back to article Whaddya mean, 'niche'?! Neo4j's chief scientist schools El Reg on graph databases

The latest version of Neo4j's graph database – 4.0 – touts new scaling features and better security. The Reg talks to self-confessed graph fanboy Dr Jim Webber about how the graph-wrangler is, at last, able to scale to accommodate large databases, and about its biggest enemy: the inertia of developers who stick with SQL no …

  1. ToddRundgrensUtopia

    Cypher?? I thought Gremlin was the "SQL" equivalent for Graph?

    1. Mark 124

      Cypher is the new name for SparQL, which is much more SQL-like than Gremlin. Remains to be seen which one turns out to be VHS and which Betamax...perhaps both will be Betamax! :-)

      1. InverseFalcon

        Not sure where you got that from, Cypher is a completely different language than SparQL (you tend to see SparQL with RDF triple stores, not labeled property graphs). And while it started off with Neo4j, OpenCypher makes it available to other graph companies to use, we've seen several adopters already.

        Cypher is also going to be one of the cornerstones of the new GQL (Graph Query Language). The GQL project is the first ISO international standard database language project since SQL. So as far as the roadmap goes, that's Cypher's future, and there's considerable support to get there.

        Rather than joining tables like in SQL, Neo4j (and other graph databases) use relationships between nodes. In many cases it's much easier to express than SQL.

        MATCH (m:Movie {title:'The Matrix'})<-[:ACTED_IN]-(actor:Person)

        RETURN m as matrix, actor.name as matrixActor

        1. Mark 124

          Ooops, I stand corrected. I'm working in Gremlin so I thought it was "the other one". My brain conflated the "Cypher to GQL" as "SparQL to Cypher"

  2. stiine Silver badge
    Unhappy

    If we are going to be able to capture clickstream data from a large retailer...

    It means you've probably been tasked with solving the wrong problem. It would help if you could also see how hard I clicked on each link or button on my sometimes fruitless quest to find information.

    I recently came across a new (this should not be surprising) error message from ESXi. I pasted the error meesage into Google Search and the first 25 pages (this should be surprising, but isn't any more) were all malware-delivery pages. I could tell, no thanks to Google since they now hide URLs because they aren't important or something, by mousing over the links, because all of the URLs were along the lines of joesflowershopboston.com or xtet4seyfrgnnhmo.com.

    1. Throatwarbler Mangrove Silver badge
      Thumb Up

      Re: If we are going to be able to capture clickstream data from a large retailer...

      Really they should also tap into your microphone to detect the nature and volume of the stream of invective.

    2. Michael Wojcik Silver badge

      Re: If we are going to be able to capture clickstream data from a large retailer...

      So cynical. I'll have you know that Joe of Joe's Flower Shop Boston is my go-to expert for ESXi problems.

  3. I Am Spartacus
    Pint

    I just tried

    and failed to convince my CTO that Neo4J was the right way to go. So now I am battling SQL Server (again).

    Someone buy me a pint, I really need it just now.

  4. ratfox Silver badge
    Megaphone

    the inertia of developers who stick with SQL no matter what

    Amen! SQL is very nice for semi-competent people who are not into programming; but it's a tragedy that developers often query databases by concatenating strings to build "queries" in a 70s syntax which is about as evolved as COBOL, only to have it parsed back at the other end. Functionally and chronologically, it's the equivalent of communicating between computers using physical punch cards.

    With all the energy spent into inventing new programming languages, I'm astonished nobody took the time to invent a declarative language based on theoretical concepts. Instead, I am stuck using nested queries four deep which look like a Proust novel rather than the equations that they should be.

    1. Benson's Cycle

      I think the attitude of people like you may have something to do with resistance to new ideas by other people, especially managements.

  5. ghp

    Perhaps the RDBMS can solve a lot of problems, whilst the graph DB is still looking for a problem it might solve?

    1. Michael Wojcik Silver badge

      Nah, there are plenty of use cases for graphical DBs, just as there are for RDBMSes, and key-value stores, and object databases, and even good old hierarchical databases and hash databases and indexed files. Horses for courses.

      Some of the obvious use cases are mentioned in the article. If a graph does a good job of capturing the data you need to model, and you want to make complex queries on it, then a graph DB is often a suitable approach. (Not always, of course, and sometimes there will be other components in play. You might be working in the sort of problem space where graph sparsification is useful, for example, and you might not want to use a graph DB until after that step.)

      That said, generalizations like "queries that explore relationships, and which would be more challenging or complex to construct in relational databases" aren't of much help. Relationships are straightforward to construct in relational databases when they can be modeled as, you know, relations.

  6. Kevin McMurtrie Silver badge

    Why we use SQL DB

    Sometimes a perfect app isn't a perfect fit.

    Relational databases remain in heavy use because companies need to perform efficient queries on small and well known relationships, or need critical data in a rigid structure so it's always perfectly understood. They can also count on SQL working long after everyone retires, even if the underlying DB changes.

    I don't know of anyone today who is still using traditional SQL to brute-force graph traversal searches. It might seem like it from the database name, but they have been updated to process graph data. I'd also say that unlimited scalability is irrelevant for a small or medium sized business that can't afford much storage or computational power anyways.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2020