Dear Chuck, your message made me think of another item some of you might enjoy.
I was a phd student at Brandeis doing a huge model of a planetary nebula which took three hours on the "big" ibm machine at MIT (1964) I could only get access once a week at midnight on friday or saturday night. But also becoming the graduate student at Brandeis doing programs for the physics professors at Brandeis, i was at mit quite a bit for seminars and such. One amusing story that set a future tradition in softwareD development was when the big IBM batch multiprogramming operational system was when it developed a bug where if someone left off the special end card that had to be added to your fortran deck, it would wipe the memory of the system and cause a need for a restart. It was report as a standard bug to IBM but a month later it happened again and someone at mit phone ibm and complained about the bug again and asked why didn't they fix it. The chap at ibm said to go look at a recent page in the current volume of fixes of which there were many. They did and found: "As an added advantage if you wish to clear memory in the computer leave off the end card from Fortran program."
This became common practice in a lot of complicated software systems where bugs were redefined as features!