Closed phochste closed 2 years ago
Hi @phochste
I have recently added Uppy S3 multipart support. It is not bug free yet.
We are using rdf-form for a headless cms on top of Apache Jena. In our use case we are using BackBlaze with their S3 compatible API. This product: rdf-form is not yet bug free, but slowly we are getting there. At the moment it is also import to note that we a refactor should be done on how translations are added. At the moment it is possible to lose types:
"22-01-2020"@en
I am thinking of how to solve that. Personally I would have loved if RDF would allow that with types.
Thanks for this!
About the date types, is not that something that doesn't need to be part of the data model but more of a separate display component, or even two separate triples a string and an inferred date?
There is difference between a string as entered by a human and the inferred ISO date. In library context that is also the case where there is, for instance, a distinction between the French Republican calendar date as seen on a book ("11 Frimaire, Year XIII") and the ISO date ("1804-12-02").
The predicate you would use for the date would inform linked data applications how to interpret the date. E.g.
<https://example.org/record/123> <http://example.org/world/clock#FrenchRepublican> "11 Frimaire, Year XIII"
Yes, the date issue we discussed was about misusing RDF.
I wanted to have localization for the predicate schema:publicationDate. Unfortunatly RDF is not made in such a way. I think conforming to spec is the best way forwar, another option might be to create a RFC for the spec. But that will take a long time.
I am closing this for now.
I've added today a challenge https://github.com/SolidLabResearch/Challenges/issues/19 for SolidLabResearch and @rubensworks mentioned pointed me to your project. It is great to see your solution! Is there a way for doing file loads via this mechanism?