Cart
[Home] [By Thread] [By Date] [Recent Entries]
Paul Hermans wrote: > Due to my background I started with the XML serialization of RDF > using XSLT to do the querying and the transformations. > But this didn't feel right, using tree based processing on graphs. > So I moved quickly to other serializations (N3) and SPARQL instead of XSLT. > Yes, it always seemed that the RDF people had just jumped on the XML bandwagon despite the bad fit. At the time, there was gossip that at least one of the leading lights had said (attributing it to Knuth) that XML was a fad. I am not against KR or any of the AI projects' many fruits: I don't want to give that impression. But almost all these systems where the model is not transparent from the markup (e.g. where there are components) it is tenuous to call them "XML" in the sense that other things are XML IMHO: I think the test is whether the developer can work productively using generic XML tools or whether they have to work with custom APIs to work on the model directly. XSD (with its "components") is one where working in the XML is terrible. Parts of OOXML get near failing the test. And RDF is certainly one. So perhaps we can amend my comment to: I have a perception of people adopting RDF (in XML) then figuring out they didn't get any extra benefit over using plain old XML or an optimized RDF notation. Cheers Rick Jelliffe
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] |
