I found the device computing schedule to be a pain. Not only can you only compute during the schedule, you can only call in and send a result during the schedule. It also closes down at a particular time, and some computing can be wasted before a checkpoint comes along. Many workunits have checkpoints every few minutes, but I have seen them as long as 90 minutes apart on a FAST machine. That could equate to every 3-4 hours on a slow system. That means that up to 3-4 hours of crunching could be wasted each day on a part-time schedule. That portion would have to be repeated the next day. Also be careful that the workunit is completed before the 2 week timeout period.

I found it preferable just to leave everything at 24/7, especially since the weekend schedule could not be set differently.

My computers have evolved just to be left on 24/7 anyway. I run with screen saver (none) and let the energy saving feature turn off the monitor after 20 minutes or so.

If you need to quit, simply turn off the computer, preferably just after a checkpoint so as little as possible will be wasted. This can happen during a lightning storm or some other flexible thing like a vacation that the fixed schedule cannot account for. Otherwise, just let it crunch. In an emergency, just quit as-is and let a little work be sacrificed.

(The checkpoint on FightAids@Home is when the green line ends and starts over. The checkpoint on Rosetta/HPF is when the percentage on the detail display jumps up a bit. This could be after a single pass of the Rosetta lines on a fast workunit. I have seen 15 and more passes on a slow workunit before the percentage jumps. Don't depend on the overall integer percentage on the main task screen. It usually only goes up a few tenths each time.)