First of I run linux on my personal machine.
Second, I shut down my work machine at the end of the day and if there is an update - let it update. The result? Not a single problem with windows updates in years! Strange, I know.
Sidenote: I always thought people were partially making fun of windows updates because you have to reboot all the time. I have to log out to switch from integrated to dedicated graphics in Linux and pretty much 90% of all updates require a reboot. And to conserve battery I have to shut down the laptop anyway, since hibernation is but a dream. But whatever, it’s not a competition.
Windows Update is just dogslow and forces a reboot. For me even a significant distro update takes much less time and it doesn’t force you to reboot (nor to update for that matter).
I don’t have to reboot after an update very often, almost never really. It’s kernel updates where I have to reboot, other stuff I can restart and avoid it that way if I still want to keep the pc on.
I know on some systems hibernation (suspend-to-disk) can be fiddly. For me it worked out of the box which was nice.
When you have a nice setup in programming (compiler, database, diverse docs, shells etc), you don’t want to shut all that down. If you can, good for you!
My dev VM is almost entirely disposable. Could be up and running again, fresh in 30-60min, not counting time to pull the repo. Why use a local db server? Seems weird to me but, I came to development through SysAdmin and support stuff, so, was used to not owning the machine that I was on. That probably has heavily influenced my workflow.
Out of curiosity, would you mind sharing a bit any the languages/frameworks and workflows that you are using? I’m mainly using Go, C++, Python, and a few others and just having trouble figuring out how I’d arrive at a situation like that. No CI/CD and test systems?
Well if I shut down visual studio, it takes some time to relaunch it, it uses a distant unix server to compile.
I usually have a bunch of explorers open on distant repos for checking traces. Some soft connected to a database (with tables open), shells open on servers, or inside a docker on that server, all that goes away at reboot.
Nothing crazy, it’s just convenient to just continue working instead of having to set it all up in the morning.
CI/CD, thats for integration and should IMO be on a server somewhere 😁 not on your PC that you shut off in the evening!
I do mostly C/C++, linux/windows. Database, gui, etc.
