User name R. Michaels
Log entry time 01:00:38 on December 1,2007
Entry number 217337
This entry is a followup to: 217330
keyword=prescale string wrong.
Jian Ping called about end-of-run deadtime 40% for run 3086.
Deadtime was actually much lower (5%) and the real problem was
the infamous "p4" instead of "ps4" problem, as reported earlier.
I've fixed the line. Odd that it happened again ...
1. No T4 in datastream for the runs with "p4" instead of "ps4"
2. Deadtime will be reported correctly by Podd analyzer and by
datamon. Same as what the trigger supervisor does with the string.
3. halog end-run scaler analysis doesn't parse the string
correctly (i.e. not like the trigger supervisor).