That's from the Internet Archive, uploaded to YouTube with ads enabled. Internet Archive version, with slightly better audio: [1]
The Linotype is dates from 1885. The basic design didn't change much for 90 years. It wasn't the first typesetting machine, but it was much better than its predecessors, and nothing that came later was significantly better. Until typesetting became computerized.
A surprisingly small number of people designed the few mechanically complex machines that worked reliably. The Linotype was designed by one of the greats of mechanical design, Ottmar Mergenthaler.
Howard Krum designed the Teletype. People had been trying to build printing telegraphs for half a century, with limited success. Krum got it right, and the basic mechanism didn't change from 1926 to 1958. A Teletype decodes serial data entirely by mechanical means, and Krum figured out how to make that work reliably.
William Burroughs built the first good adding machine, in 1884. There had been adding and multiplying machines for over a century, but his had reliability. When you pulled the handle, you were putting energy into a spring, and the release of the spring, slowed by a dashpot, powered the mechanism consistently. Earlier machines could get wrong answers if operated too fast. He also came up with the mechanism which causes adding machines and calculators to jam rather than produce wrong answers as they wear out. There's a bail that drops into a slot in the number wheels at the end of a cycle. If the number wheels haven't made it to a valid position, as can happen with a worn-out machine, the slots and bail won't line up, the bail won't drop, and the machine locks up.
Henry Maudslay built the first good lathe in 1787. It's in the Kensington Science Museum in London, where I've seen it, and it looks like a lathe from the 20th century. He got the design and feature set right, and that basic design persisted for centuries.
Few people are really good at this. It takes not only 3D design ability, but insight into how to design so that machines wear well.
Often the wear-well factor came thru iteration rather than out of the box - AT&T was famous for doing this, because it did its own remanufacturing and recycling in house - it had literal feedback on how its equipment wore in service - beyond the indirect feedback it got from the field forces doing the repairs.
Mechanical machines are really satisfying, for their tactileness and simplicity. I have a typewriter, watching real impressions appear on real paper is way more fulfilling, and the organicness of the interaction is not even nearly matched in any computer program I used: you're always detached from the circuitry, where with mechanical devices you're a part of it. But the advantages of computerised production far outweigh those of mechanical devices, so I only type on my typewriter every once in a while as a recreative activity.
I too have a typewriter at home. I don't use it often, but when I do it's very satisfying for some reason.
I find it to be a pretty useful tool as well. I like to type out initial contact or offer letters for potential clients. In a digital world, receiving something that was typed really says something - or at least I've been told. Every key stroke is done with thought, as one screw up and you'll have to start over.
Ohhh yeah, we lost a lot of things: We lost all those noises that Teletype did that drove anyone crazy.
Without the Linotype we lost all the toxic fumes that killed people, and the accidents with molten lead being expelled and burning people.
We lost the exclusivity factor of only a few companies being able to spend the enormous price those things had and the many people that were needed to repair the machines, write, physically send and receive messages.
But we also have lost the ability for regular people to look at the machines they use every day and understand them purely by inspection, with no “black boxes”. That’s no small loss.
While we can now use computer controlled machines to cut materials at higher precision and into trickier shapes than ever before, mechanical knowledge and skill is in the hands of a smaller proportion of the population. There has been a significant erosion in the broader society’s skill at physical/mechanical reasoning and invention.
We have amazingly precise calculators and displays but the elimination of slide rules and nomograms and physical measuring devices and drafting tools and the steady reduction of geometry in school curricula mean that geometric intuition and insight into the numerical relationships involved are on the retreat.
Instead of inventors, machinists, assembly workers, or repair technicians we are left with a society of salespeople and paper pushers, with the occasional computer programmer or statistician.
Even folks working in architecture, industrial design, mechanical engineering, or similar careers today spend most of their time sitting and staring at a screen.
There has been a significant erosion in the broader society’s skill at physical/mechanical reasoning and invention.
This used to be a problem with education in some countries other than the US. Now the US has it, too. Back in the 1980s, I was talking to a mechanical engineering professor at Stanford, and he was grumbling that many Indian students at Stanford had never used a screwdriver, and thought it was beneath them to use tools. That was something for the lower classes.
Now the US has that mindset. Too many Americans have never even been inside a factory, let alone worked in one. They have no idea how stuff is made. No sense of the basic manufacturing processes. No sense of how a factory works. This is a problem, and part of the decline in US manufacturing.
The last Tesla discussion showed this. Many of the people commenting didn't realize that the problems Tesla is having are normal factory startup problems. Tesla management thought they could take short-cuts and escape those problems as they scaled up for the Model 3. That didn't work out as well as expected.
We went through this with the 3D printing mania a few years back. There were people claiming you'd get a custom-made fender 3D printed at a body shop in a few hours, and this would be cheaper. No.
There are some very cheap, very effective standard processes for making things in quantity, and if you can design to use those processes, you'll get a lot of parts cheaply. Here's a video of one process, progressive stamping. [1] Complex metal parts are made from sheet metal this way. The dies are complex and expensive, but once it's working, each shift bangs out about 25,000 parts with minimal human attention. After watching this video, look at some bent metal parts around you and try to imagine the bending and punching steps that made them. Only a tiny fraction of the US population today even knows progressive stamping exists.
While every day I use machines I can't easily inspect I also have access to a wealth of very easily accessible information about all manner of physical/mechanical things, plus the ability to order just about any tool material that can be dreamt of.
On a personal level, I'm about to embark on a very ambitious DIY project that I doubt I would have the confidence to undertake if I didn't have access to thousands of youtube tutorials and an internet's worth of material suppliers.
It’s easier than ever before for someone enthusiastic to learn just about anything... and fewer people than ever bother. At least that’s my impression.
Now that's true. In the 1990s, to build electronics, I had to establish a credit account with Hamilton/Avnet just to order parts. Making a PC board at least used CAD, but AutoCAD, not an electrical CAD program. The files went to one shop that generated Gerber films, which had to be sent to the board fab shop, with about two week turnaround.
First we had people who raised their own sheep or grew or scavenged their own plant fibers, spun their own thread, made their own cloth, sewed all their own clothing together by hand, etc., and spent a significant proportion of their lifetimes on clothing production.
Next we moved to industrial production of cloth and mass adoption of sewing machines, so people could make their own clothing from purchased cloth, or hire someone local to sew the clothing.
Next we moved to industrial production of clothing, but most people could still alter and repair clothing or hire someone local to alter or repair it.
Now we are at the stage where clothing is so cheap that most people have no idea about any of the process, and buy masses of disposable clothing which they mostly never wear, donating it in bulk or throwing it out when they notice it’s been sitting in a closet for a decade.
And the same is true for housing, agricultural products, food, transportation, furniture, containers, toys, books, artificial lighting, music, storytelling, financial record-keeping, electrical gadgets, ...
But those final stages of transition to pure consumerism are I think the most alienating, involving not just lack of personal skill but lack of even the most basic concept of what goes into the products we use every day, or any understanding of the supply chains and production processes involved, or even a comprehension of how the products themselves work.
Heidegger himself is too dense and has alternate interpretations ("Tool-being" by Graham Harman has this whole emphasis on the tool-analysis that reminds me of the vibrations in Buddhist dharma), but Herbert Dreyfus' "Being in the world" emphasizes this side of "Being in time".
After "Being in time" and his brief flirtation with Nazism, Heidegger actually made his project to write a metaphysical history of Being in stages much like those you described. That too is dense and somewhat requires an acute sensibility for German lyrical poetry...
Anyway, "stockpiles of Being" is a good google keyword I venture...
And in return we have gained landfills of perfectly functional devices that no longer can be sued because some "genius" decided to replace some framework or other only a few years after the devices were shipped...
I knew a proofreader who had been a Linotypist until one day he bent over to pick up the next magazine and could not straighten up. They led him off to the infirmary, still bent at 90 degrees, and when he came back from medical leave they decided that lifting magazines should no longer be part of his job.
I was thinking in a more general sense, and about how we keep seeing a churn in software "solutions" were as the parent comment pointed out how once a mechanical solution was found it stayed in use for decades.
The documentary covers the last day that the NYT used a metal/linotype press and talks through the entire process that happened from newsroom to printing. It is a truly interesting way to spend 28 minutes.
On a more serious note, I think a lot of people today have a hard time believing that specialized machines as complex as this were ever built. It's like looking at one of those pocket mechanical calculators.
I'm not sure how they feel about drop ins these days, but they usually take part in the Open Doors Toronto days.
They still have an operational Linotype, or at least they did 8-10 years ago. I don't know if old Stan Bevington is still involved much. They used to show guests how to operate it. That and an old fly-wheel platen press.
edit: It looks like they've opened a bookstore there now. They also used to do a fall and spring launch. The fall launch they'd host at the shop and serve beer. The spring launch used to be held at a venue of some kind. Not sure about it anymore -- it's been a while.
Keep an eye out for their fall launch. They let you run the Challenge Gordon platen press to make a print. They don't let you set any type because it would be pretty time-consuming, but they do let you get close and personal with a case of lead type and a composing stick.
I think they ran the linotype when I was last there, made a row of type, or let you do it, and then tossed it back into the molten lead again.
Just make sure to catch up on some of the poets or authors they've published. You won't have to go far. Then again, IIRC Stan was not as much the literary type as he was the technical-mind. He's the founding master-printer. You can usually find him in some pub around the annex drinking too much Guinness.
I have Canadian citizenship, so I may just have to check it out. I've only been to Toronto a few times. It'd be a fine excuse to go back. My other excuse is watching the Jays play.
I grew up inside a small print shop of my parents where we got to have 3 linotypes. I never learned to work on a linotype, but at certain time, my job was to melt lead lines to be reused as bars.
I just lost an hour. Thanks that was really interesting. The New York Times video in the thread was exactly what I was wondering about. How the switch happened.
If you want to learn more about this, "Graphic Means" [0] is an excellent documentary about pre-digital graphic design production that also covers the Linotype era.
The Linotype is dates from 1885. The basic design didn't change much for 90 years. It wasn't the first typesetting machine, but it was much better than its predecessors, and nothing that came later was significantly better. Until typesetting became computerized.
A surprisingly small number of people designed the few mechanically complex machines that worked reliably. The Linotype was designed by one of the greats of mechanical design, Ottmar Mergenthaler.
Howard Krum designed the Teletype. People had been trying to build printing telegraphs for half a century, with limited success. Krum got it right, and the basic mechanism didn't change from 1926 to 1958. A Teletype decodes serial data entirely by mechanical means, and Krum figured out how to make that work reliably.
William Burroughs built the first good adding machine, in 1884. There had been adding and multiplying machines for over a century, but his had reliability. When you pulled the handle, you were putting energy into a spring, and the release of the spring, slowed by a dashpot, powered the mechanism consistently. Earlier machines could get wrong answers if operated too fast. He also came up with the mechanism which causes adding machines and calculators to jam rather than produce wrong answers as they wear out. There's a bail that drops into a slot in the number wheels at the end of a cycle. If the number wheels haven't made it to a valid position, as can happen with a worn-out machine, the slots and bail won't line up, the bail won't drop, and the machine locks up.
Henry Maudslay built the first good lathe in 1787. It's in the Kensington Science Museum in London, where I've seen it, and it looks like a lathe from the 20th century. He got the design and feature set right, and that basic design persisted for centuries.
Few people are really good at this. It takes not only 3D design ability, but insight into how to design so that machines wear well.
[1] https://www.youtube.com/watch?v=jcp60z0E3ls