Hydra W3C Community Group Telecon

Minutes for 2019-06-25

  1. TemplatedLink cannot formally have supportedOperation
Karol Szczepański
Tomasz Pluskiewicz
Karol Szczepański, Tomasz Pluskiewicz, Angelo Valtens
Audio Log
Karol Szczepański: Meeting: Hydra W3C Community Group Conference Call
Tomasz Pluskiewicz is scribing.
Karol Szczepański: Pieter isn't here so let's skip the first point on our agenda
... moving on to issue #197

Topic: TemplatedLink cannot formally have supportedOperation

Karol Szczepański: all I did was add a statement to the JSON-LD, subclassing TemplatedLink from Link
... this seems least intrusive. it should not break anything.
... alternative was to add TemplatedLink to range and domain.
Tomasz Pluskiewicz: yes, I just approved
... we did discuss this. I think it's the right thing to do
Angelo Veltens: one question
... not sure. not sure how to understand hydra:Link
... it would be something to dereference?
Karol Szczepański: in general, yes
... Hydra adds explicit term to explicitly say "this is a link"
Angelo Veltens: clients might try to dereference the templated link?
Karol Szczepański: you dereference the "right side of this relation"
... it allows you to define that the IRI may vary
Tomasz Pluskiewicz: the object is a blank node, a TemplatedLink
Angelo Veltens: I expect the object to be dereferencable. here I cannot
Tomasz Pluskiewicz: can you elaborate on domain and domainIncludes?
Karol Szczepański: TemplatedLink did not exist in those domains. now it will be able to participate in those relations
Tomasz Pluskiewicz: can you give an example?
Karol Szczepański: hydra:title has Link in domainIncludes but not TempaltedLink.
Tomasz Pluskiewicz: ok, I understand
Karol Szczepański: I wonder if it's a problem that TempaltedLink is not dereferncable
Tomasz Pluskiewicz: if that's a problem than we'd need to change the approach
Karol Szczepański: yes, I will think it over. Angelo's doubts give me pause
Tomasz Pluskiewicz: Angelo, could you comment that on the PR?
Karol Szczepański: a one line change turned out to be not a trivial
... what other issue you'd like to discuss?
Tomasz Pluskiewicz: we have #10 in the cookbook
Karol Szczepański: angelo, there were change requests on that. any news?
Angelo Veltens: ah, yes. I was busy. I will get back to it in the next week
Karol Szczepański: the cookbook says that client consults API Docs
... I would like to give impression that client use out-of-band information
Angelo Veltens: in the example API so far we have just the collection and type of collection
... there are no inline operations
Tomasz Pluskiewicz: nothing here is out-of-band. Hydra is fully accessible by the client
Tomasz Pluskiewicz: I'm sure that the Api Documentation is mandatory, a design of Hydra
Karol Szczepański: I don't agree. we risk ending up with something like Swagger etc
Angelo Veltens: back to the PR, I will clarify that bit
Karol Szczepański: I'm fine with that
Karol Szczepański: anything else?
Tomasz Pluskiewicz: I spiked this in hydra analyzer. we start building the DSL
Karol Szczepański: DSL as a Domain Specific Language? I find the example hart to follow
aventels: it will generate api docs?
Tomasz Pluskiewicz: I think that a API testing should resemble a hypermedia client. follow your nose
Tomasz Pluskiewicz: I have a blog post in the making. I will share later this week
Angelo Veltens: this will be good to test multiple implementations of the movies API
Karol Szczepański: I think this what Ruben was mentioning for a meeting
... I started porting Heracles to .NET, if you find interesting
Angelo Veltens: not .NET here
Karol Szczepański: okay, I will work on this and give more info
Karol Szczepański: I thought to postpone the calls this summer
Tomasz Pluskiewicz: unless there is a big topic
Angelo Veltens: we should decide when we continue
Karol Szczepański: September maybe
Tomasz Pluskiewicz: tpluskiewicz: +1