Dear members,
thanks to the walkthrough by Kristina
during our last TC meeting and based on our discussions around the main
rationale behind the main fields in
the DTA version of the Provenance
Type, I came to the conclusion, that
we should identify roles around the
data structure we define.
Actors are an integral part of any
structure one wants to weave into
the interactions of real life.
We have to define who fills what field
and which values are final and which may incrementally chnage and by whom.
Example:
The link from the meta to the real data
should be defined in ways that are
feasible and useful at the same time
in real use cases.
Besides consumers and producers we may want to consider defining
brokers, validators, aggregators,
proxies, documents, converters,
translators, withdrawers, superseeders, matching systems,
libraries, viewers, modifiers, as well as
other management systems.
Not all above roles will make enough
sense to be considered, but there
may be others I do not know of and
it may make sense to further differentiate some of the above listed into say basic, extended, and full flavors.
Such roles nicel derive from and
link to existing use cases as well as
allow us to include test cases in the
core specification.
Finally, such actors help formulate
actoonable conformance statements
that foster high quality
implementations of the spec in the ecosystem.
Looking forward to any feedback
and to our next steps along the way.
Cheers,
Stefan.
PS: Written on a tiny soft keyboard whild looking at a two by two inch display.