On Aug 28, 2015, at 2:27 PM, Kian-Tat Lim <[log in to unmask]> wrote:

Jacek,

I think it’d be useful to document this kind of wisdom in one place. Maybe K-T has a suggestion where. Perhaps in DM-3452?

I think it would be good to document the dataset types produced
by the various Tasks and especially the ones generated by the
transformation framework for ingest into the database.  That shouldn't
be (just) in a JIRA issue.  Ideally, each Task (or obs_* package for
now) should document its dataset types and metaserv, the SUI/T, and the
schema browser (if still needed given the others) should reuse that
documentation (via pointer or extract) to describe the tables in the
database.

As to where the "integration deployment" data access Web
services are currently running, that should also be documented, perhaps
on a Confluence page.