Dr. Dobb's Journal January 1999
This year marks the 50th birthday of the Electronic Delay Storage Automatic Computer (EDSAC), the computer developed by Maurice Wilkes and the staff of the Mathematical Laboratory at Cambridge University in the flurry of computer building that followed World War II. Every year since 1946 has marked the 50th birthday of some pioneering computer, and the first few years of the 2000s will be no different. But this particular pioneering computer deserves especially to be honored. EDSAC was the first large-scale, fully functional, stored-program electronic digital computer. When it first ran on May 6, 1949, it was fully functional, with a fully realized instruction set, paper tape for input, and a teleprinter for output. Memory was a set of mercury baths through which generated and regenerated acoustic pulses represented the bits of data.
One of the many problems Wilkes faced with little previous work to guide him was the bootstrap problem: How do you get the initial program into the machine, so that it knows how to read other programs in from the input device?
He solved the problem by putting the program in what we would today think of as the ROM. Wilkes called the initial program the initial orders, and EDSAC's ROM was a bank of stepping telephone switches. The initial orders just directed the reading of a paper tape; for example, the first few bits representing the opcode, the following bits representing the address, with no leading zeros, terminated by a special code. David J. Wheeler wrote several versions of the initial orders, and his version designed in September of that year included a very advanced feature: address modification.
Wheeler can also probably be credited with the development of the subroutine, although there were chunks of code called "subroutines" before 1949. But, Wilkes and Wheeler early saw the value of writing what we would call modular or structured code, and prided themselves on writing programs that could be understood without flow diagrams.
-- M.S.