Originally published November 2, 2013 I am old school: most of my data storage requirements used to be well met with Postgres. Now most of what I work with is graph databases of one form or another: This morning I was in my office at Google working on a Knowledge Graph schema (yeah, I know it is Saturday, so it goes...) and now that I am back "home" (*) this afternoon, I am working on a technical review for the book Neo4j In Action - Neo4j is a very fine open source (and commercial) graph database. Last night I was reviewing third party developer documentation for Facebook's Open Graph, Wolfram Alpha, and the still publicly available Freebase APIs. This is all for my "after Google" project that I have planned for later next year, if my plans don't change. I don't usually put long quotes from other people in my blog, but I saw something early this morning that really resonated with me: Darin Stewart's blog at Gartner discussing Google's K...

Neo4j is a high performance graph database that I usually use with the JRuby neo4j gem and sometimes with Clojure ( documentation here ). Neo4j is open source (AGPL v3) and is alternatively available for a reasonable fee with a commercial license (where you don't need to AGPL your project). I took advantage of the free offer to get your first Commercial Basic Server license, even though I am likely to open source my project anyway.