Ancient Web: David McQuillan Kept Unusual Computing Machines Explained

Part of the Ancient Web Collection.

There is something revealing about a computer engineer's personal site that devotes space both to an early parallel processor and to figuring out the best way to lay wooden railway track.

Visit David McQuillan's Fano homepage.

McQuillan's site is a collection of mathematical amusements, personal projects, and firsthand computing-history notes. Its links include problems about cake cutting and bouncing cannonballs, photographs of Meccano builds, an interactive tesseract, and pages about specialized machines that rarely reach popular histories.

When one processor was not the only answer

His Distributed Array Processor page describes the DAP, a machine attached to ICL mainframes. The architecture used a 64-by-64 array of one-bit processing elements, each with memory, under a master control unit. It was an alternative approach to computing-intensive problems long before GPU programming became everyday vocabulary.

McQuillan describes his own involvement with block I/O software and preserves code fragments in the DAP's Array Processor Assembly Language alongside equivalent-looking DAP Fortran examples. There is even a joke about naming a block I/O system so that the manual could describe biological transfers; somebody noticed and changed the acronym.

The page also links to a technical overview of the later AMT Gamma series, rescued from an archived website. It is a reminder that technical documentation sometimes survives only because one enthusiast knows which lost link mattered.

A different branch examines Percy Ludgate's analytical machine, designed around 1907–1909 but never built. McQuillan discusses decimal storage, rod-based mechanisms, paper-tape control, and an ingenious multiplication scheme based on transformed logarithms. His page links Ludgate's original 1909 paper and later historical studies. The author is careful to distinguish proposed operating times from an actual tested machine.

This is not a single-topic, frozen museum. Its importance comes from the combination of engineering experience and curiosity: old software notes, technical sources, physical models, and side trips into mathematics, all connected by one author's links.

Modern computing histories tend to lead from one commercial success to the next. McQuillan's pages are interested in alternate mechanisms, improbable projects, and the explanations that survive when the hardware does not.

Explore the DAP memoir and Ludgate research pages.

Part of the Ancient Web Collection.