1. Forum
  2. Usenet
  3. LINUX.DEBIAN.BUGS.DIST

  • Bug#264582: package breaks everything else

    From Mark Robinson@1:229/2 to All on Thu Aug 12 03:30:09 2004
    From: [email protected]

    Let me try and shed some light on this debacle.

    I'm assuming that the original reporter has his priority for debconf questions set to critical or none. Consequently he doesn't see the pop up window giving him
    the option to allow some dodgy patch to mess with his filesystem. The defaul (No)
    is thus selected, and the rest of the upgrade, all following packages, not just this one, breaks.

    Completely daft.

    In my case, being a bit paranoid, I select NO, BOTH times the popup window appears, as I prefer not to let a fresh package in unstable mess around with my filesystem, at least until some other poor fool has done so first. Same effect.

    The user electing to stick with the old filesystem format should not break the installation process. This is a critical bug. The popup window itself says things
    should keep working.

    regards
    Mark Robinson
    12 Aug 1863 Gunboat Avon bombs Meremere pa.


    --
    To UNSUBSCRIBE, email to [email protected]
    with a subject of "unsubscribe". Trouble? Contact [email protected]

    --- SoupGate-Win32 v1.05
    * Origin: you cannot sedate... all the things you hate (1:229/2)
  • From Domenico Andreoli@1:229/2 to Mark Robinson on Thu Aug 12 12:10:08 2004
    XPost: linux.debian.devel
    From: [email protected]

    hi all,

    i have a problem with reiser4progs package, with version 0.5.6 the disk
    format of the filesystem changed.

    the conversion of old filesystems is performed at the first filesystem
    check, operation that should not trash existent filesystem data. the
    new disk format requires a recent patch in the kernel otherwise the
    user will not be able to mount the converted partition.

    in order to not let the user unwarned of the "live" fsck she will face
    and, most important, make her aware of the required change of kernel,
    i decided to present the classical question via debconf to stop the
    install (making preinst returning 1) at the last minute.

    i recall to have seen something like this before but my users seem to
    not like it at all.

    any suggestion?

    cheers
    domenico

    On Thu, Aug 12, 2004 at 01:03:27PM +1200, Mark Robinson wrote:
    Let me try and shed some light on this debacle.

    I'm assuming that the original reporter has his priority for debconf questions set to critical or none. Consequently he doesn't see the pop up window giving him the option to allow some dodgy patch to mess with his filesystem. The defaul (No) is thus selected, and the rest of the upgrade, all following packages, not just this one, breaks.

    Completely daft.

    In my case, being a bit paranoid, I select NO, BOTH times the popup window appears, as I prefer not to let a fresh package in unstable mess around
    with my filesystem, at least until some other poor fool has done so first. Same effect.

    The user electing to stick with the old filesystem format should not break the installation process. This is a critical bug. The popup window itself says things should keep working.


    -----[ Domenico Andreoli, aka cavok
    --[ http://people.debian.org/~cavok/gpgkey.asc
    ---[ 3A0F 2F80 F79C 678A 8936 4FEE 0677 9033 A20E BC50


    --
    To UNSUBSCRIBE, email to [email protected]
    with a subject of "unsubscribe". Trouble? Contact [email protected]

    --- SoupGate-Win32 v1.05
    * Origin: you cannot sedate... all the things you hate (1:229/2)
  • From Steve Greenland@1:229/2 to Domenico Andreoli on Thu Aug 12 16:10:05 2004
    XPost: linux.debian.devel
    From: [email protected]

    On 12-Aug-04, 05:06 (CDT), Domenico Andreoli <[email protected]> wrote:
    in order to not let the user unwarned of the "live" fsck she will face
    and, most important, make her aware of the required change of kernel,
    i decided to present the classical question via debconf to stop the
    install (making preinst returning 1) at the last minute.

    i recall to have seen something like this before but my users seem to
    not like it at all.

    any suggestion?

    Ship both the old and the new version of fsck.reiser4 (or whatever it's called). Use update-alternatives in the postinst to set which one is
    actually used depending on the answer to the debconf question.

    OR

    Ship a wrapper script that checks to make sure the kernel version is
    suitable before running the new fsck.reiser4.real.

    Steve


    --
    Steve Greenland
    The irony is that Bill Gates claims to be making a stable operating
    system and Linus Torvalds claims to be trying to take over the
    world. -- seen on the net


    --
    To UNSUBSCRIBE, email to [email protected]
    with a subject of "unsubscribe". Trouble? Contact [email protected]

    --- SoupGate-Win32 v1.05
    * Origin: you cannot sedate... all the things you hate (1:229/2)