There's a lot of practical/contextual knowledge that isn't documented, and what's documented isn't guaranteed to be correct or can be misleading (terminology changes, standards change, some things are implied because they are obvious at the time, etc.)
Though knowledge of experience is sometimes wrong as well. Sure it works - unlike print where mistakes (sometimes intentional to hide a secret) can tell you something that cannot work, but that doesn't mean it is a good way to do things.
Give some examples? That kind of 'undocumented' knowledge is lost because better tech has completely replaced it. I'm talking about inventions like the Spinning Mule or the steam engine. We have superseded them with electric motors and logic boards.
First thing that comes to mind is friends father was hitting retirement age after working at a petrochemical/plastics facility for ~20 years, was in charge of maintenance of some section. I think he told me the owner had to call him twice to help diagnose problems that were causing product outages.
These things don't get built on a whim - there's risks, regulations, documentation, procedures, experts, etc. At the end of the day you have people doing the work for decades, with an intuition about how things work.
Given infinite time you can recreate anything - but by the time you're done putting the puzzle together you're out of business.
>At the end of the day you have people doing the work for decades, with an intuition about how things work.
That's exactly my point. This intuition is undocumented however it becomes superseded by new tech. Nobody can (successfully) run a plastics plant with decades of old hardware knowledge and expect to be in business another 20 years.
"In 1991, Aldo retired and Ed took over day-to-day management of the company. First Plastics began the initiative to replace and update its molding equipment with new large capacity presses in 1995. This initiative has placed First Plastics as one of the most technologically advanced injection molders in the Northeast. Today it continues to add the latest technology while increasing its manufacturing capacity."
We can't build the engines for the Saturn five rocket anymore though the engineering documents still exist. You need the documentation and skilled craftsman.
About 20 iterative engine designs from the Rocketdyne A-1 to the F-1, building experience and skills along the way.
It's probably a myth anyway, Rocketdyne's expertise continues with the RS-68 and there are probably still employees that worked on the RS-56 that was derived from the H-1, which was essentially a smaller F-1
edit: Apparently, an upgraded F-1B is part of the SLS Block 2 spec anyway, so yes, they can still make them.
You can't even document computer code accurately. And neither can I.
Gordon Ramsey can't document how to cook a steak and how to tell thats its done. 3 different cooks will read his instructions and will end up with 3 different steaks.
The idea that a lifetime of steelworking experience can be foind in books, especially by people who aren't known for love of books, is pure folly.
I call BS. Code and any process can be documented just fine.
> 3 different cooks will read his instructions and will end up with 3 different steaks
Either his instructions are shit, or the cooks are donkeys.
Sear steak at high heat, 1 minute, rotate steak 45 degrees, flip steak to sear back side for 1 minute, rotate steak 45 degrees. Move steak to indirect heat until internal temp probed at center of steak reaches 145F, remove steak and plate steak after resting for 4 minutes.
There. If you follow those instructions you can get a perfectly done steak every. single. time.
Shit is only subjective if you don’t have a repeatable way of measuring things.
You absolutely can learn steel working and foundry type of stuff from books, but you have to find ways of measuring what you are doing. E.G. heat to this temp, control cooling process by means of $X, $Y, $Z. Inspect crystal structure of coupon, X-ray for porosity.
> Code and any process can be documented just fine.
Can be? perhaps. Is documented? definitely not.
90% of code is not documented, whether it is a trojan or firmware for a soda dispenser. If I find you a random piece of code that was not maintained for the last 5 years, in a random language, what is the chance that you can get it to work? 30%?
> It isn’t easy, but it can be done
So most of the time it isn't done.
Its software engineer's job to document code, it is not the job of a machinist to document minitua of his job. In fact some proffeshions made historically a point of keeping their craft secret and passing knowledge on to an apprentice.
> Code and any process can be documented just fine.
