The release team have accepted it for the next Jessie point release:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=802851
https://release.debian.org/proposed-updates/stable.html
http://ftp.debian.org/debian/dists/jessie-proposed-updates/qpsmtpd_0.84-11+deb8u1_amd64.changes
--
Devin \
since 0.66 from oldstable has the
+opposite compatibility problem
+
+ -- Devin Carraway <de...@debian.org> Sat, 24 Oct 2015 07:27:50 +
+
qpsmtpd (0.84-11) unstable; urgency=low
* Add Japanese debconf translation; thanks to victory (Closes:
diff -Nru qpsmtpd-0.84/debian/contro
Package: release.debian.org
Severity: normal
Tags: jessie
User: release.debian@packages.debian.org
Usertags: pu
I propose a patch to qpsmtpd in jessie-proposed-updates; this is a one-line
patch to address a compatibility breaking API change introduced in
libnet-dns-perl 0.81. The effect of
It is; I'm planning to roll a new package this weekend. Sorry for the delay.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 4096R/9197B5F9: 9C64 37CD 1B7B 029D 0933 49EA 1E52 7672 9197 B5F9
signature.asc
Description: Digital signature
Package: selinux-policy-default
Version: 2:2.20140421-7
Severity: normal
unconfined_t doesn't seem able to touch gpg_agent_exec_t; this blocks its use,
but also prevents dpkg from upgrading gnupg-agent, and thus blocks installation
of gnupg2:
root@atlantic:/etc/selinux/local# id -Z
] Permission
denied: u'/etc/selinux/default/modules/active/file_contexts.local'
-- no debconf information
commit 00fb07499077baa3cb97764fdb3453ba86e13fb4
Author: Devin Carraway g...@devin.com
Date: Mon Mar 3 01:34:08 2014 -0800
Label postgresql manpages under Debian
diff --git a/policy/modules
This is possibly being triggered by an inherent quirk of ISC's libraries;
since host(1) triggers it too, even from unconfined_t:
type=AVC msg=audit(1392969683.579:1590): avc: denied { block_suspend } for
pid=14446 comm=host capability=36
Package: selinux-policy-default
Version: 2:2.20140206-1
Severity: important
On a jessie system with refpolicy 2:2.20140206-1, and allow-hotplug
set on the primary network interface, sshd is left running in udev_t,
breaking it thoroughly (and in fact flooding the logs with socket
errors until the
On Thu, Feb 20, 2014 at 10:15:54AM +0100, Laurent Bigonville wrote:
Could you please attach the AVC denials to the bug.
Sure, here you are -- this was taken in permissive mode obviously. The
openssh-server command referred to in the first ssh-related denial is
On Thu, Feb 20, 2014 at 12:28:43AM -0800, Devin Carraway wrote:
I'll test out restoring the labelling and see if there's more to this.
Slightly more -- udev_t also lost the ability to transition to initrc_t, which
it will do in the old wheezy refpolicy. Labelling /etc/network/if-*d
Package: selinux-policy-default
Version: 2:2.20140206-1
Severity: normal
The courier suite uses a series of daemons that start one another at various
stages in startup and authentication (e.g. a logger starts a tcpd which starts
an authentication wrapper which starts an actual imapd). The same
Package: selinux-policy-default
Version: 2:2.20140206-1
Severity: normal
mailman's qrunner and friends aren't being properly labelled and so aren't
transitioning properly on startup:
system_u:system_r:initrc_t:s03523 ?S 0:00 /usr/bin/python
/var/lib/mailman/bin/qrunner
: Devin Carraway g...@devin.com
Date: Thu Feb 13 21:59:54 2014 -0800
Allow ndc_t to prevent suspend on epoll events.
Capability background: http://thread.gmane.org/gmane.linux.kernel/1289986
Fixes audit denial:
Feb 13 00:46:38 a6 kernel: [ 541.076682] type=1400 audit
Package: src:linux
Version: 3.10.1-1
Severity: normal
linux-image-3.10-1-kirkwood (also reproduced with 3.9-1-kirkwood) cannot boot
on a dreamplug; the boot hangs and the watchdog fires while down in
orion_nand code, which so far as I know is not even applicable to this
hardware.
Boot was off
Although I didn't explicitly fix it in this patch release, I'm unable to
reproduce the problem with 0.84-10, uploaded a few minutes ago - please
give it a try with that release and see if you still observe trouble.
Devin
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway
tags 693181 + wheezy security
severity 693181 serious
quit
Thanks for the report. I'll roll a release and see if the release team will
grant a freeze exemption.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E 65F0 8905 2E43 5395 CA0D
severity 693103 important
quit
Marking 'important' -- this doesn't meet Debian guidelines for 'serious,'
quite.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E 65F0 8905 2E43 5395 CA0D E9AB FCD2
signature.asc
Description: Digital
Ah, version control. Introducing problems that can only be solved with yet
more version control.
I'll roll a -8 tonight.
Devin
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E 65F0 8905 2E43 5395 CA0D E9AB FCD2
--
To UNSUBSCRIBE,
On Mon, Aug 20, 2012 at 05:34:43PM -0400, David Prévot wrote:
Depending of how much « a few days » are, what do you prefer? If it's
about ten days or more, I could go now with the first proposal, but if
it's less than five days (if the fix meet the freeze exception criteria
of course), or more
forwarded 684571
http://www.nntp.perl.org/group/perl.qpsmtpd/2012/08/msg9954.html
quit
Forwarded upstream for consideration. Upstream has no BTS, linking mailing
list thread instead.
Devin
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E
Sep 17 00:00:00 2001
From: Devin Carraway g...@devin.com
Date: Tue, 24 Jan 2012 00:54:24 -0800
Subject: [PATCH] Stop linking nonexistent ARM volume lib
---
debian/rules |3 ---
1 files changed, 0 insertions(+), 3 deletions(-)
diff --git a/debian/rules b/debian/rules
index f850db9..960e85a
On Sun, Nov 20, 2011 at 02:18:14PM +0100, gregor herrmann wrote:
I've prepared an NMU for pidgin-librvp (versioned as 0.9.7-1.2) and
uploaded it to DELAYED/2. Please feel free to tell me if I
should delay it longer.
Nope, thanks for the upload. I've just moved and the build machine with my
Package: forked-daapd
Version: 0.16-1
Severity: normal
forked-daapd includes a logrotate config which runs 'killall -q -HUP
forked-daapd'. However, it doesn't depend on psmisc which actually contains
killall.
-- System Information:
Debian Release: wheezy/sid
APT prefers testing
APT
On Sun, Mar 06, 2011 at 03:04:45PM +0100, Julien BLACHE wrote:
Forked-daapd does not appear to be correctly extracting track numbers from
AAC
files, at least not those encoded within iTunes.
Can you retry with ffmpeg 0.6 and 0.12~git0.11-125-gca72ee5-4 now that
they are both available
On Sun, Mar 06, 2011 at 03:04:45PM +0100, Julien BLACHE wrote:
Can you retry with ffmpeg 0.6 and 0.12~git0.11-125-gca72ee5-4 now that
they are both available in testing?
Yep, will do.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E
Package: perl-modules
Version: 5.10.0-19lenny3
Severity: normal
Packages using SelfLoader (older ones, generally) seem to have broken recently
when taint checking is enabled. I haven't narrowed down exactly where this
was introduced, but I don't believe I saw it prior to the 5.0.8 point release
Package: forked-daapd
Version: 0.12~git0.11-125-gca72ee5-3
Severity: normal
Forked-daapd does not appear to be correctly extracting track numbers from AAC
files, at least not those encoded within iTunes. This causes albums to appear
in title-order from DAAP clients (Rhythmbox, iTunes tested),
On Thu, Dec 23, 2010 at 03:36:42PM +0100, Julien Cristau wrote:
I see this is now uploaded. Unblocked, should migrate in 5 days if no
new RC bugs appear; and removal hint added for libmail-spf-query-perl.
I submitted a freeze exception request for qpsmtpd_0.84-5. Thanks for doing
the advance
Package: release.debian.org
User: release.debian@packages.debian.org
Usertags: freeze-exception
I've uploaded qpsmtpd 0.84-5 to unstable; it includes a fix for RC Bug#607472.
Terse summary: libmail-spf-query-perl is RC-buggy; qpsmtpd was its only
dependency, used by its SPF plugin.
On Sun, Dec 19, 2010 at 12:35:32PM +0100, gregor herrmann wrote:
Right, I set #607472 to grave under the assumption that we want to
drop libmail-spf-query-perl which would leave qpsmtpd without a
dependency.
I'm cc'ing the release team to get an opinion if this plan is ok.
I've deployed a
I'll try to take a look at this this weekend.
Not severity=grave for qpsmtpd, but will let it stay on the presumption that
it's blocking another grave bug.
Devin
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E 65F0 8905 2E43 5395 CA0D
Package: release.debian.org
Severity: normal
User: release.debian@packages.debian.org
Usertags: freeze-exception
Please unblock qpsmtpd_0.84-4 for another translation update. No other
changes.
qpsmtpd (0.84-4) unstable; urgency=low
.
* Update Spanish translation; thanks to Francisco
Package: release.debian.org
Severity: normal
User: release.debian@packages.debian.org
Usertags: freeze-exception
Please unblock qpsmtpd_0.84-3 for a translation update. No other changes.
Changes:
qpsmtpd (0.84-3) unstable; urgency=low
.
* Update Czech debconf translation; thanks to
Package: release.debian.org
Severity: normal
User: release.debian@packages.debian.org
Usertags: freeze-exception
Please unblock qpsmtpd_0.84-2 for translation update; changelog follows:
Changes:
qpsmtpd (0.84-2) unstable; urgency=low
.
* Update Portuguese debconf translation; thanks to
Package: selinux-policy-default
Version: 2:0.0.20080702-6
Severity: normal
Lenny's refpolicy seems to prevent mdadm from writing to an interactive
(unconfined) tty, although I'm somewhat mystified as to the specific cause.
mdadm has its own domain, defined by the raid refpolicy module; run as a
Package: pulseaudio
Version: 0.9.21-3
Severity: normal
After a recent pulseaudio upgrade to my armel machine (sheevaplug), a
previously-working audio device (Burr-Brown Japan PCM2702, the USB DAC inside
a consumer-grade JVC amplifier) ceased working; pulseaudio would die on
startup or on device
Package: hugin
Version: 2009.4.0+dfsg-2
Severity: normal
The hugin UI's Stitcher tab allows one to set arbitrary commandline options
to the image fuser and blender apps. However, once set to a non-empty string,
the setting can't then be cleared -- reopening the options dialog and deleting
the
Package: boa
Version: 0.94.14rc21-0.2
Severity: important
Tags: security
CVE-2009-4496 describes a multi-step vulnerability whereby Boa does not escape
nonprintable characters from the request when writing to its error log. While
not a vulnerability in itself, this provides a vector for an
On Wed, Apr 07, 2010 at 10:37:31PM +0200, Stefano Zacchiroli wrote:
I did the DELAYED NMU just in case you won't have time to fix this by a
proper upload anytime soon; if this is the case, you can just let the
NMU flow in and later on do a proper upload ack-ing the NMU and adding
extra
On Sun, Mar 21, 2010 at 07:54:13AM +0100, Petter Reinholdtsen wrote:
I picked the version number when the bug was introduced according to
the changelog, to make sure it is registered against all version where
it is present.
Are you saying that it is fixed in a later version and reintroduced
On Fri, Jan 29, 2010 at 12:34:27AM +, Ben Hutchings wrote:
Is this reproducible if you plug the adapter into a PC?
Did some testing on a PC (Asus M2N-E, MCP55 controller, 2.6.32-trunk-amd64
2.6.32-5) and the problem was not reproducible there. The pegasus adapter
comes up with a valid MAC
I tried one other experiment, plugging the pegasus adapter in at various
points in the NSLU2's boot sequence, to see whether the bootloaders might be
affecting the device. Based on http://www.cyrius.com/debian/nslu2/boot.html ,
the MAC came up zeroed out if the device was plugged in anywhere
On Fri, Jan 29, 2010 at 12:34:27AM +, Ben Hutchings wrote:
Is this reproducible if you plug the adapter into a PC?
I'll try it and see.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7 199E DD1E 65F0 8905 2E43 5395 CA0D E9AB FCD2
--
To
Package: linux-image-2.6.26-2-ixp4xx
Version: 2.6.26-19lenny2
Severity: normal
I've been experimenting with a Pegasus-chipset USB ethernet adapter on
an NSLU2 unit. If the adapter is connected prior to system boot, although it
will be recognized by the kernel and udev, it starts out with a MAC
reassign 567098 linux-image-2.6.32-trunk-ixp4xx 2.6.32-5
quit
This also manifests with linux-image-2.6.32-trunk-ixp4xx 2.6.32-5 from sid;
reassigning there since it's not locked up in stable-land.
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2: 13E7
On Thu, Dec 10, 2009 at 10:51:04AM -0500, John Bazik wrote:
Package: qpsmtpd
Version: 0.40-3
Severity: normal
There's a minor error in the config script. If you don't set INTERFACES,
it is set as follows:
INTERFACES=`/sbin/ifconfig -a | \
grep
severity 540791 minor
tags 540791 +pending
quit
On Mon, Aug 10, 2009 at 01:31:39PM +0200, kaouete wrote:
Can it be possible to add support for multiple port and/or ip adresses in
qpsmtpd startup script?
For example when using the tls plugin, you can use the port 465
to do SMTP over SSL.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Package: mysql-client-5.0
Version: 5.0.51a-24
Severity: grave
Tags: security
Justification: cross-site scripting vulnerability
Upstream is tracking a security flaw in the mysql commandline client,
identified as CVE-2008-4456:
I neglected to close this bug from the changelog, but I've just uploaded
0.81-1 with a patch to address this issue in the clamdscan plugin, by using
ClamAV::Client instead; the package now depends on that perl module, so it
should start cleanly on a fresh install without further fiddling.
--
Package: qpsmtpd
Version: 0.81-1
0.81-1 included prefork, which was usable via editing the init.d script, but
couldn't be selected through debconf, because prefork didn't as of 0.81
support listening on multiple interfaces, and hence would have broken some
deployed setups.
A patch to support
On Fri, Sep 12, 2008 at 08:55:46PM +0100, Steve Kemp wrote:
OK please find attached a patch which mostly do the job for the version
of qpsmtpd which is currently available to Debian unstable, via the
introduction of a new debconf setting.
Hi Steve --
I uploaded 0.81-1 last night (and -2
On Tue, Jan 06, 2009 at 05:03:54PM +0100, Martin Michlmayr wrote:
Devin, I'm inclined to close this bug report since I believe that d-i
is doing the right thing. There are reports that the clock sometimes
gets stuck on the NSLU2, but I don't think Debian/debian-installer
can do anything about
On Thu, Nov 20, 2008 at 08:40:59AM -0200, Otavio Salvador wrote:
Could you give a try in current daily installer since it has a new
kernel version and see if RTC looks OK? If it doesn't it looks like a
kernel issue to me.
Devin, don't waste your time - nothing has changed in the kernel
Package: installation-reports
Severity: normal
-- Package-specific info:
Boot method: network (upslug2)
Image version: Unofficial nslu2 image (stock lenny image plus onboard ethernet
microcode), http://www.slug-firmware.net/d-dls.php
Date: Date and time of the install
Machine: Linksys NSLU2
On Wed, Sep 17, 2008 at 02:30:43PM -0400, Scott Kitterman wrote:
Here is a debdiff for Etch. In addition to fixing this regression, I also
switched DNS/Base.py to use the upstream fix for the DNS cache poisoning
problem. Their fix is more robust. If you'd rather just deal with this
exact
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Wed, Sep 17, 2008 at 05:28:37PM +0200, Fabio Tranchitella wrote:
transaction ID randomization. If instead of reverted to the package that
has
neither, you change 'self.tid' to '0' in line 199
of
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Wed, Sep 10, 2008 at 07:56:53PM +0100, Steve Kemp wrote:
It would be wonderful if the Debian package of qpsmtpd
could be updated a little from the SVN trunk, and allow
you to choose to use the qpsmtpd-prefork version of
the server.
It would,
A tentative fix to refpolicy is here:
http://klecker.debian.org/~devin/refpolicy/
Martin, can you test these to confirm that they address the problem and check
for trouble during the upgrade?
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \ 1024D/E9ABFCD2:
Package: clamav
Version: 0.90.1dfsg-3etch13
Severity: grave
I released clamav 0.90.1dfsg-3etch13 to fix CVE-2008-2713, the DoS exposure
via the Petite unpacker. However, I screwed up the build and didn't actually
include the patch. I'll be releasing a fixed build shortly.
This bug only applies
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Fri, Jul 11, 2008 at 10:59:13PM +0200, Bastian Blank wrote:
I'm currently not able to see the problem to push a _fix_, not a
_workaround_, through stable-security. Please explain.
Pushing a fix to stable-security is easy -- we can patch the
On Fri, Jul 11, 2008 at 08:26:53AM +0200, Martin Godisch wrote:
bind9 security update 9.3.4-2etch3 breaks named running in a selinux
enabled (enforcing) environment:
audit(1215756426.448:248): avc: denied { name_bind } for pid=16218
comm=named src=12949 scontext=user_u:system_r:named_t:s0
tags 480292 +patch
quit
Here's a patch I'm building for an Etch update to address the problem. It's
pretty close to the same one used in the first fix to this bug, except that it
adds a call to realpath() to resolve all components of the path, and fixes the
argument passing so as not to throw
On Fri, Jul 04, 2008 at 02:56:00PM +0200, Tomas Hoger wrote:
Looks like upstream patch is incomplete. Have you already notified
upstream about the problem?
Not yet -- I still need to hand verify it against a pristine upstream; it's
reproducible with 5.0.51a from Sid, but the implementation of
reopen 480292
quit
I don't believe that the patch applied to address this bug was sufficient. In
preparing the stable update I initially applied it, before finding two things:
First, fn_format() only calls readlink() once on the entire path, not on any
component thereof; hence it will only
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Thu, Apr 10, 2008 at 12:44:49AM +0200, Sylvain HITIER wrote:
However this looks like a bug against libcairo2...
I suggest that other integer-overflow vulnerabilities in malloc() calls be
the cause of it.
As you've just NMU'ed a security fix
This happens with strict refpolicy also. I think the entire dependency
detection process is broken by the ricci module, whose dependencies aren't
satisfiable at all with the module packages installed.
Run with the ricci module includes, semodule_deps produces no graphviz output
at all:
$
Meh. Now that I patch that and actually install the module, I find that it
doesn't do anything, because somewhere along the chain of patching and
checking the Debian-compatible paths got dropped off.
I submitted a ticket to upstream's trac, but in the meantime if you plan on
staying with the
Package: selinux-policy-refpolicy-dev
Version: 0.0.20071214-1
Severity: normal
0.0.20071214-1's -dev package installs debian/build.conf.targeted as
/usr/share/selinux/refpolicy-targeted/include/build.conf, which is then
used by the usual Makefile. However, this build.conf file explicitly
Package: refpolicy
Version: 0.0.20071214-1
Severity: normal
Tags: patch
refpolicy 0.0.20071214-1 includes an Exim policy, but doesn't install it
on a fresh refpolicy installation, because the module package is
exim.pp, while Debian calls its exim package 'exim4'.
Attached is the obvious patch.
Package: dpkg
Version: 1.14.15
Severity: normal
On upgrading to dpkg 1.14.15 without upgrading dpkg-dev (APT held it
back because of the implied install of lzma), package builds began
failing thusly:
dpkg-source -b quelcom-0.4.0
compression is not defined in %Dpkg::EXPORT_TAGS at
tags 459126 +unreproducible
quit
On Fri, Jan 04, 2008 at 01:56:28PM +0100, Lucas Nussbaum wrote:
During a rebuild of all packages in sid using /bin/dash as /bin/sh, your
package failed to build.
I'm unable to reproduce this behavior with 0.4.0-12 (just uploaded), with
either dash or bash as
[ upstream author here, not the maintainer ]
The xmms-xf86audio plugin is tightly tied to xmms. While it's possible that it
could be rewritten to work with another audio player, most others already
have the feature this plugin provides. I know audacious does, and last I
looked none of the XMMS2
On Fri, Aug 03, 2007 at 08:29:39AM +0200, Christian Perrier wrote:
Please find attached the french debconf templates update, proofread by the
debian-l10n-french mailing list contributors.
Uploaded in 0.40-2 (I forgot to put the Closes: in the changelog). Thanks!
--
Devin \
On Mon, Aug 06, 2007 at 09:37:15PM +0200, Helge Kreutzmann wrote:
Please find the updated German debconf translation for qpsmtpd
attached.
Uploaded in 0.40-2 (forgot to include the Closes: in the changelog). Thanks!
--
Devin \ aqua(at)devin.com, IRC:Requiem; http://www.devin.com
Carraway \
On Fri, Aug 24, 2007 at 09:03:52PM -0400, Ari Pollak wrote:
I plan to NMU gaim-librvp package to fix this bug. Attached is the patch
to do so.
Go ahead, thanks. Been totally slagged under with work and everytime I try to
detach it gets worse. Sorry if I held up the migration.
--
Devin \
Thanks for pointing those out. I've fixed most of them. Ordering of commas
with respect to quotations is debated in English composition; the usage in the
current template (that is, commas inside the quotation) represents MLA style,
the nearest thing to a standard for such things, at least here
Package: dante-server
Version: 1.1.18-2.1
Severity: normal
I haven't ascertained the exact circumstances, but danted crashes quite
readily when reporting configuration-related errors. Here's a crash that
arises from failure to open a logfile specified in its configuration:
Program received
block 423869 by 405151
tags 405151 +pending
quit
On Mon, May 14, 2007 at 06:13:04PM +0200, Olaf Zaplinski wrote:
# /etc/init.d/qpsmtpd start
Starting qpsmtpd: eval Can't locate Clamd.pm in @INC (@INC contains: lib
/etc/perl /usr/local/lib/perl/5.8.8 /usr/local/share/perl/5.8.8
tags 414565 +unreproducible
usertag 414565 +only-in-ubuntu
quit
On Mon, Mar 12, 2007 at 04:02:32PM +0100, Luca Falavigna wrote:
qpsmtpd should create /var/run/qpsmtpd directory with proper permissions in
order to preserve files between reboots.
Attached debdiff implements this.
This is an
Package: gnome-pilot
Version: 2.0.15-0.1
Severity: important
Tags: patch
The gnome-pilot file conduit cannot be safely used on amd64 due to an
upstream pointer-size bug (GNOME Bugzilla#410666). If one attempts it,
the entire gnome-pilot setup for that user will crash on sync until the
file is
On Fri, Feb 09, 2007 at 12:07:26PM +0100, Lars Lindner wrote:
It could be worth to backport it. The simple patch is available from
here (md5.patch):
http://sourceforge.net/tracker/download.php?group_id=87005atid=581684file_id=213406aid=1636563
I also attached the file.
Yugh. Why do
When using GtkHTML2 or Gecko?
From the reports upstream I got the impression that the problem
is solved for everyone when Gecko rendering is used. But I got only
feedback from a few users.
That most recent crash was 1.0.27-1 from sid, with the MD5 integer size
patch hand-applied, built with
FWIW, liferea 1.2.5 was released recently; its changelist cites the fixing of
a crash on 64-bit platforms. I've been using it for the past five or six days
and haven't crashed it yet, which is better than 1.0.27-1 ever did.
This deep in the Etch freeze it's probably not reasonable to try to do a
Here's udevinfo from an etch/sid install with an old Adaptec 2940 (since Dann
asked):
looking at device '/block/sr0':
KERNEL==sr0
SUBSYSTEM==block
SYSFS{stat}== 000000
00000
Package: selinux-policy-refpolicy-targeted
Version: 0.0.20061018-2
Severity: normal
The targeted refpolicy has a number of issues that prevent courier's IMAP and
SSL servers from working:
* courier.fc specifies labelling for most of courier's files, but virtually
all of these have the --
Package: wnpp
Severity: wishlist
Owner: Devin Carraway [EMAIL PROTECTED]
* Package name: libclamav-client-perl
Version : 0.11
Upstream Author : Julian Mehnle [EMAIL PROTECTED]
* URL : http://search.cpan.org/dist/ClamAV-Client/
* License : GPL, Artistic
Package: selinux-policy-refpolicy-targeted
Version: 0.0.20061018-2
Severity: normal
The current refpolicy doesn't allow clamd (in clamd_t) to search var_lib_t.
This yields audit errors like this one when clamd starts:
Dec 28 06:48:39 atlantic5 kernel: audit(1167317319.154:167): avc: denied {
Package: selinux-policy-refpolicy-targeted
Version: 0.0.20061018-1
Severity: normal
Tried the targeted refpolicy on dcc-client/dcc-common tonight. Many of the
files in these packages are overlooked when labelling files, because
refpolicy's dcc module stipulates paths not consistent with the
tags 398125 +pending
quit
On Sat, Nov 11, 2006 at 04:33:31PM -0500, Ari Pollak wrote:
With the upload of gaim 2.0.0beta5, gaim-librvp will not build or run
due to a gaim ABI change. Attached is a patch which will fix the
problem.
Splendid timing for it. Thanks for the warning, I've
tags 389025 + fixed-upstream pending
quit
A fix for this issue has been checked in upstream. The upstream maintainers
were planning on a release within a week or so; if this doesn't happen shortly
I'll backport the fix and upload a patched version.
Thanks for the report, and sorry for the delay
In case upstream doesn't issue a new release as intended, I've prepared
packages with a backport of the same fix, and am testing them now. You can
find copies here if you'd like to test it yourself (look for qpsmtpd_0.32-5):
http://devin.com/debian/
--
Devin \ aqua(at)devin.com,
Package: gaim-librvp
Version: 0.9-2
Severity: minor
Preemptory bug filing: as of 0.9-2, the gaim-librvp package can't be cleaned
after it's built -- the patch stamp must be removed before the build will
succeed.
This is caused by Bug#387103 in cdbs, and should fix itself once that's
addressed.
Package: compiz
Version: 0.0.13+git20060928-2
Followup-For: Bug #390503
The problem is probably conneced to this error, of which compiz (not the
decorator) emits several any time gtk-window-decorator is run:
/usr/bin/compiz.real: No GLXFBConfig for depth 32
Presuming the non-free nvidia
Package: selinux-policy-refpolicy-targeted
Version: 0.0.20060813-2
Severity: normal
(as usual, making a guess as to the likeliest package involved)
After upgrading selinux-policy-refpolicy-targeted to 0.0.20060813-1 tonight,
I've started seeing errors from everything which checks file_contexts
On Thu, Aug 17, 2006 at 05:07:20PM -0500, Manoj Srivastava wrote:
You just need a newer version of refpolicy (see the version
uploaded t Sid, for example)
Yup. Pulled that one down last night and it builds just fine; I presume that
checkpolicy/libsepol just got stricter about rule
Package: checkpolicy
Version: 1.30.10-2
Severity: normal
After upgrading to checkpolicy 1.30.10-2, refpolicy no longer builds;
the build fails when checkmodule is run on the strict policy's base.conf:
m4 -D strict_policy -D enable_mcs -D distro_debian -D direct_sysadm_daemon -D
I think this may be caused by an undeclared build dependency on gawk; accoding
to the changelog, gawk was added as a build-dep in 20060224-1 by Erich, but
doesn't appear there now. Without gawk installed, the build includes this
error, which isn't treated as fatal:
echo
I tried forcing the build over to use a plain awk command (mawk, on my
machine), but this produces the same build failure -- mawk produces no output
when run in the build, while gawk does. I don't know enough awk to sort this
out, so here's the obvious patch to build-depend on gawk again.
--
Poking at it a little more, a simpler test case seems to be to get checkmodule
to read its own output:
$ checkmodule -m -M -o test.mod test.te ; echo $? ; checkmodule -b test.mod ;
echo $?
checkmodule: loading policy configuration from test.te
checkmodule: policy configuration loaded
This may have been fixed by the upload of checkpolicy-1.30.10 last night,
which build-depends on libsepol1 1.12.24. According to the buildd logs, the
previous upload was built against libsepol1 1.12-1. In that version,
avrule_block_write() was missing an endian conversion (where num_decls is
1 - 100 of 136 matches
Mail list logo
