Hello,
I was typing away for the October2015 Podling status update and Stephen ping'd me and was like "TinkerPop doesn't have to do one this month." I had written up a nice report so I decided to share the contents with everyone as I think what we have done (and are doing) is quite impressive.
------------
TinkerPop's adoption is growing. Nearly every popular graph system vendor is "TinkerPop enabled." Interestingly enough, even RDF graph systems (historically in another area of graph computing) are starting to provide TinkerPop connectivity. I had breakfast with Kendall Clark (in Santa Fe of all places) on Sunday and he was talking about Stardog's TP3 integration and it is impressive what they are doing with TinkerPop for Stardog4 (being released this week). I will be helping them on a blog post about Stardog4+TinkerPop3.
* Neo4j
* OrientDB
* Stardog (RDF)
* Titan
* Blazegraph (RDF)
* IBM BlueMix Graph
* Sqlg
* Apache Spark
* Apache Giraph
* Apache Hadoop
Furthermore, there are many graph systems that are still TinkerPop2-enabled in the process of migrating to TinkerPop3.
Amazon (the providers of DynamoDB) just announced Titan integration for DynamoDB and that supports TinkerPop2. With Titan 1.0 just released, we will see Amazon supporting TinkerPop3. As a cloud service provider that enables a "flip of the switch" to get a DynamoDB cluster up and running, there will be lots of Tinker-and-Popping on AWS.
DataStax (the commercial providers of Apache Cassandra) just announced that they will be providing a graph system called DSEGraph whose sole interface will be TinkerPop3. They are committed to Apache TinkerPop and are banking on it for the graph aspect of their business.
With the recent publication of two articles on the "Gremlin virtual machine" we hope to see a growth in the number of languages that compile to Gremlin.
Right now, there exists SPARQL-Gremlin (proof-of-concept) which really helps to blur the distinction between the RDF and Property Graph worlds, thus expanding TinkerPop into another area of developers, consumers, promoters, and the like.
When I was at Cassandra Summit last week, I spoke with Ted Wilmes who used Apache Calcite to compile SQL to TinkerPop2. He has some initial plans to use Apache Calcite to compiled SQL to Gremlin3 and thus, we may be opening the doors to the entire SQL world to use graph technology for both OLTP and OLAP graph processing. That could be huge.
I've stated that I would like to see SPARQL-Gremlin or SQL-Gremlin ultimately as a 3rd reference language implementation merged into TinkerPop3. As such, I'm keeping a close eye on both projects to see how they evolve and see where we can help take them.
There have been 3 TinkerPop3-focused conference presentations recently:
Look at #3 above. Gremlin is the keynote at an ACM conference. This means that the academic community is realizing the benefits of TinkerPop not only from a "we can build stuff"-perspective (as we get in industry) but from a "that is a theoretically trippy concept"-perspective (as we get in academia). I have a new article I am working on for an upcoming conference that will hopefully get us tapped into another space of academics (beyond "just graphs."). The ideas in the upcoming article will be presented at GraphDay (January of 2016). I plan to demonstrate what I believe to be the craziest concept to hit the graph space yet.
It will fry brains…be there or be normal.
And that, my fellow TinkerPoppers, is my interpretation of the major accomplishments of our work here at Apache TinkerPop.
The future looks bright for TinkerPop with 3.0.2 and 3.1.0 releases coming over the remainder of this year.
Take care,
Marko.