Re: FW: Data to be collected in the 'master?' database

[Tom Droege <tdroege@veriomail.com>]

Arne writes:

At 11:58 AM 8/6/01 -0700, you wrote:
>   However, for the pipeline, no modification to the fits
>file is made.  The program name and version number need
>to be added to the starlist header, not to the fits file.

OK, I am out of my field here, but I thought we were going to do starlists 
in fits list format.  So the fits header I was referring to is the header 
for the star list, or for any subsequent operation.  I agree 100% with 
Arne.   We should have an audit trail for every operation.

We are doing science that never ends.  Many years from now, something that 
we did today might suddenly become interesting.  Some previously 
uninteresting star can blow up and now it is interesting to study the 
earlier measurements and see how the blow up developed.  Then one will want 
to be able to reconstruct everything.  Thus I wan to keep the raw data and 
a record of everything done to it.

My understanding is that there is a fits format for everything we might 
want to do.  If so I hope that we stick with fits all the way.

I think part of every output file at every stage of the operation should be:

We started with data file xxxx operated on it with programs yyyy ver 234, 
zzzz ver 256, to produce the data of this file.  We did this on yyddmm on 
computer "fred"...

Programs of every version should be stored somewhere where everyone can get 
at them.  Since the software is mostly small, even in source form compared 
to the data, I like putting it in the output file directory.  Then it never 
gets lost and there is no question about what software was used.

Tom Droege