Babbage makes his Engines (RAR!)

[Zor]

This fellow is Charles Babbage (1791 to 1871), an English mathematician and inventor. Among his most notable claims to fame was designing two devices, the Difference Engine (a mechanical computer for crunching numbers) and the Analytical Engine (A programmable mechanical computer).

Though his designs were never completed in his lifetime, a functional version of one of Babbage's difference engine was built in the 1990s using materials and to within tolerances available at the time. It worked as designed.



Lets say that someone decided to provide Mr Babbage with enough funding to actually get these machines built in his lifetime. What would be the consequences of said development?

Zor

[SCRawl]

Did you read "The Difference Engine"? Because that was the plot.

[Simon_Jester]

Honestly? Not much, I suspect. Difference engines don't miniaturize worth a darn, and when you try to build a less capable smaller one it's going to converge on the kind of adding machines and other similar mechanical computing technologies that already existed.

The most likely consequences are going to be marginal improvement in the accuracy of astronomical calculations, the precision with which logarithms and trigonometric functions are published in look-up tables, and other things that were historically common applications of very early computers like cryptography and artillery firing tables.

[Jub]

I wonder if bombardments get less carpet saturation oriented as gunnery calculations get more precise or if military inertia means that one bit of new tech doesn't change anything?

[Sidewinder]

1990 technology could build the difference engine's parts within the tolerances necessary to MAKE IT WORK. 18XX technology could not. Even if you showered Babbage with money, the difference engine would not work until technology had sufficiently advanced to manufacture the its parts within the necessary tolerances.

[Zor]

1990 technology could build the difference engine's parts within the tolerances necessary to MAKE IT WORK. 18XX technology could not. Even if you showered Babbage with money, the difference engine would not work until technology had sufficiently advanced to manufacture the its parts within the necessary tolerances.

As I said what I gathered the ones the historians/engineers took that into consideration and the Difference Engine they did build were made with bits in the tolerances which could be achieved at the time at which it was designed/when Babbage was alive.

Zor

[Zixinus]

Electric computers and the digital revolution happens earlier. When making electric computers Babbage's work wasn't well-known about and electric computer makes spent time re-making the ideas that Babbage already made. Now computers would evolve from the mechanical computer to the electric in more gradual steps. Clockwork logic may have a stronger influence on modern computers. But not much beyond that.

Aside that, not terribly much. Computers were a specialized machinery until the home-computer revolution.

[HMS Sophia]

I wonder if bombardments get less carpet saturation oriented as gunnery calculations get more precise or if military inertia means that one bit of new tech doesn't change anything?

You might see mechanical fire control computers on Ironclads maybe? I honestly don't know much of the capabilities of the difference engine. If they could work like this, you might see an earlier switch over to a uniform battery on capital ships sooner than 1910 (HMS Dreadnought) to take advantage of fire control for the larger guns. Not that this would change a huge amount, Britain could still outproduce everyone massively at this point.

[Simon_Jester]

I wonder if bombardments get less carpet saturation oriented as gunnery calculations get more precise or if military inertia means that one bit of new tech doesn't change anything?

There are other obstacles than military inertia. One is communication- without radios you can't communicate between forward observers and artillery, which means you're still firing blind into a target area. If the position of the target was accurately surveyed in advance you may hit it more often, but you still hit it. If enemy troops are dug into an area, you still have to saturation-bombard that area to suppress them, let alone destroy their fortifications.

Also, none of this really becomes relevant until you get precision-machined guns and shells capable of reaching out to indirect-fire ranges, starting in the very late 19th or very early 20th century.

1990 technology could build the difference engine's parts within the tolerances necessary to MAKE IT WORK. 18XX technology could not. Even if you showered Babbage with money, the difference engine would not work until technology had sufficiently advanced to manufacture the its parts within the necessary tolerances.

As I said what I gathered the ones the historians/engineers took that into consideration and the Difference Engine they did build were made with bits in the tolerances which could be achieved at the time at which it was designed/when Babbage was alive.

Well, key question here: in what era? Today, turning out ten thousand very precisely cut gears might be routine, while in 1845 it might have taken the combined efforts of half the watchmakers and gearcutters in Europe for all I know.

A Difference Engine that is too expensive will accomplish very little if making even one of the things is too much work to justify the uses to which it can be put. It's like, if you could build a computer with all the modern capabilities, but any such computer cost a hundred billion dollars, there MIGHT be demand for one of them. That one would be elaborately time-shared for numerous enterprises... but there damn sure wouldn't be demand for them among anyone except the very biggest national governments and maybe some huge corporations. So it wouldn't change nearly as much as it would if the things cost a reasonable sum of money.

Electric computers and the digital revolution happens earlier. When making electric computers Babbage's work wasn't well-known about and electric computer makes spent time re-making the ideas that Babbage already made. Now computers would evolve from the mechanical computer to the electric in more gradual steps. Clockwork logic may have a stronger influence on modern computers. But not much beyond that.

Aside that, not terribly much. Computers were a specialized machinery until the home-computer revolution.

Well, inventing them earlier has some consequences. But for a lot of applications, Babbage's machine may not offer much actual advantage over a punch-card machine or an analog mechanical computer that runs on cams.

I wonder if bombardments get less carpet saturation oriented as gunnery calculations get more precise or if military inertia means that one bit of new tech doesn't change anything?

You might see mechanical fire control computers on Ironclads maybe? I honestly don't know much of the capabilities of the difference engine.

I'm not sure they'd offer much advantage over an analog computer like the ones that were historically built. The advantage of the Babbage engines was that they were programmable and could perform arithmetic fast- or rather, in massive parallel. It was not that they could calculate the complicated trigonometric functions involved in naval gunnery fast. For that purpose, an analog computer that can literally return the correct value of f(x) for arbitrary input values of x by turning a crank to point to x has some advantages.

Moreover, a Difference Engine has so many moving parts that it would be very hard to maintain on a warship that experiences a lot of vibration and shocks.

If they could work like this, you might see an earlier switch over to a uniform battery on capital ships sooner than 1910 (HMS Dreadnought) to take advantage of fire control for the larger guns. Not that this would change a huge amount, Britain could still outproduce everyone massively at this point.

Well, historically the switch to uniform battery was made in 1905-6, yeah. However, if you go back much before 1905, you lose some of the other key technologies needed for central battery control, like electrical circuits to relay the control signals. You also get smaller, more cramped warships that don't have as much room for a four ton computer... and guns that simply can't throw shells as far, making it less advantageous to be able to elevate them for indirect fire at extreme range.

The first naval battle fought at such extreme ranges was Tsushima, which was right on the brink of the Dreadnought Revolution as it is. In cases like Manila Bay, only seven years earlier, there was basically no effective fire into enemy ships from more than a few thousand yards away. Go back to 1882 and the bombardment of Alexandria and while the fire control may be better than in Nelson's day, the gun ranges are still down to "line of sight, point, elevate slightly, and shoot." And a difference engine might not change that very much.

[Purple]

It's like, if you could build a computer with all the modern capabilities, but any such computer cost a hundred billion dollars, there MIGHT be demand for one of them. That one would be elaborately time-shared for numerous enterprises... but there damn sure wouldn't be demand for them among anyone except the very biggest national governments and maybe some huge corporations. So it wouldn't change nearly as much as it would if the things cost a reasonable sum of money.

Er... you might want to familiarize your self with a company called IBM. Obviously their setup did not cost as much as your hypothetical but the logic behind it was pretty much what you describe.

[Channel72]

Well for one, something like the Church-Turing thesis would have been proposed earlier... perhaps by Babbage himself. His engine had conditional branching, so it was Turing complete, which is pretty remarkable. In actual history, it took until the 1930s for this to be really appreciated and absorbed by mainstream computer science - if Babbage's machine worked, no doubt this idea would have been widely appreciated much earlier.

The immediate concern would of course have been to improve the machine so it would work faster.

[Eleventh Century Remnant]

I have read The Difference Engine, but to be honest have forgotten most of it.

First application that comes to mind is fluid dynamics. Complex, iterative calculations far more easily and quickly done by mechanical means than by hand; gunnery is the obvious application, true, but ship design could be transformed by it, and aircraft brought on that much sooner.

Theoretical mathematics will leap ahead, with practical applications following; first you need the theories to realise what you have to be calculating, but testing hypotheses should be much easier. We are basically looking at computer aided design coming into play in the 1860's, even if the output is on sheets of paper that have to be couriered from office to office.

Perhaps we could short circuit thirty or forty years of the worst of the industrial revolution, slums, squalor and Victorian social stasis; handing an age that, perhaps because it was impelled by it's own ugly side to aim for some kind of redemption, still believed in progress tools of that kind- there would be change, and there would be feedback.

[Elheru Aran]

Don't get too much ahead of yourself. The big barrier with powered aircraft wasn't aerodynamics, it was making a strong and light engine. Materials science will need to advance before you can see serious changes to society.

Likewise, there's not going to be any kind of visual display for some time. The best we can probably expect is some form of printer.

Given the size, complexity, and expense of these machines, I would probably expect them to be much like the 1950s conception of computers-- large, building-size machines capable of brute-forcing calculations and then spitting out data on sheets of paper or tape. Not so much the sit down, turn on the computer and dick about on a monitor. You'll be able to do plenty of equations on it, but you might still have to sit down and render drawings by hand; the only thing that will changed is you'll have numbers to refer to rather than having to knock them off yourself.

[Simon_Jester]

Yes, but you have numbers to support the theory that they'll work. A lot of things got drawn that would never stand up, float, or fly as applicable... because of inability to do even the crudest form of finite element analysis.

It's like, if you could build a computer with all the modern capabilities, but any such computer cost a hundred billion dollars, there MIGHT be demand for one of them. That one would be elaborately time-shared for numerous enterprises... but there damn sure wouldn't be demand for them among anyone except the very biggest national governments and maybe some huge corporations. So it wouldn't change nearly as much as it would if the things cost a reasonable sum of money.

Er... you might want to familiarize your self with a company called IBM. Obviously their setup did not cost as much as your hypothetical but the logic behind it was pretty much what you describe.

Except that the machines they did successfully market were very different than Babbage's analytical engines. In some ways a lot less capable- but cheaper, easier to maintain, and better tailored to what a business actually needs.

Also, they weren't founded until 1911...

[Elheru Aran]

Yes, but you have numbers to support the theory that they'll work. A lot of things got drawn that would never stand up, float, or fly as applicable... because of inability to do even the crudest form of finite element analysis.

Of course. That would provide incentive to improve your materials in order to hit the numbers proposed. However, until that happens, it's strictly theoretical rather than practical. Ironically, pretty much the same situation Babbage was in with the original engines...