Learning Blueprints, should I move directly to Tinkerpop 3? [closed]

Question

I need to learn Gremlin etc. and possibly use it within a public facing production-ready system I'm working on. As any release of my system is, at a minimum, 6 months away (say end Q1 2015), should I switch directly to the new, still pre-release, Tinkerpop 3 implementation, now?

I'd rather not mis-use development time writing code against Tinkerpop 2 if it's all going to change within a year or so, however, I'm also wary of using untested code, for production purposes.

Any feedback or ideas would be very helpful ;-)

Waylonwayman

Consider posting this question on the gremlin-users mailing list: groups.google.com/forum/#!forum/gremlin-users – Puerperium

Thanks, I've posted over there, I got a quick and precise answer from Marko Rodriguez, one of the Tinkerpop lead developers, see my answer below. – Waylonwayman

I probably could have answered for you here - just trying to comply with Stackoverflow rules. Glad you got your answer. – Puerperium

Answer 1

OK, I went over to the Tinkerpop Google Group, as suggested by @stephen-mallette, and Marko Rodriguez, one of the Tinkerpop lead developers, had this to say regarding moving to Tinkerpop 3 development:

(As of 16th Sept 2014)

Here are my thoughts.

1. TinkerPop3 M2 is days away so start coding to that when it comes out.
2. If you are using Neo4j, you will be a in great position for TinkerPop3 usage -- as TinkerPop3 GA is then maybe 1 month away and Neo4j is a reference implementation of TP3.
3. If you are using any other backend, you will most likely see their TP3 implementations coming by years end. ?? (up to the vendors)
4. If you are using Titan, then Titan 0.9 will come out within 1-2 months with TP3 support.

In short, avoid using TP2 if you can.

HTH, Marko.

http://markorodriguez.com

Waylonwayman