Re: [unrev-II] Re: Intel's new XML network devices

From: Eric Armstrong (
Date: Wed May 10 2000 - 14:38:16 PDT

  • Next message: Eric Armstrong: "Re: [unrev-II] Slashdot thread; License type"

    > Bill Bearden wrote:
    > Bias alert... I am interested in Java development. If the rest of the
    > group isn't, these aren't the way to go.
    Thanks for the info.
    My bias is towards Java, as well. After 20 some languages though,
    I'm always interested in something better.

    Python is a realistic possibility. Its easy to use. But its
    dynamic nature means that many errors are found at run time.
    That's going to require a lot of code-path testing and/or a
    willingness to live with a downtime. I don't see downtime as
    something a user community is going to live with in a communications

    Java doesn't eliminate the need for testing, but I am often
    amazed at the number of bugs I catch at compile time that I
    would have spent a week looking for back when I was coding C
    or even Pascal.

    Then there are the latest HotSpot numbers. Doing runtime optimization
    allows them to do many things a static compiler just can't do:
      --modify the code for specific processors
      --modify the code for a multi-processor system
      --unwind loops when the loop index is small
      --optimize the hell out of frequently called routines
        and ignore "one-time" code.
      --aggressively inline code
      --write error messages as though they came from the unoptimized
        version, and even debug interactively

    Despite its problems, it still looks like the best vehicle out
    there for serious hackery.

    Remember four years of good friends, bad clothes, explosive chemistry

    Community email addresses:
      Post message:
      List owner:

    Shortcut URL to this page:

    This archive was generated by hypermail 2b29 : Wed May 10 2000 - 15:06:27 PDT