1. 8

  2. 3

    Hm.. more directly related to the title than the post content, the future of Databases sometimes is discussed by a Mike Stonebraker and a group of people he invites for this specific reason every few years, publishing a report afterwards. The Seattle Report on Database Research is believe is the latest.

    1. 2

      The first reference on the ACM page points to a broken URL which did not contain the full report. A quick search gives the following URL which allows you to access the article without a requirement for ACM membership or payment: http://seattle-dbresearch-assessment.cs.washington.edu/

    2. 2

      “You can solve any problem by adding a layer of abstraction, except for the problem of having too many layers of abstraction.”

      In this case, I suspect a cross-database abstraction layer could conflict with scalability.

      I also wonder how one abstracts away the large differences in data models and schema between multiple database types like relational, document/JSON, graph…

      1. 1

        Cool article.

        One thing that immediately comes to mind is using Posgres as a job queue, its pretty common. The only thing that is (currently) missing is a better interface for it. This gist has been around forever and shows how fast it can really be and it’s only gotten faster due to the newer PG features being introduced, SKIP LOCKED most notably.

        If there were more client libraries and a better interface for leveraging your existing postgres instance as a queue maybe the feature would get used a lot.