Re: Refurbished draft thoughts...

Thanks both a lot for your replies. As I understand (and I'm not an expert)
you seem to be talking, in one side, about the server side of the question
(ontologies, representation 'formats') and, in the other side, about the
client side (browsers, representations 'rendering'). This may be, perhaps,
the 'inherited' way of thinking about a distributed *document* centric Web
in which the third component: protocols end up building the cake.

I don't know if such a scheme would be much constraining for a distributed
*data* centric Web in which linked resources are dynamic by nature and
where data or their schema or their interpretations / representations are
not to be known beforehand but by the means of a discovery process.

So, I ended up thinking that each client (browser) or each linked data
source (server) should have similar capabilities in terms of sharing their
'views' of knowledge and, if HTTP should be the predominant protocol, it
has to be via 'reactive streams' (dataflow) and maybe RESTFul web services,
being each node a potential producer / consumer of (data) events.

This way it should be easier to build a SW 'cake' in which interested
consumers produces content for interested producers via a 'dynamic'
criteria / pattern based discovery mechanism allowing for building
server/client applications which are rendering (UI frameworks or another
protocol, ie.: SOAP) agnostic for being them implemented as 'Nodes' also.

I think this shares resemblance with the original vision of the document
oriented Web where each home folder of a user had a 'public_html' folder or
where things as WebDAV existed to allow 'user edited content' from the
beginning but where not so popular when the web became mainstream.
On Jul 29, 2017 6:20 AM, "Dave Raggett" <dsr@w3.org> wrote:

> Thanks for sharing.  You may also want to look at W3C’s work on data sets
> and abstraction layers for services.  The Data Exchange Working Group [1]
> is refining the existing DCAT Recommendation to make it easier to discover
> and interoperate with catalogs of data sets published the Web. The Web of
> Things Interest and Working Groups [2] are focusing on enabling open
> markets of services through an abstraction layer for things that stand for
> physical or abstract entities. Both activities reduce the effort, cost and
> risk for dealing with the diversity of approaches, standards and
> technologies.
>
> [1] https://www.w3.org/2017/dxwg/wiki
> [2] https://www.w3.org/WoT/
>
>
> On 29 Jul 2017, at 07:05, Sebastian Samaruga <ssamarug@gmail.com> wrote:
>
> It's just thoughts about making what HTML did for 'web' applications but
> for SW (RDF) instead for 'whatever' applications...
>
> My point should be that SW should not have as much 'trouble' as HTML did
> for being useful having 'real' business applications (beside academia)
> running on top of it.
>
> We could skip much of the cumbersome work that was trying to build use
> cases over a distributed *document* edition system (with all the
> 'amendments' it imposed) trying to build applications over a distributed
> *data* edition and linking framework.
>
> At least for what I know we only have (lots of) protocols...
>
> https://docs.google.com/document/d/1OqsVn6uo0cr6qruzWj9yRASrmvAIA
> f4HsHuLS2aRSy8/edit?usp=drivesdk
>
> Best regards,
> Seb.
> <Business.pdf>
>
>
> Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett
> W3C champion for the Web of things & W3C Data Activity Lead
>
>
>
>
>
>