Yes, I know the difference between kernel updates, which do require restarts, and application updates, which (generally) don't. Moreover, you don't need to comment on my lack of awareness re: software programming. Thanks, but no.
In any case, I was simply talking about focusing on each active program in userland and requiring the user to terminate that process. In other words, save or quit. It's not hard; it wouldn't be complicated. Hell, query the last input; if the user hasn't been active in n seconds, go for it, sure. If s/he has, recognize that someone's there and allow them to wind down their workflow.
The ticking timebomb countdowns signaling restarts, which I experience regularly at work (yes, I blame the IT department) are unconscionable.
If you give users the ability to go back and revisit a program with the idea of letting them save their work and quit, they will, without a doubt, abuse that privilege and pretend that they don't have to restart.
You put a time limit on that and threads like these will still exist, just instead with people complaining that Windows only gave them ten minutes to save everything before restarting.
Plus, if you do the thing where you check if the user has been active for the past several minutes, you'll have the same people saying "I left my computer on with my work open and it restarted wtf!!!!!" because you can't assume a computer not actively being used is one you can safely restart. (It's why they introduced the active hours thing--that's a far better way to judge if a computer's safe to restart because the user can define when not to restart the computer, but they're not allowed to set the entire day).
Really, complaining about the ticking timebomb--all software on the scale of Windows where most of the users fail to comprehend on how it works and how best to organize workflow to both stay secure and at high efficiency, you
have to make the assumption that the user cannot be trusted and must be given as little ability as possible to create loopholes around stuff like this. Heck, I work at a division of IBM where we handle server software, and over the last few weeks we've had division heads continually reminding people saying "hey we still have servers with SMBv1 enabled why is this happening we thought we told you guys to disable it weeks ago" and
those are people who write server management software for a living.
Like this is why I questioned your familiarity with software programming, you can never trust the user to work with you and install updates when they should or not insert sql commands into your online forms, etc, etc. That is something you learn very early on.
Is that really so much to ask tho? It's so nice to has machines that can stay up and update without restarting for as long as I want. (Which has been up to a few months at a time.) I'm used to using them without those kind of interruptions now. Seems like a basic feature.
I don't think there's a single OS out there that can update every part of itself without restarting. Google's introducing something in the next version Android that might be able to do something like that? Not sure.
If you choose to get Windows 10, it's pretty much your own fault. I'll stick with Windows 7 as long as possible for instance and will see what I do afterwards.
the win7 stans are almost as bad as the winxp stans back when win7 first hit the scene. fine the last windows release was a little rough, but refusing to update now that there's a better windows version out that brings forth updates you probably actually would make use of doesn't make you smart, it just makes you stubborn.