Hacker News new | ask | show | jobs
by aperiodic 5452 days ago
People might have that misconception due to hearing about the Apollo flight and guidance computers, which were pretty janky. They had a really difficult to use interface, and used core-rope memory[1] that was woven together by old ladies.

[1] http://en.wikipedia.org/wiki/Core_rope_memory

4 comments

"Janky"? You're right, they should have used Ruby on Rails. And, ermm, AJAX 2.0.

I'm sick of people slamming old tech with 50 years of hindsight. They were the smartest people in the world doing the best work they could with what they had available to them. And that's what they came up with - and it got the job done.

And those IBM mainframes are massively more reliable than the crap (software and hardware) we use today.
Which made for hellishly long compile times. Core-rope memory was used because of the weight and volume savings. On Apollo every pound counted. If you look at the Wikipedia article you find a link to HAL/S which used to program the shuttle computers. A similar language (the same?) was used to program the Apollo computers and I believe the ground support computers. I do remember finding 3 punched cards per statement to be a bit bizarre. Yes, I was there at MIT Instrumentation Lab, but some of the details have become hazy. :-)

Edit: The compile times for core-rope memory were pretty long, so there were lots and lots of simulations. Going to core was an infrequent occurrence. The computers had a small amount of conventional core so that patches could be applied. Bug fixes half way to the moon were possible.

> They had a really difficult to use interface

AFAIK their Noun-Verb keyboard interface was deemed quite novel and particularly easy to use at the time.

good lord the best part: "Some NASA programmers nicknamed the finished product LOL memory, for Little Old Lady memory."

normally I don't post this sort of thing on HN, but LMFAO.