OpenBSD Journal

System Test Suites

Contributed by jose on from the beat-it-til-it-dies dept.

OpenBSD has a small regression testing framework in /usr/src/regress . However, it's not a complete environment and it doesn't get used by many outside of a few core developers. Other systems, however, have more rigorous regression testing frameworks. The Posix Test Suite , for example, may reveal a few more bugs in OpenBSD. The Linux Test Project can also be adapted to find and isolate bugs in the OpenBSD system. Coding tests teached you a lot about standards, library calls, and system quality.

Is anyone working on more extended tests for OpenBSD than those that already exist?

(Comments are closed)

  1. By Todd Fries () on

    It sounds like a good solution to me. I've heard rumors about integrating posix test suites (where applicable) and other things. So long as the test code itself has an acceptable license, I would be surprised to see something outside of regress being declared 'more useful'.

  2. By Anonymous Coward () on

    What could be a better test than releasing your software to thousands of people?

    It doesn't help if you ignore bug reports, and don't fix any of the problems, so maybe that's why they need testing tools.

    1. By djm () on

      You are an idiot. The point of regress tests is to detect bugs during development cycles.

    2. By Anonymous Coward () on

      Umh, if you are talking about security, then even if you are releasing your software to thousands of people, it still needs skilled people who are willing to look at your code. Not all of thousands people are skilled and not all skilled people are willing to look at your code. And it is always better to release better quality software in the first place to get wider acceptance of your code.

  3. By Anonymous Coward () on

    I would welcome a little more light on regression testing. Maybe something in the FAQ, that told how you could install/upgrade your system, then run a standard set of tests that would help you feel that all was well. I remember once doing an upgrade and finding that netcat had a problem, that I didn't discover until later. If I had been depending on it in a cronjob, I wouldn't know until something else got messed up.

    After all, we'll never get to the Star Trek "level 5 diagnostic" or what ever it was, until we do more testing along the way. But of course, around here it seems that no one sends in their dmesg's, so who would report that regression tests failed?

Latest Articles


Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]