Adrià's reply to Nick
Gesendet: Freitag, 20. September 2019 um 13:16 Uhr
Von: "Adrià Mercader" <>
An: "Nicholas Car" <>
Cc: "Stefan Oderbolz" <>, "Rob Atkinson" <>, "" <>
Betreff: Re: CKAN DCAT plugin - Conneg by P
Thanks for reaching out Nick,
It does sound useful and we are keen to keep ckanext-dcat up to speed with the latest DCAT developments even with limited resources.
Perhaps when you start work if would be good to summarize your implementation plans on the issue you created in the repo so we can flag potential issues early
I can also help reviewing and providing guidance.
Best regards,
On Fri, 20 Sep 2019 at 09:43, Nicholas Car <> wrote:
Hi Stefan,
Thanks for looking into the spec! Yes, not a huge job but I’m keen to have the plugin maintainers on-board before I start so there’s no surprises!
I’ll let you know when I get into the technical work (likely in about a fortnight).
Nicholas Car
Data Systems Architect
SURROUND Australia
0477 560 177

On 20 Sep 2019, at 4:47 pm, Stefan Oderbolz <> wrote:
Hi Nick
That sounds like a good idea, I quickly skimmed the draft spec and it should not be too complicated to implement on top if the existing code.
I'd be happy to assist with reviewing and/or testing the new functionality.
Best regards
On Wed, Sep 18, 2019, 05:09 Nicholas Car <> wrote:

Hi Adrià  & Stefan,


The Dataset Exchange WG within the W3C has just released a final Editors Draft of “Content Negotiation by Profile” [1]. This specification describes how to negotiate for content according to profiles, rather than say by format or language. The Dataset Exchange WG is also responsible for the revised version of DCAT [2] which is also in final Editor’s Draft stage.


For the last W3C standards process, it’s important for implementations of standards to be noted and tested. I’m keen for CKAN to support Conneg by P as much as possible both to demonstrate an implementation of that specification but also so that CKAN implementors can get their content out according to old and new versions of DCAT.


I think that the ckan-dcat plugin’s most of the way there – the fact that new profiles can be defined by users and used to export data etc. – so the only real work is to ensure that the API matches the new Conneg spec. This situation – that there’s a system that actually can do all the things Conneg by P’s talking about but which is using an existing API that’s not aligned – was anticipated by us, the Conneg by P editors, and now we’re keen to test out retrofitting the API to an existing system.


If I was to start work on enhancing ckan-dcat to work as per Conneg by P, would either of you two be available over the next 6 weeks to review the enhancements? I’ll be exiting the extended version on a CKAN catalogue that I’m using day-to-day for the Geological Survey of Queensland in Australia [3].