RE: Hydra Status

I have good experience with data streams and microservices, beside REST
APIs obviously. My focus is on data infrastructure more than data modeling,
so please forgive me in advance if I write something too opinionated or not
completely compliant to modeling/design approach.

My opinion about the state of the draft is that the top-down part of the
job has been done. RDF is a huge standard but I have to modestly say that,
as far as I understand, has been a little over-engineered for the purpose
of research; as a Pythonist I appreciate JSON-LD and Fragments because they
make things much more readable and allows access to RDF. As an RDF novice I
can only suppose there is some way of "sub-setting" the RDF to encompass
basic logic (Boolean? Math?) operations that could cover the vast majority
of minimal usecases (is it basically what have been done as far as I
understand?).
Apart these guessings on my side, we should think more functionally, in
terms of operations more than classes compared to RDF (I would say with a
joke: more Scala types-thinking and less Java classes-thinking; or from
another point of view: take aside science for a while and embrace
engineering) I suppose. HYDRA is meant to be used in the wild so it should
be sensitive to trends in different industry-level ecosystems.

@Adam and @Graham: if you are interested I can start a G document with the
simple use case I have ready. It's a randomly generated database of IoT
devices. So we can all start commenting and adding pieces to that.
The usage of the API in the beginning can be just serving plain objects and
allow properties comparisons and basic math operations on these properties
to train the API from the very basics. Classes are already defined using
the basic JSON-LD features and linked to well-know RDF repositories, so
everybody can start following. In parallel the RDF experts can QA and test
and propose features.

*If we all have a little consensus in proceeding this way, I start writing
the document and then share the link. Please cast your opinion.*

As Markus pointed, the direction is focusing on IoT networks, I think
communication and higher level applications will come naturally afterwards.

Thanks to everybody for your attention. I am really proud of contributing
to this effort.

Lorenzo Moriondo, from mobile
https://it.linkedin.com/in/lorenzomoriondo