reopen 605127 thanks On Samstag, 27. November 2010, Holger Levsen wrote: > I'm using this code in the post_purge_exception script for testing file-rc > on piuparts.d.o now: > > echo "Debug: this files don't exist when sysv-rc is installed by > debootstrap, they are just created when sysv-rc is installed later. See > #605127." rm -fv /var/lib/update-rc.d/* > > (The piuparts test first creates a chroot with debootstrap, then file-rc > installation is done there, then file-rc is removed and sysv-rc is > reinstalled, thats when those files in /var/lib/update-rc.d/ are created. > As they didnt exist when the chroot was created, piuparts would otherwise > detect them as new files.) I dont think this is the right approach anymore, /var/lib/update-rc.d seems to be generally populated on upgrades from lenny to squeeze, so I now think piuparts should just ignore /var/lib/update-rc.d(/.*) completly. Comments? See http://piuparts.debian.org/lenny2squeeze/fail/whereami_0.3.34-0.3.log for an example, or: holger@piatti:/org/piuparts.debian.org/master$ grep -l "/var/lib/update-rc.d/" */fail/* lenny2squeeze/fail/buildbot_0.7.12-2.log lenny2squeeze/fail/file-rc_0.8.12.log lenny2squeeze/fail/gfs-tools_3.0.12-2.log lenny2squeeze/fail/lcdproc_0.5.2-3.log lenny2squeeze/fail/netscript-2.4_5.2.2.log lenny2squeeze/fail/rgmanager_3.0.12-2.log lenny2squeeze/fail/whereami_0.3.34-0.3.log It also seems to be rather new behaviour, as I havent noticed this before and those logs are max. 2 days old. I'd appreciate feedback. cheers, Holger
Description: This is a digitally signed message part.