Slashdot Mirror
65 Years Ago, Manchester's 'Baby' Ran Electronically Stored Program
hypnosec writes that the first ever practical implementation of the stored program concept took place 65 years ago, "as the Manchester Small Scale Experimental Machine aka 'Baby' became the world's first computer to run an electronically stored program on June 21, 1948. The 'Baby' was developed by Frederic C. Williams, Tom Kilburn and Geoff Tootill at the University of Manchester. 'Baby' served as a testbed for the experimental Williams-Kilburn tube – a cathode ray tube that was used to store binary digits, aka bits. The reason this became a milestone in computing history was that up until 'Baby' ran the first electronically stored program, there was no means of storing and accessing this information in a cost-effective and flexible way."
Wow. It's easy to forget that the entire industry of programmable computers is younger than a lot of ordinary people walking around today. It makes me wonder what entirely new industry I might see develop from nothing over my lifetime.
In one sense it wasn't missed. Machines like EDSAC and LEO followed shortly afterwards but the US had a booming economy by comparison and it was a lot easier for US businesses (with the much larger internal market as well) to grow big on the back of that.
When I first started in this industry, I worked with Chris Burton who'd worked on Baby (and later led the team which rebuilt it); he had known Turing, as had another man I worked with later. Our team was led by Charlie Portman, who gets a credit in The Mythical Man Month. It's pretty amazing how close we are - two generations away - from the legendary figures who founded our industry, who built the first computers.
Chris was famous in our team because we had some new Mannesman Tally inkjet printers, which could only print ASCII, and we needed them to print bitmaps. The processor in the printers was one that no-one in the team had any experience of. So Chris took the datasheet for the printer, the datasheet for the processor, a dump of the printer ROM, and a square ruled pad home with him on the train, and came back in the morning on the train with code for a new ROM for the printer, written not in assembler but in the actual opcodes (hexadecimal), in pencil on the pad. We blew them into the ROM and it worked first time printing perfect bitmaps, no errors, no bugs to fix.
That's how good the first generation programmers were. I am still in awe of that. And he was a very modest man, very generous with his experience. I'm proud to have learned from him.
I'm old enough to remember when discussions on Slashdot were well informed.