The thing that strikes me as funny about this, technical-background-wise, is that the lessons learned have basically been common software dev knowledge since, good gods, the early sixties _at least_ .
The whole story reads like the familiar story of everyone who's ever had to work with vertical software in general, in which due to limited markets, low budgets, and lack of alternatives it - well, it's all cheaped-out pure shit, in a nutshell.
Which leads me to wonder - anyone know if Litton were on a fixed-price or cost-plus contract?
Well it could be worse. It could be like those German warships that still use floppy discs
I still have both 3½ and 5¼ drives squirreled away in my office just in case...
I hear they are planning to replace the floppy drives with solid-state floppy emulators
Way back in the day, I had a contract at a place which had some business-critical control software which only ran on the 80s-era BBC Micro. Which they had been running on an old PC, using a very specific BBC Micro emulator (that could tunnel Beeb hardware-fu out using serial protocol to a hardware widget made by an out-of-business company) But the old PC had died, and they needed someone to make it work on a new one.
Guess what I left them with?
If your answer was "a new PC running an emulator for an old PC running the Beeb emulator tunnelling hardware-fu over serial over USB to a USB serial emulator to the hardware widget to the device it was supposed to control", you win a hearty exclamation of DEAR GOD HOW COULD ANYONE THINK THIS IS A GOOD IDEA.
(The amount of industrial machinery I've seen controlled by things painfully similar to this particular stack of crazy really make me wonder how the hell we succeed in manufacturing anything at all.)
