Bug#1036457: flash-kernel: please add BeagleBone AI-64 support

Package: flash-kernel Version: 3.107 Severity: wishlist X-Debbugs-Cc: sascha-debian-bugs-flash-kernel-20230...@silbe.org Dear Maintainer, I successfully tested flash-kernel on a BeagleBone AI-64 running Bookworm with the following entry in /etc/flash-kernel/db: === Begin === Machine:

Bug#1010605: pytest: please package pytest 7.1.x

Source: pytest Version: 6.0.2-2 Severity: wishlist X-Debbugs-Cc: sascha-debian-bugs-pytest-2022-05...@silbe.org Dear Maintainer, pytest 7.0.0 introduced a couple useful new features. It would be great to have the latest upstream version (7.1.2 at the time of writing) in Debian. Thanks in

Bug#1010391: afew: missing dependency on python3-dnspython

Package: afew Version: 3.0.1-1 Severity: normal X-Debbugs-Cc: sascha-debian-bugs-afew-2022-04...@silbe.org Dear Maintainer, after a fresh installation afew fails to start because it cannot find the "dns" module: === Begin === Traceback (most recent call last): File "/usr/bin/afew", line 33,

Bug#1008762: debos: does not work with Linux 5.13+

Package: debos Version: 1.0.0+git20201203.e939090-4+b3 Severity: important Tags: upstream X-Debbugs-Cc: sascha-debian-bugs-debos-2022-03...@silbe.org Dear Maintainer, debos does not work on systems running a 5.13 or newer kernel because the 9pnet_virtio and virtio_pci modules cannot be loaded:

Bug#999772: limesuite-udev: replaces entire /dev/serial directory with single symlink

Package: limesuite-udev Version: 20.10.0+dfsg-2 Severity: important X-Debbugs-Cc: sascha-debian-bugs-2021-11-16-limesuite-u...@silbe.org Dear Maintainer, the udev rules contained in limesuite-udev cause *all* USB serial adapters using the default vendor / product id of several FTDI chips

Bug#995444: bsdutils: "script" no longer prints output file name at the end

Package: bsdutils Version: 1:2.36.1-8 Severity: normal Tags: upstream Dear Maintainer, since Bullseye, probably caused by upstream commit d805688afc0c709154c9c6f29383b175c05ffc92 ("script: report also timing file, do it only once"), "script" no longer outputs the file name at the end. I'm using

Bug#994093: libgradle-android-plugin-java: tries to use removed method setDependencyCacheDir()

Package: libgradle-android-plugin-java Version: 2.2.2-3 Severity: grave Justification: renders package unusable X-Debbugs-Cc: sascha-debian-bugs-libgradle-android-plugin-java-2021-09...@silbe.org Dear Maintainer, the Android Gradle Plugin version in sid (2.2.2-3) does not work with the Gradle

Bug#971696: linux-image-4.19.0-11-armmp: network hangs, oops in ath9k_htc after upgrade

Sascha Silbe writes: > since the last kernel update (from linux-image-armmp:armhf > 4.19+105+deb10u5 / linux-image-4.19.0-10-armmp 4.19.132-1 to > linux-image-armmp 4.19+105+deb10u6 / linux-image-4.19.0-11-armmp > 4.19.146-1) our BeagleBone Black systems with AR9271 based

Bug#971696: linux-image-4.19.0-11-armmp: network hangs, oops in ath9k_htc after upgrade

Package: src:linux Version: 4.19.146-1 Severity: important Dear Maintainer, since the last kernel update (from linux-image-armmp:armhf 4.19+105+deb10u5 / linux-image-4.19.0-10-armmp 4.19.132-1 to linux-image-armmp 4.19+105+deb10u6 / linux-image-4.19.0-11-armmp 4.19.146-1) our BeagleBone Black

Bug#950508: debos: broken in Buster: libresolv.so.2: cannot open shared object file

Package: debos Version: 1.0.0+git20190123.d6e16be-1 Severity: grave Justification: renders package unusable Dear Maintainer, after the upgrade to Buster debos stopped working. This happens even with the minimal example from the man page. The output when run with --show-boot suggests a shared

Bug#944518: linux-image-4.19.0-6-armmp: please enable CONFIG_INPUT_TPS65218_PWRBUTTON

Source: linux Version: 4.19.67-2+deb10u1 Severity: wishlist Dear Maintainer, currently it's impossible to use the power button on the BeagleBone Black when running the Debian kernel. Please enable CONFIG_INPUT_TPS65218_PWRBUTTON. Kind regards, Sascha Silbe -- Package-specific info: ** Version

Bug#944256: linux-image-4.19.0-6-armmp: please enable cpufreq for AM335x (BBB)

. For starters enabling CONFIG_ARM_TI_CPUFREQ will be required; not sure if other options are missing, too. Kind regards, Sascha Silbe -- Package-specific info: ** Version: Linux version 4.19.0-6-armmp (debian-ker...@lists.debian.org) (gcc version 8.3.0 (Debian 8.3.0-6)) #1 SMP Debian 4.19.67-2 (2019-08

Bug#846175: gnupg-agent: Unable to use SSH key with empty passphrase since upgrade to Stretch

Package: gnupg-agent Version: 2.1.18-8~deb9u2 Followup-For: Bug #846175 Dear Maintainer, after the upgrade to Stretch we're hitting this bug, too. We have an SSH key that's shared between a group of users and used by automated processes, too (so it cannot be password-protected). The OpenSSH

Bug#901675: notion: Unconditionally uses /bin/sh instead of user shell

Package: notion Version: 3+2015061300-2+b1 Severity: normal Dear Maintainer, I'm defining a couple of shell functions in my ~/.bash_profile and export them so I can use them in xterms started from within notion. This worked fine in jessie, but broke with the update to stretch. Apparently in

Bug#596334: dnsutils: Still happens in Stretch

Package: dnsutils Version: 1:9.10.3.dfsg.P4-12.3+deb9u3 Followup-For: Bug #596334 Control: found 596334 1:9.10.3.dfsg.P4-12.3+deb9u3 Dear Maintainer, I'm not the original reporter but I can still reproduce this bug in Stretch: sascha@vorratskeller3:~$ dig +sigchase +dnssec +topdown -t MX

Bug#868352: linux: Please enable USB_SERIAL_CONSOLE

Source: linux Version: 4.11.6-1 Severity: normal Dear Maintainer, on some amd64 systems no native or PCI(e) serial ports are available resp. possible. USB serial adapters are the only option to get a serial console for debugging boot problems in this case. Unfortunately the Debian amd64 kernels

Bug#862083: firejail: crashes when using --allow-debuggers ("free(): invalid pointer")

Dear Reiner, Reiner Herrmann writes: [...] > I was able to reproduce the issue in a jessie chroot with the bpo package. > After some debugging I found that it is a memory corruption in fs.c. > Fortunately it has also already been fixed upstream [1], which is > already part

Bug#862218: mumble: spams console with "warning: Unknown speex_preprocess_ctl request: 26" (and 35)

Package: mumble Version: 1.2.8-2 Severity: normal Dear Maintainer, on all the ARM based systems that I've tried (OLPC XO-1.75 (*), Wandboard Quad, OpenRD Base) mumble spams stdout or stderr with these lines: === Begin === warning: Unknown speex_preprocess_ctl request: 26 warning: Unknown

Bug#862085: spice-client: Doesn't start at all: crtc_overlap_test: unsupported partial overlap

Package: spice-client Version: 0.12.5-1+deb8u4 Severity: important Dear Maintainer, not sure exactly when it broke (upgrade to jessie or some later update), but spicec doesn't work at all anymore on my system: === Begin === sascha.silbe@twin:~$ spicec -h localhost -p 1234

Bug#862083: firejail: crashes when using --allow-debuggers ("free(): invalid pointer")

Package: firejail Version: 0.9.44.8-1~bpo8+1 Severity: normal Dear Maintainer, when passing --allow-debuggers to firejail to enable strace or gdb to work inside firejail (in order to figure out why the sandboxed application doesn't work), firejail crashes immediately on start-up:

Bug#861272: fdroidserver: Backport of fdroidserver does not work with aapt from Jessie

Package: fdroidserver Version: 0.7.0-1~bpo8+1 Severity: important Dear Maintainer, the Jessie backport of fdroidserver has only an unversioned recommend on aapt, so the Jessie version gets pulled in. "fdroid update" chokes on the output of "aapt dump badging ", particularly the

Bug#858048: [Pkg-tigervnc-devel] Bug#858048: tigervnc: Outdated build dependencies

Dear Ola, Ola Lundqvist writes: > Thank you for the report. > > Just a check. These two problems you reported, they are specific to jessie, > right? I suppose so. Haven't tried building on Stretch or Sid. But since both ship recent enough versions of the relevant packages, it

Bug#858048: tigervnc: Outdated build dependencies

Source: tigervnc Version: 1.7.0+dfsg-6~bpo8+1 Severity: minor Dear Maintainer, while building a local backport of tigervnc to jessie, I noticed that some of the build dependencies were too lax. At least the following are now required: === Begin === x11proto-core-dev (>= 7.0.31),

Bug#857031: cabal-install: Please update backport to 1.24.x

Package: cabal-install Version: 1.22.6.0-2~bpo8+1 Severity: normal Dear Maintainer, the versions of cabal-install in jessie and jessie-backports contain severe memory leaks, causing systems with "only" 512MiB RAM (e.g. ARM based servers, some VMs / virtual servers) to run out of memory during

Bug#854499: git-daemon-sysvinit: unconditionally enables --verbose, causing IP addresses to be logged

Package: git-daemon-sysvinit Version: 1:2.1.4-2.1+deb8u2 Severity: normal Dear Maintainer, /etc/init.d/git-daemon always passes the --verbose option to git-daemon. Since there is no corresponding option like --quiet that would be able to counter it, there's no way to disable IP address logging

Bug#854313: bleygraph needs python-matplotlib, cannot handle dbconfig-common.conf

Package: bley Version: 2.0.0-2 Severity: normal Dear Maintainer, tried to run bleygraph; ran into two issues: 1. bleygraph needs python-matplotlib, but the bley package doesn't depend on (or even just suggest) it: sascha@outpost:~$ sudo -u bley bleygraph -d /tmp/bley Traceback (most

Bug#853216: rsyslog: mmanom does not anonymise IPv6 addresses

Package: rsyslog Version: 8.4.2-1+deb8u2 Severity: normal Dear Maintainer, rsyslog ships the mmanom module which is useful for stripping IP addresses from log messages for services that do not offer log format customisation themselves. Unfortunately it doesn't support IPv6, so it's of no use for

Bug#853185: gatling: No simple way to disable logging of IP addresses

Package: gatling Version: 0.13-5+b3 Severity: normal Dear Maintainer, there's no way to disable logging of IP addresses in gatling itself as all the logging is hard-coded. As logging is done to a custom log file rather than via syslog, filtering via the syslog daemon isn't possible either. The

Bug#548880: closed by Julien Cristau <jcris...@debian.org> (Bug#548880: fixed in debootstrap 1.0.85)

Dear Ansgar, dear Julien, Debian Bug Tracking System <ow...@bugs.debian.org> writes: [...] > debootstrap (1.0.85) unstable; urgency=medium [...] >[ Ansgar Burchardt ] [...] >* Error out when seeing short options. (Closes: #548880) Kudos! Sascha -- Softwareentwicklu

Bug#825861: freecad: After update to 0.16 design contents are "invisible", ViewProvider2DObjectPython on console

Package: freecad Version: 0.16+dfsg2-1~bpo8+1 Severity: important Dear Maintainer, thanks for providing a backport for 0.16! Unfortunately after updating, I cannot load my document created with the 0.15.4671 backport anymore. When loading my "old" document, the 3D view continues to show the

Bug#581039: tightvnc: Version 2.7.10 available by now; fixes key bindings for Mac OS X vnc server

Dear Ola, Ola Lundqvist o...@inguza.com writes: I'll see what I can do. The package is up for adoption as I have little spare time (kids...), but I'll see if I can get some free time some day. TL;DR: TightVNC 2.x is Windows-only. Cannot send Option key with any non-Apple VNC viewer to

Bug#581039: tightvnc: Version 2.7.10 available by now; fixes key bindings for Mac OS X vnc server

Source: tightvnc Followup-For: Bug #581039 Dear Maintainer, it would be really great to have the latest upstream release (version 2.7.10 from 2013-07-24) packaged. Besides tons of other changes compared to 1.3.9, it supposedly fixes sending Option key presses (Windows key on the local side) to

Bug#787001: pgrep/pkill: incorrect return code on syntax error

Package: procps Version: 1:3.3.3-3 Severity: normal File: /usr/bin/pgrep Dear Maintainer, the manual page states that in case of a syntax error, pgrep / pkill with return with an exit status of 2. This is useful to distinguish between proper operation (but no match found) and incorrect usage.

Bug#775953: openntpd if-up.d hook can cause system boot to hang indefinitely

Hello Dererk, Dererk der...@debian.org writes: I'm trying to reproduce this, and trying to understand your working environment, which seems to be an armel architecture of an ARMv5tel device. Is your device being able to be emulated by qemu or some equivalent tool? What physical device is

Bug#775953: openntpd if-up.d hook can cause system boot to hang indefinitely

Package: openntpd Version: 20080406p-4 Severity: critical Justification: breaks the whole system Dear Maintainer, when a) openntpd is configured to listen on some interface and b) openntpd is configured to step the time on start-up and c) the DNS servers are not reachable for any reason, the

Bug#775953: openntpd: no longer happens with wheezy-backports version

Package: openntpd Followup-For: Bug #775953 Dear Maintainer, after upgrading to the openntpd version from wheezy-backports (no other changes), the indefinite hang no longer occurs. Kind regards, Sascha -- System Information: Debian Release: 7.7 APT prefers stable-updates APT policy: (500,

Bug#766867: python-gevent: SSL support broken with current Python in Jessie (2.7.8)

Package: python-gevent Version: 1.0.1-1 Severity: normal Dear Maintainer, with current python2.7 (2.7.8-10) from Jessie, the SSL support in python-gevent is broken for at least two reasons: 1. References no longer existing ssl._ssl.sslwrap() gevent.ssl.SSLSocket.connect() tries to use

Bug#747871: mkinitramfs: breaks with LUKS root on MMC (mkinitramfs: for root /dev/dm-0 missing mmcblk /sys/block/ entry) and MODULES=dep

Package: initramfs-tools Version: 0.109.1 Severity: normal File: /usr/sbin/mkinitramfs Dear Maintainer, on a system with the rootfs living inside LUKS on an SD card, mkinitramfs (and thus update-initramfs) fails with: === Begin === root@mimosa:/boot# update-initramfs -k

Bug#738556: libaqbanking: Please package and backport 5.3.6beta

Package: libaqbanking Version: 5.3.5beta-2 Severity: important Dear Maintainer, please package 5.3.6beta and provide a backport of it for wheezy. Aqbanking 5.3.6beta [1] fixes a bug that prevented interoperation with some banks when using RDH-10. In addition, the aqbanking 5.2 and 5.3 series

Bug#717904: w3m: doesn't support IPv6 link-local addresses

Package: w3m Version: 0.5.3-8 Severity: normal Dear Maintainer, w3m works fine with URLs containing an IPv6 address of global scope. However, when adding the interface specifier required for link-local addresses, w3m completely mis-parses the URL. This is apparent when using a proxy: w3m

Bug#696176: No problem here

Filipus Klutiero chea...@gmail.com writes: I also use a Radeon HD 7660D (on A10-5700) and do not experience this issue on my wheezy install (using Linux 3.8). Can you reproduce with current versions? Yes, I can still reproduce this using xserver-xorg-video-radeon 1:6.14.4-8 and

Bug#696176: [Fwd: Undelivered Mail Returned to Sender]

Michel Dänzer daen...@debian.org writes: Not very nice... sascha-debian-bugs-xserver-xorg-video-radeon-2012-12...@silbe.org: host b.mx.chost.de[87.106.8.89] said: 534 Message header size, or recipient list, exceeds policy limit. (in reply to end of DATA command) Not sure how you've

Bug#710672: python-webdav: Unsolicited use of persistent connections causes HTTP/1.0 clients to hang

Package: python-webdav Version: 0.9.8-3 Severity: normal Dear Maintainer, pywebdav uses persistent connections even for HTTP/1.0 GET requests without a Connection: Keep-Alive header. This causes the client to hang waiting for the connection to close. RFC2616 explicitly states HTTP/1.1 servers

Bug#710690: python-webdav: recursive allprops only yields a subset of properties

Package: python-webdav Version: 0.9.8-3 Severity: normal Dear Maintainer, when all properties for a resource and its children are requested, pywebdav only returns the value of those properties that are also set on requested collection itself. It does not return the properties that are only on

Bug#710697: python-webdav: no support for properties that are not plain XML-safe character strings (binary data, XML nodes)

Package: python-webdav Version: 0.9.8-3 Severity: wishlist Dear Maintainer, there's currently no way to return property values from dav_interface.get_prop() that are anything else than plain characters strings with no XML-unsafe characters. While PROPFIND.mk_prop_response() would accept a

Bug#707564: minicom: please show config or device name in status bar

Package: minicom Version: 2.6.2-1 Severity: wishlist Dear Maintainer, I'm often using multiple minicom instances to work on multiple serial ports (connected both to different devices and to different ports of the same device) in parallel. Currently, it's hard to distinguish these instances as

Bug#707564: minicom: please show config or device name in status bar

Adam Lackorzynski a...@os.inf.tu-dresden.de writes: Starting minicom with -T will show the device name instead of the online time. I guess that's what you're looking for. Thanks for the hint; I wouldn't have expected the option to have this effect, based on the description in the manual page

Bug#699205: [3.7.1-3.7.3 regression] ps, pgrep and pkill segfault

Version: 3.8.5-1~experimental.1 This doesn't happen with linux-image-3.8-trunk-amd64 3.8.5-1~experimental.1 (currently available from experimental). The problematic kernel version (linux-image-3.7-trunk-amd64 3.7.3-1~experimental.1) isn't available from experimental anymore, so I'm considering

Bug#706518: python3-serial: rfc2217 support broken (cannot even import serial.rfc2217)

Package: python3-serial Version: 2.5-2.1 Severity: important Dear Maintainer, the RFC2217 support of pyserial, at least when used with Python 3, is completely broken. Even just importing serial.rfc2217 fails: === Begin === sascha.silbe@twin:~$ python3 Python 3.2.3 (default, Feb 20 2013,

Bug#706518: python3-serial: rfc2217 support broken (cannot even import serial.rfc2217)

Sascha Silbe sascha-debian-bugs-pyserial-2013-05...@silbe.org writes: the RFC2217 support of pyserial, at least when used with Python 3, is completely broken. Even just importing serial.rfc2217 fails: [...] I managed to fix that part by replacing serial.serialutil.to_bytes

Bug#699914: python3-serial: TCP mode: data loss, timeout handling broken

Package: python3-serial Version: 2.5-2.1 Severity: normal Dear Maintainer, when used in TCP mode ('socket://address:port'), python3-serial will loose most of the received characters. This is because SocketSerial.read() _replaces_ the buffer instead of appending to it: [...] data =

Bug#699218: procps: FTBFS with test suite error: testsuite/lib.test/fileutils_badfd.sh missing

Package: procps Version: 1:3.3.3-2 Severity: serious Justification: FTBFS Dear Maintainer, in order to help diagnosing #699205, I need to rebuild procps with debugging options as there's no pre-built -dbg package (hint, hint). Unfortunately that fails with a test suite error, despite nocheck

Bug#699205: linux-image-3.7-trunk-amd64: ps, pgrep and pkill segfault after upgrade from 3.7.1-1~experimental.2 to 3.7.3-1~experimental.1

Jonathan Nieder jrnie...@gmail.com writes: after updating the linux-image-3.7-trunk-amd64 from 3.7.1-1~experimental.2 to 3.7.3-1~experimental.1 today and rebooting, the tools ps, pgrep and pkill (package procps) are always segfaulting: Just to check: if you downgrade to 3.7.1 again, do the

Bug#699205: linux-image-3.7-trunk-amd64: ps, pgrep and pkill segfault after upgrade from 3.7.1-1~experimental.2 to 3.7.3-1~experimental.1

Package: src:linux Version: 3.7.3-1~experimental.1 Severity: important Dear Maintainer, after updating the linux-image-3.7-trunk-amd64 from 3.7.1-1~experimental.2 to 3.7.3-1~experimental.1 today and rebooting, the tools ps, pgrep and pkill (package procps) are always segfaulting: === Begin ===

Bug#697624: minicom: can't use /dev/serial/by-path/*usb* device names

Package: minicom Version: 2.6.1-1 Severity: normal Dear Maintainer, 123456789012345678901234567890123456789012345678901234567890123456789012 minicom cuts off the name of the serial device file at the first colon, so by-path persistent serial device names can't be used; at least not for USB

Bug#696989: policykit-1: refuses all actions if user is member of a large number of groups

Package: policykit-1 Version: 0.105-1 Severity: important Dear Maintainer, PolicyKit refuses all actions for my own user account, but works fine for other accounts: === Begin SSH session as sascha.silbe === sascha.silbe@twin:~$ cat

Bug#696989: policykit-1: refuses all actions if user is member of a large number of groups

of groups before, barring all access to users with a large number of groups. . Dynamically allocate the list instead, potentially resizing it based on what getgroupslist() tells us. Author: Sascha Silbe sascha-...@silbe.org --- The information above should follow the Patch Tagging Guidelines

Bug#692009: bup: should Recommend python-pyxattr and python-pylibacl

Package: bup Version: 0.25~git2011.11.04-5 Severity: minor Dear Maintainer, bup meta requires python-pyxattr and python-pylibacl in order to back up and restore extended attributes resp. ACLs, so they should be listed in Recommends or at least Suggests. === Begin === sascha.silbe@twin:~/baz$

Bug#689499: python-notmuch: Please install Python 3 bindings

Package: python-notmuch Version: 0.13.2-1 Severity: normal Dear Maintainer, notmuch ships with Python 3 bindings, but the Debian package doesn't install them. Some third-party tools (that use the notmuch bindings) work better with Python 3 than with Python 2.x. There may even be some that don't

Bug#687250: scdaemon: spamming syslog if reader cannot be found

Package: scdaemon Version: 2.0.19-1 Severity: important Dear Maintainer, If scdaemon cannot find the configured reader, it will flood syslog: Sep 11 09:17:42 twin pcscd: winscard.c:241:SCardConnect() Reader Gemalto USB Shell Token V2 01 00 Not Found Sep 11 09:17:42 twin pcscd:

Bug#685365: mpd: init script shows start-create-db in help message, but does not support it

Package: mpd Version: 0.16.7-2 Severity: normal Dear Maintainer, The help output for /etc/init.d/mpd includes the start-create-db action: root@twin:/var/lib/mpd/music# /etc/init.d/mpd --help Usage: /etc/init.d/mpd {start|start-create-db|stop|restart|force-reload} However trying to execute it

Bug#683335: xpdf: can't disable reverse video on the command line anymore

Package: xpdf Version: 3.03-10 Severity: normal Dear Maintainer, after upgrading from Squeeze to Wheezy, the +rv option that previously disabled reverse video (when it had been enabled by X resources) doesn't work anymore: sascha.silbe@twin:~$ xpdf +rv time_based_workflow.pdf error: +rv file

Bug#683343: conkeror: extensions screen never finishes loading

Package: conkeror Version: 1.0~~pre+git120527-1 Severity: important Dear Maintainer, after the update to Wheezy, the extensions screen (invoked via Alt+x extensions) shows only a constantly spinning Loading... message. It doesn't use the CPU much and doesn't seem to be doing anything on disk

Bug#642903: nodm: config failure still happens, makes dist-upgrade to Wheezy fail

Package: nodm Version: 0.11-1.2 Followup-For: Bug #642903 Dear Maintainer, this still happens and causes the distro upgrade from Squeeze to Wheezy to fail. I was able to recover from the dist-upgrade failure, but it wasn't exactly a smooth experience (due to a combination of factors, the

Bug#565764: python-dogtail: Please package 0.8.0 as well

Package: python-dogtail Severity: normal In the meantime (over two years), version 0.8.0 has been released. This is the version required to test GTK3 based systems. Upstream will continue maintaining 0.7.x for testing GTK2 based systems. AIUI, both versions would need to be installable in

Bug#675513: libgeier0: loaded xmlsec library version is not compatible

Package: libgeier0 Version: 0.13-1 Severity: grave Justification: renders package unusable Followup-For: Bug #675513 Dear Maintainer, the purpose of libgeier is to send VAT reports to the IRS (or rather the German equivalent). The xmlsec incompatibility issue makes that impossible, so I'm

Bug#668997: bup: breaks if PYTHONOPTIMIZE is set

Package: bup Version: 0.17b-2squeeze1 Severity: important bup uses the Python assert statement for regular operations, not just for additional sanity checks that go beyond what should be done during normal operation (i.e. debugging assertions). Python disables assert when running in optimised

Bug#648724: libgconf2-4: GCONF_DEFAULT_SOURCE_PATH support broken

Excerpts from Sascha Silbe's message of 2011-11-29 14:52:52 +0100: My patch has landed upstream without changes as commit c129898afaa562ffc38f434e5e0c607f525101b8. Is there a chance of cherry-picking the fix into Debian? It's already in upstream git, but there hasn't been a release yet and

Bug#593815: bugs.debian.org inaccessible when IPv6 is enabled but not routed

Package: squid3 Version: 3.1.6-1.2+squeeze2 Severity: normal During an IPv6 outage today I hit this bug again and tried out the patch [1] from Kusanagi Kouichi sl...@ac.auone-net.jp (thanks, Kusanagi!). Since my IPv6 upstream was working again before the squid build was completed, I used

Bug#662683: w3m: incorrect URLs for GET forms with a fragment identifier

Package: w3m Version: 0.5.2-10 Severity: normal Trying to reply to a comment on a Trac instance doesn't work because w3m appends parameters to the end of the URL even though it contains a fragment identifier. To give an example, this is the HTML code for the first Reply button on [1]: form

Bug#631807: segfault in libcap-ng0 is back on armel - filecap , bluetoothd etc

Package: libcap-ng0 Version: 0.6.6-1 Followup-For: Bug #631807 Dear Maintainer, I'm afraid I can also confirm this bug. Took me some time to realise that it's not gnome-keyring-daemon's fault that it crashes (with a segfault) on every invocation, even --help. Fortunately, as Thomas Maass

Bug#652747: listadmin: --remove-member fails silently

Package: listadmin Version: 2.40-4 Severity: normal listadmin --remove-member appears to succeed (prints Ok, return code 0), but listadmin -l shows that the member remains on the list: sascha.silbe@twin:~$ listadmin -l a@l.a.c |grep r@l.o.a r@l.o.a sascha.silbe@twin:~$ listadmin --remove-member

Bug#652480: wget: Fails to verify SSL server certificate

Package: wget Version: 1.13.4-1 Severity: normal Dear Maintainer, wget fails to connect to https://patchwork.sugarlabs.org/, claiming the certificate is untrusted: (wheezy-jhbuild)sascha.silbe@twin:~/sugar-jhbuild/source/sugar$ wget -d -O - https://patchwork.sugarlabs.org/patch/1084/mbox/

Bug#648724: libgconf2-4: GCONF_DEFAULT_SOURCE_PATH support broken

tags 648724 + patch fixed-upstream thanks My patch has landed upstream without changes as commit c129898afaa562ffc38f434e5e0c607f525101b8. Sascha -- http://sascha.silbe.org/ http://www.infra-silbe.de/ signature.asc Description: PGP signature

Bug#648724: libgconf2-4: GCONF_DEFAULT_SOURCE_PATH support broken

Package: libgconf2-4 Version: 3.2.3-1 Severity: important Dear Maintainer, the upgrade from gconf 2.32.4-1 to 3.2.3-1 broke the support for GCONF_DEFAULT_SOURCE_PATH. The latter had been introduced to allow tools like jhbuild to tell GConf about additional places to search for defaults and

Bug#648724: [PATCH] Fix #648724

-GCONF_DEFAULT_SOURCE_PATH-support-.patch: +Patch to fix GCONF_DEFAULT_SOURCE_PATH support when building with D-Bus +backend. Closes: #648724 + + -- Sascha Silbe sascha-...@silbe.org Mon, 14 Nov 2011 20:24:27 + + gconf (3.2.3-1) unstable; urgency=low [ Jeremy Bicha ] diff --git a/debian/patches

Bug#647313: xdg-user-dirs-update uses incorrect config file paths

Package: xdg-user-dirs Version: 0.14-1 Severity: important Dear Maintainer, xdg-user-dirs-update tries to load the system-wide configuration files from /etc, rather than from /etc/xdg where they have been installed by the package: (wheezy-jhbuild)sascha.silbe@twin:~$ strace -f -s 256

Bug#637823: [Debian-olpc-devel] Bug#637823: uses deprecated python-evince

Excerpts from Michael Biebl's message of Sun Aug 14 23:34:47 +0200 2011: sugar-read-activity uses python-evince, a python binding for libevince which is part of gnome-python-desktop 2.32. python-evince will no longer be buildable, as soon as we upload evince 3.0 to unstable (it's currently

Bug#644010: python-hippocanvas: Needs to be rebuilt against Python 2.7

Package: python-hippocanvas Version: 0.3.1-1 Severity: grave Justification: renders package unusable Dear Maintainer, the default Python version in Wheezy has been changed to 2.7 and consequently python-hippocanvas broke as it only ships a Python 2.6 extension module: Traceback (most recent

Bug#640222: nfs-common: init script failure during installation, leaving package half-configured

Package: nfs-common Version: 1:1.2.4-1 Severity: normal For an unknown reason /etc/init.d/nfs-common start failed during installation of nfs-common, leaving the package in half-configured state: sascha.silbe@mimosa:~$ sudo aptitude install nfs-common The following NEW packages will be

Bug#638716: etckeeper: please ignore shadow backup files (/etc/{passwd, group, shadow, gshadow}-)

Package: etckeeper Version: 0.48 Severity: wishlist The shadow backup files /etc/{passwd,group,shadow,gshadow}- duplicate data already tracked by etckeeper. Please add them to the default ignore list. -- System Information: Debian Release: 6.0.2 APT prefers stable-updates APT policy: (500,

Bug#638718: debootstrap: please add loopback configuration to /etc/network/interfaces

Package: debootstrap Version: 1.0.26+squeeze1 Severity: wishlist I'm filing this against debootstrap because the system was installed using debootstrap and /etc/network/interfaces isn't owned by any package, so my best guess is that the file was created by debootstrap. Feel free to reassign this

Bug#638723: network-manager: uses group netdev in D-Bus policy files, but does not create the group

Package: network-manager Version: 0.8.4.0-2 Severity: normal On a brand new system installed using debootstrap, the following error occurs during installation of NetworkManager: Setting up dbus (1.4.14-1) ... Starting system message bus: dbusUnknown group netdev in message bus configuration

Bug#638723: Acknowledgement (network-manager: uses group netdev in D-Bus policy files, but does not create the group)

This appears to have been just a sequencing issue: after the installation was finished, it worked fine: root@mimosa:/etc# /etc/init.d/dbus reload Reloading system message bus config...done. root@mimosa:/etc# I needed to restart installation after running into Debian#630750 and can't rule out

Bug#573518: mime-support: outdated for several years, Fedora git repo

Package: mime-support Version: 3.48-1 Severity: normal mime-support apparently hasn't been synced with upstream (IANA) for several years now. The MIME type I've noticed as missing (application/vnd.olpc-sugar) has been registered at IANA since at least 2007-06-07 [1]. Having proper file

Bug#635289: [PATCH sugar-base] Add support for IPython 0.11+

Silbe si...@activitycentral.com Reviewed-by: Sascha Silbe si...@activitycentral.com --- src/sugar/logger.py |8 +++- 1 files changed, 7 insertions(+), 1 deletions(-) diff --git a/src/sugar/logger.py b/src/sugar/logger.py index 275c57d..21cd2c9 100644 --- a/src/sugar/logger.py +++ b/src

Bug#634265: d-feet: GTK warnings on start-up

Package: d-feet Version: 0.1.10-3 Severity: minor d-feet prints the following warnings on start-up: === Begin === /usr/lib/pymodules/python2.6/dfeet/_ui/uiloader.py:38: GtkWarning: Unknown property: GtkAction.enabled self.ui.add_from_file(ui_dir + '/' + file)

Bug#631384: python-webdav: Incomplete PROPFIND response

Excerpts from Daniel Baumann's message of Sun Jul 10 15:26:36 +0200 2011: you forgot to attach the patch. Oops, sorry. Here it is. Sascha -- http://sascha.silbe.org/ http://www.infra-silbe.de/ propfind-ns-fix.patch Description: Binary data signature.asc Description: PGP signature

Bug#631384: python-webdav: Incomplete PROPFIND response

Package: python-webdav Version: 0.9.4-1+squeeze1 Severity: normal Tags: patch python-webdav returns only the properties of a single namespace, not all of them. That's because the two loops in DAV.propfind.PROPFIND.mk_propname_response() are sequential instead of nested. Patch attached. --

Bug#628484: python-xpcom: 'print' method from nsIWebBrowserPrint doesn't work because it triggers a syntax error

Package: python-xpcom Version: 1:0.0~hg20100212-5+1 Severity: normal When trying to access the 'print' method of the nsIWebBrowserPrint interface, python-xpcom generates Python code on the fly and fails to evaluate it because 'print' is a reserved keyword in Python:

Bug#627450: paprass available only in french

Package: paprass Version: 2.05-1 Severity: normal paprass is written in french without translation to any other language, not even to english. It's very hard to use simply because I'd have to look up every string in a dictionary first. This should be mentioned clearly in the package

Bug#620345: libcap2-bin: setpcaps, sucap and execcap are missing

Package: libcap2-bin Version: 1:2.19-3 Severity: important libcap2-bin is lacking several of the utilities mentioned in the package description: execcap, sucap and - this is the problematic part - setpcaps. execcap and sucap were intended as work-arounds for the lack of support for file-based

Bug#620419: psi: causes 10Hz wake-ups even if not connected

Package: psi Version: 0.14-2 Severity: normal powertop shows psi causing 10 wake-ups per second even if it's offline. This wastes energy; in particular it can prevent some hardware from entering lower power states. -- System Information: Debian Release: 6.0.1 APT prefers squeeze-updates APT

Bug#619643: gnome-session: please provided updated package (at least 2.91.4)

Source: gnome-session Version: 2.30.2-3 Severity: wishlist gnome-session 2.91.4 changed the way sessions are defined, allowing non-Gnome desktop environments to use gnome-session as-is (by passing a CLI option). It would be great to have an updated Debian package for gnome-session (with all the

Bug#619126: python-gtk2: please update PyGtk to at least 2.22

Package: python-gtk2 Version: 2.17.0-4 Severity: wishlist The current upstream version of PyGtk is 2.26 (released on 2010-09-27). It would be great to have at least PyGtk 2.22 since it wraps API that was added in GTK 2.20 (most notably GtkOffscreenWindow). -- System Information: Debian Release:

Bug#617350: etckeeper init in existing VCS checkout chokes on file names containing special characters

Package: etckeeper Version: 0.48 Severity: normal Running etckeeper init on a checkout of a git repository managed by etckeeper chokes on file names containing special characters: sascha.silbe@twin:/tmp/sascha_silbe/tmpbox.7kmjhKHhbT$ git clone flatty:git/etc-xo15-sascha Cloning into

Bug#617350: Acknowledgement (etckeeper init in existing VCS checkout chokes on file names containing special characters)

tag 617350 + patch thanks From 39f0dab095eb1a72264920762e2d4d7617b1fd86 Mon Sep 17 00:00:00 2001 From: Sascha Silbe sascha-...@silbe.org Date: Tue, 8 Mar 2011 12:54:40 +0100 Subject: [PATCH] restore-etckeeper: don't choke on special characters Include quotes in the to-be-evaluated string so

Bug#612428: libtext-wikicreole-perl: links without description are converted to double a href=... on output

Package: libtext-wikicreole-perl Version: 0.07-1 Severity: normal If the input contains a link without additional text, creole_parse() will output two nested a tags: === Begin Input === [[https://sascha.silbe.org/]] === End Input === === Begin Output === pa href=https://sascha.silbe.org/;a

Bug#611633: python-xpcom: please provide updated package linking against xulrunner-1.9.2

Package: python-xpcom Version: 1:0.0~hg20100212-5 Severity: wishlist AFAICT xulrunner-1.9.2 fixes a lot of bugs that remain unfixed in xulrunner-1.9.1, so it would be great to get python-xpcom updated to support xulrunner-1.9.2. -- System Information: Debian Release: 6.0 APT prefers testing

Bug#605767: git-email: UTF-8 content in To: causes Subject: to also be RFC2047 encoded

Package: git-email Version: 1:1.7.2.3-2 Severity: minor If To: contains a non-ASCII character, the Subject: header is RFC2047 encoded even if it contains only ASCII characters: === Begin === From: Sascha Silbe sascha-...@silbe.org To: =?UTF-8?q?St=C3=A9phane=20Magnenat?= stephane.magne

  1   2   3   4   >