Re: SETI like tass project

[<aah@nofs.navy.mil>]

I agree with most of the posters:  doing the first-step
processing of extracting starlists from images is much
easier to do at the telescope site than to ship images
somewhere else.  Not only is this the most bit-effective
method, but often the local site knows something about
the weather conditions for that night, can delete images
where errors occurred, etc.
  A night's worth of data should take much less than a day
to process to starlists.  At worst, you can have a second
computer available so that it can process data 24hrs/day.
Normally, even if a 12-hour night takes 14 hours to process,
you always have an equal number of cloudy nights on which
no data is being collected that can be used to catch up.
  Tom is in the position of having lots of data, but has
not processed that data.  He can either wait until he has
a local pipeline and process the data himself, or ship
the archived data elsewhere for processing.  The latter
option is what he has been doing lately, but he is not getting
a lot of feedback.  This lack of feedback is partly my
fault (since he has sent me some of the disk sets as well),
but I am just too busy these days and cannot get to even
high priority projects like this.
  On the other hand, as Doug and others have mentioned, once
you have starlists the real task begins.  Not only can you
search each and every star for variability, but you can start
making master catalogs of real photometry.  These tasks are
*not* small, and a SETI-like setup for doing them makes sense,
though I wonder if the effort to do such a setup is not better
spent at this stage in making good analysis pipelines.
  Enough.  I have to get busy for those folks that are
paying my salary!
Arne