For the past 29 hours, my United Devices agent is stuck at 71%. It says there is no current molecule to display hasn't moved except for the time.. I tried restarting once but it just went back to the last save... I need to restart my comp soon but that's 29 hours down the drain

That is Exactly the same problem I had. What I had to do was use the "panic button". What you need to do is leave your UD on then end the task (Ctrl,Alt And Del) called ud_#####. (# means there will be a number there) It won't be able to connect to the server and will download you a new task and your on the way.

But my hours would leave right?

You'll just start a new work unit so the one your on now will be erased and a new one will be downloaded from the server. You'll keep your OVERALL points and Total CPU time. (Thats what happened with me.)

OH Bye Bye 29 hours

If it is at all possible for you to let a work unit(W/U) run, then let it run. The longer it runs, the more points you will recieve when it is finished. The point system is based on clock time, not CPU time. You probably already knew this, but I wanted to encourage you to let them run. Unless you catch one early enough, you will just keep throwing time out.

One thing I do is, I have figured out that SuperOxide Dimutase(SOD) gives long W/Us. I throw them out if I see it in its first 5 hours. VGEF goes both ways(long{but never as long as SOD} and short.) I throw this one out if I see it in its first 3 hrs. I like the RAS ones. They are never long. Done in less than 4hrs usually. I haven't had a TK yet, but I hear the are out there now.

Happy Conformer Crunching!

eek, same thing here, only with 17 hrs.

Most of my targets are SOD (including this one), and this one is after 5 hours in 99% .

UD is well aware of the problem of long running molecules, hopefully there will be a solution allowing for user determined saves soon. It's unfortunate that there are cases where you lose processing time. It happens to me whenever Win95 crashes, which is expected.

In terms of the percentages, you cannot attribute it directly to time. The percentages seemed to be based on the number of molecules that need to be examined, and the number of conformers for that particular molecule. Some work units are more complex than others, and will take more time to complete.

If you crash or reboot you will lose the results for molecules since your last save. You can use UD Monitor to help track when your last save was.

In terms of "throwing out" molecules. I personally don't see the difference if I'm running SOD, TK or RAS. Each one can be either short or long. Regardless of the time it takes to process a molecule, the points I would generate over a specific time frame are the same, regardless of how many work units are processed. Besides, I'm not in it for points/prizes (other than reaching a new avatar level here). It's more important to be making a contribution that might just make a difference one day.

If you are running the agent, and it has been a Loooooooong time since the last save...suck it up, and let the molecule finish. Ctrl-Alt-Del'ing your computer and deleting the workunit accomplishes little. Let the little bugger finish on its own.



-Tacks

The SOD's (The ones I have had) give pretty quick work units. What I read on the FAQ, that there are usually long molecules on the Think task.

I work on a PII233 with 322mb of RAM. RAS is a smaller, easier protien, so it finishes slightly quicker. I'm on a Win95 also, so I can't get a 40 or 50 hour unit working without a reboot. All my SOD protiens go more than 60 hours and can camp down on a particular molecule for over 20 hours sometimes.

Yes it is about fighting cancer, but UD sends out hundreds of copies of the same W/U, so me CTRL+ALT+DELing will not ruin the final results of the project. And yes, I also did join with UD for the possibility of winning one of the prizes. I need the money.

If you have a faster computer that can do several W/Us a day, then do not 'throw out' units. I encourage all to keep them if you can. I can't. If I keep one, it will crash and might not make one % before it does again.