Your message dated Tue, 25 Nov 2008 19:54:17 +0100
with message-id <[EMAIL PROTECTED]>
and subject line Re: Bug#506840: exim4-daemon-heavy: relative symlinks may have
negative effects (/usr/lib/exim4/exim4)
has caused the Debian Bug report #506840,
regarding exim4-daemon-heavy: relative symlinks may have negative effects
(/usr/lib/exim4/exim4)
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [EMAIL PROTECTED]
immediately.)
--
506840: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=506840
Debian Bug Tracking System
Contact [EMAIL PROTECTED] with problems
--- Begin Message ---
Package: exim4-daemon-heavy
Version: 4.63-17
Severity: minor
(Report is about 4.63-17 but as far as I see it si present in 4.69-9 as
well.)
There are some relative symlinks in the package tree:
debian:/storage/tmp# ls -la /usr/lib/{exim4,sendmail}
lrwxrwxrwx 1 root root 13 2008-10-30 10:28 /usr/lib/sendmail -> ../sbin/exim4
/usr/lib/exim4:
total 256
drwxr-xr-x 3 root root 52 2008-10-30 10:38 .
drwxr-xr-x 402 root root 131072 2008-11-21 13:48 ..
lrwxrwxrwx 1 root root 16 2008-10-30 10:28 exim4 -> ../../sbin/exim4
-rwxr-xr-x 1 root root 65300 2008-09-30 20:27 eximon.bin
drwxr-xr-x 2 root root 6 2004-03-28 19:49 local_scan
debian:/storage/tmp#
On a particular machine there are some symlinks around:
particular:/usr/lib/exim4# ls -la /usr/
total 64
drwxr-xr-x 10 root root 4096 2008-05-20 14:13 .
drwxr-xr-x 18 root root 4096 2008-05-20 14:08 ..
drwxr-xr-x 2 root root 20480 2008-09-02 15:09 bin
drwxr-xr-x 2 root root 4096 2008-05-20 12:56 doc
drwxr-xr-x 2 root root 4096 2008-05-20 12:56 games
drwxr-xr-x 37 root root 8192 2008-08-14 11:38 include
lrwxrwxrwx 1 root root 13 2008-05-20 14:13 lib -> /var/usr2/lib
drwxr-xr-x 3 root root 4096 2008-05-20 12:57 lib64
lrwxrwxrwx 1 root root 15 2008-05-20 14:09 local -> /var/usr2/local
drwxr-xr-x 2 root root 8192 2008-09-02 15:09 sbin
drwxr-xr-x 125 root root 4096 2008-05-20 14:13 share
lrwxrwxrwx 1 root root 13 2008-05-20 13:40 src -> /var/usr2/src
drwxr-xr-x 3 root root 4096 2008-05-20 12:55 X11R6
This causes exim not to start, since doing ../../sbin on /usr/lib/exim4 ends up
in /var/usr2/sbin which in turn does not exist.
Now, I do not want to argue about this being a _bug_, because frankly I cannot
really
tell.
First, it should be a bug since the package should not depend on
assumptions outside his scope (namely that a directory is or isn't a
symlink).
Second, this should not be a bug since a manually unchanged Debian
should never contain symlinks on base directories.
Third, this should not be a bug since Mr. Problem-between-keyboard-and-chair
should
not selectively symlink the directories (which, I'm happy to share: wasn't me).
Still, if there wasn't a good reason to have it relative, you might want to
make them absolute symlinks, because it would not possess this problem.
Thanks for reading my medium-length essay. ;-)
(Symlinks are fun things.)
--- End Message ---
--- Begin Message ---
On 2008-11-25 Peter Gervai <[EMAIL PROTECTED]> wrote:
> Package: exim4-daemon-heavy
> Version: 4.63-17
> Severity: minor
> (Report is about 4.63-17 but as far as I see it si present in 4.69-9 as
> well.)
> There are some relative symlinks in the package tree:
[...]
> /usr/lib/exim4:
[...]
> lrwxrwxrwx 1 root root 16 2008-10-30 10:28 exim4 -> ../../sbin/exim4
[...]
> On a particular machine there are some symlinks around:
> particular:/usr/lib/exim4# ls -la /usr/
> lrwxrwxrwx 1 root root 13 2008-05-20 14:13 lib -> /var/usr2/lib
[...]
> This causes exim not to start, since doing ../../sbin on
> /usr/lib/exim4 ends up in /var/usr2/sbin which in turn does not
> exist.
> Now, I do not want to argue about this being a _bug_, because
> frankly I cannot really tell.
> First, it should be a bug since the package should not depend on
> assumptions outside his scope (namely that a directory is or isn't a
> symlink).
> Second, this should not be a bug since a manually unchanged Debian
> should never contain symlinks on base directories.
> Third, this should not be a bug since Mr.
> Problem-between-keyboard-and-chair should not selectively symlink
> the directories (which, I'm happy to share: wasn't me).
> Still, if there wasn't a good reason to have it relative, you might want to
> make them absolute symlinks, because it would not possess this problem.
> Thanks for reading my medium-length essay. ;-)
Hello Peter,
I have got a fourth point of data. ;-) Policy 10.5
http://www.at.debian.org/doc/debian-policy/ch-files.html#s10.5
I think especially in case of uncertainity going with policy is the
right thing to do.
cu and- bind mounts should work better for you than symlinks -reas
--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'
--- End Message ---
