High-quality News submission. All claims sourced and verified verbatim against the snapshots. The verdict is APPROVE_WITH_CORRECTIONS solely because 5 of the 6 source domains are not in the allowlist (Tom's Hardware is the only one that is) — a technical compliance issue rather than a content issue. The corrections file documents this gap. The article itself is publishable as-is.