Re: who compiled mutt, and where?
On Wed 23 Dec 1998, Christopher C Chimelis wrote:
> On Wed, 23 Dec 1998, Paul Slootman wrote:
> > The last couple of mutt versions have a postinst that call
> > suidregister with numeric values for user and group instead of
> > the names. This results in the nightly suidregister-check run
> > to complain that mutt_dotlock was 0.8 and now root.mail ...
> > I tried building mutt myself, and _that_ postinst has the names,
> > not the numbers.
> > Maybe you're using a buggy version of debhelper / whatever?
> Hmmm...I think I built the last one on a.d.nl. I wonder how that
> happened. I'll check the version of debhelper on there today. In the
> meantime, care to do an NMU? :-)
Actually, I tested it on a.d.nl myself :-) That works fine.
$ ls -lc /usr/bin/dh_suidregister
-rwxr-xr-x 1 root root 1831 Dec 22 08:49 /usr/bin/dh_suidregister
Looks like debhelper was upgraded on a.d.nl yesterday morning!
I don't think that it's that important to do an NMU; I'm sure there'll
be another version of mutt to frozen before the deepfreeze :-(
There are other things that also need NMU's for simple recompiles to
eliminate obsolete dependencies; libguile-1.2 is gone, and so is
libguile-1.3 (or is almost). Any others? Anyway, I think that we
shouldn't get into a panic about this, seeing the rate of uploads into
frozen for i386; most will probably be rebuilt in the course of things
Everybody just make sure that they have the correct (latest) -dev
packages installed, before compiling stuff! (Yes, I've been guilty
Chris, are you going to do an upload of libc6.1-dev? If so, could you
fix the following line in /usr/include/netinet/ip_fw.h:
u_int32_t fw_pcnt, fw_bcnt; /* Packet and byte counters */
The u_int32_t should be unsigned long; that's what it is in the kernel
source. This is what's screwing ipfwadm.
home: firstname.lastname@example.org | work: email@example.com | debian: firstname.lastname@example.org
http://www.wurtel.demon.nl | Murphy Software, Enschede, the Netherlands