t only that, it also breaks if the conffile has been moved to another
package. Today's systemd upgrade surprised me by deleting two conffiles
of systemd-timesyncd:
,
| systemd (247.9-2) wird eingerichtet ...
| Entfernen des veralteten Conffiles /etc/dhcp/dhclient-exit-hooks.d/timesyncd
...
Package: dpkg
Version: 1.20.9
Severity: normal
X-Debbugs-Cc: Niels Thykier
Hi,
the new remove-on-upgrade logic from #822462 breaks the use-case of
replacing a conffile by a non-conffile version.
In postgresql-common, I'm trying to remove
/etc/apt/apt.conf.d/01autoremove-postgresql as a
the packages.
I also include example patches to base-passwd and mawk (gawk and
original-awk need almost identical changes to mawk)
The core feature is to add support for a flag "write-once" for
conffiles. In most circumstances this operates identical to a regular
conffile, however, it disables all
Your message dated Fri, 08 Jan 2021 04:33:38 +
with message-id
and subject line Bug#822462: fixed in dpkg 1.20.6
has caused the Debian Bug report #822462,
regarding dpkg should automatically clean up obsolete conffiles (make
rm_conffile unnecessary)
to be marked as done.
This means that you
Author: Guillem Jover
AuthorDate: Fri Jan 8 00:29:49 2021 +0100
Support remove-on-upgrade conffile flag via DEBIAN/conffiles
Add support for a new "remove-on-upgrade" flag in DEBIAN/conffiles to
mark that the package wants to remove a given conffile (without having
Author: Guillem Jover
AuthorDate: Tue May 5 05:30:06 2020 +0200
t-deb-conffiles: Add new non-relative conffiles case
The conffiles file must not contain relative pathnames. This happens
to work because when parsing the file the dpkg filesystem lookup
function prepends
Author: Guillem Jover
AuthorDate: Tue May 5 05:26:47 2020 +0200
t-deb-conffiles: Rename packages and feature macro
Make these more descriptive.
---
Feature.mk | 2 +-
t-deb-conffiles/Makefile | 12
Author: Guillem Jover
AuthorDate: Sat Dec 14 02:43:45 2013 +0100
t-deb-conffiles: New test case
---
Makefile | 1 +
t-deb-conffiles/Makefile | 15 +++
t-deb-conffiles/pkg-conff-dupe/DEBIAN/conffiles
Author: Guillem Jover
AuthorDate: Thu Oct 11 06:41:24 2012 +0200
Verify that conffiles got correctly obsoleted on package replaces
---
t-conffile-replaces-existing-and-upgrade/Makefile | 6 ++
t-conffile-replaces-upgrade/Makefile | 6 ++
2 files changed, 12 insertions
Author: Guillem Jover
AuthorDate: Thu Oct 4 17:53:00 2012 +0200
t-multiarch: Check reinstallation of coinstallable packages with conffiles
Bug-Debian: http://bugs.debian.org/684776
---
t-multiarch/Makefile | 12
1 file changed, 12 insertions(+)
diff --git a/t-multiarch
Author: Guillem Jover
AuthorDate: Thu Oct 4 17:52:28 2012 +0200
t-multiarch: Check for proper updates of Conffiles fields
---
t-multiarch/Makefile | 4
1 file changed, 4 insertions(+)
diff --git a/t-multiarch/Makefile b/t-multiarch/Makefile
index cff190266..45b08b77b 100644
--- a/t
Author: Guillem Jover
AuthorDate: Tue Apr 26 06:37:45 2011 +0200
Add conffiles with spaces test-case to t-conffile-prompt
---
t-conffile-prompt/pkg-conff-prompt-0/DEBIAN/conffiles | 1 +
t-conffile-prompt/pkg-conff-prompt-0/DEBIAN/postinst | 1 +
t-conffile-prompt/pkg-conff
Hello,
Guillem Jover, le lun. 14 déc. 2020 01:51:58 +0100, a ecrit:
> and then conditionally run rm_conffile in speech-dispatcher
> iff speech-dispatcher-kali is not present?
Right, that seems to be doing the job.
Thanks!
Samuel
On Tue, 2020-12-08 at 01:25:30 +0100, Samuel Thibault wrote:
> I moved a configuration file kali.conf from the speech-dispatcher
> package to the speech-dispatcher-kali package (as well as others, but
> let's keep that example only).
>
> The thing is: speech-dispatcher does not depend on
>
Hello debian-dpkg,
I moved a configuration file kali.conf from the speech-dispatcher
package to the speech-dispatcher-kali package (as well as others, but
let's keep that example only).
The thing is: speech-dispatcher does not depend on
speech-dispatcher-kali (and cannot: the former is in main,
Author: Guillem Jover
AuthorDate: Wed Sep 2 04:23:11 2020 +0200
dpkg-deb: Fix single-instance memory leak on missing conffiles control file
This gets leaked just once while checking the conffile control file.
Warned-by: gcc ASAN
---
dpkg-deb/build.c | 4 +++-
1 file changed, 3
(HEAD -> master)
Author: Guillem Jover
AuthorDate: Tue May 5 05:47:22 2020 +0200
dpkg-deb, dpkg: Do not accept relative pathnames in DEBIAN/conffiles
The pathnames in the binary package conffiles control file must be
absolute. This has not been a problem in dpkg because the loo
7737483b10a49ab7db316b158728c161ea37ab53
Author: David Kalnischkies
AuthorDate: Sat Jul 23 10:07:53 2016 +0200
dpkg-maintscript-helper: Support DPKG_ROOT for conffiles
As dpkg-maintscript-helper "works around known dpkg limitations in
maintainer scripts" it should behave as if it were a mainta
Processing commands for cont...@bugs.debian.org:
> retitle 163657 dpkg: Support taking over conffiles from other packages
Bug #163657 [dpkg] dpkg: Should support conffile hijacking
Bug #678674 [dpkg] Provide a way to transfer config files between packages
Changed Bug title to 'dpkg: Supp
TS schrieb/wrote:
> Hello,
> % dpkg-query -W -f='${Conffiles}\n' | grep -E '(grub|archive).*obsolete$'
> /etc/apt/trusted.gpg.d/debian-archive-wheezy-stable.gpg
> 64d549adf06d734bb947d742898d9a19 obsolete
> /etc/apt/trusted.gpg.d/debian-archive-wheez
Hello,
>>> since fix for #910959 grub2-common marks kernel update scripts falsely as
>>> obsolete conffiles.
>>>
>>> % dpkg-query -W -f='${Conffiles}\n' | grep 'grub.*obsolete$'
>>> /etc/kernel/postrm.d/zz-update-grub 536d9d45e3e547638db3c5d58a925b6c
Hello,
>>> since fix for #910959 grub2-common marks kernel update scripts falsely as
>>> obsolete conffiles.
>>>
>>> % dpkg-query -W -f='${Conffiles}\n' | grep 'grub.*obsolete$'
>>> /etc/kernel/postrm.d/zz-update-grub 536d9d45e3e547638db3c5d58a925b6c
TS schrieb/wrote:
> Hello,
> % dpkg-query -W -f='${Conffiles}\n' | grep -E '(grub|archive).*obsolete$'
> /etc/apt/trusted.gpg.d/debian-archive-wheezy-stable.gpg
> 64d549adf06d734bb947d742898d9a19 obsolete
> /etc/apt/trusted.gpg.d/debian-archive-wheez
On 2018-10-28 15:03 +, Colin Watson wrote:
> On Sun, Oct 28, 2018 at 03:42:16PM +0100, TS wrote:
>> since fix for #910959 grub2-common marks kernel update scripts falsely as
>> obsolete conffiles.
>>
>> % dpkg-query -W -f='${Conffiles}\n' | grep 'grub.*obsolete$'
On Sun, Oct 28, 2018 at 03:42:16PM +0100, TS wrote:
> since fix for #910959 grub2-common marks kernel update scripts falsely as
> obsolete conffiles.
>
> % dpkg-query -W -f='${Conffiles}\n' | grep 'grub.*obsolete$'
> /etc/kernel/postrm.d/zz-update-grub 536d9d45e3e547638db3c5d58a
Your message dated Thu, 18 May 2017 03:18:40 +
with message-id <e1dbbxy-00071m...@fasolo.debian.org>
and subject line Bug#861217: fixed in dpkg 1.18.24
has caused the Debian Bug report #861217,
regarding dpkg fails to unpack m-a:same instance with conffiles over removed
but not
erence for shared conffiles
The code in charge of inferring the digest for a conffile was
checking the owning package status twice, and ignoring conffiles
with a status lower than "unpacked" even if they had been configured
previously and their md5sums were valid.
Package: dpkg
Version: 1.18.23
User: helm...@debian.org
Usertags: rebootstrap
Hi Guillem,
it seems that dpkg exhibits a strange behaviour with Multi-Arch: same
and conffiles. If you have one instance removed, but not purged, and try
to install another instance, dpkg errors out with an unpack
ILE $PRIOR_VERSION $PKGNAME --
>> "$@"
>>
>> in the package's prinst, postinst, and postrm maintainer scripts.
>>
>> Instead, dpkg should be aware of the dropped config file and handle it
>> automatically.
>
> Are you suggesting that dpkg would diff t
le $CONFFILE $PRIOR_VERSION $PKGNAME --
> > "$@"
> >
> > in the package's prinst, postinst, and postrm maintainer scripts.
> >
> > Instead, dpkg should be aware of the dropped config file and handle it
> > automatically.
> Are you suggesti
's prinst, postinst, and postrm maintainer scripts.
>
> Instead, dpkg should be aware of the dropped config file and handle it
> automatically.
Are you suggesting that dpkg would diff the prior .conffiles with the in-.deb
control file, and perform the equivalent of dpkg-maintscript-helper
Your message dated Mon, 06 Mar 2017 06:18:56 +
with message-id <e1cklyy-0006kl...@fasolo.debian.org>
and subject line Bug#854417: fixed in dpkg 1.18.23
has caused the Debian Bug report #854417,
regarding dpkg-dev: improve deb-conffiles manpage
to be marked as done.
This means that you
Package: dpkg-dev
Version: 1.18.22
Severity: minor
Dear Maintainer,
The section of Debian Policy where the handling of conffiles is
documented [0], gives more information than what is currently available
in deb-conffiles (5) man page :
'The filenames should be absolute pathnames, and the files
Processing commands for cont...@bugs.debian.org:
> severity 847357 wishlist
Bug #847357 [src:dpkg] dpkg: dpkg does not reinstall removed/lost conffiles
unless forced to
Severity set to 'wishlist' from 'serious'
> tags 847357 wontfix
Bug #847357 [src:dpkg] dpkg: dpkg does not reinstall r
Your message dated Thu, 8 Dec 2016 01:24:58 +0100
with message-id <20161208002458.dwkcltuddbng5...@gaara.hadrons.org>
and subject line Re: Bug#847357: dpkg: dpkg does not reinstall removed/lost
conffiles unless forced to
has caused the Debian Bug report #847357,
regarding dpkg: dpk
>
> > After upgrading to sid the conffiles don't seem to be installed any longer?
> > Examples are bash, passwd, basefiles and libpam-runtime. Especially the
> > last one cost me a day debugging to find out why logins crashed. What is
causing this, do I have some settings disab
This is an automated email from the git hooks/post-receive script.
guillem pushed a commit to branch master
in repository dpkg.
commit 4af3d6777f86226d46260910cedccf22815f0991
Author: Guillem Jover <guil...@debian.org>
Date: Sun Jul 31 16:09:38 2016 +0200
debian: Add deb-conff
This is an automated email from the git hooks/post-receive script.
guillem pushed a commit to branch master
in repository dpkg.
commit 90e33b3495cb60c2008673aedf467a796ff85a41
Author: Guillem Jover <guil...@debian.org>
Date: Thu Jul 7 19:55:10 2016 +0200
man: Add new deb-conffiles(
iggers for conffiles on purge
The code has never activated triggers for conffiles on purge, the code
before commit 65ade6390b47fe3ec6a0e2ba341f3d553bf4 was activating them
on removal, which was obviously wrong.
Stable-Candidate: 1.17.x
Reported-by: Helmut Grohne <hel...@s
o safe matching of directories containing
conffiles
By using a variable for the grep match we are exposed to metacharacters
acting as part of the regular expression.
Proposed-by: Carsten Hey <cars...@debian.org>
---
debian/changelog | 4
Package: dpkg
Version: 1.18.4
Severity: normal
Control: user debian-d...@lists.debian.org
Control: usertag -1 declarative-packaging
Currently, if a package drops a config file, the package maintainer
has to invoke:
dpkg-maintscript-helper rm_conffile $CONFFILE $PRIOR_VERSION $PKGNAME -- "$@"
k when unpacking conffiles
We keep a queue of conffile filenodenames, and never free it. We should
be using instead the obstack allocator, so that when we are done with
this package the list entries get all released at the same time.
---
debian/changelog | 1 +
src/archives.c | 2 +-
2
Your message dated Wed, 15 Jan 2014 18:18:47 +
with message-id e1w3v39-0007mi...@franck.debian.org
and subject line Bug#108196: fixed in dpkg 1.17.6
has caused the Debian Bug report #108196,
regarding [CONFFILE] dpkg: missing newline in conffiles causes wrong error
message
to be marked
Your message dated Sat, 27 Jul 2013 04:03:03 +
with message-id e1v2vih-0002dm...@franck.debian.org
and subject line Bug#716948: fixed in dpkg 1.17.0
has caused the Debian Bug report #716948,
regarding bootlogd: prompting due to modified conffiles which were not modified
by the user: /etc
Processing control commands:
affects -1 syslog-ng-core
Bug #690051 [dpkg] symlink-conffiles: dpkg stores md5sum of target file in its
database and does not update it on upgrades
Added indication that 690051 affects syslog-ng-core
--
690051: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug
Package: dpkg
Version: 1.16.8
Severity: important
User: debian...@lists.debian.org
Usertags: piuparts
Control: affects -1 syslog-ng-core
Hi,
the syslog-ng-core is one of the few packages using symlink conffiles,
causing trouble for piuparts and debsums because dpkg does not properly
handle
Processing control commands:
forcemerge 421344 -1
Bug #421344 [dpkg] dpkg: does not gracefully handle symlink conffiles
Bug #690051 [dpkg] symlink-conffiles: dpkg stores md5sum of target file in its
database and does not update it on upgrades
Severity set to 'normal' from 'important'
421346
Control: clone -1 -2
Control: reassign -2 syslog-ng-core 3.3.5-2
Control: found -2 3.3.6-1
Control: severity -2 serious
Control: retitle -2 symlink conffiles are not supported, causing problems for
dpkg on upgrade/removal and incorrect debsums reports
On 2012-10-09 17:10, Guillem Jover wrote
Processing commands for cont...@bugs.debian.org:
# trying again after hitting #684542
clone 690051 -1
Bug #690051 [dpkg] symlink-conffiles: dpkg stores md5sum of target file in its
database and does not update it on upgrades
Bug #421344 [dpkg] dpkg: does not gracefully handle symlink conffiles
Processing commands for cont...@bugs.debian.org:
# # trying again after hitting #684542
# trying again after Failed to clone 690051: Bug is marked as being merged
with others. Use an existing clone.
unmerge 690051
Bug #690051 [dpkg] symlink-conffiles: dpkg stores md5sum of target file in its
Processing control commands:
block -1 by 689836
Bug #689790 [msva-perl] msva-perl: modifies conffiles during upgrade from
squeeze (policy 10.7.3): /etc/X11/Xsession.d/70monkeysphere_use-validation-agent
689790 was not blocked by any bugs.
689790 was not blocking any bugs.
Added blocking bug(s
, but it leaves an old md5sum entry in the Conffiles entry of
old-pkg in the status file.
This causes subsequent debsums runs to report modified files (for a
package which no longer owns the file).
Seen in grub-* (moved something to grub-common) and libopenal1 (moved
/etc/openal/alsoft.conf to libopenal
Control: reassign -1 debsums
Control: retitle -1 ignore obsolete conffiles which are not owned by the package
On Wed, 03 Oct 2012, Andreas Beckmann wrote:
seems to work, also if the conffile gets replaced by an updated version
on the way, but it leaves an old md5sum entry in the Conffiles entry
Your message dated Mon, 19 Mar 2012 09:17:33 +
with message-id e1s9yib-0001qp...@franck.debian.org
and subject line Bug#658854: fixed in dpkg 1.16.2
has caused the Debian Bug report #658854,
regarding dpkg-maintscript-helper, local package rebuilds and orphaned conffiles
to be marked as done
Hello,
On Mon, 06 Feb 2012, Sam Morris wrote:
And the conffile will not be removed.
Is the user in the wrong here? debian-faq says no, as it specifically
advises the user of dch -l when performing a local build. The user
cannot know when he's rebuilding the package that a subsequent version
Processing commands for cont...@bugs.debian.org:
retitle 658854 dpkg-maintscript-helper, local package rebuilds and orphaned
conffiles
Bug #658854 [dpkg] dpkg-maintscript-helpee, local package rebuilds and orphaned
conffiles
Changed Bug title to 'dpkg-maintscript-helper, local package
Your message dated Fri, 11 Nov 2011 14:36:39 +0100
with message-id 2011143640.ef26ede4.th...@math.uni-paderborn.de
and subject line Not a bug - sorry
has caused the Debian Bug report #648360,
regarding dpkg-maintscript-helper.sh does not remove obsolete conffiles in some
cases
to be marked
Package: dpkg
Version: 1.15.8.11
Severity: normal
Tags: patch
During a dist-upgrade from Lenny to Squeeze I noticed some obsolete conffiles
not being removed.
For example the package desktop-base contains a conffile /etc/kde3/kdeglobals,
which should be removed as the package is upgraded
On Mon, Oct 31, 2011 at 08:49:42AM +0100, Raphael Hertzog wrote:
Well, dpkg certainly handles conffiles moving from one package to the
other if you use the proper Replaces field like for any other file.
Currently console-setup-linux doesn't Replace console-setup.
If this was the real cause
Your message dated Fri, 23 Sep 2011 05:17:21 +
with message-id e1r6y8x-00033m...@franck.debian.org
and subject line Bug#638291: fixed in dpkg 1.16.1
has caused the Debian Bug report #638291,
regarding dpkg: Hardlinks problem with conffiles
to be marked as done.
This means that you claim
already...
can you merge your test in that one?
It already tests hardlinks on various type of files, you'd just add a
test-conffile-link{0,1} and mark test-conffile-link{0,1} as conffiles.
The testcase in the attached patch slightly differs from the original
one in that the hardlink
just noticed/remembered that we have t-unpack-hardlink already...
can you merge your test in that one?
Done in the attached patch.
It already tests hardlinks on various type of files, you'd just add a
test-conffile-link{0,1} and mark test-conffile-link{0,1} as conffiles.
The testcase
(since unchanged conffiles are not
replaced).
Cheers,
Sven
From 6d1aa097ea1f0c204f74bb744ad72e06667e7a7b Mon Sep 17 00:00:00 2001
From: Sven Joachim svenj...@gmx.de
Date: Mon, 22 Aug 2011 18:44:01 +0200
Subject: [PATCH] New test t-conffile-hardlink
Test if a package with a hardlink
2011/8/20 Sven Joachim svenj...@gmx.de
The reason is that conffile processing takes place after
unpacking all non-conffiles, so for hardlinks to conffiles you hit the
ENOENT bug. If a file is both a conffile and a hardlink in the package,
it is treated as the latter.
Do I understand
Hi Ondrej,
On Wed, 18 May 2011, Ondřej Surý wrote:
I'm just not sure what the correct solution is. Instead of purge_dir maybe
the right answer is manage-manual-conffile and/or
manage-manual-file. And it would drop the file on removal/purge and try
to remove the parent directories if they
On Sun, 2010-12-05 at 17:44 +0100, Julian Andres Klode wrote:
Package: packagekit
Version: 0.6.8-2
Severity: normal
In addition to the debconf support, PackageKit must also
support conffiles.
Dear dpkg maintainers,
for PackageKit support we need a way to handle conffiles. PackageKit
allows
Hi,
2011/5/16 Raphael Hertzog hert...@debian.org:
Hi,
On Mon, 16 May 2011, Ondřej Surý wrote:
(small) ping with question...
Are you going to consider implementing purge_dir to
dpkg-maintscript-helper or should I do it directly in
php5-sapi.postinst scripts? (Both options are OK for me.)
Hi,
On Mon, 16 May 2011, Ondřej Surý wrote:
(small) ping with question...
Are you going to consider implementing purge_dir to
dpkg-maintscript-helper or should I do it directly in
php5-sapi.postinst scripts? (Both options are OK for me.)
I'm definitely interested into providing a
(small) ping with question...
Are you going to consider implementing purge_dir to
dpkg-maintscript-helper or should I do it directly in
php5-sapi.postinst scripts? (Both options are OK for me.)
O.
--
Ondřej Surý ond...@sury.org
http://blog.rfc1925.org/
--
To UNSUBSCRIBE, email to
wrote:
your patch doesn't fix the problem, and even introduces more problems,
because as it is written it leaves all directories which have
conffiles on the disk.
That's not correct, it behaves as it should be, but that's not changed
by my patch, that's the current behaviour already
not remove the
conffiles from the package files list when it unlinks them, and they
linger around when we are removing the leftover directories on purge.
On entry removal_bulk_remove_leftover_dirs() should really only
contain directories (or symlinks to directories) nothing else. I'll
fix that up
the
conffiles from the package files list when it unlinks them, and they
linger around when we are removing the leftover directories on purge.
On entry removal_bulk_remove_leftover_dirs() should really only
contain directories (or symlinks to directories) nothing else. I'll
fix that up.
Cool
Hi Guillem,
your patch doesn't fix the problem, and even introduces more problems,
because as it is written it leaves all directories which have
conffiles on the disk.
This section:
static void
removal_bulk_remove_files(struct pkginfo *pkg)
{
[...]
if (namenode-flags fnnf_old_conff
- the file on disk is our version
- the new pristine version is their version
... which leads to the next patch.
In cases with renamed conffiles, one might want something more
complicated, but that can wait for later.
src/configure.c | 14 --
1 files changed, 12 insertions(+), 2
that if the required conffiledb data isn't there for any reason that
it falls back to the old behavior/options.
In cases with renamed conffiles, one might want something more
complicated, but that can wait for later.
Yeah, that's another topic that we didn't reach, but doesn't present
any problems
sean finney wrote:
Yeah I don't think we got to the point where we were considering
issues like that. I'd guess that the menu should be made more flexible
so that if the required conffiledb data isn't there for any reason that
it falls back to the old behavior/options.
Yep, agreed. A
On 6.5.2011, at 6:28, Guillem Jover guil...@debian.org wrote:
Hi!
On Wed, 2011-05-04 at 11:11:10 +0200, Ondřej Surý wrote:
The attached patch keeps the directory in the pkg.list (aka add it
to the leftover) if the children directory is found on the leftover
list.
The problem is that the
2011/5/6 Raphael Hertzog hert...@debian.org:
On Fri, 06 May 2011, Ondřej Surý wrote:
If you want this fixed, just ping me, it's easy to add something like
(-name[namelen] == \0 or ...-name[namelen] == '/'), and
I'll fix it at both places.
Ah nice catch! I've pushed a fix for this
=UTF-8
Content-Transfer-Encoding: 8bit
When a directory is kept during removal to be later dealt with during
purge, it might be due to the directory containing conffiles from the
same package, it not being empty, etc. In any case we should keep all
its parent to make sure when the subsequent trial
, to avoid
getting in piuparts's way, without regressing in the current cases
where directories are being correctly removed at removal time.
At the moment the only solution I see is what I have suggested - keep
ownership of whole tree up-to-the root for directories holding
conffiles and try
Ondřej Surý wrote:
At the moment the only solution I see is what I have suggested - keep
ownership of whole tree up-to-the root for directories holding
conffiles and try to remove them at purge time...
dpkg --remove php5-cli
# ownership kept for /., /etc, /etc/php5, /etc/php5/cli
Yes
php5-cli):
# cat php5-cli.list
/etc/php5/cli
I guess this is something not a php5 specific and could happen in more
packages which put more config files into subsubdirectories in /etc/.
I think the correct behaviour of dpkg should be to leave all
directories leading to all left conffiles after
forcemerge 316521 625241
quit
Hi,
Ondřej Surý wrote:
I don't know whether it is real dpkg bug or not, but that's something
I have found in php5 piuparts testing.
Both php5-common and php5-cli (and other SAPIs) owns /etc/php5
(/var/lib/dpkg/info/*.list), now after
dpkg --remove php5-cli
Processing commands for cont...@bugs.debian.org:
forcemerge 316521 625241
Bug#316521: dpkg: incomplete cleanup of empty directories
Bug#625241: dpkg: stale directories when packages install conffiles to
subsubdirectories of /etc
Bug#348133: dpkg: incomplete cleanup of empty directories
Bug
Your message dated Wed, 27 Apr 2011 05:39:42 +0200
with message-id 20110427033942.ga18...@gaara.hadrons.org
and subject line Re: Bug#351900: dpkg remove should remove unedited conffiles
has caused the Debian Bug report #351900,
regarding dpkg remove should remove unedited conffiles
to be marked
guess, might be checking if the file has the obsolete
flag, if it does not then it was installed by a buggy dpkg.
$ dpkg-query -W -f '${Conffiles}\n' bash | grep bash_completion
And an unversioned Replaces in bash-completion would be the correct way
to handle that. Fixing that in bash
bash: /etc/bash_completion
This is with bash 4.0-7.
The message is in tarobject(). I think dpkg 1.13.14~19 (Improve
processing of disappearing conffiles, 2006-02-10) was supposed to deal
with this case:
If the file to be unpacked is (1) a conffile in the new package and
(2
This is with bash 4.0-7.
The message is in tarobject(). I think dpkg 1.13.14~19 (Improve
processing of disappearing conffiles, 2006-02-10) was supposed to deal
with this case:
If the file to be unpacked is (1) a conffile in the new package and
(2) a regular file rather than a symlink
Your message dated Thu, 29 Jul 2010 09:32:22 +
with message-id e1oepts-0003pm...@franck.debian.org
and subject line Bug#102609: fixed in dpkg 1.15.8
has caused the Debian Bug report #102609,
regarding dpkg: --force-confask - replace conffiles with no new version
to be marked as done
The option --overwrite-conffiles is added to dpkg. It allows to install
the package maintainer's configuration files. If the file has been
modified by the user, an backup is stored as .dpkg-old.
--force-confmiss does now work with less code.
Signed-off-by: Daniel Bock daniel.b...@web.de
Hi,
sorry, I forgot the use of cfo_newconff, like in the old patch.
+if (strcmp(currenthash, newdisthash) != 0 f_overwriteconf
+|| !strcmp(currenthash, NONEXISTENTFLAG) fc_conff_miss) {
+if (useredited) {
+what |= cfof_prompt;
+fc_conff_new = 1;
+
Your message dated Wed, 21 Apr 2010 03:34:27 +
with message-id e1o4qhn-0006d0...@ries.debian.org
and subject line Bug#514316: fixed in dpkg 1.15.7
has caused the Debian Bug report #514316,
regarding Should provide shell functions to handle conffiles
to be marked as done.
This means that you
On Mon, Jan 04 2010, Russ Allbery wrote:
Umm, directories can't be conffiles, no? So not considering
directories, if symlinks can be conffiles, and regular files can be
conffiles, I think users should be allowed to change one conffile into
another. And the code must then handle
On Sun Jan 03 19:57, Russ Allbery wrote:
Symlinks in /etc pointing to files not in /etc are used now, so I'm not
sure they should be Policy violations. /etc/nologin is the canonical
example. Depending on how and whether Debian adopts upstart, we may have
other cases.
Shouldn't they be files
On Sun, Jan 03 2010, Russ Allbery wrote:
Manoj Srivastava sriva...@ieee.org writes:
On Sun, Jan 03 2010, Russ Allbery wrote:
I don't believe that listing symlinks as conffiles works properly at
the moment. See #421344. It doesn't make any sense to list a
directory as a conffile. I think
the behaviour.
Oh, agreed. My proposal for Policy would be to state, for the time being,
that only regular files can be conffiles and that neither symlinks nor
directories (the latter being somewhat obvious, but worth stating
explicitly) can be conffiles.
If at some point symlinks as conffiles
Do conffiles have to be regular files? Policy does not seem to be
explicit about this (although I'd be happy to be proven wrong on this),
it seems to just talk about files.
--
Saludos,
Felipe Sateler
--
To UNSUBSCRIBE, email to debian-dpkg-requ...@lists.debian.org
with a subject
On Sun, 2010-01-03 at 14:50 -0800, Russ Allbery wrote:
Felipe Sateler fsate...@gmail.com writes:
Do conffiles have to be regular files? Policy does not seem to be
explicit about this (although I'd be happy to be proven wrong on this),
it seems to just talk about files.
I don't believe
of management of a directory implied by the rules for a conffile is
basically done already, automatically, provided that the contents of the
directory are all marked as conffiles or are other directories.
Policy does specifically talk about files, and conventionally directories
are not considered
On Sun, Jan 03 2010, Russ Allbery wrote:
Felipe Sateler fsate...@gmail.com writes:
Do conffiles have to be regular files? Policy does not seem to be
explicit about this (although I'd be happy to be proven wrong on this),
it seems to just talk about files.
I don't believe that listing
1 - 100 of 220 matches
Mail list logo
