To display time reolution plots and get out the values themselves:
(go to the directory where you have the Time resolution runs)

java -Xmx1024m -jar /home/testdaq/bin/fatdisplay.jar

Read the detailed instructions below on how to get the resolutions.
You may skip the 1500Hz and only process the 15000Hz TimeRes Testdaq runs.
You can combine the ATWD0 and ATWD1 runs in the "FATDisplay Run List" window.

You need to have a file called 'nicknames.txt' in the testdaq/bin directory.

That file has the format:it has the following format:

| mbid                      | thedomid    | thename                | dor      |
| 4D5B7EE18822   | TV4P0030 | Neptune               | NULL |
etc.

you get it by doing:

loginsql
use fat;
select mbid,thedomid,thename,dor from  domtune;


(Instructions from the author C. Wendt):
It gives you two windows, "FATDisplay Run List" and
"DOM by DOM Display".

You have to use the "Run List" window to pick which data will be used to make the plots.
First you set the "Master Directory" to something that contains all the individual run directories,
then the program will scan through the directory tree rooted at that point, down to some level.
For example, on pub.icecube.wisc.edu, we set the Master Directory to
/data/exp/IceCube/2005/TestDAQ/PSL-FATDAQ1 or you can drill further
down to the month or day of interest.
You can specify more than one run if you want, the chosen runs will be lumped together.
As the data files get processed, the progress is shown in a progress bar.  You can start
looking at plots right away though, i.e. you don't have to wait for the whole data files to be
processed.

After you get the runs started, switch to the "DOM by DOM Display" window.
It has a list of which plots you want, which starts out empty.
Push the "New..." button to add a plot to the list.  The only type currently known is the
delta-t plot for time resolution (others coming soon).

There have been a few glitches with the display which may or may not appear on your system,
because I made a mistake in the display programming.
I fixed the problem in the "Run List" window but I still need to fix the "DOM by DOM Display"
window.   But you can probably use it as is.

There is a very crude numerical output feature, if you push the "Export" button on the
"FATDisplay Run List" window (after the processing is complete or mostly complete), it will spit
out a list of numbers for each DOM.  It is supposed to get put into a text file but for now it just
goes to standard output.  You see there the time resolution for each DOM (sigma), plus some
quantification of how many events were outside the gaussian peak.
Like there's "tailfrac5ns" which says what fraction of events were more than 5 nsec away from
the nominal position.
Mostly you're just looking for DOMs that appear different.  We've only had two DOMs ever that
had mildly bad tails and slightly worse time resolution (like 6 nsec instead of 3 nsec).

The other thing you can use this program for is to tell if the data are completely screwed up.
That's why I made it interactive with almost-instantaneous feedback. Sometimes the GPS
synchronization fails and the hub(s) have to be rebooted.  That happened during our last FAT
but it wasn't noticed until I ran the time resolution code.