Last modified: 2011-03-03 08:22:07 UTC
A bunch of output gets reported, including startup of processes, number of revisions processed at any given time, errors encountered when retrieving revisions, etc. Right now these all go into the ether since the whole thing runs in a screen session. They should get written into a log file so that we can see when something dies instead of having any relevant messages long since scrolled off the virtual screen.
fixed in my branch and deployed; specified by command line option to worker.py. Not included (but probably ok): debugging output from CommandManagement.py