Re: The state of hydra

Am 05.12.18 um 18:32 schrieb Lorenzo Moriondo:

> 
> Hydra Ecosystem is part of the HTTP-APIs working group, it is endorsed
> by Markus Lanthaler to use the specs in the context of Google Summer of
> Code; it can be seen as the "operational" side of Hydra Community (which
> main focus is on defining the specs and the W3C draft), it is meant to
> create tools that can ease the adoption of Hydra-based solutions among
> present and future API developers (in this early phase especially
> university students thorough Google Summer of Code).   

Thanks for clarification!

> [...]
> The best people that can talk about Heracles are the contributors to the
> code as they appear in Github or in "Github Issues" page for that
> particular repository.

Ok, I will try by filing an issue there.


> hydrus is a good example of how anybody can deploy an API server
> starting form the a Hydra ApiDoc document.

Therefore even with hydrus I would need an ApiDoc document first? But
this is where the questions begin. How do I set it up correctly? In my
concrete example: How do I describe collections correctly? How do I use
"hydra:operation" within a resource. HydraConsole seems not to
understand what I have done. Other Hydra clients I could test against
are not really available.

What I do now is using the Hydra ontology as I see it fitting, and build
my own mainly json-ld/RDF-based client to reason about what it gets.

> You can possibly quite easily
> port the code to Java 8 using the functional tools provided (as hydrus
> is developed in a functional-first Web frameworks called Flask; probably
> you will get a better fit using Scala). For the client, there is no Java
> implementation afaIk, but you can try starting from hydra-py to
> understand how to write your own Java client by leveraging RDF libraries
> provided for JVM-based languages.

Yeah, ok, I will take a deeper look at the code and be inspired by that.
But at some point in time, the assumptions and conventions that are
coded into hydrus / heracles, or the stuff I do, have to be formalized
in the spec. Is this, what you / the community group are going to do?

> 
> If you need particular answers, you should try asking questions by
> attaching code examples with problems you are encountering and I suppose
> anybody would be glad to help.

Unfortunately this is not the case until now. If anything is unclear
about what I described in "Hydra console and hydra:operation" someone
let me know and I will clarify.

> Your issues are adoption issues, unfortunately is quite common for any
> early-stage project to have such barriers like lack of examples and
> support or stable packaged libraries;

It is not an early stage project. Hydra is around since 2012. If it is
not usable until now, I wonder if it will ever get there?

> that's why the effort of the Hydra
> Ecosystem to create tools and standard practices; for the moment, as
> Python volunteers, we are doing this using Python but I hope as anybody
> that other developers will join to widen the spectrum and provide
> support for multiple technologies.

I really appreciate your efforts! I am one of those other developers.
Actually it's me and two colleagues. But we cannot act in a vacuum, we
need exchange and discussion with people working on the spec.

Best regards,
Angelo

Received on Wednesday, 5 December 2018 21:01:27 UTC