On Monday 12 April 2010 16:14:56 Georgi Kobilarov wrote: > [...] > So after your work on SQUIN, what is your conclusion? Does the distributed > querying approach work well enough that you can imagine a developer > building an application/website using it?
First of all, the query approach implemented by SQUIN and the underlying Semantic Web Client library (SWClLib) is very different from distributed query processing. I call this new approach link traversal based query execution. Re your question: building an application based on this query approach is very easy, in particular, if you use a tool like SQUIN that can be accessed as if it were a SPARQL endpoint. What might be a bit tricky for beginners is the development of the queries. We implemented two real LD based application using SQUIN: * http://researchersmap.informatik.hu-berlin.de/ * http://linkeddata.informatik.hu-berlin.de/altmed/ These applications, indeed, use Linked Data from multiple sources. > An application for users who have no appreciation for the underlying > technology at all and hence no tolerance for hickups cause by it? That's a different story. SQUIN and SWClLib are prototypes that don't work very efficiently and that still lack several features. Hence, I wouldn't recommend them for real applications that require UI-friendly response times, yet. However, the query approach in general, implemented with suitable caching mechanisms, can serve specific LD based applications quite well. Greetings, Olaf
