Re: [Iotsi] interactive vs. programmatic IoT

[Dave Raggett <[e-mail]>]

> On 25 Mar 2016, at 18:31, Michael Koster <[e-mail]> wrote:
> 
> I think this defines a use case for what we are doing: 
> 
> "Service providers have royalty-free use of shared information models that provide access to resources hosted in various "ecosystems", where ecosystems include but are not limited to: OCF, ZigBee, Bluetooth, IPSO, OneM2M, Z-Wave and emerging ecosystems like the "native web" thing ecosystem."
> 
> What this means is we are not defining a new framework for people to build things to, rather a way forward to model common things that we can all share on a royalty-free basis to promote semantic interoperability across connected things and services.
> 
> This, practically by definition, enables application logic to be decoupled from thing specifications.

If everyone uses different frameworks for representing models, the challenge will be how to map between them. The suggestion is to use RDF as an interlingua as a way to avoid having to define direct mappings of terms between all modelling languages. This means that terms used in modelling languages need to be mapped to RDF URIs. We want to encourage people to use the same URIs for terms with the same meaning.  However, people can define their own URIs if appropriate.

—
   Dave Raggett <[e-mail] <mailto:[e-mail]>>