- From: Armando Stellato <stellato@info.uniroma2.it>
- Date: Wed, 4 Dec 2013 11:42:58 +0100
- To: "'Ivan Herman'" <ivan@w3.org>, "'Richard Light'" <richard@light.demon.co.uk>
- Cc: "'W3C LOD Mailing List'" <public-lod@w3.org>
> >> http://www.w3.org/TR/rdf-api/ > > Why did it die? > > Lack of interest:-( There were no real uptake in the idea neither by users nor > by implementers. It really was heading for a paper-only specification. It > seems that this direction was not what the community wanted at large. ..though maybe it is just the specific instance which didn't fly ...I personally feel the need for a standard Java implementation (as others may have for other languages obviously), and whether it is the implementation of an abstract interface or not, it would be much welcome, instead of seeing various middlewares (Sesame and Jena mostly, but there are others) and adapters between them, which are not always guaranteed to be updated with the versions of the things they adapt. Regarding the abstract interface, I think it is not a matter of more Python-like interfaces; simply in DOM the objective of a very general interface ended up into something which is very heavy and not simple to use for most practical cases (cit. the Java Tutorials: " As you can see, when you are using DOM, even a simple operation such as getting the text from a node can take a bit of programming"). For this reason, even in Java some people prefer to use other DOM-oriented interfaces but not the standard one. Against this, my personal experience and point of view is: 1) I was still happy about having one standard. I personally use in many cases some short-cut/facility methods over standard DOM, and still my software is compliant with the standard, can switch its implementations if new better performing ones are available etc.. 2) the above issues with DOM (for XML) IMHO do not represent failure stories which should be necessarily repeated with RDF Cheers, Armando
Received on Wednesday, 4 December 2013 10:43:38 UTC