Yeah, I think JSON-LD is intended for shipping around smallish packets of linked data formatted as JSON in a standard quality-preserving way. I do not get the sense that JSON-LD is intended for large query results.
If you want large query results as JSON, then the SPARQL JSON format is already defined and implemented. It can be consumed by the typical front-end developer, and they can reconstruct parts of the graph as needed from the tabular format, just as they would do if they were getting JSON back from any tabular data source, but with nice name spaced identifiers and standard semantics. On Fri, Jul 19, 2013 at 4:51 AM, Neubert Joachim <[email protected]> wrote: > For some use cases, the tabular SPARQL select query output formats are not > optimal. That's true when the results are in fact tree-shaped, such as > multiple skos concepts with their relations and their respective labels. > Therefore, it would be highly useful if I could offer the output of a SPARQL > construct query in a format which a web developer without any knowledge of > RDF can immediately cope with. > > From Andy's mail on this list > (http://mail-archives.apache.org/mod_mbox/jena-users/201306.mbox/%[email protected]%3E) > I understand that there are issues re. the scalability of json-ld and its > suitability for updates. However, I imagine that for a large number of use > cases these issues would not matter, and an integration of jena-jsonld into > Fuseki (perhaps with a warning hint in the documentation) could be > tremendously helpful for people using Fuseki just out of the box. > > What do you think? > > Cheers, Joachim >
