James Watt: A Twelve-Year Flash of Genius

A good place to start is 1825, with the Stockton and Darlington Railway. Every railroad before that was a test or demo, but the Stockton and Darlington was 25 miles long, had several locomotives, and carried useful loads. They still had to use cable systems to get up some hills, though. Think of that as the public beta of railroads. The Liverpool and Manchester Railway (1830) can be thought of as the first production version - everything was moved by steam locomotives, and for the first time, there were tickets and schedules.

That's a key moment in history. Up until then, technology was a niche item - clocks, a few clunky steam engines here and there, some textile machinery. The life of most people hadn't changed much in the previous thousand years. Most people never got further than 50 miles from their place of birth. Suddenly, in 1830, that all started to change. Fast. Over the next 30 years, railroads went everywhere. Ordinary people could travel great distances. "Railroads will only encourage the common people to move about needlessly." - Lord Wellington.

Next, steel. Read the history of the Bessemer process. Until the 1860s, steel was an exotic material, used mostly for knives, swords, and some gun barrels. The early industrial revolution and the early railroad era were built with iron, not steel. It's amazing how much was done with really crappy metal. Cast iron is strong in compression but weak in tension, and brittle. Wrought iron is reasonably strong in tension but easy to bend. Steel is isotropic, equally strong in compression and tension. The railroad industry adopted steel rapidly, and locomotives went from wimpy little things to monster machines. Other steel processes obsoleted the Bessemer process, and steel became widely available and cheap. Steel is about $100 a ton right now.

The history of the rolling mill is also important. No more banging out sheet from hot bar stock on an anvil with a hammer. With multiple stands of rolling mill, ingots went in one end and miles of thin sheet came out the other. It took a while to get that working right, but once it did, sheet metal was everywhere.

Then, machining. Few academics write about the history of machine tools. Lathes and drills go back to antiquity, but a lathe was not a precision machine until Maudslay built one in 1800 with a slide rest and leadscrew. He also came up with gauges and techniques for precision turning. This was the beginning of precision machining. The metal planer was developed around 1810. Now you could make precision round things and precision flat things. Casting was well understood, so you could make arbitrary shapes by hand-carving a wood master, using it to make a negative in sand, and pouring in molten metal. Casting isn't precise and won't get you a smooth surface, though; if you need a precision surface, you have to follow up with planing or turning.

Look at machines made in the 19th or early 20th century. You'll see cast or hammer-forged parts with a little finish machining, and the machining will almost always be a flat or circular surface. This is why steam locomotive parts look the way they do.

The general-purpose milling machine wasn't developed until 1932. The Bridgeport milling machine was such a great design that it's still manufactured and widely used. Now, you could make a huge range of shapes on one machine.

Along the way, cutting tools got better. You usually cut steel with steel, which requires clever metallurgy, cutting fluids, edge treatments, suitable speeds and feeds, and other details people don't think about much unless they do machine shop work. Getting that all worked out took decades, and all the details are in a thick book called "Machinery's Handbook", which can be found in any machine shop.

The first half of the twentieth century was when most of the clever tricks for making stuff in quantity were worked out. Stamping, progressive stamping, the automatic screw machine, the four-slide machine, and lots of other techniques were developed for making vast numbers of identical parts. For the first time in history, the ability to make stuff outstripped consumption. This is a key point. Throughout all of human history until then, the big problem was making enough stuff. During the 20th century, that problem was solved in a big way. Writers in the 1920s commented on the deluge of banal objects. (Sheet metal stamping had really taken off, and vast amounts of cheap decorative crap stamped out of thin sheet steel were everywhere. See a 1920s Sears catalog for a good selection.)

That's enough to give a sense of where to look for background. I'll stop at 1950, although manufacturing technology certainly hasn't. I've ignored the whole energy side (electricity, oil, etc.) because that's better known and well covered in Discovery Channel shows.

I find the Google play policy on books weird.

I can download all the apps i like for free, but to grab a free book i suddenly either need a gift card or a credit card.

Not sure if the Korean peninsula is the best example, given the massive trade embargo going on.

Frankly we may be looking at a variant of learned helplessness on a national scale, in that if NK leadership is faced with a dilemma they can throw a proverbial tantrum and get the rest of the world to step in to calm them down "before the nukes start flying".

That is, the peninsula did not develop in a vacuum. they were part of a larger power struggle between ideologies.

I recall a interview in recent years with James Burke, about his Connections tv series.

At one point he mentions that he basically had to ignore certain aspects to fit things into a clean narrative, as things were just so darn interconnected.

And even then you get some inkling of that in some episodes, where he basically has to tell the viewers to keep some event in mind while he rewinds and starts on a different branch that will eventually reconnect to the event to drive the story onwards.

And from a technical aspect you can sometimes see this in old telecom setups. Where now retired box A was used to bootstrap box B, that then bootstrapped box C, that was used to bootstrap box A+1. Except that if you then have a complete power failure you find that you can't bring up any of it because none of them can be booted on their own.

Makes me think of my musings on corners cases in computer programming. As of late it feels like whenever a corner case is "fixed", a couple of new ones sprout at the edge of said fix. End result is something akin to zooming in on a fractal, and finding ever new ones.