kcbrown wrote in post #10559519
Today, it's
accepted that computers crash. Why? Because they just do. People don't know any differently. They think it's
normal. But it doesn't have to be, and shouldn't ever have been. Microsoft, with their junky operating systems and slick marketing, have managed to convince people that it is normal for computers to be unreliable and unstable. That it's normal to have to periodically reboot computers to make them work "properly". When the truth is that those things are necessary because Windows is full of bugs. It is defective, and always has been.
But the truth is that there is no good reason for an operating system to require periodic reboots in order to avoid crashes. How do I know? Because I (and many others, as well) have run my Linux-based systems continuously without a reboot or power cycle for
well over a year at a time. When they crash, it's because of a
hardware fault, and the hardware gets replaced. When they "reboot", it's because the power goes out and stays out long enough to run the batteries in my uninterruptible power supply dry. But they otherwise stay up and running, continuously. It can be done. It
has been done, routinely, even. But people have been trained through experience to think it's normal to not be able to do that. They have adapted to mediocrity and been trained to believe that it's normal and acceptable.
Huh?
I don't accept that computers crash, and in my experience they don't crash almost ever, and when they do it is frequently because of bad hardware, not bad OS.
I have a PC at work that I built myself running Windows 7 and overclocked like a mother. It has been running Folding@Home on all 4 real cores and all 4 virtual ones continually for a month now without rebooting. I use it for work and gaming too, so it isn't just running one app all day. I guarantee that if it weren't for the occasional voluntary updates, like newer video card drivers, the system would stay up for over year without breaking a sweat.
Our security camera PC at work which was cheaply built at every level and is running Windows XP stayed up from November 1998 until last week when the PSU finally gave up the ghost. Simple cleaning and PSU swap and it is right back up and is probably good until this new PSU fails.
Back in the Windows NT 4.0 days at my company we only rolled out the major patches. In one 3 year period in the early 00's between 6 servers running 4.0 we had about 10 total reboots. 6 for patches, 4 for hardware issues. None for software problems. This was completely normal for ANY company that purchased quality hardware, like high-end Compaq servers. Those who bought crappy hardware needed lots more reboots, but that was really primarily a hardware issues, not OS.
We had a laptop sitting in our server closet running Windows XP continually for over a year (a remote user connected to a couple times a week for testing) that never needed rebooting.
Even back in the Windows 95 days we had a few dedicated systems at work that ran for months at a time without reboots, and when they did get rebooted it was usually because some user did something stupid and farked something up that called for a reboot to fix.
Don't get me wrong. Some Linux distro's are more stable than any version of Windows released to this point (if you only bought hardware with good drivers available), but only by a slim margin. The vast majority of the apparent stability advantage actually comes from the manner in which it is used, the kind of people that use it, the kind of software available for it, the low amount of malware written for it, and its tendency to be installed on more reliable hardware. It is kind of like how Apples OSs were always considered more "Secure and Virus Free" than Windows PCs. It wasn't because they are actually more secure, it was because there were so few Apple computers in use that nobody bothered writing viruses for them.
