The Lost Relic of Computing's Dawn

In 1975, a 4KB BASIC interpreter for the MITS Altair 8800 microcomputer didn't just run code—it catalyzed the entire PC software industry. Co-created by Bill Gates, Paul Allen, and the lesser-known Monte Davidoff, Altair BASIC was Microsoft's first product, a foundational pillar of modern computing. Yet, for over two decades, its source code vanished into obscurity, becoming the Holy Grail for tech historians. As Gates himself admitted in a 2000 interview, "We were digging around for this recently and found it. We will get the source code up on the Web." But the path to its discovery was far from straightforward—a tale of broken promises, accidental archaeology, and a disassembly project that finally pieced together tech's origin story.

Article illustration 1

The Decades-Long Hunt

By 1987, with the Altair relegated to museums, French journalist André Warusfel petitioned Gates for the source code, hoping to preserve a critical artifact. Gates agreed, but the code never materialized. The quest was later taken up by David Mery, another French journalist and programming historian, who spent years lobbying Microsoft. Mery emphasized its cultural weight: "The PC software industry has a short history, and we should have access to the milestones of its evolution." Despite Gates' assurances, the code remained elusive—until serendipity intervened.

Accidental Discovery at Harvard

The breakthrough came not from Microsoft's archives but from Harvard University, where Gates had studied. In the late 1990s, Harry Lewis, a computer science professor and Gates' former tutor, stumbled upon a printed listing of the 4K BASIC source code behind a filing cabinet. This copy, one of several presumed lost, was deposited in Harvard's Pusey Library. Ian Griffiths, a UK programmer who examined it, noted the irony: "Bill's total input to the preservation project was his act of absent-mindedly dropping a listing somewhere in Harvard in 1975." Griffiths documented key details, including a revealing comment in the source that credits each creator:

00560 PAUL ALLEN WROTE THE NON-RUNTIME STUFF.
00580 BILL GATES WROTE THE RUNTIME STUFF.
00600 MONTE DAVIDOFF WROTE THE MATH PACKAGE

This snippet not only confirms Davidoff's pivotal role—often overshadowed by Gates and Allen—but also highlights the collaborative ingenuity behind early software. Davidoff, amused by renewed interest, later reflected on the project's significance during interviews, though Microsoft maintained its "look, don't touch" policy, forbidding copies.

Article illustration 2

Disassembling History

The story didn't end in Harvard's archives. Reuben Harris, a London-based programmer and associate of Griffiths, took on a parallel mission: disassembling an early Altair BASIC binary to reconstruct the source. This painstaking reverse-engineering effort, which Davidoff reportedly found "tickling," aimed to verify the authenticity of the Harvard find and provide a public reference. Harris' work, completed weeks after the discovery, underscored a broader truth—without proactive preservation, critical digital history relies on luck and hacker tenacity.

Why This Code Still Matters

Altair BASIC's rediscovery isn't just nostalgia; it's a cautionary tale for today's developers. In an era of cloud ephemerality and proprietary silos, this 4KB artifact reminds us that software shapes civilization. Its loss would have erased insights into optimization tricks for limited hardware—skills now resurgent in IoT and edge computing. Moreover, Davidoff's belated recognition challenges the myth of solitary genius, emphasizing teamwork in innovation. As open-source advocates push for transparent archives, this saga proves that preserving code isn't merely academic—it's safeguarding the blueprints of our digital future. After all, the next revolution might be hiding behind another filing cabinet.

Source: Adapted from original reporting by Andrew Orlowski at The Register.