Re: 2 cents

From: Henning Meier-Geinitz (hmg-ml@gmx.de)
Date: Fri Nov 03 2000 - 10:00:47 PST

  • Next message: Henning Meier-Geinitz: "Re: detection problem"

    Hi,

    On Thu, Nov 02, 2000 at 10:57:11AM -0600, Arno A. Karner wrote:
    > split into two packages majorver.api-version.patch level
    > keep even api versions stable, odd ones development-beta
    > thats been working for the linux kernel since i can remember
    > this allows you to output relativly stable test stuff because it will be
    > up to the developers when to cut a release on odd numbers,

    I don't think that this is necessary for SANE. Release cycles are
    relatively short (compared to Linux kernel) and we shouldn't split our
    (short) time on too many targets. I think the current model works
    pretty well. However, we will probably have two different source trees
    for a while when (if?) SANE 2 development starts.

    > people who
    > test won't have to deal with cvs, and will give them a solid tag for bug
    > reports. instead of i think i checked it out at noon, noon ? what time
    > zone are you in?

    Is this really a problem? CVS checkins aren't that often and I don't
    remember reading a report dealing with this kind of problem. However,
    we may start releasing snapshots again.

    > i was using 0.73 on redhat 5.2 was working fine, upgraded my boxes to
    > 6.2 now that 7.0 is out, got 1.0.3 ran configure, it compiled, it
    > installed, both xcam and xscanimage core dump with nasty gtk error
    > messages

    Could you send the error messages? Maybe your device wasn't found. In
    this case something about GTK main loop is printed.

    You may also try xsane.

    Ciao,
      Henning

    --
    Source code, list archive, and docs: http://www.mostang.com/sane/
    To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
    



    This archive was generated by hypermail 2b29 : Sat Nov 04 2000 - 01:54:04 PST