The book details the life of Charles Babbage (1791-1871) and his attempt to build a machine to execute a well-defined task: numeric evaluation of polynomials using the Method of Differences... Today, that would be called a 'special-purpose computer'.
Why would one need such a thing? Back then, human computers (the term actually used) worked out numerical tables for statistics and navigation. There were of course lots of errors. The use of a mechanism offered the hope of streamlining the process.
As Babbage did not have access to valves or transitors, the idea was to use mechanical means - basically build a hugely complex clockwork - and that in an age that did not yet have any tradition of mass production. Do I hear 'Good Luck' shouts?
How it all began, how Babbage tried to secure the funds and find the engineering skill, and how he ultimately failed through basically bad PR, bad marketing and lousy project management, not only once but twice, is a tale that will be recognizable to anyone doing complex and risky engineering tasks today. It's all a little bit of history repeating...
What remained was an unfinished difference engine no.1, the detailed plans (complete with bugs) for a more elegant and simpler difference engine no.2 and details for an 'analytical engine' - something that was not too far away from a general-purpose mechanical computer (no program store though, sorry). In the end, no-one of the modern computing pioneers seems to have been influenced by Babbage's legacy, so, he represents only a dead branch of the 'computing engine development tree'.
The second part of the book treats us to the attempt of the British Science Museum to actually implement the plans of difference engine no.2 -- the author, Doron Swade, having been the project manager of that undertaking, we really get the inside view. Again, it's a battle to secure the funds and find the engineering skill, with outside funding sometimes on, sometimes off (I especially liked the part of IBM promising to bankroll the whole projet at the serious risk of destroying it by wanting to take over management, then (luckily?) turning chicken once the yearly results came in). The engine was just barely finished after 17 years and >250'000 GBP for the bicentenary of Charles Babbage, in 1991.
It works. And I plan to check it it out next time I'm in London.
The book puts paid to the notion that the Engine was not built because Victorian mechanical engineering was not up to the task. But it leaves the question: Would Babbage's Difference Engine, if built, have been actually useful and economically viable? Well, I guess no one knows. But the answer is probably no...