Operation Last Night

[Tom Droege <tdroege@veriomail.com>]

Last night I wrote 2 GBytes of noisy data.  This is because we were having 
a party and I just rushed upstairs at dusk and started the program without 
looking at the data first.  The result was that I ran all evening at 10 
C.  The data would have been very nice, but now it has noise levels of 400 ADU.

Those of you working on real time code, there needs to be built in 
checks.  These programs operate differently from what most astronomers 
do.  One just turns on and runs all night.  This is quite a bit different 
from taking and image and looking at it, then taking another.  The later 
process is self correcting, the former is not.

I think I would like to see the data collection program put up a small 
piece of the image.  It should also compute things like the noise level and 
display them prominently.  One should be able to glance at a screen from 
time to time and have some assurance that things are going OK.  Remember 
the Mark III code did this.

Tom Droege