Does anyone have experience finding what kind of objects are taking up the most memory in Python from just a core dump (and the executables of course)?
My point here isn't that rebooting a machine is a bad troubleshooting step, it's that rebooting a machine and not investigating the issue further is unacceptable.
@ajroach42 Before that, in my memory the most common failure mode of malfunctioning software was "computer becomes totally unresponsive", in which case, what other option do you have?
@ajroach42 This was definitely the normal everyday Windows 95 / Windows 98 user experience I remember. You could often keep using the system after a BSOD but it was rarely a good idea. Not to mention the forced (& sometimes multiple!) reboots after every program / driver install. I used to reinstall Windows overtop of itself every six months or so when it became too unstable to use and it usually helped.
this behavior is decades older than that, going back to the beginnings of the home computer revolution when almost every computer had a momentary-press reset button specifically for taking the computer back to a known good state. it just got more inefficient when manufacturers started leaving off reset buttons and forcing users to twice-toggle the power switch instead. the fact that using the reset button preserved no data for troubleshooting implied that none was necessary; best to move forward rather than look back.
At what point in the history of computer science did "I don't know what causes it, but a reboot solves the problem" become an acceptable answer that merited no further investigation?
People in this case means me and some co-workers. This in this case means throwing axes at planks of wood such that they embed themselves in the center of that wood.