- From: Martynas Jusevicius <martynas@graphity.org>
- Date: Mon, 10 Mar 2014 19:17:39 +0100
- To: Jeremy J Carroll <jjc@syapse.com>
- Cc: Michael Brunnbauer <brunni@netestate.de>, semantic-web <semantic-web@w3.org>
Hey all, for me it helps to illustrate the advantage of RDF universality using 2 idioms: - hub-and-spoke that leads to linear integration costs: http://en.wikipedia.org/wiki/Spoke-hub_distribution_paradigm - pivotal data conversion with RDF in the center: http://en.wikipedia.org/wiki/Data_conversion#Pivotal_conversion RDF is like Unicode of data models. Since graph data model can accomodate any type of data and RDF is the only standard in this area, other similar models will be isomorphic but non-standard reinventions at best. So lets focus on RDF and build it even deeper into the software ecosystem, so we can finally produce some user-friendly yet generic applications. Martynas graphityhq.com On Mon, Mar 10, 2014 at 6:37 PM, Jeremy J Carroll <jjc@syapse.com> wrote: > > On this excellent thread, there is little else to say ... but still I will > have a go > > On Mar 9, 2014, at 7:48 AM, Michael Brunnbauer <brunni@netestate.de> wrote: > > The claims regarding interoperability and semantics are a bit exaggerated, > IMO. > > If we had something like annotated portable RDB schemas, would they carry > less > meaning and would applications built with them be less interoperable than > with > MLHIM? > > > > Maybe an advantage of RDF is that we can partially align schema, whereas in > a more traditional world, either we do the work to do a full alignment, or > we don't do it at all. Since, often, the full alignment is cost prohibitive > we end up with data silos > > So my contribution to this discussion is the word 'partial' > > Jeremy >
Received on Monday, 10 March 2014 18:18:06 UTC