Hacker News Reader

Comment by CharlieDigital | Hacker News Reader

Glad they used Cypher. One of the best and most intuitive query languages I've used personally.

I had to use neo4j on a project some years ago and hated it (it didn’t help that the client forced its use for really dumb and bad reasons, wasting tons of time in the process) but Cypher is great. Only thing about it I liked. It’s very good.

Yeah Cypher was so good and made everything so intuitive. Loved working with Neo4J for a prototype and then had to face the licensing issues and huge costs and had to abandon it.

I got bitten by the same, we were looking at switching to memgraph (which is also Cypher, but C++ rather than Java), sadly abandoned for other reasons ...

What reasons? Would love to know for future ref.

If your org is amenable to Google Spanner or already using Spanner, it has preview support for Cypher and graph semantics: https://cloud.google.com/spanner/docs/graph/overview

Unrelated to the tech, just the graph-based algorithm (recommender stuff) that we had in mind did not give us the results we were hoping for ... and other things were queued up to be tried ...

There are open source alternatives to Neo4J, and also very good scalable open source graph databases with query languages similar to Cypher. I think we are in agreement that it is a good idea to choose platform that scales both for compute and all other costs, like licensing costs.

Can you list them?

Not sure if you everyone reading is aware, but Neo4j did end up turning Cypher into the openCypher spec and that allowed a lot of other graph stores to adopt Cypher as a query language.

I agree with you about Cypher. For a couple of decades I was a RDF and then RDF + SPARQL advocate, but I am a fairly recent convert to property graphs and Cypher.

I was going to say, this is what SPARQL was designed for and it's much better at than SQL.

Reply on news.ycombinator.com