We have just started the Clean Energy Project (CEP) and already we have identified a number of bugs. I will list them here so people will be aware of what is going on.

1) Some work units error out on every machine they are sent to. Apparently this can happen at any stage of progress. The project scientists are looking into this.

2) Sometimes a work unit will end early but return a non-error result code. This can cause anybody else working on the same work unit to get a very low number of points. Again, everybody is looking into this.
[Added] Sekerob suggests that I include a sample line from the error file on the Results Status page for one of these non-error errors:
'[ERROR] Failed to open either source or destination files while copying wcgrestart.rst to ../../projects/www.worldcommunitygrid.org/E000042_595A . . . . . . . .'

3) The CHARMM molecular mechanics package developed by the Karplus group at Harvard University and used by CEP is an enormous Fortran program, much larger than the HCC program. Just like the original HCC program, it is running a very large number of page faults under 'some' circumstances. Running more than one instance on a computer appears to exacerbate this problem. But because the program is so large, tracking these inefficiencies down and trying to correct them is probably going to be a last-priority issue for a while.

4) Figuring out work unit length is not easy. It seems to be very variable. Perhaps we shall have a blinding flash of insight that shows us how to determine the length of a work unit. Perhaps not.

There may be more errors. Reading the complaints posted indicate more possibilities but these 4 problems have been confirmed.

On a personal note, I am running all projects right now to minimize the aggravation caused by a new project that is just starting to get the bugs identified. Some people are running CEP-only, either in order to help locate the bugs quickly or to earn a badge quickly. CEP will be running for a long time since we will not start Phse 2 using Q-CHEM until much later.

Kawrence