Publishing of the RDF/OWL or JSON-LD version with resolvable URIs and content negotiation on EU server

[jgmikael]

Hi,

I'm afraid I have to request the re-opening of my closed issue #43 since the answers you have given in issue SEMICeu/Core-Location-Vocabulary#28
are not satisfactory.

There still isn't a functioning way of pointing at an URI of a class or attribute in the Core Business Vocabulary through using the namespace http://data.europa.eu/m8g

In out Core Vocabularies tool on the national Interoperability Platform, I can for instance point at classes and attributes in

• vocabulary.uncefact.org (the UN/CEFACT BSP JSON-LD Vocabulary)
• https://ontology.tno.nl/logistics/federated/BusinessService# (the FEDeRATED ontology for the logistics domain)
• the Italian governments Core Person Vocabulary Italian Application Profile - Person Ontology at https://w3id.org/italia/onto/CPV

These are examples of standard Linked Data publishing = the classes and attributes and associations are directly accessible from our modeling tool, no offline copying needed.

If you still please could look into this issue or at least clarify why Core Business has been left totally out of the loop (developments taking place only in the Core Location Vocabulary), thank you in advance!

[bertvannuffelen]

@jgmikael I am not sure what your concern is:

At this moment the behaviour is as follows:

1. individual URIs of the Core Vocs e.g. http://data.europa.eu/m8g/PublicOrganisation, are human and machine readible available.
   in your browser http://data.europa.eu/m8g/PublicOrganisation is redirected to https://semiceu.github.io/CPOV/releases/2.1.0/#PublicOrganisation
   curl -L -H "Accept: text/turtle" http://data.europa.eu/m8g/PublicOrganisation returns the turtle representation.
2. As full collection only the aggregates RDF file is available at the moment.

So what are you expecting, or looking for? Can you make it very concretely?

[jgmikael]

Problem is that I cannot personally since I rely on my tech team and they've said it is enough to ask for "resolvable URIs and content negotiation" > this request worked with the UN/CEFACT team.

I'll ask for someone on our tech side to drop a more concrete line to this issue, thanks for your patience!

[amiika]

I think the issue is that the response content-type is text/html for:

curl -L -I "Accept: text/turtle" http://data.europa.eu/m8g/

Request is done by accepting multiple types of RDF from the namespace with priorities like:

Accept: application/rdf+xml;q=1,application/turtle;q=0.8,application/x-turtle;q=0.8,text/turtle;q=0.8,application/ld+json;q=0.7,text/rdf+n3;q=0.5,application/n3;q=0.5,text/n3;q=0.5

If the response does not return the correct content-type that has been found and instead gives wrong content-type like text/html systems must infer the content-type which in this case fails for some reason.

[bertvannuffelen]

@amiika are you responding on behalf of @jgmikael tech team?

If not, then I still wait for them to make their comments as I would like to understand their usage pattern.

[jgmikael]

Hi, yes Miika is on the team, any additional questions can be directed to him, thanks! br Mikael Den tors 19 okt. 2023 13.20Bert Van Nuffelen ***@***.***> skrev:

…

@amiika <https://github.com/amiika> are you responding on behalf of @jgmikael <https://github.com/jgmikael> tech team? If not, then I still wait for them to make their comments as I would like to understand their usage pattern. — Reply to this email directly, view it on GitHub <#46 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7ET3VRJGK77EHP4AEQ55TYAD5IXAVCNFSM6AAAAAA6DRHIQSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONZQGUYTKMBUGA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[bertvannuffelen]

@jgmikael thanks for the clarification.

@amiika as written earlier: today there is for the full download of the RDF file available.
I agree that the content negotation for that whole document could be improved.

Aside from that what is your usage pattern? How does content negotation fits in your usage?

[amiika]

@bertvannuffelen At the moment, the Finnish interoperability platform is resolving data models only trough namespaces. This was done to ensure that no-one is claiming or polluting namespaces that they don't actually own. Ideally it works, but often ends up with issues like this.

I didn't find any other issue with the content negotiation other than it is not returning the right Content-type for the response. I think this issue could be fixed by using updated Jena library to guess the Content-type better, but it would take some time to do the updates to the new upcoming version.

However, in ideal world, you would get the Content-type from the response and the requester would know which type of format to parse.

[EmielPwC]

Hello @amiika @jgmikael,

The issue should be resolved now could you test and verify our current fix in your system?

Kind regards,
The SEMIC team

[RiittaA]

Hello @EmielPwC and @jgmikael,
as the superuser of our FI-platform I can confirm, that your fix seems to be working now. We tested it in our test environment and can see the resources properly, i.e. as resolvable and thus reusable.

[jgmikael]

Hi, thanks Emiel for the fix - however, I was under the wrong impression that the "m8g" namespace would cover all EU Core Vocabularies - this doesn't seem to be the case (actually I don't even understand what the m8g covers, the classes and attributes don't seem to be that interesting. Therefore I've opened a new issue: #49 Hopefully there could be a quick fix for this also? best regards, Mikael Den ons 17 juli 2024 kl 10:14 skrev Emiel Dhondt ***@***.***>:

…

Hello @amiika <https://github.com/amiika> @jgmikael <https://github.com/jgmikael>, The issue should be resolved now could you test and verify our current fix in your system? Kind regards, The SEMIC team — Reply to this email directly, view it on GitHub <#46 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7ET3WFSN2WO4HJ2RYHTG3ZMYKTZAVCNFSM6AAAAABLAABZMWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZSGU4TIOBVGY> . You are receiving this because you were mentioned.Message ID: ***@***.***>