Here is a list of features Eric implemented in the deployment we just pushed out.
change job stats tables according to a discussion I had with Clark Barrett; hopefully what we have now will make better sense. You should be able to see a tool tip when you hover over a column, to see what the semantics is now.
put preprocessor in job details page
putting all pairs in job info csv instead of only complete ones
allow users to get only complete pairs in the csv optionally
put expected value attribute in csv
delete job logs older than 30 days with a periodic task
put input restrictions in help pages
starexeccommand prints ids for all commands where that makes sense.
Also, we shrank the queue size from 800 to 200 temporarily, because small queues (with just a few nodes) were getting bogged down with queue size 800. We will put in place a more dynamic arrangement where we set the queue size based on some factor times the number of nodes in the queue, later this week.
There is more information in the help files, but unfortunately the per-page nature of the help is broken temporarily (so when you click on Help, you will just see all the help pages, whereas before we had it so that you would see help for that page if such is available). We will restore that functionality hopefully later this week.
Aaron
StarExec 
Leave a comment