Announcement

Collapse
No announcement yet.

PostgreSQL 11 Is Continuing With More Performance Improvements, JIT'ing

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • PostgreSQL 11 Is Continuing With More Performance Improvements, JIT'ing

    Phoronix: PostgreSQL 11 Is Continuing With More Performance Improvements, JIT'ing

    PostgreSQL 11 is the next major feature release of this open-source database SQL server due out later in 2018. While it's not out yet, their release notes were recently updated for providing an overview of what's coming as part of this next major update...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    I was baffled when I found out Postgres doesn't have an issue tracker.

    Other than that, a great project.

    Comment


    • #3
      I wish PostgreSQL had a real, standard open source license instead of a custom vanity license.
      I wish PostgreSQL had a sane userspace which was intuitive, coherent and predictable and didn't clutter the file system.

      It would be cool to be able query JSON using XPath 3.1.
      Last edited by uid313; 23 May 2018, 10:25 AM.

      Comment


      • #4
        Originally posted by uid313 View Post
        I wish PostgreSQL had a real, standard open source license instead of a custom vanity license.
        I wish PostgreSQL had a sane userspace which was intuitive, coherent and predictable and didn't clutter the file system..
        I do hear what you are saying, I also like things kept simple. Maybe there is value in keeping the binaries individual from each other, which I suppose would require a name for each. Certainly if they were named consistent, prefix all with "pg_" as the name, not all the non-intuitive names they currently use (yes some binaries do prefix with "pg_" already, but others do not - I always thought postmaster was for a mail server!!) But the current situation is probably a historical artifact, and that is what we have. I suppose they could also go the route like git where where have the main git command with sub-commands, called as arguments/parameters to the main git command.

        In either case, if they were going to make the switch, they would still need to symlink from the current commands to whatever they changed to.
        Last edited by ehansin; 23 May 2018, 01:40 PM.

        Comment


        • #5
          Originally posted by uid313 View Post
          I wish PostgreSQL had a real, standard open source license instead of a custom vanity license.


          Sorry PostgreSQL license is not vanity there are quite a few projects that do use the PostgreSQL licence.



          Postgresql license was based off the BSD license. Do note at the time Postgresql was release the BSD 4 term license was in usage but the project started when the first BSD license was in usage at the university. Both of the early BSD licenses have issues for those providing commercial support.


          Most people don't see that Postgresql comes out of Berkeley.

          So the PostgreSQL license is a hang over of a historic license issue. The one thing about Postgresql license is there is no on going veneration its the same today as what was published in 1996. Even projects that use the Postgresql license have taken it exactly as is. Postgresql license is one of the most stable open source licenses.

          Postgresql license is a real standard open source license. Fedora project guy who deals with legal issues in the license audit on fedora project was in fact how shocked how often Postgresql License is used and how stable it is. MIT and BSD licenses have hundreds of variations each with their own quirks. Yes bigger number of projects are under MIT or BSD license collectives but if you break that down by variation there is more project under Postgresql license than are under any of the BSD/MIT individual variations. Postgresql License is most likely the most successful variation of BSD/MIT licenses.

          Open source licensing is one hell of a mess. Yes there are vanity licenses but Postgresql License is not one of them as there was legal reasons for different license create back when the project started. Changing license is not a simple process and can be highly disruptive.

          I would say on the license thing you could call Postgresql Licence legacy hang over due to historic legal problems with existing licenses at the time. The number of open source licensing in 1996 was not as large as today.

          A pure vanity license you cannot find any clear legal requirement for the different license to be created when it started. I will admit there are quite a few filling out open source pool of licenses that are vanity.

          There are a heck load of MIT and BSD licenses variations in this class of vanity where the wording of the license has been modified without altering legal meaning that are basically pure vanity licenses.

          Originally posted by uid313 View Post
          I wish PostgreSQL had a sane userspace which was intuitive, coherent and predictable and didn't clutter the file system.
          When postgresql is embedded in programs it manages not to clutter file system. So some of the clutter file system is distribution choices for backwards compatibility so this is a more than two to tango problem. Very much a hurd of cats problem and attempting to move the hurd will be a fast way to get scratched to death.

          Comment


          • #6
            Originally posted by cen1 View Post
            I was baffled when I found out Postgres doesn't have an issue tracker.

            Other than that, a great project.
            Could this be considered an "issue tracker"? https://www.postgresql.org/list/pgsql-bugs/
            Last edited by rudregues; 25 May 2018, 04:04 PM.

            Comment

            Working...
            X