Celebration of David Wheeler (1927–2004)
Martin Campbell‐Kelly · The Computer Journal · 2005
On 18 April 2005 the Computer Laboratory at Cambridge University celebrated the life and work of David Wheeler FRS, who spent his working life—from research student to professor emeritus—in the Laboratory. The 170 attendees included Laboratory alumni, academic friends and colleagues, and his family. They ranged from fresh-faced research students to eminences gris with walking sticks—a telling indication of the universal appeal of Wheeler's friendship and reputation. Wheeler had, in fact, retired from the Laboratory in 1994, but as professor emeritus he continued to cycle in daily until his death from heart failure last December. Fittingly, the proceedings were opened by Professor Andy Hopper, a former research student of Wheeler and currently head of the Laboratory; they were closed by Professor Sir Maurice Wilkes, director of the Laboratory 1946–1980—and Wheeler's PhD supervisor. The main proceedings began with two of Wheeler's colleagues from the early days in the Laboratory, Dr David Hartley (one-time head of the Cambridge University Computing Service and a past president of the BCS) and Professor David Barron (founding professor of computer science at Southampton University, now emeritus). Their presentations were interspersed with evocative video clips from the EDSAC Film made in 1951 and a lecture Wheeler gave nearly 50 years later at the EDSAC '99 conference. Hartley and Barron both joined the laboratory in the mid-1950s, and for them Wheeler was already ‘up there with the gods’. Wheeler had become a research student in the Laboratory in 1948 when the EDSAC computer—the world's first stored-program computer to go into regular service—was still under construction. Wheeler was given the task of designing the programming system. His reputation was secured by the invention of the closed subroutine and an assembly system known as the Initial Orders—concepts that are still cornerstones of programming technology. Barron recalled that at the Summer Schools in programming, students would be shown examples of subroutines from the EDSAC library, much of which had been written by Wheeler. Every subroutine was a programming masterpiece in miniature. Another example of Wheeler's programming virtuosity was the 768-instruction fixed store for EDSAC2, a read-only memory that contained an assembly system and basic subroutines. The EDSAC2 was the first machine to use a microprogram, and Wheeler wrote that also. Both had to be hand-wired onto magnetic cores. The fixed store routines were then ‘potted’ in epoxy resin—a point from which there was no return; they worked the first time and faultlessly.