Jack moved the linux box (hacweb5.jlab.org) that is controlling that crate from near the crate to the labyrinth recently (this morning maybe?).
Unfortunately the java gui thread that talks to that crate is now locking up, leaving the control pane for the beamline crate unresponsive. The software on hacweb5 seems to be running OK, so I suspect that there is some intermittent line noise creeping in. The Java GUI can be pretty sensitive to that kind of thing (seen it in the past with the stock firmware in the unmodified HV crates too).
It could be that the finest, cheapest, integrated-as-an-afterthought HW serial port driver that Dell can buy just isn't up to pushing a signal over 100' of serial cable. Or the cable is routed next to a big, sparking fan -- hard to say.
A serial repeater might be worth a shot if we have one, otherwise I'd like to try moving hacweb5 back into the Hall and playing with a few cables to see if I can isolate the problem. Talk to you tomorrow Jack.