A Cambridge thesis puts noughts and crosses on a tube

What happened

EDSAC existed to do mathematics for the university, and it carried a set of cathode-ray tubes used for inspecting memory. Douglas, working on a thesis about how people and machines might interact, pressed one of those tubes into service as an output device. A grid of dots formed the board and a rotary telephone dial selected a square. The program then played its own move, correctly, since the game has a known solution. Reading the display was a modest technical trick, but it made the machine's state visible to a person instead of printing it on tape.

Almost nobody played it. There was exactly one EDSAC, its time was booked for research, and the program could not be copied to another machine because no compatible machine existed. That isolation is why the claim usually made for it, that it was among the earliest graphical computer games, needs both halves of the phrase. The graphics were real and the game was real, yet it functioned as a research artifact demonstrating an interface idea, and it vanished along with the computer it was written for.

What Douglas was actually studying outlasted the game. His subject was the exchange between a person and a machine, at a time when computing meant handing over work and collecting output hours later. Choosing a game, dialing a move and watching an immediate answer appear on a screen described a way of working that would not become normal for another twenty years. It anticipated the interactive terminal far more than it anticipated an entertainment industry.

The world at the time

Cambridge's EDSAC had been running for a couple of years and was among the first practical stored-program computers anywhere. Programs went in on paper tape, results came out on a printer, and machine time was rationed. Display tubes were diagnostic instruments rather than interfaces. That a computer might respond to a person moment by moment was a research question, not an expectation.

What it changed

The program had no descendants, since each early game of this kind was written from scratch for a single machine. Its lasting contribution is evidentiary: it is the reference point historians reach for when arguing about which screen game came first. Douglas went on to a long career in computing, and the interaction problem he chose to study eventually became a field of its own.