Package: network-manager-openconnect
Version: 1.2.10-3+b1
Severity: normal
Dear Maintainer, proper MTU is not being set on the tunnel device when
connecting to a cisco anyconnect server. Value of 1396 is always being set.
This does not produce any ill effects as long as actual MTU is equal or
Source: xdg-terminal-exec
Version: 0~20230318-1
Severity: normal
Dear Maintainer, in recent months the proposed spec was refined and the
implementation
was subjected to intensive optimizations. It can now consume stock desktop
entries
in most cases and supports optional cache.
Context:
If plymouth password prompt is used during boot (i.e. root volume decryption
from initramfs), just the patch above will not work, also need to add
plymouth-quit.service to 'After' property in console-setup.service
Package: wnpp
Severity: wishlist
* Package name: niri
Version : 0.1.2
Upstream Contact: Ivan Molodetskikh
* URL : https://github.com/YaLTeR/niri/
* License : GPL
Programming Lang: Rust
Description : Scrollable-tiling tiling Wayland compositor
Windows
A bit more refined variant without ps:
--- a/lib/console-setup/console-setup.sh 2024-02-09 17:45:38.0 +0300
+++ b/lib/console-setup/console-setup.sh 2024-02-14 01:15:24.413177808 +0300
@@ -16,6
One can repurpose sysetmd XDG Autostart filter:
Either exlusive method:
ExecCondition=/lib/systemd/systemd-xdg-autostart-condition "" "GNOME:KDE"
Or inclusive:
ExecCondition=/lib/systemd/systemd-xdg-autostart-condition
"wlroots:sway:Wayfire:labwc" ""
It might depend on what SpaceFM uses to mount local devices.
Do you have udevil installed and does it work?
udevil mount /dev/sdd1
udevil umount /dev/sdd1
Perhaps /etc/udevil/udevil.conf needs tweaking.
Package: wnpp
Severity: wishlist
* Package name: qt-heif-image-plugin
Version : 0.3.4
Upstream Contact: Jake Harr
* URL : https://github.com/jakar/qt-heif-image-plugin
* License : LGPL-3.0
Programming Lang: C++
Description : Qt Plugin for HEIF Images
Package: game-data-packager
Version: 76
Severity: wishlist
Dear Maintainer, please add an option to ignore file size/checksum expectations.
For cases where exact data files are inaccessible.
-- System Information:
Debian Release: trixie/sid
APT prefers testing
APT policy: (900, 'testing'),
Package: fuzzel
Version: 1.8.2-2
Severity: wishlist
Dear Maintainer, New versions of fuzzel were released since 1.8.2: up to 1.9.2,
including l10n support introduced in 1.9.0.
-- System Information:
Debian Release: trixie/sid
APT prefers testing
APT policy: (900, 'testing'), (300,
Package: qimgv
Version: 1.0.2+git20230615-1
Severity: normal
Dear Maintainer, Video support in qimgv is not working, because package installs
player_mpv.so into /usr/lib/x86_64-linux-gnu/qimgv/player_mpv.so, but qimgv
searches for it in different locations:
Could not find "player_mpv.so" in
Check if systemd is present to avoid errors if it is not.
--- /lib/console-setup/console-setup.sh 2023-02-09 01:33:51.198998809 +0300
+++ /lib/console-setup/console-setup.sh 2023-02-09 01:59:48.786418782 +0300
@@ -16,6 +16,13 @@
-nt /etc/default/keyboard ] || do_configure=yes
Still reproducible on testing.
Here is the deal:
Console setup is supposed to rely on cached scripts in /etc/console-setup
(copied to initramfs). But plymouth starts in initramfs, occupies tty1, which
causes setupcon to detect graphical mode and skip tty1.
When scripts exist in
Package: wayfire
Version: 0.7.5-1~exp1
Severity: wishlist
Dear Maintainer, please consider packaging plugins and software related to
wayfire:
https://github.com/WayfireWM/wayfire-plugins-extra
https://github.com/WayfireWM/wcm
https://github.com/timgott/wayfire-shadows
A script that syncs EFI volume data to mountpoints with numbered suffixes '.N'
and adds boot entries via efibootmgr#!/bin/sh
if [ "$1" = "--write" ]
then
COMMAND_PREFIX=''
RSYNC_N=''
elif [ "$1" = "--dry-run" ]
then
COMMAND_PREFIX='echo will run:'
RSYNC_N='n'
else
echo "Usage: $(basename
Installing ESP on software raid may not be a good idea, even if raid metadata
is at the end of partitions.
Perhaps, a better approach would be to add procedural ESP duplication via hooks
in grub and whatever else needs to write to /boot/efi (potentially
debian-installer?).
Basic concept of a
Package: wpasupplicant
Version: 2:2.9.0-22+b1
Severity: important
Dear Maintainer, wpa_supplicant can not use hardware token again if it was
unplugged at some point after previous use by wpa_supplicant, requires service
restart.
(other applications do not experience such problems)
The other
Forcing clocksource back to tsc with tsc=nowatchdog fixes the performance
regression.
Seems to be related: 5.14 chooses hpet clocksource, while 5.10 uses tsc:
kernel: Linux version 5.10.0-8-amd64 (debian-ker...@lists.debian.org) (gcc-10
(Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2) #1
SMP Debian 5.10.46-4 (2021-08-03)
kernel: tsc: Fast TSC
Package: src:linux
Version: 5.14.9-2
Severity: normal
Dear Maintainer, after upgrade from 5.10 to 5.14 I noticed severe pefrormance
regression when running
GZDoom with content above some threshold of complexity.
More info here: https://github.com/coelckers/gzdoom/issues/1479
It seems to be
Confirming, the cause is partial mesa upgrade.
Got the full mesa set from unstable today, no longer having this issue.
Package: qimgv
Version: 1.0.2-1
Severity: normal
Dear Maintainer, qimgv crashes when trying to display video with mpv (mpv on
itslef works fine)
Seems to be AMD-related:
Cannot find target for triple amdgcn-- Unable to find target for this triple
(no targets are registered)
Segmentation
Thank you!
As there may be other even older revisions out in the wild (and I now know of
some),
wouldn't it be more robust to have a highest known affected ('<= 0x0013')
condition?
gt;
> Le 06/09/2021 à 12:16, Vladimir K a écrit :
>
>> Hello.
>
> Hello,
>
>> safenetauthenticationclient maintains some sort of binary cache in
>> /var/tmp/eToken.cache/.
>> It masks the problem after libccid upgrae for a while.
>> I was able to fully
Package: libccid
Version: 1.4.35-2
Severity: important
Dear Maintainer, upgrading libccid to 1.4.35 or 1.4.36 results in non-working
SafeNet token
The problem is not immediate, appears apparently after some cache refresh
events of either
pcscd or the middleware (safenetauthenticationclient).
I haven't used suspend for about a year now due to #911411 and remote work.
No issues with hibernation though.
Package: jgmenu
Version: 4.3.0-1
Severity: normal
Dear Maintainer, jgmenu-lx backend is missing from the build.
It generates a menu closest to XDG spec with XDG_MENU_PREFIX and
directory structure support. Ideally it should be the default backend.
-- System Information:
Debian Release:
model=aspire-headset-mic has similar effect.
Upgraded to pulseaudio 14.2 from unstable.
analog-output-speaker and analog-output-headphones now mutually trading between
"availability unknown" and "not available" on jack state. At least auto
switching now works for output.
How does pulseaudio determine "availability" and how to affect it?
Seems to be fully fixed by adding:
options snd_hda_intel model=alc255-acer
to a file in /etc/modrobe.d/
If there is a mechanism for assigning this model automatically based on dmi
data, it might be a good idea to apply it.
Nope, it keeps happening even when unloading r8169.
And there is no pattern. Sleep can work reliably for a month straight, then
fail almost every other time for a week.
It may be that in my case the problem occurs during getting into suspend.
System appears to stop in every way but power indicator.
It also MAY be caused by r8169 module.
I configured unloading r8169 for suspend, and haven't had any freeze since
that. Although due to rarity of those freezes,
It's ffmpeg packages. Downgrade to 7:4.2.2-1+b1 fixes it.
I can confirm this. Chromium tab crashes on Youtube video pages, also on MS
Teams web client.
Chromium was not upgraded recently, but other packages were. Some of these may
be causing crashes.
Start-Date: 2020-06-27 10:10:00
Upgrade: libqmi-proxy:amd64 (1.24.12-0.1, 1.26.0-0.1),
There are some performance/stability issues currently with picom. From my
experience of using compton and picom I would recommend packaging both compton
5.1 (https://github.com/yshui/picom/releases/tag/v5.1) and current picom.
Package: mc
Version: 3:4.8.24-2
Severity: normal
Tags: upstream
Dear Maintainer, in the newest version of mcedit syntax highlight for YAML
inline block highlight is started after '>' or '|' directives, but does not end
until EOF.
Instead it should end after encountering any of:
- an empty
So linkage of user and session keyrings is apparently done by pam_keyinit
module.
There is a line:
sessionoptional pam_keyinit.so force revoke
in /etc/pam.d/login, but not in /etc/pam.d/lightdm.
One intriguing thing about 2.5 is that it provides Secret Service now!
I've been waiting for a long time to replace DE keyring tools with something
standalone.
Upstream was renamed to picom:
https://github.com/yshui/picom
Latest release:
https://github.com/yshui/picom/releases/tag/v7.5
Package: gcr
Version: 3.33.4-3
Severity: important
Dear Maintainer, gcr-viewer.desktop contains literal '${exec_prefix}' string in
its Exec value.
Because of that, gcr-viewer can not be launched normally via a desktop entry,
thus breaking
ability to open relevant files with gcr-viewer from file
Hi.
I have a similar problem sometimes. It happens rarely, but frequency has
increased since upgrade to 5.2 kernel.
Failure to wake up from suspend, last message in journal is:
PM: suspend entry (deep)
How can I extract any useful info?
keyctl link workaround works for me.
I'm using lightdm and slimmed down LXQt session.
What system component should link keyrings by design? DM, session lead, PAM,
basic login scripts? I feel that this bug should be redirected accordingly.
Please consider weighting in on topic "Spec to define the default terminal?" on
XDG mailing lists and supporting this proposition:
https://github.com/Vladimir-csp/xdg-terminal-exec
There is a slew of bugs: #520843, #612306, #906136, #920451, #301677, #281852,
#445444 that more or less have one thing in common: they could benefit from
inclusion of XDG file association mechanisms.
The current situation is sometimes counterintuitive: sensible-browser and other
utils are
I've tested the patch for mypaint, built local versions for mypaint and
libmypaint as described by Jacob Nevins.
Packages installed without problems, both Mypaint and GIMP seem to work fine,
including working l10n for Mypaint and mypaint brushes in GIMP.
The fix works,
Thank you!
Tried booting into old kernels from Debian Snapshots, down to
linux-image-4.14.0-2-amd64. Problem persists. Either I just didn't notice it
before, or the problem originated from something other than kernel.
Package: src:linux
Version: 4.19.28-2
Severity: normal
Dear Maintainer, since an unknown moment possibly in the past several months
I'm experiencing constant
90-100% CPU load on one core by a kworker/X:2+events process.
At least versions 4.19+101, 4.19+102, 4.19+104 are affected.
According to
I couldn't find relevant upstream bug, so I've reported one:
https://github.com/systemd/systemd/issues/11825
Hi.
Yes, this is still reproducible with buster's 2:4.9.4+dfsg-2
At least removing zram on service stop should be a simple fix:
ExecStop=/usr/sbin/zramswap stop
Package: firmware-atheros
Version: 20190114-1
Severity: important
Dear Maintainer, ath10k_pci crashes on wakeup from suspend.
Hardware: Qualcomm Atheros QCA6174 802.11ac Wireless Network Adapter (rev 32)
Usually device recovers automatically, but sometimes manual action is needed.
Attached
There was nothing in 3.6.1-8 that is related to this request.
The bug wasn't fixed upstream, it was moved to
https://gitlab.freedesktop.org/xdg/xdg-user-dirs/issues/16
> Bug #594427 [xdg-user-dirs] xdg-user-dirs: "Update standard folders" should
> add language links to folders, not rename them
> Added tag(s) fixed-upstream.
Package: browser-plugin-freshplayer-pepperflash
Version: 0.3.9-1
Severity: normal
Dear Maintainer, the new version of the package unconditionally depends on
libavcodec58,
instead of libavcodec58 | libavcodec-extra58, therefore it can not be upgraded
on systems
with libav-extra packages
I can confirm that something broke somewhere after upgrade of systemd packages
to v239 or v238.
I use LXQt and Lightdm. Restarting logind during upgrades never caused any
problems until recently.
Now suspend authentication breaks, policykit auth is required on every attempt
of suspending after
Adding just this to NetworkManager environment seems to be a sufficient
workaround:
$ cat /etc/systemd/system/NetworkManager.service.d/openvpn-workaround.conf
[Service]
Environment="NM_OPENVPN_CHROOT="
At least the connection survives SIGUSR1 to openvpn process.
I presume the
This still happens with:
network-manager/testing,unstable,now 1.10.6-2 amd64 [installed]
network-manager-openvpn/testing,unstable,now 1.8.2-1 amd64 [installed]
openvpn/testing,unstable,now 2.4.5-1 amd64 [installed]
I've found a potentially less clumsy workaround, but didn't try it yet:
Latest master branch now builds with l10n if 'autoreconf -i' is rum explicitly
before build. Also it now has extended Russian and German translations.
Considering https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=888254 the
following line also may be added:
OnlyShowIn=OPENBOX
Package: obsession
Version: 20140608-2+b1
Severity: normal
Tags: patch
Dear Maintainer, I propose a desktop entry for obsession-logout:
[Desktop Entry]
Version=1.0
Type=Application
Name=Logout
Name[ru]=Выход из системы
Comment=Dialog for logout, shutdown, reboot, etc.
Package: xsane
Version: 0.999-5
Severity: normal
Tags: l10n patch
Dear Maintainer, there are some inconsistencies in xsane.desktop
Since it is provided mostly by debian patch, I would suggest replacing it with
the attachment.
Fixed issues:
- added GenericName fields
- distributed content
Related: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=888254
Further improvements make use of XDG_MENU_PREFIX.
Since gnome-menus package lacks dependencies except python3, we can construct
relatively DE-agnostic menu file for openbox using directories from gnome-menus
Proposal for menu file is attached.
openbox-applications.menu
Description: Binary data
Package: openbox
Version: 3.6.1-6
Severity: normal
Tags: patch
Dear Maintainer, I would like to propose some improvements to openbox-session
mechanism based on my experience with maintaining a simple custom desktop
enviromnent.
First, make some additions to environment vars, to beef it up closer
As far as I can tell, XDG_CURRENT_DESKTOP is used to determine autostart
process, not DESKTOP_SESSION.
Either way, both environment variables are not set by openbox.
DESKTOP_SESSION might be even contaminated by DM, i.e.
DESKTOP_SESSION=lightdm-xsession
Same goes for yet another var,
Hi.
I am also seenig this issue in initramfs-tools 0.130 with swap on LVM. Provided
patch fixes it.
Package: openbox
Version: 3.6.1-6
Severity: normal
Dear Maintainer, I've done some simple testing of obamenu and found out
that it does not survive a broken symlink desktop entry (third party
pakcaging may leave some nasty surprises, so I sometimes test things for
that):
$
Package: extace
Version: 1.9.9-7+b1
Severity: normal
Dear Maintainer, the desktop entry extace.desktop has ISO-8859-1 encoding
instead of UTF-8 mandated by the spec, or ASCII.
-- System Information:
Debian Release: buster/sid
APT prefers testing
APT policy: (900, 'testing'), (400,
The thing is, as far as I remember when network units were in place, even
before reboot, the system tried to apply them to any newly appeared interface,
i.e. created by tinc or pptp, despite MAC filter. Anyway, now it does not, so
it fixed.
> Please make sure to update your initramfs (via update-initramfs -u).
> Those link and udev rules files are embedded in the initramfs, so you
> need to ensure it is rebuilt.
>
> Can you please report back what happens if you do that?
I do not remember for certain if I did that earlier.
Anyway,
Package: wnpp
Severity: wishlist
Version: 0.7.4
URL: https://github.com/johanmalm/jgmenu
License: GPL-2
Description: jgmenu is a simple standalone X11 menu. It is highly customizable,
can be used as a pipe-menu or XDG applications menu. Plays well with tint2 and
openbox, can optionally use some
Sorry, for some reason I thought that fonts with the same name would be chosen
alphabetically. At least that was what motivated me to purge v1 while finally
installing v2 in attempt to make text look sane again after recent fontconfig
changes. A faulty memory of some old discussion perhaps.
Package: flightgear-data-base
Version: 2017.2.1+dfsg-2
Severity: normal
Dear Maintainer, the package flightgear-data-base has a hard dependency on
fonts-liberation. Since having both fonts-liberation and fonts-liberation2 on
the system may lead to undesired results, consider changing dependency
Now there are also fixes for RCE and DoS vulnerabilities.
http://blog.talosintelligence.com/2017/07/vulnerbility-spotlight-freerdp-multiple.html
https://github.com/FreeRDP/FreeRDP/commit/1648deb435ad52206f7aa2afe4b4dff71d9329bc
https://bbs.archlinux.org/viewtopic.php?pid=1718106#p1718106
Apparently, NM silently started to require >=2 level domains. I see no info on
this in upstream changelog.
But luckily they reconsidered:
dchlient.conf is stock from the package.
With dhcp=internal Network manager ignores option 119 completely and does not
even list search domains via `nmcli con show`
If I manually launch dhclient on the interface, it successfully writes search
domains to resolv.conf/
Also there is a fix for clipboard:
https://github.com/FreeRDP/FreeRDP/issues/3983
Package: system-config-printer
Version: 1.5.7-3
Severity: normal
Dear Maintainer, printing applet (started in openbox session via XDG autostart)
does not exit upon logout from openbox session. This leaves a process:
/usr/bin/python3 /usr/share/system-config-printer/applet.py
and a dangling
Reproduction steps from fresh state:
- add a library with at least one track
- switch player window layout to "with lists"
- try typing in main list
> This looks like a kernel problem. There is a related redhat bug log[1]
> and a kernel one[2], that suggests disabling the wireless interface
> might avoid the problem. Could you try that?
>
> [1] https://bugzilla.redhat.com/show_bug.cgi?id=1262957
> [2]
Package: pulseaudio
Version: 10.0-1
Severity: normal
Dear Maintainer, sound is interrupted occasionally when pulseaudio is installed.
Every 1-10 minutes there is an audible glitch. Playback source does not matter,
it happens with mpv, firefox, wine and other applications.
The problem does not
Package: gmusicbrowser
Version: 1.1.15~ds0-1
Severity: normal
Dear Maintainer, Invoking search in any list in gmusicbrowser crashes the
program.
Stderr for different situations:
In main list:
invalid iter -- stamp -1875915568 does not match requested 94319900871888 at
Here is an idea: a set of complementary packages utilizing dpkg diversions.
I.e. apulse-firefox which would move firefox to firefox-orig and replace it with
#!/bin/sh
exec apulse firefox-orig "$@"
May be others. But firefox is the only app that demands pulseaudio.
Unfortunately, developers' decision to break the sound seems to be very strong.
#775098 should be given some attention.
The it is alive, buildable, and works.
Tried installing pulseaudio and found that while it made some progress, after
all these years it still can not output sound without occasional stutter. On i7
CPU. Without load.
So I'm also in the list of people who do not want to lay any bloatware over
rock-solid ALSA soundsystem that just
> Vladimir, please would you post the 'scanimage -L' output from the
> server.
It shows among others:
device `hpaio:/net/HP_LaserJet_M1536dnf_MFP?hostname=printerhostname' is a
Hewlett-Packard HP_LaserJet_M1536dnf_MFP all-in-one
I'm using locally patched version of the packages, so
Any news on this? localOnly check is still active in current testing.
https://github.com/l3ib/nitrogen/issues/93
The cause was an empty structure in /usr/local/share/icons/hicolor/ with
supposedly broken cache.
This bug can be closed.
With 1.6.0 I have a working semi-working DHCP with NM and nm-applet.
But there is another problem with DHCP in new NM. It reports a duplicated
domain:
IP4.DOMAIN[1]: example.comexample.com
DHCP4.OPTION[20]: domain_name = example.comexample.com
This bug may be closed now as gtk2 style is packaged in qt5-style-plugins.
With this package installed and the following vars in the environment Qt now
correctly renders GTK themes.
QT_STYLE_OVERRIDE=gtk2
QT_QPA_PLATFORMTHEME=gtk2
I've tested it again.
Removed 70-persistent-net.rules, created two units in /etc/systemd/network:
10-lan1.link:
[Match]
MACAddress=xx:xx:xx:xx:xx:xx
[Link]
Name=lan1
10-wan0.link:
[Match]
MACAddress=yy:yy:yy:yy:yy:yy
[Link]
Name=wan0
(Apparently MAC values in quotes are
> ppp0 being renamed to rename13 sounds suspiciously like the old network
> interface naming is still being used.
> Can you attach your /etc/udev/rules.d/70-persistent-net.rules if present
I do not remember for sure, but I probably moved or commented out this file
back when I tested the new
I haven't observe the crash in upstream Thunderbird yet.
I seem to able to increase the probability of the crash by vigourously browsing
through mail in Icedove, caught it a few times. But upstream Thunderbird is
pretty stable with the same profile.
Crash in Icedove while browsing through mail, gdb log attached.
I've downloaded upstream Thunderbird and testing it now. But due to randomness
of the crashes, it might take a while to be sure if it is reproducible there.MOZILLA_FIVE_HOME=/usr/lib/icedove
Fixed with version 2.2.4-7, nicely done!
Confirming on current Debian testing.
03.09.2016, 13:26, "Dmitry Shachnev" :
> In Qt 5.7 the GTK+ style has been removed from the default installation.
>
> It is available in qtstyleplugins repository (home for abandoned styles),
> but the new version is not yet released or packaged.
>
> As I mentioned earlier in
I'm trying new Qt packages from experimental and found that GTK integration
does not work.
I've installed qt5-gtk-platformtheme and testing with:
QT_QPA_PLATFORMTHEME=gtk3 screengrab
App looks like it uses Fusion.
Could not find any docs on how to properly use qt5-gtk-platformtheme.
1 - 100 of 246 matches
Mail list logo