Minutes of SHARE Meeting, Aug. 15, 2001 in N420

Present: Niall Gaffney, Dorothy Fraquelli, Anton Koekemoer, Howard Bushouse,
         Daryl Swade, Mark Dickinson, Megan Donahue, Jerry Kriss

1. Placed real-time evaluation of Science-ordered and Tractability lists on
   the board:

	Science			Tractability
	-------			------------
	WCS			CUSTOM
	IMGCOMB			QUALITY
	OBJECTS			WCS
	CATALOG			CATALOG
	PHOTOZ			
	SPCOMB			- indeterminate -
	TIME			
	QUALITY			
	CUSTOM			PHOTOZ
	CUBES			CUSTOM

   There was a wide dispersion in these orderings among the respondents,
   but the top and bottom 2-3 items are fairly representative.  The mid range
   ordering is not well determined.

2. Daryl suggested that an additional criterion that we should use is the
   added value that STScI can provide that is unique, that is, it is easier
   and more valuable for us to do a particular task than it is for the user.
   Example, image combination and catalogue cross-correlation is easier and
   high added value; customized reprocessing that duplicates the capabilities
   of the existing distributed pipeline s/w is low value, and it is just as
   easy (if not easier) for the user to do the reprocessing themselves.

3. Niall made the additional point that high science value need not be judged
   on the intrinsic science value of the product, but also on its utility as
   a tool to providing enhanced science.  Example, a catalog may not satisfy
   the criteria of a rigorous science result in and of itself, but it may
   facilitate the search for appropriate data to carry out a study.

4. Long discussion of the value of catalogues in various forms.  Mark and Megan
   note that one corollary for understanding a catalog is making HST filter
   throughput curves easily accessible.

5. Discussion of "data quality".  Agreed that it is useful for making more
   effective archive queries.  Quality flags can be highly instrument
   dependent, however. Example: NICMOS could flag how close an observation
   happened relative to the last SAA pass.  At the simplest, most general
   level, it would be good to have some quality flag that noted when a data
   set might have a significant problem.  One could use this as a selection
   criterion in a search.

6. Assigned writeups on various topics to the following:

        WCS     	Anton, Mark, Dorothy, Howard
        IMGCOMB		
        OBJECTS		
        CATALOG		
        PHOTOZ		Mark (short, dismissive)

	SPCOMB		Jerry
	CUBES

	TIME		Mark

	QUALITY		Daryl, Dorothy, Mark

	CUSTOM		Jerry, Daryl

   These writeups should lay out the science case for undertaking a reprocessing
   enhancement that implements these capabilities.  Emphasize the unique
   value that providing this service at STScI can bring.  Shoot for 1-2 pages.

   *** Drafts due before meeting on Wed., September 12.  ****


********************************************************************************
*** Next Meeting: Nominally in two weeks, August 29, 9 am, N420.  Stay tuned. **
********************************************************************************