Minutes of SHARE Meeting                                        01/23/02



1. Present: Mark Dickinson, Gerhardt Meurer, Anton Koekemoer, Daryl Swade,
	Bill Sparks, Howard Bushouse, Jerry Kriss.

2. Megan (FASST chair) met with Rodger Doxsey, Chris Blades, Daryl Swade,
   and Marc Postman to discuss highest priority tasks of SHARE and FASST.
   Rodger would like to work on defining resources needed in the HD for
   implementing the highest priority recommendations over the next year.
   Needs our final report by mid-February.

3. There will be some synergy between SHARE & FASST recommendations and
   Treasury programs and large proposals accepted for Cycle 11 (e.g., GOODS).
   Should involve those teams in planning to avoid duplication.

4. Distributed the attached table summarizing SHARE tasks and min and max
   resource estimates.  Max estimates are not deltas over the min, but are
   the full resources and time required for implementation of max.

5. Several people noted that all tasks have assumptions and prerequisites that
   are not included in their resource estimates and timescales.  For example,
   a min goal of "Source segmentation and Classification", which is to
   produce "engineering-grade" catalogues, would be needed both for WCS
   improvements and image combination.

6. Mark suggested that a diagram showing the interrelationships among the
   tasks would be helpful.  Jerry will prepare this.

7. Most estimates so far have ignored DSD pipeline implementation work
   (Daryl will look over the lists and provide something), and
   archive work related to interfaces and ingestion of products.

8. Gerhardt asked about policies on ingesting new data products into the
   archive.  These are normally done on a case-by-case basis.  May see
   some more formal development as products from new Treasury programs
   and Large proposals are integrated into the system.

9. Discussion of Jim Rose's suggestion, which Jerry called "Customized
   Post-processing".

	* Concerns similar to those expressed before about Customized
	  Reprocessing about how this could potentially use a lot of institute
	  resources, and steps backwards from distributed computing to
	  centralized computing.
	* Analogies to Treasury programs---it produces products that
	  would be ingested by the archive.
	* To do proper resource allocation, assure science quality and utility
	  of products, and assure quality and safety of implementation, it
	  might be best to regulate access via the proposal process.
	  Analogous to how one now requests supercomputer time.
	  An internal technical review of the proposal also needs to be done by
	  the archive and DSD.
	* This idea is an example of an enabling technology or infrastructure.
	  Developing it would be useful for implementing some of the
	  more specific SHARE recommendations, e.g. photo-z or spectral
	  combination.
	* A suggested development path might be to solicit ideas for internal
	  use, e.g., doing instrument trending by following some parameter.
	  This could develop and test structures and procedures internally
	  in a useful way before opening up to the general community.

	* What is our ranking?  Probably somewhere near the middle, at
	  about the same level as Customized Reprocessing.

10.  ACTION ITEMS

	1. Jerry will produce the diagram showing task dependencies.

	2. Daryl will provide pipeline development resource estimates for tasks.

	3. Megan will produce archive development resource estimates for tasks.

	4. Jerry will draft a final report

	5. Jerry will put all products on a web site for easy access.


NO FURTHER MEETINGS! (Unless we decide to have a discussion prior to issuing
			the final report.)