rawhide report: 20060714 indent scriptlet error

Tom Brinkman tbrinkman at sbcglobal.net
Tue Jul 18 13:00:41 UTC 2006


On 17 8:40:S, Jim Cornette wrote:
> Tom Brinkman wrote:

> >   Yes, that's what I've decided for updates with a new kernel,
> > do the kernel separately, and I have for the last two.  Just
> > now I checked for updates an there were 17, no kernel. 'yum
> > update' was successful, except for....
> >
> >   Updating  : indent             #######################
> > [13/34] install-info: /usr/share/info/indent.info.gz: empty
> > file error: %post(indent-2.2.9-13.fc6.i386) scriptlet failed,
> > exit status 1   .... and there was no 'Cleanup' statement for 
> > indent
> >
> >   Then running the script to chk for dups.....
> >
> >  # yumdups
> > Duplicates were found:
> > indent-2.2.9-12.3.1
> > indent-2.2.9-13.fc6
>
> That is the trouble with %post scriptlet failures. %post leaves
> trails and %pre only goes up to where it fails and does not
> install the rpm.
>
> I avoided indent, m4, kdebase, gnome-vfs2 and a few other
> packages where there are dependency errors. It's good that you
> noticed the failure with indent and the scriptlet. Is there a bug
> report filed against this package?
>
> Jim

    No.

    While I've used Linux for over a decade, test distros for much 
of that (an BZ's), I'm fairly new to FC-test. I'm not about to file 
a bug, either very quickly, or on every packaging error. Better to 
see if the next indent update fixes it. I did post the same indent 
message to the -devel list tho in a thread discussing causes of 
dups.

    I haven't had dep errs on the rpms you list, an particularly not 
for indent. Mine are currently elf\*, ltrace, an sysfsutils.  An 
they've lasted a few days more than normal, even for rawhide.

    I should say tho that FC devel is way ahead of, better organized 
an run, an with fewer 'niggles' than I'm used to with other 
distros.  Should'a come here sooner ;)
-- 
    Tom Brinkman                   Corpus Christi, Texas




More information about the fedora-test-list mailing list