Paul McJones, who worked as John Backus's assistant at IBM starting in 1974, has assembled a draft history of Backus's functional programming project for the Software Preservation Group, preserving reports, code, and correspondence from 1969 until Backus's 1991 retirement.

Motivated by the late-1960s software crisis, the FORTRAN creator set out to attack programming itself. His 1972 report introduced Red, a variable-free language where functions are built from combining forms rather than parameter descriptions. His 1974 IBM annual report coined the "von Neumann bottleneck" and claimed conventional languages could never slash programming costs — a claim IBM manager Patricia Goldberg politely half-rejected.

The 1977 Turing Award lecture, "Can Programming Be Liberated from the von Neumann Style?", presented FP, an algebra of programs with rewrite laws for proving equivalences, and applicative state transition systems. It also gave the world point-free style: functions composed without ever naming data. By 2026 the paper had over 2,000 citations, most of them apparently just for the phrase "von Neumann architecture."

Reactions were lively. Dana Scott and Robin Milner sent congratulatory letters that gently suggested typed lambda calculus instead. Edsger Dijkstra's EWD692 called the lecture a progress report suffering from "aggressive overselling," later admitting his review had been a "political pamphlet" defending axiomatic semantics.

John H. Williams joined in 1978, followed by Edward Wimmers, Peter Lucas, and Alex Aiken; the effort culminated in the FL language. Outside IBM, FP spawned Berkeley FP (shipped with BSD Unix), Illinois FP, Biagioni's FPC translator, an OCaml implementation, reduction machines at GMD and UNC, and research threads from lazy evaluation to VLSI design.

The site welcomes donations of further materials — the assessment section is still an empty outline.