Re: Mad idea: Programming language based on RDF

Thanks, David. "Some pretty cool ideas" is how I would characterize the
early years of the Linked Data movement. It was a kind of Cambrian
explosion of possibilities, some of which settled down and connected with
real world needs. Although this is probably the first time you have heard
of Ripple, I am sure you have heard of its cousin Gremlin [1]. TinkerPop,
Gremlin, and property graph databases like Neo4j and OrientDB became
popular because they gave developers the essential power of the Semantic
Web -- knowledge representation and integration, knowledge-based queries --
without the baggage of rigid and complicated standards.

However, as I argued in a recent talk at Data Day Seattle, the graph
database landscape keeps moving further in the direction of the Semantic
Web, albeit with the same disregard for standards. Every major graph
database vendor has some notion of a schema, which is basically a
lightweight ontology. SPARQL-like graph pattern languages, notably Neo4j's
Cypher, have become popular. Has anything like Linked Data been emerging in
the graph DB space? I have not seen it, but I think it should come as no
surprise when we do, and a lot of what was discovered to be true of Linked
Data may be applicable there.

It should also come as no surprise when developers start compiling their
property graph queries "into the graph" for later retrieval, execution, and
analysis. That is one important feature that never made it into Gremlin,
but which may be a worthwhile goal for TinkerPop 4. It's really interesting
to see all of the recent work in this area.

Josh

[1] https://en.wikipedia.org/wiki/Gremlin_(programming_language)


On Thu, Nov 9, 2017 at 8:13 AM, David Booth <david@dbooth.org> wrote:

> On 11/09/2017 02:13 AM, Joshua Shinavier wrote:
>
>> A general-purpose programming language based on RDF? Good idea (no,
>> really). [ . . . ]
>>
>> https://cdn.rawgit.com/joshsh/ripple/develop/doc/screencast/index.html
>>
>
> Nice demo!   And some pretty cool ideas.  Thanks for sharing.
>
> David Booth
>
>