Re: Extended types in JSON(-LD)

Note that data types in RDF are really just borrowed from XML Schema, so extending the data type model is not strictly dependent on RDF at all. Depending on what you want to model, you may want to use a more complicated node value for data types, such as schema.org StructuredValue https://schema.org/StructuredValue.

But really, for any scalar datatype, the @type definition can come from either a pre-defined list such as XML Schema, or something you’ve described yourself, it’s just that RDF/OWL provides some best practices for doing such modeling. 

Gregg Kellogg
Sent from my iPhone

> On Aug 6, 2018, at 10:05 AM, Manu Sporny <msporny@digitalbazaar.com> wrote:
> 
>> On 08/06/2018 10:47 AM, Leonard Rosenthol wrote:
>> I agree with you that for JSON-LD a full
>> understanding/implementation of RDF
> 
> There are several ways to interpret this statement. I'm concerned about
> the one where folks may be insisting that most people will have to learn
> RDF to use JSON-LD.
> 
> Yes, there are some cases where you may need to understand RDF in order
> to do the right thing with JSON-LD... but if we make this the standard
> case, we'll quickly see JSON-LD shelved by developers for the same
> reasons other RDF serializations have been largely ignored by the
> general developer population.
> 
> Most JSON-LD will happily do full round trip serialization to RDF and
> back. That was by design. However, understanding why that works or the
> RDF data model is not central to authoring JSON-LD in any way shape or form.
> 
> Folks mistakenly keep thinking that JSON-LD needs RDF to exist, but
> rather it's that it exists in spite of RDF. JSON-LD can be losslessly
> round-tripped to other data models as well (YAML, for example). JSON-LD
> is an attempt to get people to use a semi-sane graph-based data model
> without knowing that they're doing it.
> 
> It's the same way that many of us use cryptography libraries without
> knowing anything about the mathematical basis for any of it. We have to
> ensure that this stays true for JSON-LD.
> 
> Insisting that people understand anything about RDF for JSON-LD to be
> useful to them is a path that leads to nowhere (and is why many RDF
> serializations failed to gain mainstream traction).
> 
> I know many of you already know all this... just repeating it in the
> event that we have new people that didn't hear it the first time we said
> this sort of thing many years ago... or have since forgotten one of the
> key design criteria for JSON-LD. :)
> 
> -- manu
> 
> -- 
> Manu Sporny (skype: msporny, twitter: manusporny, G+: +Manu Sporny)
> Founder/CEO - Digital Bazaar, Inc.
> blog: Veres One Decentralized Identifier Blockchain Launches
> https://tinyurl.com/veres-one-launches
>