torsdag den 2 december 2010 klockan 15:38 skrev Alexander Reichle-Schmehl detta: > Hi! > > Am 02.12.2010 15:26, schrieb Niko Tyni: > > >>> The conffile modification fix for -7 doesn't actually invoke ucf > >>> to install /etc/webfsd.conf at all: > [..] > >> I just installed webfs_1.21+ds1-7 on a squeeze/kfreebsd-amd64 system > >> that never has seen webfs before. It installed perfectly and get up > >> and running as it should. Similarly, my older test system > >> lenny/linux-i386 also installs the packag perfectly in long > >> runs of purges, installs and possibly upgrades. > > Weird. Does ucf really get run for you? I assume you don't have an > > executable file named "ucf" in the root directory? :) > > Sorry that this slipped through when I reviewed the package for > sponsoring the upload. > > I can reproduce the problem in a clean pbuilder chroot on sid amd64:
Why did it work out on kfreebsd-amd64? Because I did it manually from the command line? > > # apt-get install webfs > [..] > Setting up webfs (1.21+ds1-7) ... > ************************************ > All rc.d operations denied by policy > ************************************ > invoke-rc.d: policy-rc.d denied execution of start. > # ls /etc/w > ls: cannot access /etc/w: No such file or directory > > # test -x ucf > # echo $? > 1 > # test -x /usr/bin/ucf > # echo $? > 0 Yes, I see the point and I agree. The amusing thing is that my original formulation used the full path in the test clause, as well as in the execution call. Howover, lintian complained about the use of a full path. Then I clumsily removed both of them. There will be a new packaging this evening. In my ignorance, I worry about the version number. Pushing a new version into unstable, while the present version is never allowed to reach testing, the handling of this is beyond my experience. -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org