Re: What are the minimal tests that should be performed when building packages?

From: Mark Gilbert (webmaster@dofty.zzn.com)
Date: Wed Jun 05 2002 - 20:25:04 EDT

  • Next message: Martin Sevior: "Zeroth order table Table support."

    Greetings.
    I've been toying with such test suite off and on for some time, with
    automation a primary goal. It still doesn't amount to much (hey, at
    least it can detect a missing lib). Unfortunately, it is one hundred
    percent not portable beyond unix.
    Basically, for manual win32 testing, the list should include but not be
    limited to:
    1) Did it build cleanly?
    2) Does it run?
    3) Can it imp/exp a simple document in X format?
    4) Is there goo?
    5) Asserts? (I'm not sure how this works on win32)
    6) Are significant known bugs from the last version fixed? (This part
    of regression testing should be done by hand if possible)

    And I'm sure there are others for win, I just can't think of them
    easily.

    Regards
    -MG

    On Wed, 2002-06-05 at 17:08, Kenneth J.Davis wrote:
    > Hello all,
    >
    > This sorta goes along with the regression testing discussion a while back.
    > As I occasionally build for Windows, I would appreciate if someone could
    > help direct me to [or suggest] a list of specific tests I should perform before
    > making the build available. I am referring more to a minimal set of tests
    > [after all, I only have so much time] and not a complete suite that
    > would best be done by 3rd party QA testers (aka AbiWord users)
    > [though if it does exists then I am interested in it as well, or could
    > make my own minimal set from it].
    >
    > Also, was an abiword-qa (or similar) mailing list ever created? or is there
    > any plan to create such a list?
    >
    > Off topic, but hurray! :-) tinderbox is back! :-)
    >
    > Thanks,
    > Jeremy Davis
    > jeremyd@computer.org



    This archive was generated by hypermail 2.1.4 : Wed Jun 05 2002 - 20:26:30 EDT