Re: Introduction and some comments ;-)

[Tom Droege <tdroege2@earthlink.net>]

Chris is mostly right below.  There are features about tass that make it 
different.  We have larger pixels than the other all sky photometry 
projects that I know of.  Read Andrew Bennett's last note for an example of 
something that probably does not bother ASAS or MACHO but is a pain for 
us.  So while many of the scrips are available, there is some hard core 
creative thinking needed before we will get all we can from the data.

It may require serious changes in the way we collect data.  That is why I 
am anxious to get to work on the "engineering run" data.  We may want to do 
things like defocus the images (I really think not, so please don't start a 
thread on defocusing) to solve a problem.  We will want to do something to 
improve the data quality, I bet.  I just don't yet know what it is.  Trying 
to process all this data should tell us how to take it right.

I am afraid that we need to "hand craft" the data until we really 
understand what is going on.  Then we want to start a "lights out" 
factory.  This will process all the data in exactly the same way.  That 
will make it consistent.  The real problem today is to start taking data 
with the right procedure and the right format.  Then we can process it many 
times until we get it right.

Tom Droege

PS, and frustrating it is!

At 03:58 PM 1/30/02 -0800, you wrote:
The current state of the art with TASS is that data reduction is a
"craft" that is done by a few people by hand, each using his own
methods and bits of borrowed code.  What we need is a modern
"lights out" factory.  The technical work required to go from the
"18th century arts and craft" stage to "21st century lights out
automation" mostly involves editing a pile of small scripts and
configuration files.  Nothing like heavy duty software engineering
is needed.  I think because everyone knows it is not a huge
engineering job is what makes not having it done so frustrating.