severity 383201 critical
clone 383201 -1
reassign -1 gnome-commander
stop
Hi,
I'm rebumping the severity to prevent this package to transition to
testing before this bug is fixed. The proposed patch I've sent to the
bug will force a scrollkeeper database rebuild to all people who are
upgrading from a version bigger or equal to the borken version (since
there's no way to know whether they had the broken version installed at
one point of time): if you let this version enter testing without the
patch, and add the patch later on, all users of testing will witness a
scrollkeeper rebuild; however, if you do apply the patch now, only
unstable users will witness the rebuild.
On Sun, Aug 20, 2006, Martín Ferrari wrote:
> Hi, I'm assigning severity important to this bug, as I think very few
> people could be affected by it, as it was only one day in the archive
> and never reached testing, so it could not affect etch's release.
> Besides, scrollkeeper has a monthly cronjob to rebuild its database.
> And finally, if this required for tomboy, the same should happen with
> gnome-commander which had this same bug for more time.
The GNOME team probably maintains the biggest number of GNOME packages;
these usually provide Help pages; we get random bugs of "Help doesn't
work" when packages in the archive overwrite the scrollkeeper database
like this.
I agree with your point that "if this is required for tomboy, the same
should happen with gnome-commander", but I don't see how this applies
to the severity of the bug against tomboy.
I also agree that the impact is mitigated by the cron and the time the
package has spent in the archive, but this is a bug nevertheless, and
applying the proposed postinst snippet soonish would have mitigated the
impact even more.
I hope the rationale above is convicing, and I'm willing to NMU on the
basis of the proposed patch; please let me know whether this wouldn't
be welcome. I hope you're not upset with the severity change which I
consider appropriate, and won't touch it anymore if you disagree
with my points.
--
Loïc Minier <[EMAIL PROTECTED]>
