Slashdot Mirror
What If Babbage Had Succeeded?
mikejuk writes "It was on this day 220 years ago (December 26 1791) that Charles Babbage was born. The calculating machines he invented in the 19th century, although never fully realized in his lifetime, are rightly seen as the forerunners of modern programmable computers. What if he had succeeded? Babbage already had plans for game arcades, chess playing machines, sound generators and desktop publishing. A Victorian computer revolution was entirely possible."
The concept of huge mechanical computers fulfilling any purpose that seems hard for us to comprehend today.
Yet huge mechanical computers for specialized use were in actual deployment in several industries, not the least of which were "fire control computers" on US and British Battle Ships and Heavy Cruisers in the pre WWII era. These were initially fairly huge mechanical beasts that were originally developed around the time of the first World War, and which were initially totally mechanical in nature. By the Second World War they were electro-mechanical (solenoids and relays and stepper motors), and were enclosed in battle hardened enclosures.
Still 1920-to-1945 is hardly 1833, and the size and complexity of such devices taxed the manufacturing capabilities of the day, and the size and complexity of the problems they could solve was probably more easily worked out on paper than set (programmed) onto the machine.
Having worked out the concepts, one wonders how far Babbage could have progressed with a large budget and a larger machine shop to build his engines. There were precious few problems to which you could apply this technology in that day. But its a chicken and egg problem. Its hard to know what computations would have been attempted had such equipment been available. The calculation problems any society tackles tend to be near the limits of the computing capabilities available to the task.
A man before his time.
Sig Battery depleted. Reverting to safe mode.
"The point is that Babbage did succeed, except it was through his inspiration, which took his ideas and better manufacturing processes and newer knowledge of materials and a refined computing model."
This is a common misconception based on earlier analyses. In fact, portions of his engines have been built from the original plans, using techniques available in his day, and it has been determined that it would indeed have worked if only it had been built.
Contrary to popular belief, the two biggest problems that Babbage faced were: (1) his inability to convince investors in the worth of his invention, and (2) his insistence on constant refinement rather than freezing the plans at some viable point, in order to make a working device.
Historically computing has never been a processing problem, but a storage problem. Or all computing, from embedded stuff to supercomputers, pretty much seems to revolve around turning a computation bound problem into a storage bound problem, and waiting for storage to improve so you can roll out faster processors to make use of it.
Try it yourself, if you have the skills. I had a pretty decent bitslice ALU design for a relay CPU down a total of 20 relays per bit slice, not just a wimpy bare adder but a pretty full featured design complete with comparator and roller/shifter unit. An 8 bit processor is well within my entertainment budget at a couple bucks per relay, and if I package each bitslice into something the size of a ream of paper, which is probably pretty pessimistic, the entire 8 bit CPU is only about the size of a box of bulk laserprinter paper. I figured for about $500 total all costs of all parts I can get a decent reliable relay based 8 bit CPU operational.
But a couple hundred bytes of relay based ram to run some "real programs" is way outside my budget, both financial, storage, and power. Even tradeoffs don't work, like using latching relays saves me considerable (cheap) power at a cost of roughly twice as much per bit. Inevitably you get into weird dynamic electrolytic capacitor designs, strange attempts at homemade core memory... Cheating and using modern sram isn't cool. Hundreds of latching relays at lets say $5 per bit isn't gonna fly if I "need" a K or so of memory to have fun, that would be $40K just of storage relays to say nothing of the address decode logic etc. Also that would be well in excess of 8000 relays for a K of memory, vs a mere 160 relays for the processor. About 80 times bigger. So that goes from a small box sized CPU to basically a room of my house.
This has interesting MTBF implications, in that any "non-trivial" relay computer is going to mostly fight memory breakdowns, not processor failures.
To an amateur, calculating is the hard part. To a pro, storage is where the real problem lies.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
It almost seems possible that it could have been done much earlier. Automatic sources of rotary motion were known (waterwheel, steam engine) along with gearing mechanisms.
One of the first "programmable toys" was a cart controlled by a winding string
But it wasn't until the Industrial revolution in the 1850's, that the use of punched cards for storing instructions and input data that made mathematical calculating machines possible. That's one important factor. The other one is the use of mathematical notation for expressing algebra that can be converted into instructions.
What if he had got both these engines working by 1849? Would he have moved onto more advanced calculations or extended the use of mechanical computation to commerce like Hollerith punched cards did in 1889? If so, that would have advanced computing by 40 years.
First documented geared calculating mechanism (Antikythera) 150 - 100BC
First documented use of waterwheels - 300BC
First documented Steam Engine - 1AD
First use of punched cards - 1725AD
Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
Much as I like the steampunk concept, Babbage's machine was at the upper end of what was buildable as an expensive prototype. Bear in mind that even consistently-good, moderately priced steel wasn't available until the 1880s. That's why fine machinery was made of brass until the 20th century.
The commercial history of mechanical calculators is not what you'd expect. Leibniz built the first mechanical multiplier in 1694. The commercial version, the "Arithmometer", wasn't produced until 1851. (It took a very long time to commercialize technology before there was industrial infrastructure.) Adding machines came later, because an adding machine is only a marginal improvement over an abacus, but a multiplier is a huge win.
The first high-volume mechanical arithmetic device was the cash register. When, in 1884, cash registers first got tape printers, for the first time merchants had some real mechanical bookkeeping assistance. By then, good steel was available, and stamped parts could be made in volume. That's the point at which something like Babbage's machine might first have been a commercial success.
Which it was. Hollerith's first punched card machines were used for the 1880 census. The Computing-Tabulating-Recording Company manufactured Hollerith machines commercially. The CTR became the International Tabulating and Recording Company, which became International Business Machines, which is today's IBM.
By 1880, there was enough manufacturing infrastructure to make stuff, and there was continuous year to year progress in mechanical calculation. The peak in purely mechanical systems was probably the Burroughs Sensimatic, in 1953, which was essentially a spreadsheet program made out of gears. IBM tabulators were more advanced, but they were electromechanical.
Babbage was not "a man before his time". He didn't need more money. He didn't need a larger machine shop. He blew it!
He had the money.
The people in 1800's Britain knew a good thing when they saw it. And when small prototypes were demonstrated the British Government committed to build the difference engine. And guess what, they wanted to use it for gunnery on ships! They invested *big*. How much? One fully kited out battleship's worth. One of these: http://en.wikipedia.org/wiki/HMS_Warrior_(1860) (more or less). That is a huge amount of money.
The skills were available.
Have a look at a British clock from this period. Very intricate work and at a lot smaller scale than Babbage required. Sure, what he was doing was on a large scale, but the skills and tools were out there. Indeed, Babbage teamed up with them and had the money to do it.
But he committed the cardinal sin. Babbage was forever changing the design. Yes Mr Babbage, your analytical engine idea is nice but we are paying you for the difference engine! He could not stay focused to build what was paid for and required. Falling out with the machinists capable of building it hardly helped maters. He did not deliver. As a result he blew not only his own reputation but that of the whole idea, killing it for the best part of a century. That is how bad he was.
You can be the most talented man in the world, but if you are so disorganised and uncivil that nobody wants to work for you it is all for nothing. A lesson we can all still learn form.
http://en.wikipedia.org/wiki/The_Difference_Engine - by Bruce Sterling and William Gibson is a fascinating and complex exploration of exactly this concept: namely that Babbage succeeded. The key historical difference - the premise of the book - is that England's backing of the American Civil War succeeded, due to cryptography in part. Towards the end of the book it's made clear that the continued war between France and England has turned "cold" and thus much effort is dedicated to sneaking obfuscated "divide by zero" algorithms into the opposing side's Difference Engines. this book is one of the only sci-fi books (out of over 500 that i've read) that i actually found it hard to understand even 50% of what was going on. still made a damn good story, though.
From my study of Catholic cultural intervention (the various crusade activities) I realized that we would have had our current level of technology about 1000 years ago had they not meticulously stamped out all attempts to gather and exchange knowledge (particularly the Albigensian era). To the minds of most educated people I think this trend kicked off at the two burnings of the library of Alexandria.
I object to power without constructive purpose. --Spock
But it wasn't until the Industrial revolution in the 1850's, that the use of punched cards for storing instructions and input data that made mathematical calculating machines possible. That's one important factor. The other one is the use of mathematical notation for expressing algebra that can be converted into instructions.
What if he had got both these engines working by 1849? Would he have moved onto more advanced calculations or extended the use of mechanical computation to commerce like Hollerith punched cards did in 1889? If so, that would have advanced computing by 40 years.
Yes, it would have advanced computing by about 40 years. But computing had reached a plateau in the 1940's (and arguably before then, there just wasn't any impetus to make a digital computer before WWII), and couldn't really advance any further than it had at that point until the invention of the transistor... the transistor itself arose from a chance discovery in late 1947, and wasn't readily available until the mid 1950's. Similarly, the integrated circuit wasn't available until the mid 1950's, either. In the absence of those technologies, it's arguable how far computing could have advanced beyond how far it had already advanced by the late 1940's, and neither IC's nor the transistor arose from people researching how to improve computers.
It really is debatable how far computing could have gone if Babbage had succeeded, considering that the computer revolution really didn't take off until integrated circuits made miniaturization possible in the 1960's.
The difference is that there was no popularly conceived need for such tools back then. Would you rather spend the modern equivalent of millions on a tabulating machine of some sort, or hire several accountants?
Bookkeeping wasn't nearly as complex than as it is today. There was negligible need for anything like this: society at large moved much slower, and there was time to do the basic arithmetic necessary to meet their needs. (Even today, most people don't need anything much more complex than a calculator around tax time...)
Cryptography was the first demonstrated use for modern computing (during WWII). Now, consider cryptography during the US Civil War. It basically didn't exist: they used cipher disks which utilized simple substitution ciphers and what we might today call a seed (by means of a visual or auditory cue). "Something you know and something you have". Imagine how complex, expensive, and precise the machinery needed to perform WWII-era ciphers would be if it were purely mechanical. It would also have to be fairly single-purpose.
The sad fact is, there's really little practicality to computers until you get to electronics. Even with electronics, it was a long time coming until they were practical for common use, and were only significantly used by governments and large corporations for one-off massive computation (code breaking, report generation, number crunching). The IC really was the bottleneck that needed to be beaten to make them generally practical (in terms of time and money vs. the results).
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
I'm no historian, but I seem to recall that in WWII, when the Allies bombed Germany, the targets generally were the factories; and the Germans deliberately put their factories in the middle of populated areas. (Were they relying on us being less willing to bomb them for fear of collateral damage?)
The allies carpet-bombed Hamburg, and fire-bombed Dresden. Between the two campaigns, both cities were almost completely destroyed, and about 100,000 victims. Only about 40,000 civilians were killed during the German bombing of England. It wasn't because the Germans deliberately put their factories in civilian areas (the English did the same), it was because the allies were bombing from very high altitude and there was a great deal of luck involved in actually hitting your target from that altitude.
And in preparation for D-Day, they completely levelled some cities in northern France. (seriously... the ground is now more than 1m higher in Caen than it was before WWII, and the cathedral was the only building left standing in that city following the allied bombardment... and even the cathedral was partly destroyed... they even destroyed the citadel of William the Conqueror, because of the risk that it could be used by German soldiers to hide out... and Caen had no industrial complex to speak of, it was just a military stronghold).
Bombs, at least the kind of bombs they were using in WWII, were not a precision instrument. The primary targets were factories, railroads, things like that. But the standing order for most aircrews was "drop any remaining ordnance wherever you want before coming home".
was Renaissance the cause or the effect of advanced technology? Adaptation of arabic numerals (the technology which enabled then-modern math) was almost concurrent with Thomas Aquinas' logical "proof" of existence of God (the act which is thought to have started the Renaissance by broaching the idea that reason is above God). Since the two developments were essentially concurrent, I wouldn't credit one as more important in causality of the events that followed than the other. It is entirely possible that the technical advances is what forced social changes on society. Gauss, for example, claimed that not introducing a decimal system was Greek's greatest failure. He went so far as to suggest that Dark Ages would not have happened at all if either the Greeks or the Romans had introduced a decimal system. Before knee jerking an answer with some facts you learn in school, think twice whether the level of your own insight on what influences math and scientific development is above that of Gauss.
Any guest worker system is indistinguishable from indentured servitude.
I am confused, what does the burning of the Library of Alexandria have to do with the Catholic Church? There are several stories of the burning of the Library of Alexandria, none of them involve the Church of Rome. The first story says that it was burned by Julius Caesar in 48 BC. The second story says that it was burned by Emperor Aurelian (who reigned before Constantine and attempted to make the Sun God the chief deity of the Roman Empire) when he put down a revolt in Egypt during his reign. The third story says that it was burned by Theophilus, Patriarch of Alexandria (who considered himself at least equal to the Bishop of Rome). The fourth story says that it was burned by the Muslims when they conquered Egypt.
The truth is that all men having power ought to be mistrusted. James Madison
neither IC's nor the transistor arose from people researching how to improve computers.
However, supposing there were mechanical computers, and they were useful; there would be an impetus to improve upon the capabilities / efficiency of the mechanical computers.
The availability of computers to scientists could have had a great effect on their studies, and it's difficult to predict what the results of that would have been.
The transistor could have been discovered a lot earlier, if inventors were looking for options, in efforts to find ways of adapting computation mechanisms into electric circuits; once all cost-effective improvements had been considered, computing would stagnate.
The transistor might have been discovered much earlier. LEDs might have been discovered, without a lightbulb ever having been invented -- Tesla might have never come to the US, AC may have not been discovered or put to practical use, even to date....
World War I and II might not have happened; because technology had a role in events leading up to them. With no real moving pictures or radio to use as propaganda tools; low-speed media communication, WWII and WWI are simply unlikely to have sparked...
With no world wars, no cold war, therefore no real investment in space travel, no NASA; no satellites, rockets, GPS, many technologies not existing.
Today, there might be no such thing as personal computers or Cell phones... no radio, no wireless communication, no television, but no PCs = no internet, just a small global network of big companies' number crunchers at most.
I'm surprised no one has mentioned this yet, but I think the biggest deficiency of Babbage's design was the base-10 numbering paradigm. Sure, he had the computer architecture down, to what we would now call the Von Neumann architecture, with the load, compute, store instructions. But making it all work in base-10 was incredibly messy, and I would think that is mostly why it was so difficult for him to implement.
It was not until 1854 that George Bool invented what we now call Boolean Algebra.
Boolean logic allowed us to simplify computing circuitry, improving it's efficiency and size. Take a look at this famous YouTube video, it shows a mechanical calculator built with marbles, where a marble indicates a one and no marble indicates a zero. AND and OR gates are incredibly simple lever mechanisms, and it is powered by gravity and the weight of the marbles. What if Babbage had thought to use marbles and base-2 numbering instead of gears and base-10 numbering to do computations? He couldn't have because Bool's idea had not been thought of until some 30 years after his death, and even after that, it was not until Alan Turing (120 years after Babbage) that anyone was clever enough to realize that Boolean logic, as any other logic, could be used to program a computer. Before Turing, Boolean logic was more or lest a reasoning language for testing the logical soundness of true/false propositions.
So, architecturally, Babbage was ahead of his time, and perhaps had his idea succeeded, it may have encouraged research and development leading to the use of Boolean logic in computing much earlier. But that wasn't the case. It is fun to think of what may have happened though: we may have seen immense computing factories powered by mills which lifted grounded marbles into a giant bin above the factory, and all of that weight would filter through the mechanisms of the computer to produce results. Such a thing would have been unbearably noisy, but fast, simple, easily reparable, and effective. And it would have continued that way until someone thought of using electrical charges instead of marbles.
In all, I think if Babbage's design had succeeded, it may have made the computer revolution happen 30 or 40 years earlier, in which case, I would have been born in the the mostly ignorant generation of kids comprising the social-networking and internet revolution, and not in the more down-to-earth generation of 8-bit gaming, Q-BASIC and assembler-programming, personal computer revolution folk.