I might suggest that 30+ years of working with computer operating systems and programs might make me somewhat experienced. I was a beta tester for many years so I know more about bugs and changes that break things than a lot of folks, I would think. I admit, too, that there have been some problems that I have not been able to figure out. That said, I'm confidant enough that once I get a system built and all the software configured it works until someone changes something and if that happens, almost every time I was able to fix it with a software change by either uinstalling, rebooting then reinstalling the buggy software or figuring out program A does not play well with program B, and showed the user it was better to use progam A with program C instead.
Here I am in this thread simply saying I was lucky this time. What's the crime in that? Do you hate it that some people are lucky? Do you assume you know much more about operating systems and programs than I do? Do you assume I'm clueless? It matters not, but I do think I may be just the exception that proves the rule.
And as far as knowing it aint broke: If I can boot the machine day in and day out and it boots the same way successfully and I can use the programs I use all day for 12 to 16 hours a day and it all works, then I turn it off at night and turn it back on the next day to do it all again without any errors. That said I've had programs freeze on me. A reboot has almost always fixed whatever glitch occurred.
I do know how fragile a software/hardware plaform can be. Memory is not error-correcting; memory is unified and the OS does not run in it's own memory space like a mainframe does, so no program can crash the OS since it can't even seen the memory the OS is running in. Even the most critically tested and debugged software in the world has shown that a bug can still hide. The example of that is Apollo 11. When the lander was descending, the radar was not working for some reason. With great trepidation, the astronaut turned the breaker off for that radar and back on again. The radar then worked. Afterwards sofware engineers determined that a bug had the radar focused at infinity and stuck like that and that is the reason that the radar was not giving the critical distance detection they needed.
Another thing, if I can leave a system on 24/7 for months on end and nothing crashes, can I say it aint broke? There are millions of servers around the world that run like that for years without ever being shut down and they run reliably. Hardware becomes so obsolete that the systems, when they are shut down, are ripped out and new ones put into their place.
Of course, personal computers are a different story. Cosmic rays can flip a bit in a memory chip on a memory module. If that bit is part of the OS, it can cause a blue screen. Only a lead shield would protect a computer from cosmic rays. Some hardware issues are with systems running at such speeds that signal crosstalk exists on a motherboard and just every once in a while one bit is crossing and another trace picks it up and flips a bit on an unrelated memory path. Some bugs like that can't be found easily. In software/hardware testing some types of bugs with a known signature don't show for 1000 hours of running. You can reliably trigger the bug, but it takes literally running a script for 1000 hours before its triggered. On a home computer you get a freeze, crash or whatever and you think it's magic.
I have a four year old cell phone. It froze on me once. It's been on 24/7 for all that time, otherwise. It's software in that phone. Can I say that it aint broke?
If I can run Prime 95 24 hours, most folks would say a system is stable. Would you agree to that? Would you say it aint broke?
What exactly makes a system unreliable? It's not magic, its simple yet its frustrating. It's only data being passed from one device to one chip to another to the processor and back to memory then back to another chip that back to the device. Data runs all over. It's all ones and zeros, but it's all there is. Yet a computer can amaze me even today, when I can fix it if the hardware fails, down to the component level. I mean, even removing surface mount parts and replacing them. I did it for a living.