Was FORTRAN buggy 4333

Yes, I understand this. That's why the Jupiter was an emulation.

Right. However, I keep running into people who believe that the emulator is always equal to the real thing. There were people who believe the Jupiter existed becuase they "saw" it run. I read a post from some guy in the military who believed that certain weaponry existed and worked because it worked on his TTY screen.

Do you understand why I'm objecting a tad here? An application is NOT the gear. There exists physical laws that cannot be broken with a piece of code.

Right. However, this can lead to complacency if one person in the group is not vigilant. Look, I am not saying that this kind of process is bad; but look at Charlton's atbreastude again. I'm buttuming this is a window into the group's approach. They aren't paranoid enough. :-) That's why I'm poking him.

Some days you can only use bandaids because you haven't found the real bug. JMF did this a lot. Then years later while working on some code, he discovers the real problem, figures out why the bandaid worked, and then fixes it while he doing the development work. I'm sure TW did similar things; I just never heard about them as much.

TW had to put a permanent bandaid into TOPS-10 because BLISS-10 had a bug but it was impossible to fix it (where fixing included getting the fix out into the field to all sites). It's still there and will never be taken out.

JMF had to put a permanent bandaid into the Alpha code because BLISS-nn had a bug. Even if the compiler people agreed it was a bug and a fix should be done, the fix wouldn't get out until two years after his Alpha code had to be shipped. Now think about this one carefully. His bandaid had to work with broken compiler and it also had to work whenever the code got fixed; and he couldn't know how that bug would get fixed. So, how do you code a workaround that will continue to work after unknown code fix has been shipped?

He spent three days on that one.

BAH