Tinkerpop is a group of projects for manipulating a graph database. It's a mix of API and implementations, since the underlying graph storage may vary (Titan DB, OrientDB, ...) which in turn may use different storages, e.g. key-value DBs Cassandra or BerkleyDB.

In 2015, Tinkerpop was donated to the Apache Foundation.

Since then, the version 3.x got to 3.1.1-SNAPSHOT but has not been released yet, probably because it's still in the incubator.

So what's new in 3.x?

This is what the documentation says:

The thought too much to bear as he approached his realization of The TinkerPop. The closer he got, the more his world dissolved — west is right, around is straight, and form nothing more than nothing. With each step towards The TinkerPop, less and less of his world, but perhaps because more and more of all the other worlds made possible. Everything is everything in The TinkerPop, and when the dust settled, Gremlin emerged Gremlitron. It was time to realize that all that he realized was just a realization and that all realized realizations are just as real. For The TinkerPop is and is not — The TinkerPop.

Well, that's quite useful, if you want to achieve nirvana, but for practical consideration wheter to upgrade, not that much.

Here is the slideshow for a presentation called What's new in Tinkerpop 3, but that seems to be only comprehensible if you could listen to the speaker.

Here's is the Changelog, with all the detailed changes.

Edit: I found these slides which explain a lot: http://events.linuxfoundation.org/sites/events/files/slides/ApacheCon2015TinkerPop3.pdf

So, what's really new in Tinkerpop 3?

I don't really know yet   I am going to find out during my next task. But from what I can tell:

1. The Pipes API has changed to Traversal API (or maybe called Process API?), which is easier to extend. Could help us in Windup where we sometimes need to interlace the queries with data from outside of the graph. In which case I assume the optimization may suffer a lot. See this video.
2. Gremlin is now more central part. Instead of being a query language part, it's more of an environment through which you can control the server.
3. The Structure API - Graph, Vertex, Edge, Element, Property, Transaction - has moved from com.tinkerpop.blueprints to org.apache.gremlin.structure.
4. More supported databases. This rather comes with time, as more databases finished their implementation of Blueprints.
   • Bluemix Graph Data Store - IBM's OLTP graph database as a service.
   • elastic-gremlin - Elasticsearch reprsented as a graph with OLTP support.
   • Hadoop (Giraph) - OLAP engine using Giraph.
   • Hadoop (Spark) - OLAP engine using Spark.
   • Neo4j - OLTP graph database.
   • Sqlg - RDBMS OLTP implementation with HSQLDB and Postresql support.
   • Stardog - RDF graph database with OLTP and OLAP support.
   • TinkerGraph - In-memory OLTP and OLAP reference implementation.
   • Titan - Distributed OLTP and OLAP graph database with BerkeleyDB, Cassandra and HBase support.
   • Titan (Amazon) - The Amazon DynamoDB Storage Backend for Titan.

1. Some more algorithms implemented in Furnace.

Do you know more news? Let me know in comments, thanks.

By the way, thanks the Tinkerpop team for all the work.