Package: matrix-synapse
Version: 1.64.0-2
Severity: minor
After upgrade to synapse 1.64.0, the service failed to restart with no
visible error message. Manually trying to launch synapse printed this:
ERROR:root:Synapse 1.64.0 needs canonicaljson<2.0.0,>=1.5.0, but got
canonicaljson==1.4.0
Package: git
Version: 1:2.35.1+next.20220211-1
Severity: minor
Tags: patch
I'm working on a large Unity project where some of out scene and prefab
files reach six digit line numbers. When using git add -p on such a file
and selecting the go to hunk option, these errors appear:
Negative repeat
Hi,
What remains to be done for this bug? Is any help needed? All bugs
which are marked as blockers have been fixed.
It's causing all sorts of subtle breakage in Blender, and at least twice
now I've bothered open source developers with spurious bug reports when
the error turned out to be
Package: libstdc++-11-dev
Version: 11.2.0-9
Severity: normal
If libtbb-dev is installed ( is present on the system), any
program using the header from C++17 or otherwise causing
to be included fails to link with this error:
/usr/bin/ld: /tmp/cc8b8juf.o: in function
Package: blender
Version: 2.93.4+dfsg-1
Severity: normal
Blender 2.93.4 is missing most options in color management settings. I only
have "sRGB" for display device, "Standard" for view transform, "None" for look
and "sRGB" and "Linear" for sequencer. A build downloaded from blender.org
shows
Package: kernel-package
Version: 13.018+nmu1
Severity: normal
I was trying to build a new kernel (5.4.35) for an old system and had an issue
where the build process would get stuck at syncconfig. Eventually I found out
that the Debian rules were inconsistent in passing the ARCH parameter to
On 29.9.2020 11.49, Nathan Willis wrote:
Actually, don't mean to make assumptions — did you already try reaching out
to upstream? Worth a shot, IMO.
I hadn't before your message. I did so shortly afterwards, but have yet
to receive a response. I need to get a publication using these fonts
Package: fonts-adf-tribun
Version: 0.20190904-2
Severity: normal
Recent versions of the ADF fonts contain an rlig (required ligatures) table
which replaces the letters ae with æ and oe with œ. This is incorrect for at
least Finnish and English and probably many other languages as well. I know
On 19.8.2020 13.38, Christian Göttsche wrote:
There is a proposed fix[1], which is claimed to resolve the issue[2].
Could you try to verify whether this change resolves the issue for you.
There is an amd64 package available at [3], from a merge-request
against the packaging[4].
Confirmed, the
Package: dovecot-gssapi
Version: 1:2.3.11.3+dfsg1-1
Severity: grave
Justification: renders package unusable
After upgrading dovecot to 1:2.3.11.3+dfsg1-1, GSSAPI login no longer works.
I'm using Thunderbird as client and it says the Kerberos ticket was not
accepted. A downgrade back to
Package: libkadm5srv-mit11
Version: 1.17-9
Severity: important
Upgrade to 1.17-9 broke my Kerberos configuration. At first I got an error
about missing database (see #962519). Downgrading all packages with krb5 in
their name back to 1.17-7 allowed me to start the KDC again, restoring user
Package: krb5-kdc
Version: 1.17-9
Severity: normal
With the recent upgrade, the kerberos servers fail to start with these
messages in log:
Jun 9 06:10:42 capybara krb5kdc[14954]: Cannot open DB2 database
'/var/lib/krb5kdc/principal': No such file or directory - while initializing
database for
Package: matrix-synapse
Version: 1.12.4-1
Severity: normal
Debian's homeserver configuration contains a https listener with certificate
files stored under /etc/matrix-synapse. However these files are not supplied
nor generated by the package and there's no instructions on how to generate
them.
Package: matrix-synapse
Version: 1.12.4-1
Severity: normal
I was trying to set up synapse and kept getting the error:
Starting matrix-synapse: matrix-synapseTLS certificate file not found ...
failed!
Turns out the synapse.config does not like certain files being missing, or
an invalid database
Package: abiword
Version: 3.0.2-9+b1
Severity: wishlist
Most programs make the cursor visible when a key is pressed and only blink
it when idle. Abiword does this too when letters are typed but ignores
arrow keys. As a result the cursor is periodically invisible when moving
around the document.
Source: gvfs
Version: 1.42.1-3
Severity: wishlist
There's a dependency chain gvfs -> gvfs-daemons -> udisks2 -> libpam-systemd ->
systemd-sysv, which is preventing me from installing gvfs. Gtk's show_uri
function uses gvfs to open a browser, and it seems to me that that should be
achievable even
Package: krb5-admin-server
Version: 1.17-4
Severity: normal
Version 1.17-4 of the kerberos 5 packages removes single DES support. If
such encryption types are enabled in kdc.conf, the admin server will error
out with:
kadmind: Required parameters in kdc.conf missing while initializing
It is
Package: librrd8
Version: 1.7.0-1+b1
Severity: important
I installed collectd-core on my system and noticed that it created directories
and rrd files with strange and insecure permissions such as 0777 and 0177. A
bit of detective work later I managed to track the problem down to this issue
in
Package: python3-certbot
Version: 0.22.0-1
Severity: serious
Justification: Policy 3.5
The dependency on python3-distutils was recently dropped from
libpython3.6-minimal and the package was automatically removed from my
system as unused. Since then certbot fails with:
Traceback (most recent
Package: borgbackup
Version: 1.1.4-2
Severity: serious
Justification: Policy 3.5
The dependency on python3-distutils was recently dropped from
libpython3.6-minimal and the package was automatically removed from my
system as unused. Since then borgbackup fails with:
Traceback (most recent call
Package: php-common
Version: 1:60
Severity: minor
The script /usr/lib/php/sessionclean goes through installed php versions and
runs the associated executable to obtain session variables. The versions are
obtained by inspecting the directory /usr/lib/php. The sessionclean script
also checks for
Package: kicad-common
Version: 4.0.7+dfsg1-1
Severity: wishlist
The 3D models take up nearly three gigabytes of space, which is nearly a
fourth of my entire installed system size. Many hobbyists do not need the
3D models and kicad is perfectly usable without them, so I suggest splitting
this
Package: tigervnc-standalone-server
Version: 1.7.0+dfsg-7
Severity: normal
I am running a graphical IRC client in a VNC server. Previously I used
realvnc, but since it got removed from Debian I switched to tigervnc around
the start of the year. Since then I've noticed that after the VNC server
of packages kernel-package suggests:
ii libncurses5-dev [libncurses-dev] 6.0+20161126-1
pn linux-source
-- Configuration Files:
/etc/kernel-pkg.conf changed:
maintainer := Mikko Rasa
email := t...@tdb.fi
priority := Low
debian = $(version)-1
-- no debconf information
On 25.02.2016 01:11, Manoj Srivastava wrote:
Thanks for the report. This should be relatively easy to fix,
and I’ll try getting to a new release perhaps this weekend.
Did you ever get to fixing this? The bug is still present, and in fact
kernel-package version hasn't changed at all
Package: gitweb
Version: 1:2.11.0-1
Severity: serious
Justification: Policy 3.5
After running full-upgrade today I noticed that my gitweb server started
returning internal server errors. Apache's error.log contained many
repeats of this error:
[Mon Dec 12 17:03:04.782271 2016] [cgi:error] [pid
-mkimage
Versions of packages kernel-package suggests:
ii libncurses5-dev [libncurses-dev] 6.0+20160213-1
pn linux-source
-- Configuration Files:
/etc/kernel-pkg.conf changed:
maintainer := Mikko Rasa
email := t...@tdb.fi
priority := Low
debian = $(version)-1
Since these emails managed to escape my mail filters and catch my
attention, I'll butt in with my opinion. Apologies if this has been
said already.
Have you considered in-source copies of the modified libraries? Perhaps
as git submodules? If you take full control of building and installing
On 22.10.2014 09:48, Andreas Beckmann wrote:
On 2014-05-29 11:49, Gleb Golubitsky wrote:
Please try 331.67.
The problem is still reproducible on the lastest nvidia-driver: OpenGL
contexts painted over other windows.
And what about the 340.xx series (testing/unstable) or 343.xx
Package: debian-installer
Version: 20140802
Severity: normal
Tags: d-i
The iso-scan step contains a loop that goes through all directories in a
partition. Line 165 contains this command:
find $dir $opt -name *.iso -o -name *.ISO
Bacause the $dir argument is not quoted, a directory name with
Package: glusterfs-common
Version: 3.5.1-1
Severity: minor
I found that I had a 1.6 gigabyte logfile for one of my bricks, containing
two years worth of logs. Adding /var/log/glusterfs/bricks/*.log to the
logrotate config file fixed this. The modified file is included at the end
of this
Package: nvidia-driver
Version: 331.49-1
Followup-For: Bug #740016
I've found that running a composition manager (such as xcompmgr) makes this
problem go away most of the time. However, occasionally a single frame still
gets painted on top of everything else. It does not happen very frequently,
Package: glusterfs-common
Version: 3.4.1-1
Severity: minor
The postrotate script in /etc/logrotate.d/glusterfs-common does not send
SIGHUP to glusterfs clients. As a result, log files for mounted filesystems
can grow indefinitely unless the machine is restarted.
An example logrotate file for
Package: libegl1-mesa-dev
Version: 9.1.3-6
Severity: minor
This bug report actually concerns version 9.1.6-2+b1, the installation of which
I'm putting off due to the reason described here.
libegl1-mesa-dev depends on libegl1-mesa-drivers, which in turn depends on
libgl1-mesa-dri. This is a
On 31.03.2013 06:45, Guillem Jover wrote:
I managed to put together a Python script which changes the
architecture of an installed Debian system from i386 to amd64. I
didn't even need to use any --force; processing packages in correct
order by dependencies is enough, since dpkg does less
On 30.03.2013 03:51, Daniel Hartwig wrote:
On 30 March 2013 06:33, Mikko Rasat...@tdb.fi wrote:
Package: apt
Version: 0.9.7.8
Severity: normal
Some background: I'm migrating my home server to new hardware and considering
changing the OS to 64-bit as well. Since apt now supports multiarch, I
Package: apt
Version: 0.9.7.8
Severity: normal
Some background: I'm migrating my home server to new hardware and considering
changing the OS to 64-bit as well. Since apt now supports multiarch, I thought
I'd do this by installing a 64-bit kernel first and then updating userspace bit
by bit as
Package: icedove
Version: 17.0-1
Severity: normal
My Icedove address book got corrupted for an unknown reason, and this
caused the program to segfault hard, without any kind of useful error
message. I eventually had to install the icedove-dbg package and use
gdb to figure out what was wrong.
Package: udev
Version: 175-5
Severity: critical
Justification: breaks the whole system
Recently, bug #654282 was fixed by not attempting to include unix.ko in
initrd images. I run custom kernels with a high degree of modularization,
so a side effect of this change was that my systems became
Package: duplicity
Version: 0.6.18-2
Severity: wishlist
Duplicity can use large amounts of resources when restoring files (see bug
#676840). In extreme cases this can cause backups to become practically
unrestorable. Apparently such long chains are not officially supported.
However, this is not
Package: duplicity
Version: 0.6.18-2
Severity: minor
Duplicity has a --tempdir option, but it affects only some temporary files.
A significant amount of temporary files is still written to /tmp, and can
prevent backup restoration in some cases (see bug #676840). The same applies
to the
On 10.06.2012 04:31, Alexander Zangerl wrote:
On Sat, 09 Jun 2012 23:49:54 +0300, Mikko Rasa writes:
I've been running daily incremental backups for about 10 months.
what ratio of full-to-incremental backups are you using?
how many incrementals are in the backup chain that you are trying
Package: duplicity
Version: 0.6.18-2
Severity: normal
I've been running daily incremental backups for about 10 months. Recently I
tried to restore a file I accidentally deleted, only to find that duplicity
ran out of space. Some investigation revealed that duplicity leaks several
kinds of
Package: abiword
Version: 2.9.2+svn20120215-2
Severity: normal
After using the borders and shading dialog to apply a border to a paragraph,
trying to open the dialog again produces a stream of errors from gobject:
(abiword:23342): GLib-GObject-WARNING **:
Package: abiword
Version: 2.9.2+svn20120215-2
Severity: minor
Scroll wheel is not working when the pointer is on top of the document area.
It does work when hovering on top of the scroll bar, but having to move the
pointer there defeats the purpose of having a scroll wheel in the first place.
In
Package: abiword
Version: 2.9.2+svn20120215-2
Severity: normal
The icon view in the clip art window is about 10 pixels high, making it
impossible to see any item in full. It's still possible to select an item
and insert it into the document. The view will resize horizontally if the
dialog is
Package: abiword
Version: 2.9.2+svn20120215-2
Severity: minor
When adding a border to a paragraph, it will add small margins on all sides,
even if not all borders are selected. In particular, if only top or bottom
border is selected, there will be margins to the left and right of the
paragraph.
On 21.09.2011 18:38, J.H.M. Dassen (Ray) wrote:
Hello Mikko,
On Mon, Sep 19, 2011 at 20:58:01 +0300, Mikko Rasa wrote:
The file in question contains names and addresses of people, so I do not
wish to publish it on the Internet. However, I can send it to the package
maintainer in exchange
Package: gnumeric
Version: 1.10.17-1
Severity: normal
When opening a particular .ods file gnumeric crashes with this message:
GLib-ERROR **:
/build/buildd-glib2.0_2.28.6-1-i386-A3fp41/glib2.0-2.28.6/./glib/gmem.c:239:
failed to allocate 2147483648 bytes
aborting...
Aborted
The number happens
Package: geeqie
Version: 1:1.0-8
Severity: minor
When opening the rename dialog, the base part of the filename (everything up
to but not including the dot) is selected. However, if there are multibyte
UTF-8 characters before the dot, some characters beyond the dot get selected.
It seems like the
Package: darktable
Version: 0.8-3
Severity: minor
Darktable uses a custom dark-colored GTK+ theme irrespective of the selected
system theme. However, the backgrounds of the TreeViews in the file selection
dialog are not themed, resulting in light gray text on white background.
-- System
Package: vim
Version: 2:7.3.035+hg~8fdc1210-1
Severity: minor
In GCC 4.5 error reporting was changed to report character position on
the line in addition to the line number. For the most part Vim handles
this just fine, but seems to get confused by the lack of an actual
error message on
Package: vim
Version: 2:7.2.436+hg~e12b9d992389-1
Severity: normal
Trying to list swap files with vim -r causes vim to immediately segfault,
without listing any swap files. This happens whether or not there are any
swap files in the current directory, and even if the directory is empty.
While
Upstream version 0.5.3 from June 2009 fixes this problem by removing the
need for usblcd. It's not advertised on the website, but can be found
from Sourceforge:
http://sourceforge.net/projects/lcdproc/files/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a
Package: gxmms2
Version: 0.7.0-2
Severity: grave
Justification: renders package unusable
The gxmms2 package still depends on libxmmsclient4, which has been superseded
by libxmmsclient5 in the 0.6DrMattDestruction release. Furthermore,
libxmmsclient5 conflicts with libxmmsclient4, and the older
Package: xserver-xorg-input-tslib
Version: 0.0.5-6
Severity: important
When running an X server without RandR support (for example, fbdev driver with
the Rotate option set), the tslib driver segfaults. A backtrace with gdb
revealed the location:
Program received signal SIGSEGV, Segmentation
This seems to be due to a change in the locale.dir file. Doing grep
fi_FI\\.UTF-8 on lenny's version (1.1.5-2) shows:
en_US.UTF-8/XLC_LOCALE fi_FI.UTF-8
en_US.UTF-8/XLC_LOCALE: fi_FI.UTF-8
While the same command for sid (1.2-1) shows:
fi_FI.UTF-8/XLC_LOCALE
Package: dovecot-common
Version: 1:1.1.11-4
Severity: minor
I had a custom SSL certificate from cacert.org installed for dovecot. After
upgrading packages last night, I discovered that I could no longer connect to
my IMAP server - dovecot was spewing errors about a mismatch in the SSL key
file.
Package: devil
Version: 1.7.5-1
Severity: important
The 1.7.5 release of DevIL added support for VTF (Valve Texture Format) files,
but the associated functions are missing definitions. Any program trying to
link libIL.so fails with the following errors:
Package: gedit
Version: 2.22.3-1+b1
Severity: minor
Tags: l10n
I have LC_CTYPE set to fi_FI.utf8, and the following items in the Tools menu
appear in Finnish:
Käännös (Build)
Avaa pääte tähän (Open terminal here)
Poista ylimääräiset tyhjät (Remove trailing spaces)
Suorita komento (Run command)
It appears that I accidentally made a diff from the upstream version to
one that contained both Debian's and my patches. Attached is a
corrected version containing only my changes.
--
Mikko Rasa The DataBeaver
--- xf86-input-tslib-0.0.5/src/tslib.c 2008-10-05 13:28:02.0 +0300
+++ xf86
Package: xserver-xorg-input-tslib
Version: 0.0.5-1
Severity: normal
Tags: patch
When the variance module is enabled in ts.conf, the tslib input driver will
emit a ButtonRelease event only when the pen is pressed the next time. This
is because the variance module caches samples and will produce
Package: iceape-browser
Version: 1.1.11-1
Severity: normal
Iceape acquires new X connections over time but does not close them properly.
Over time (on the order of days or weeks) this eventually leads to getting an
error Maximum number of clients reached when trying to start any X
application.
Package: blender
Version: 2.45-5+b1
Severity: normal
Occasionally, blender enters a select() with a NULL timeout, waiting for
events from the X server. If this happens when it's inactive for a
prolonged time, autosaves get queued up and are processed at once when
the select finishes (due to
Cyril Brulebois wrote:
Mikko Rasa [EMAIL PROTECTED] (18/06/2008):
In my opinion, the problem is not in the infinite-timeout select but
in the fact that blender seems to increment the next autosave time
instead of setting it to current time plus interval.
Hi,
thanks for your analysis. Before
I have figured out a possible source of this problem. On my system,
there was the following path:
/var/lib/oprofile/samples/current/{root}/usr/bin/env/{dep}/{anon:/home/...
I do not know how such a path came to be (possibly an earlier
differently buggy version of oprofile?), but when
Package: wine
Version: 0.9.44-1
Severity: wishlist
Wine 0.9.49 was released on Nov 9, two weeks ago. Since the most recent
version in Debian is 3 months old, it would be really nice to get an
update.
-- System Information:
Debian Release: lenny/sid
APT prefers unstable
APT policy: (500,
Package: xorg-server
Severity: wishlist
The Kdrive-based Xfbdev server is useful on embedded systems like the
Sharp Zaurus. Since it is built anyway, I see no reason not to include
it in the distribution.
-- System Information:
Debian Release: lenny/sid
APT prefers unstable
APT policy:
Package: debtags
Version: 1.7.2
Severity: important
debtags 1.7.3 depends on libapt-pkg-libc6.5-6-4.4, which is provided by
apt 0.7.3
debtags 1.7.3 also depends on libept0 (= 0.5.10)
libept0 0.5.10+b1 depends on libapt-pkg-libc6.6-6-4.4, which is provided
by apt 0.7.6
Thus, I cannot upgrade
Package: mdadm
Version: 2.5.6-9
Severity: normal
Tags: patch
In Monitor mode, mdadm fails to detect raid6 devices with --scan, and
thus the monitor daemon started by the init script does not monitor such
arrays. The following patch fixes the problem:
*** begin patch ***
--- Monitor.c.orig
Package: libgtkmm-2.4-dev
Version: 1:2.8.8-1
Severity: minor
When compiling applications with -Wshadow enabled, GCC gives the
following warning:
/usr/include/glibmm-2.4/glibmm/helperlist.h:40: warning: declaration of
‘gparent’ shadows a member of 'this'
The following patch gets rid of this
Package: gnome-applets-data
Version: 2.14.3-4
Severity: wishlist
I installed the gnome-applets package to get the system and battery
monitor applets. It took 40MB of disk space, and a closer examination
revealed this:
-rw-r--r-- 1 root root 20877053 2007-01-03 20:56
Package: hal
Version: 0.5.8.1-4
Severity: normal
If I unplug a USB storage that is mounted, HAL gets stuck and doesn't
respond to any more USB plugging events, even if the device is unmounted
after that. It does react to the unmounting event itself, as well as
events produced by plugging in a
Package: libgl1-mesa-dri
Version: 6.4.2-1.1
Severity: grave
Justification: renders package unusable
I have recently been getting serious rendering problems with Mesa. For
example, glxgears looks like this:
http://tdb.fi/~tdb/glxgears-bug.png
My own applications show their graphics shifted
I have recently been getting serious rendering problems with Mesa. For
example, glxgears looks like this:
http://tdb.fi/~tdb/glxgears-bug.png
This particular issue is a glxgears bug which has been fixed upstream in
Mesa 6.5. It only happens with some window managers or when running
without a
Package: libwxgtk2.6-0
Version: 2.6.1.2
Severity: minor
The GTK file chooser dialogs created by wxgtk do not have a default
button specified. This makes it necessary to manually activate the
button, which is slightly annoying when trying to work with keyboard
only.
-- System Information:
Package: coreutils
Version: 5.93-5
Followup-For: Bug #340236
I hit this problem as well today. Poking around with strace, I noticed
several things:
1. cp executes utimes before closing the file. This might be a problem,
although it seems to work on local destinations.
open(baz,
Package: libavifile-0.7c102
Version: 1:0.7.43.20050224-1
Severity: wishlist
The default C++ compiler in unstable recently changed to g++ 4.0, which
by default links programs against libstdc++6. However, libavifile is
still linked against libstdc++5, which makes it unusable with g++ 4.0.
Package: fluxbox
Version: 0.9.12-1
Severity: minor
When changing from on style to another, the menu title texture only updates
if its size changes. For example, when changing between styles which only
differ in colors, the menu title texture isn't updated. This doesn't affect
any other
Package: memtest86+
Version: 1.55.1-1
Severity: normal
I'm running memtest86+ on a machine with 3GB of RAM installed and I noticed
that the progress bar only goes up to 20%. I suspect this has something to do
with the range of signed integers. I'm not sure if this affects the amount of
memory
Package: libxine1
Version: 1.0-1
Severity: normal
The Ogg demuxer plugin requires libtheora to be installed. However, libxine
does not depend on that package. The user won't even get any kind of useful
error when running xine normally. I wondered a few minutes why it wouldn't play
my ogm
81 matches
Mail list logo
