Currently a distribution within a dataset isn't required to have an IRI (@id). This might cause problems for services which need to identify a specific distribution for processing purposes (eg. like Network of Terms and the RCE terminology sources which have 2 Sparql endpoints for 1 dataset / @sdevalk ).
Currently a distribution within a dataset isn't required to have an IRI (@id). This might cause problems for services which need to identify a specific distribution for processing purposes (eg. like Network of Terms and the RCE terminology sources which have 2 Sparql endpoints for 1 dataset / @sdevalk ).
Adding the requirement to have an @id for each distribution has an impact on current implementations by Spinque (https://data.spinque.com/ld/data/netwerkoorlogsbronnen/ and https://data.spinque.com/ld/data/vangoghworldwide/), Atlantis (https://lit.hosting.deventit.net/AtlantisPubliek/Opendata/download-set?name=catalog&format=RDF), Triply (https://data.netwerkdigitaalerfgoed.nl/). Sites likes https://www.openarch.nl/datasets/ and https://data.overheid.nl/datasets/ also don't have @id with their distributions of datasets. The CKAN DCAT extension (used by Picturae) does have an IRI for each distribution.