Subject: Re: pkg/36256: security/seahorse fails to install properly (NetBSD 3.0)
To: None <pkg-manager@netbsd.org, gnats-admin@netbsd.org,>
From: Robert Elz <kre@munnari.OZ.AU>
List: pkgsrc-bugs
Date: 05/02/2007 11:40:02
The following reply was made to PR pkg/36256; it has been noted by GNATS.
From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: pkg/36256: security/seahorse fails to install properly (NetBSD 3.0)
Date: Wed, 02 May 2007 18:38:32 +0700
Date: Wed, 2 May 2007 09:30:02 +0000 (UTC)
From: Thomas Klausner <wiz@NetBSD.org>
Message-ID: <20070502093002.9F85263B8BD@narn.NetBSD.org>
| seahorse-daemon is in the PLIST, so it probably should exist. Please
| check why it isn't built.
It looks to be being built OK (make output from scripts that are full
of shell control statements are never really easy to comprehend, without
knowing which path through the maze was taken, anything is possible).
But at least, it certainly looks to be going through a path which could
build seahorse-daemon and not complaining anywhere there.
However, there is no sign of any attempt to install it, before the
attempt (that failed) to chmod the (not) installed file.
I notice that seahorse-daemon is (apparently) compiled in
.../seahorse/default/seahorse-1.0.1/daemon
which looks to be a very plausible location for it to live.
On the other hand, the chmod that fails is in
.../seahorse/default/seahorse-1.0.1/agent
Something in the install there is attempting to chmod seahorse-daemon.
seahorse-daemon may just not be installed yet (there may be some sequence
error in the makefiles or install recipes), or there might be a typo in
the seahorse-agent (assuming there is such a thing, I really know nothing
about it) install script, and it should be chmoding seahorse-agent at
that point, or ...
This is about all I know about it, though it is fairly easy to build it
all again (and at least as long as my system stays up and /tmp remains,
I do have the build logs from several attempts to build this - I tend to
keep trying to build things over and over hoping someone else will notice
the problem and I won't need to send a PR ...)
Because I use pkg_comp, which always does "make clean" (succeed or fail)
I don't have the corpse from the make attempt, but I could build it
manually if no-one else can reproduce what I am seeing.
kre