Hacker News Reader

Comment by cjohansson | Hacker News Reader

Funny that discovering how they did concrete 2000 years ago can be seen as progress..

Well, imagine how much knowledge gets lost through the ages. If there aren't contemporary specialists who value or have enough time to study a technology or a cultural work, it could eventually just disappear. Add to that that different ages and societies build and create from different frameworks, then its conceivable that there are valuable pieces of information that our current society may never uncover or discover, just by virtue of working within a completely different context.

Yeah. It is a mistake to believe that technology can only progress forward. If a generation dies without passing on certain knowledge, it's lost and has to be rediscovered the hard way, and that can certainly happen to technological know-how just like anything else.

It is surprising how easy that can happen. I returned to a client to help with the next generation of a product they were developing. I helpwd with the S/W for the earlier generation. The people working on it when I returned had the source code but none of the supporting documentation. The documentation wasn't extensive and without my help, it might have been difficult to understand anyway. (Or maybe I underestimate the folks working on it at the time.) The bigger problem was that they had no idea what was missing, let alone where to find it. Much of it included descriptions of why things were done the way they were.

On my return, I knew what was missing but had no idea where it was. In the ensuing decade, the company had been through three generations of documentation archives and I had to ask to get access to anything in the new system. I didn't know what to ask for. Further, they gave me a new login and could not provide access to my old email account. (Much of the documentation was shared via email attachments.)

I got lucky and ran across another engineer who had worked on the project and had saved a bunch of stuff. He was now with IT, had helped with the document migration and knew where I could find it. I found that a pretty tenuous way to propogate technical history.

That was over a ten year period and many of the participants were still alive and working there. Any break in the chain and they would have had to start the next generation from existing code and the embedded comments (which would be cryptic if you didn't understand how and why the system worked the way it did.)

This is why even though the US already built a few Saturn V rockets, building the next one would be really hard.

I'm just thinking right now about how we have terrible knowledge silos at work and how screwed we'd be if these people won the lottery. I'm pretty sure NASA had the same problem and we can't rebuild a Saturn V even if we wanted to. Didn't we also forget how to build nuclear bombs or something?

I think you're talking about "fogbank"?

https://news.ycombinator.com/item?id=20115211

No, sadly there's money in building nuclear bombs. But we might be at risk of losing track of where the nuclear waste is, and how to keep track of it. So here's the standard for that:

https://en.wikipedia.org/wiki/Long-time_nuclear_waste_warnin...

There was this about a component called "fogbank" : https://inmilitary.com/lost-technology-a-weapon-so-classifie...

Discussed here this: https://news.ycombinator.com/item?id=20115211

> jbay808: * If a generation dies without passing on certain knowledge, it's lost and has to be rediscovered the hard way, and that can certainly happen to technological know-how just like anything else. *

> cbanek: Building the next [Saturn V] would be really hard

Leading edge technology isn't known by that many people to begin with, so it's easier to lose. Even though the organization that built Saturn V rockets was huge and lasted several years, there are still details that will take, as I understand it, a bit of archaeology to get at. If a technique doesn't become routine, can technology be said to have progressed up to that technique?

I'm not an expert, but from what I understand, the hard part wouldn't be the cutting edge rocket science (choice of fuels, valve design, etc) but the fabrication techniques.

The 1960s had a generation of highly skilled welders, machinists, and sheet metal workers that had honed their craft in the factories of the second world war. The USA today doesn't seem to have the same quantity of skilled workers and NASA has to design parts differently to allow automated tools to do the work.

But those fabrication techniques probably would have been considered routine at the time, although maybe above-average in complexity. So even a routine technology can be, in a sense, lost if it falls out of demand.

Yup that's what curious droid's video on this is saying as well [1]. Basically the plans all were off a little bit and each rocket engine was hand-tailored to the demands of the individual rocket by these skilled people you mentioned. They did document their changes but only on scrap notes that since were lost. The plans themselves that were archived don't contain those notes and even if they had them, you couldn't find people with such skills any more.

[1]: https://www.youtube.com/watch?v=ovD0aLdRUs0

This is part of the theme of Johnathan Blowes talk here was worth watching https://m.youtube.com/watch?v=pW-SOdj4Kkk

Roman engineering was just ingenious, even measured by today's standards. And a big part of Roman domination, especially roads and aqueducts.

that is a little funny you're right. It reminds me of something interesting I read, don't know if it's true or not. I heard that in the "middle ages" (or dark ages) it was taken for granted that the ancients -- meaning the Greeks and maybe even the Romans, not sure -- had been way wiser and more knowledgeable than them. (You know, as they read Euclid's elements and Socrates and stuff.) As I remember it, all the way through the Renaissance that was the thinking, that the ancients had known way more.

Imagine thinking that the ancients knew more than we ever will, and trying to study their works to get at just some of the wisdom they had held. I think few if any people read ancient authors that way today.

however my memory on this subject may be shaky.

You might be thinking about the Sapiens book.

And today, the internet has fuelled the Dunning-Kruger effect so that Joe Bloggs feels free to dump on anyone who isn't in his Facebook group.

There was this excellent video about the misconception that technology and knowledge advances over time.

It does not. It fades, gets lost. Where are the Steve Wozniaks and bill gates? Where are the people truely understanding this stuff. Where are the people who can optimize code like back in the days?

I think you are conflating the notion that, at a given time slice, a majority of the popular sentiment about the future may not have always been positive, with your own opinions of modern engineering.

For vast swaths of human mind, popular opinion tended to see the future as bleak and the past as filled with greater men, accomplishments, etc - the good days are behind us. Post industrial revolution thinking and adoption of science has mostly changed that opinion to hope - the good days are ahead of us.

As for people who can optimize code like back in the day: I am not sure where you work, but I bet if you look at the best engineers in the best tech companies today (google/ Facebook / Netflix / amazon / Uber / palantir) you’ll find a lot to be impressed with.

I think it is more about why it works than how they did it. It’s probably going to be a challenge to apply that knowledge to a modern process that produces the quantity of concrete we need today.

Or it might be the case that now that we know that they needed volcanic ash we are going to figure out a way to produce that stuff at industrial scale and pump tons and tons of it.

And I complain about Lisp...

Reply on news.ycombinator.com