I was here about 9 years ago. A lot of my work units had errors, so I had to stop processing work units for CPDN. I was able to contribute to other BOINC projects.

I've since improved cooling on my computer, so it can handle more processing. (BTW, I do not overclock.) But, the main reason for my high error rate was likely that I turned my computer off every night. CPDN wanted users to leave computers on 24/7. My question is this: Is that still the case? Will shutting down my computer every day cause problems using BOINC for processing CPDN tasks?

If they're just turned off, yes.
For this project, they need to be shut down gently; my method is to Suspend each task, wait until they've had a chance to stop and save all the files, then shut down BOINC, then turn off the computer.

But there hasn't been any work for months, and the last lot was for Linux and Mac.
Until the Linux failure rate got too high, when those were removed from the queue.

The next lot may be a new type of Linux model, with massive resource requirements.
But that may be a long way off yet.

Thanks.

I do use Linux.

When new work units become available, I'll wait a couple of months before attempting to rejoin.

I don't remember what I did 9 years ago when shutting down my computer. But, in recent years, I've used the BOINC Manager GUI to suspend each project before shutdown.

Suspending a task makes things better sometimes, but there are still plenty of tasks that fail on task suspend/resume.

My advice would be to not try and run CPDN tasks on something that can't run the tasks uninterrupted from start to end - though this is from the process point of view.

You can sleep/resume the system just fine with CPDN tasks. I do this regularly - a bunch of my compute boxes are solar powered, so they run during the day and sleep at night (though lately I can leave one running 24/7 out there). Sleep/resume of the system doesn't bother the tasks. But if I try to suspend a task, instead of the system, it's about a coin toss if it will come back properly or crash on resume. There are a lot of workunits I've seen with suspend/resume errors in other tasks from the WU.

Thanks. In that case, I'll not try to rejoin.