Re: Should I report a failure to boot?
-------- Original-Nachricht --------
> Datum: Fri, 21 Sep 2007 12:47:53 +0530
> Von: Kapil Hari Paranjape <kapil@imsc.res.in>
> An: Michael Burschik <michael.burschik@gmx.de>
> CC: debian-qa@lists.debian.org
> Betreff: Re: Should I report a failure to boot?
> Hello,
>
> On Fri, 21 Sep 2007, Michael Burschik wrote:
> > if my Debian system fails to boot, should I report this as a bug? If
> > so, against which package should I file the bug and which severity would
> > be appropriate?
>
> As long as you can clearly identify that this is not due to "operator
> error" (:-)) it is indeed a bug against some package/packages.
The kernel and other updated packages were installed by synaptic, and at the point at which the boot fails, the only thing I have done is to press the power on switch. For this reason, I consider an "operator error" unlikely.
> However, in order to file the bug correctly it helps if you can
> identify the package which is causing the failure. I admit that in
> the case of a non-booting system this may not be entirely trivial.
>
> > As the kernel is unable to find its root device
> > due to the fact that device names change in a semi-random manner, I
> > filed a bug against the kernel, but it was summarily dismissed on the
> > grounds that the kernel does not guarantee persistent device names.
>
> Which indicates that the bug is not with the kernel.
I am willing to concede that it is not a kernel bug. However, I maintain that it is inappropriate for the owner of the kernel package to dismiss the bug report. If Debian wants to receive bug reports, the maintainers need to become more cooperative.
> > However, as a Debian user, I really don't care whether this is,
> > strictly speaking, a kernel bug or not.
>
> But to file a bug report which can result in appropriate corrective
> measures (and indeed that is the purpose of filing a bug report ---
> though occasionally one succumbs to the temptation to file a bug
> report based one's annoyance at that %@#-thing that refuses to
> function properly), it is necessary to identify correctly the source
> of the problem.
I would claim that most users are unable to correctly identify the source of most problems without active help from maintainers. I once filed a bug report against kooka and xsane, because they did not work. The maintainer realized that the bug was not due to the applications themselves, but to libsane. He sensibly reassigned the bug report. This is the kind of behaviour I would expect from a responsible maintainer. At the very least, the maintainer should suggest which package might be the real culprit.
In this particular case, I am unable to identify the source of the problem. Is it the kernel? Why does the kernel change device names on reboot, which never happened before? Is it yet another jmicron problem? Is udev responsible? But I never configured udev, because I saw no need to do so. Did the installer configure udev incorrectly? Does the installer ignore the fact that device names can change on reboot?
> > If it is not the kernel, then it must be
> > udev, which should ensure that device names do not change randomly
>
> Which is does *if* an appropriate configuration file for persistent
> names has been created.
I'm not claiming that udev doesn't work, only that the installer did not set it up correctly (if it must be taken for granted that disk names can change on reboot).
> > the installer, which should have configured the system in such a way
> > that it continues to function if device names change on reboot
> > (provided that this behaviour must be expected).
>
> If indeed the partition to which the system was installed was created
> by the Debian installer and the boot mechanism was installed by the
> installer then this does seem to point to a bug in the installer.
>
> > I do not expect you to fix my system. That is my responsibility. But I
> > don't see why I should waste my time filing bug reports that are not
> > taken seriously.
>
> Isaac Asimov (I think) once wrote an entire short story based on the
> theme:
>
> To ask the correct question one must already know most of the
> answer.
>
> This happens to also apply to bug reports. However, please be patient and
> try again --- we will all benefit as a result of your patience and
> effort.
>
> Regards,
>
> Kapil.
> --
What am I supposed to try again? File a bug report against the kernel? File a bug report against some other component? If so, which one?
Thank you for taking the time to reply to my complaints. I am willing to pursue the bug further, but I do need some suggestions how to do this.
Regards
Michael Burschik
--
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail
Reply to: