does anybody know what keeps getting wu's stuck?

I suspect that there is more than one reason.

Shortly before IBM left, they tried to get them going again by reducing the TimeStep for stuck units. This worked to some extent, but not for all. They never said what the problem was but I suspect it was caused by the original program being unable to cope with some local geography.

I don't think they said what else might be the problem. I believe that Krembil have since tried the TimeStep method but not consistently. I don't believe anyone has tried anything else. The problem with the TimeStep method is that it takes proportionately longer as more calculations are needed.

Mike

ARP is my favorite project too (I'm a weather nerd). It is a more complicated project to be honest though.
My best try at deciphering the notes
- work on validation of a huge backlog of MCM is the current project.
- start up MAM again. MAM is similar to MCM, so this should be an easy next step once MCM is figured out.
- lots of testing on validation of ARP. maybe test batches go out, maybe a small amount of real work. Not sure how they will approach this. My hope is that once this is working well that the new server system they have will allow them to feed the ARP WUs to us quickly.

So in summary, they can't give us a date as the unknown issues are , well, unknown. But there are steps along the way that may tell us when the time is getting closer.