The task in this ticket is to define a proposed environment within the LSP for the retrieval of LSST's released data products in their LSST Python forms, by normal users of the Science Platform at the DACs. (I.e., this is not specifically about things like the Commissioning Cluster environment.)
A key issue in this regard is that the tabular data products should be assumed to be in databases (e.g., Qserv) rather than in, say, afw.table-style FITS-binary-table files.
The resulting memo (proposed to be a DMTR) should address the role of the Butler, of the DAX services, and what it would mean to provide this service inside the Notebook Aspect, as compared to providing it remotely.
The task in this ticket is meant to be "science-user-flavored" - that is, it is meant to be a proposal for the functionality of the environment, not a specific implementation design. That will have to follow, and implementation issues may end up providing back pressure on the final adopted design for the functionality.
This ticket was created to re-capture a long-standing DMLT Confluence task for Robert Lupton and myself (Gregory Dubois-Felsmann), originating at the May 2017 DMLT in-person meeting. At the 2017-10-09 DMLT meeting it was noticed that this task had slipped through the cracks.
(The original Confluence task specification: "write specifications for user interface for reconstituting Python objects (including, e.g, HeavyFootprints) from a database-backed Butler.")