Bug#1019035: systemd-cron: prerm fails in unbooted systems
Package: systemd-cron Version: 1.15.19-1 Severity: minor Tags: patch The package fails to remove when the system is not booted. I think this is a bug; I think any package should be removable, regardless of the boot status of the system. #v+ Spawning container 220903 on /backup/220903. Press ^] three times within 1s to kill container. root@220903:~# apt remove systemd-cron Reading package lists... Done Building dependency tree... Done Reading state information... Done The following packages will be REMOVED: cron-daemon-common* systemd-cron* 0 upgraded, 0 newly installed, 2 to remove and 1 not upgraded. After this operation, 211 kB disk space will be freed. Do you want to continue? [Y/n] (Reading database ... 58695 files and directories currently installed.) Removing systemd-cron (1.15.19-1) ... System has not been booted with systemd as init system (PID 1). Can't operate. Failed to connect to bus: Host is down dpkg: error processing package systemd-cron (--remove): installed systemd-cron package pre-removal script subprocess returned error exit status 1 dpkg: too many errors, stopping Errors were encountered while processing: systemd-cron Processing was halted because there were too many errors. E: Sub-process /usr/bin/dpkg returned an error code (1) #v- Now as you can see in the above reproduction transcript that when I experienced the error I was nspawn'ed into a copy of the live system. This was to perform some "trial surgery" on the copy without fear of breaking the real one. But please consider also the case of performing maintenance in rescue or emergency mode. I think packages should remain removable even in such instances. The fix is simple: do not call deb-systemd-invoke without first checking for /run/systemd/system. This allows the package to be removed as expected. It is worth noting the automatically added debhelper section has the same check. #v+ --- prerm.orig 2022-07-16 20:42:24.0 +0100 +++ prerm 2022-09-03 10:06:10.254464031 +0100 @@ -1,7 +1,7 @@ #!/bin/sh set -e -if [ "$1" = "remove" ]; then +if [ "$1" = "remove" ] && [ -d /run/systemd/system ]; then deb-systemd-invoke stop cron-update.path fi #v- Thus, the package may now be removed as expected. #v+ root@220903:~# apt remove systemd-cron Reading package lists... Done Building dependency tree... Done Reading state information... Done The following packages will be REMOVED: cron-daemon-common* systemd-cron* 0 upgraded, 0 newly installed, 2 to remove and 1 not upgraded. After this operation, 211 kB disk space will be freed. Do you want to continue? [Y/n] (Reading database ... 58695 files and directories currently installed.) Removing systemd-cron (1.15.19-1) ... Removing cron-daemon-common (3.0pl1-149) ... Processing triggers for man-db (2.10.2-2) ... (Reading database ... 58652 files and directories currently installed.) Purging configuration files for systemd-cron (1.15.19-1) ... Purging configuration files for cron-daemon-common (3.0pl1-149) ... #v- Thank you in advance for your consideration. I hope you will agree this is a bug and apply the (trivial) patch, not send me away with an admonishment that this is user error. :) -- https://rjy.org.uk/ --- prerm.orig 2022-07-16 20:42:24.0 +0100 +++ prerm 2022-09-03 10:06:10.254464031 +0100 @@ -1,7 +1,7 @@ #!/bin/sh set -e -if [ "$1" = "remove" ]; then +if [ "$1" = "remove" ] && [ -d /run/systemd/system ]; then deb-systemd-invoke stop cron-update.path fi
Bug#1007887: xxd autoremoved - please recommend or suggest
James McCoy wrote: >I'm not sure how many people that use Vim actually know or care about >xxd. For those that do, manually installing the package (now that it >has its own) is pretty easy. > >Other than a menu item that uses xxd, there's nothing else in Vim that >directly uses it. > >I guess I can add the Recommends to (neo)vim so there's a common >baseline, and people can uninstall if they don't want it. Fair enough. If that is the case I defer to your better judgement. May I withdraw the request? I see now it is unnecessary. Please do not add any extra unnecessary package relationships. I would feel terribly guilty if you did. Sorry to waste your time. Thank you for your work on vim packaging.
Bug#1007887: xxd autoremoved - please recommend or suggest
Hi, this morning I was surprised to see xxd being removed by the latest
apt upgrade (because it was marked as automatically installed - I have
since marked it as manual).
I agree it should not be a "Depends:" but would it make sense to retain
a "Recommends:" or "Suggests:" relationship between the packages to
avoid surprising users?
[ At least those who have not switched the settings of
APT::AutoRemove::{Recommends,Suggests}Important away from their default
values :) ]
--
https://rjy.org.uk/
Bug#915766: wget: pcre regex support lost
Package: wget Version: 1.19.5-2 wget has lost pcre regex support: % wget --regex-type=pcre wget: --regex-type: Invalid value 'pcre'. This is presumably not intentional. (Running the same command on a Stretch system produces the expected "missing URL" error.) This can also be seen from the output of wget --version which no longer mentions -lpcre in the Link command line, and --help for --regex-type which only lists posix (stretch: posix|pcre). Finally, buildd logs show "checking for PCRE... no" "checking pcre.h usability... no" "checking pcre.h presence... no" "checking for pcre.h... no" during configuration. I think this is likely due to the switch to new libpcre2 in 1.19.5-2. I note in #909084 (a similar report regarding loss of pcre on switching from old pcre3 to new pcre2 in a different package, zsh) it was stated "[...] pcre2 and pcre3 are not exchangable without code changes [...]". -- https://rjy.org.uk/
Bug#908518: sicherboot: avoid uuid-runtime dependency by using kernel directly
Package: sicherboot
Version: 0.1.5
Severity: wishlist
Tags: patch
Hi. Something I've been meaning to suggest for some time: avoid the
dependency on uuid-runtime by using /proc/sys/kernel/random/uuid instead
of uuidgen. Patch attached.
--
https://rjy.org.uk/
>From 9b9eda0672a0a75120c7e3d5fc27bc13bc117337 Mon Sep 17 00:00:00 2001
From: RjY
Date: Mon, 10 Sep 2018 18:21:42 +0100
Subject: [PATCH] Obtain random uuid from /proc/sys, remove uuid-runtime
dependency
Avoid the dependency on uuid-runtime by asking the kernel for a random
UUID via the /proc/sys/kernel/random interface. uuidgen and uuidd are
excessive unless time-based UUIDs are required in abundance, while
sicherboot needs only a single random UUID created at initial setup.
(We may assume the /proc/sys interface always exists, as sicherboot is
less useful on non-Linux kernels due to its hard dependency on systemd.)
---
debian/control | 2 +-
shippable.yml | 2 +-
sicherboot | 2 +-
3 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/debian/control b/debian/control
index dad6226..b572ec2 100644
--- a/debian/control
+++ b/debian/control
@@ -10,7 +10,7 @@ Vcs-Browser: https://github.com/julian-klode/sicherboot
Package: sicherboot
Architecture: all
-Depends: ${misc:Depends}, ${shlibs:Depends}, efitools, binutils, systemd, uuid-runtime
+Depends: ${misc:Depends}, ${shlibs:Depends}, efitools, binutils, systemd
Enhances: dracut, systemd, initramfs-tools
Description: systemd-boot integration with UEFI secure boot support
sicher*boot manages kernels and systemd-boot on a secure boot
diff --git a/shippable.yml b/shippable.yml
index 176f9b6..81aafbe 100644
--- a/shippable.yml
+++ b/shippable.yml
@@ -7,5 +7,5 @@ build:
pull: true
ci:
- touch /etc/kernel/cmdline
-- apt-get install -y -qq efitools binutils systemd fakeroot uuid-runtime
+- apt-get install -y -qq efitools binutils systemd fakeroot
- run-parts -v tests
diff --git a/sicherboot b/sicherboot
index 8b47b4a..7a8617f 100755
--- a/sicherboot
+++ b/sicherboot
@@ -205,7 +205,7 @@ generate_keys() {
chown root:root "${KEY_HOME}"
chmod 700 "${KEY_HOME}"
cd "${KEY_HOME}"
-uuidgen > "${KEY_HOME}/uuid"
+cat /proc/sys/kernel/random/uuid > "${KEY_HOME}/uuid"
_generate_key PK PK
_generate_key KEK PK
_generate_key db KEK
--
2.19.0.rc2
Bug#797067: apt: please restore upgrade auto-autoremove
Package: apt Version: 1.6~alpha5 Severity: wishlist % grep AutomaticRemove /etc/apt/apt.conf APT::Get::AutomaticRemove "true"; % sudo apt upgrade [...] The following packages were automatically installed and are no longer required: [...] Use 'sudo apt autoremove' to remove them. May apt upgrade go back to honoring APT::Get::AutomaticRemove "true"; ? This used to work, but was broken, apparently deliberately, in #855891. Having to ^C upgrade and rerun with --autoremove is... "suboptimal" :).
Bug#883437: rsnapshot dependencies seem wrong
Richard James Salts wrote: >rsnapshot's default is to use /usr/bin/logger to log to syslog, however >there is no dependency on bsdutils. I think that the dependency on >logrotate should be downgraded to recommends and that there should be a >new dependency on bsdutils. bsdutils is "Priority: required" and "Essential: Yes", thus AFAIK no dependency is needed, cf. policy 3.5. I agree logrotate could be moved to Recommends. -- https://rjy.org.uk/
Bug#882740: hostapd installation error
Andrew Shadura wrote: >By 'running' I meant running in terms of systemd, that is the unit is >somehow started already. Why this matters: the user might have edited >the unit file or written one of their own, which potentially uses a >different configuration file from a different location. If we just >mask it, the user may say, hey, why do you mess up with my config, I >want it this way. Fair enough, sorry for misunderstanding. >Please have a look at the latest hostapd package version and let me >know if that works for you. It does. 2:2.6-10 installed fine for me. I removed any "systemctl mask"s and "update-rc.d disable"s on hostapd.service beforehand, and it was masked again as expected by postinst, without error. Thank you very much, and sorry for the trouble. -- https://rjy.org.uk/
Bug#882740: hostapd installation error
Andrew Shadura wrote: >On 28 November 2017 at 10:24, RjY <r...@users.sourceforge.net> wrote: >> I'm still observing this on upgrade from 2:2.4-1.1 to 2:2.6-8. [...] >> As workaround I did "sudo update-rc.d hostapd disable" (Wasn't sure what >> the systemctl equivalent is) then "sudo dpkg --configure hostapd" >> unbroke dpkg. Apparently the call to "systemctl mask hostapd.service" in >> postinst had no effect? (or was not performed) >Well, that was something I haven't thought about carefully. Probably I >need to mask the unit every time when (during upgrades or >reinstallation) we find out /etc/defaults/hostapd is unconfigured >*and* hostapd isn't already running. I think unfortunately it needs to be more than just "hostapd isn't running" - in my case hostapd *was* already running, started via ifupdown: iface netw inet static hostapd /etc/hostapd/hostapd.conf address xxx.xxx.xxx.xxx network xxx.xxx.xxx.xxx netmask xxx.xxx.xxx.xxx broadcast xxx.xxx.xxx.xxx up start-stop-daemon -S -o -p /run/udhcpd.pid -x /bin/busybox -- udhcpd -S /etc/hostapd/udhcpd.conf down start-stop-daemon --signal KILL -K -o -p /run/udhcpd.pid -x /bin/busybox (which is probably a bit strange, but works well enough for me, so I've yet to attempt to port it to a systemd unit) So I guess it needs to be "mask the unit if /etc/defaults/hostapd isn't configured and hostapd isn't already started by a sysv init script or systemd unit". Or just simply "mask the unit if /etc/defaults/hostapd isn't configured" - does it matter if it is already running or not? -- https://rjy.org.uk/
Bug#882740: hostapd installation error
I'm still observing this on upgrade from 2:2.4-1.1 to 2:2.6-8. rjy@vile /tmp % systemctl status hostapd ● hostapd.service - Advanced IEEE 802.11 AP and IEEE 802.1X/WPA/WPA2/EAP Authenticator Loaded: loaded (/lib/systemd/system/hostapd.service; enabled; vendor preset: enabled) Active: failed (Result: exit-code) since Tue 2017-11-28 09:17:50 GMT; 28min ago Tasks: 0 (limit: 4915) CGroup: /system.slice/hostapd.service Nov 28 09:17:50 vile systemd[1]: Starting Advanced IEEE 802.11 AP and IEEE 802.1X/WPA/WPA2/EAP Authenticator... Nov 28 09:17:50 vile hostapd[3728]: Configuration file: Nov 28 09:17:50 vile hostapd[3728]: Could not open configuration file '' for reading. Nov 28 09:17:50 vile hostapd[3728]: Failed to set up interface with Nov 28 09:17:50 vile hostapd[3728]: Failed to initialize interface Nov 28 09:17:50 vile systemd[1]: hostapd.service: Control process exited, code=exited status=1 Nov 28 09:17:50 vile systemd[1]: hostapd.service: Failed with result 'exit-code'. Nov 28 09:17:50 vile systemd[1]: Failed to start Advanced IEEE 802.11 AP and IEEE 802.1X/WPA/WPA2/EAP Authenticator. As workaround I did "sudo update-rc.d hostapd disable" (Wasn't sure what the systemctl equivalent is) then "sudo dpkg --configure hostapd" unbroke dpkg. Apparently the call to "systemctl mask hostapd.service" in postinst had no effect? (or was not performed) -- https://rjy.org.uk/
Bug#882104: debian-goodies: "checkrestart -n" throws TypeError: a bytes-like object is required, not 'str'
Great, thanks for the bug cloning magic + swift fixes :)
Bug#882080: debian-goodies: checkrestart from debian-goodies 0.77 finds no files to restart
Just wanted to add I am also seeing this. Also wanted to note using
checkrestart with -n (I wondered if not using lsof might produce
different results) instead gave a python exception/backtrace.
--
% sudo lsof -n | grep -E 'DEL|deleted' | grep /lib/
mpv2765rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv2765rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
mpv/termi 2765 2769 rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv/termi 2765 2769 rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
mpv/lua2765 2770 rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv/lua2765 2770 rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
mpv/cache 2765 2774 rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv/cache 2765 2774 rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
mpv/demux 2765 2775 rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv/demux 2765 2775 rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
mpv/ao 2765 2778 rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
mpv/ao 2765 2778 rjy DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
inetd 3864 root DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
sshd 3871 root DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
login 4085 root DEL REG8,2
3014687 /lib/x86_64-linux-gnu/security/pam_systemd.so
apt9909rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
systemd 21060rjy txt REG8,2
16657602886230 /lib/systemd/systemd (deleted)
systemd 21060rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
systemd 21060rjy DEL REG8,2
2883762 /lib/systemd/libsystemd-shared-235.so
(sd-pam 21061rjy txt REG8,2
16657602886230 /lib/systemd/systemd (deleted)
(sd-pam 21061rjy DEL REG8,2
3014687 /lib/x86_64-linux-gnu/security/pam_systemd.so
(sd-pam 21061rjy DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
(sd-pam 21061rjy DEL REG8,2
2883762 /lib/systemd/libsystemd-shared-235.so
dbus-daem 29391 messagebus DEL REG8,2
2883608 /lib/x86_64-linux-gnu/libsystemd.so.0.19.1
systemd-l 29392 root txt REG8,2
2192722886245 /lib/systemd/systemd-logind (deleted)
systemd-l 29392 root DEL REG8,2
2883672 /lib/x86_64-linux-gnu/libudev.so.1.6.7
systemd-l 29392 root DEL REG8,2
2883762 /lib/systemd/libsystemd-shared-235.so
% sudo checkrestart
Found 0 processes using old versions of upgraded files
% sudo checkrestart -n
Traceback (most recent call last):
File "/usr/sbin/checkrestart", line 798, in
main()
File "/usr/sbin/checkrestart", line 157, in main
toRestart = procfilescheck(blacklist = blacklist, excludepidlist =
excludepidlist)
File "/usr/sbin/checkrestart", line 435, in procfilescheck
data = re.split('\s+', output.strip('\n'), 3)
TypeError: a bytes-like object is required, not 'str'
--
https://rjy.org.uk/
Bug#872368: gpgme: please adjust libgpgme11 dependency on gnupg package
RjY wrote: >It seems in recent unstable the gnupg package has turned into a >metapackage pulling in the whole gpg suite. Thus the dependency chain > mutt -> libgpgme11 -> gnupg -> [lots of stuff] >is pulling in a lot of stuff I don't need. As a test I created an empty package to subvert the dependency and see if anything breaks. The control file was thus: Package: no-full-gnupg-install Maintainer: RjY <r...@users.sourceforge.net> Architecture: all Version: 1 Depends: gpg (>= 2.1.23-2) Recommends: gpg-agent (>= 2.1.23-2) Provides: gnupg (= 2.1.23-2) Description: No full GNUPG install This package "provides" gnupg, to avoid libgpgme11 pulling in the full gnupg suite, until 872368 is fixed. This allowed installing only gpg and gpg-agent. So far, things still seem to be working as expected: mutt appears fine, I can still use gpg to encrypt/decrypt backups, and so on. It would still be good to get this fixed in src:gpgme, though! Thanks in advance. -- https://rjy.org.uk/
Bug#872368: gpgme: please adjust libgpgme11 dependency on gnupg package
Package: src:gpgme1.0 Version: 1.8.0-3 It seems in recent unstable the gnupg package has turned into a metapackage pulling in the whole gpg suite. Thus the dependency chain mutt -> libgpgme11 -> gnupg -> [lots of stuff] is pulling in a lot of stuff I don't need. I use mutt and don't want to uninstall it, so could the dependency be reduced? It seems splitting up the gnupg package into smaller pieces is pointless if commonly used applications force the whole lot to be pulled in. Many thanks in advance for your consideration. -- https://rjy.org.uk/
Bug#860763: imagemagick: /etc/imagemagick-6/policy.xml useless limits settings
Oh so this explains why IM started refusing to create/handle images more than a few thousand pixels tall despite having gigabytes of free memory. Yes, please remove the limits policy, it is annoying, especially when you don't know about it, and stuff that used to work, stops working, with no obvious explanation. Or at least use debconf, to ask root if he wants the policy.xml on package installation, as he will know whether the machine is going to be a server handling remote requests, or merely a single-user desktop. Thanks in advance! -- https://rjy.org.uk/
Bug#868890: sicherboot: tries to access nonexistent files, fails upgrade
Package: sicherboot
Version: 0.1.3
Severity: important
Tags: patch
I installed sicherboot after its recommendation in #826045 for
systemd-boot integration (instead of a tediously manually updated
syslinux-efi configuration file). As I don't use secure boot, I answered
'n' to the question about enrolling keys, as hinted. This seemed to
produce a bunch of warnings, but on inspection the contents of /boot/efi
appeared to be correct, and rebooting worked as expected.
However now I find apt upgrade is failing and leaving things
unconfigured:
Setting up linux-image-4.11.0-2-amd64 (4.11.11-1) ...
/etc/kernel/postinst.d/dracut:
sicherboot: Installing 4.11.0-2-amd64 to ESP
objcopy: cannot open: /boot//initrd.img-4.11.0-2-amd64: No such file or
directory
run-parts: /etc/kernel/postinst.d/dracut exited with return code 1
I think what's happened here is that the new kernel (with ABI bump, so a new
filename) has been unpacked, then sicherboot install-kernel has been called
(via /etc/kernel/postinst.d/dracut) before any call to update-initramfs -u to
make initrd.img-4.11.0-2-amd64. Thus the hook exits 1 and apt/dpkg dies.
I increased the bug severity to "important" because of this breakage.
To try to fix things I ran update-initramfs manually, but sicherboot doesn't
seem to like not having keys enrolled:
% sudo update-initramfs -u
update-initramfs: Generating /boot/initrd.img-4.11.0-1-amd64
warning: data remaining[23685632 vs 23693957]: gaps between PE/COFF sections?
warning: data remaining[23685632 vs 23693960]: gaps between PE/COFF sections?
Can't load key from file '/etc/sicherboot/keys/db.key'
140154209566080:error:02001002:system library:fopen:No such file or
directory:../crypto/bio/bss_file.c:74:fopen('/etc/sicherboot/keys/db.key','r')
140154209566080:error:2006D080:BIO routines:BIO_new_file:no such
file:../crypto/bio/bss_file.c:81:
run-parts: /etc/initramfs/post-update.d//zz-sicherboot exited with return code 1
As I had answered 'n' to the sicherboot setup question about enrolling
keys, no keys were generated. This isn't handled by sign_image, which
passes nonexistent files as parameters to sbsign, causing it to fail.
The following patch fixes this and allows update-initramfs to complete
successfully.
--- /usr/sbin/sicherboot.distrib2017-07-04 10:52:58.0 +0100
+++ /usr/sbin/sicherboot2017-07-19 14:18:24.830787198 +0100
@@ -100,6 +100,11 @@
exit 1
fi
+if [ ! \( -e "${KEY_HOME}/db.key" -a -e "${KEY_HOME}/db.crt" \) ]; then
+echo "No db.key, skipping sign_image."
+return 0
+fi
+
local image="$1"
local out="$2"
Hope this makes sense. Thank you for writing sicherboot.
--
https://rjy.org.uk/
Bug#805322: nginx: please default to keeping logs for less time
Matt Taggart wrote: >Currently the logrotate config file that is shipped with nginx specifies: > > daily > rotate 52 > >so it's keeping 52 days of logfiles (I wonder if daily used to be weekly >and it was keeping a year? that would make more sense). Strange to see this again. I thought the same thing, reported it, and it was apparently fixed, three years ago. I wonder what happened. Oh well. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=696440 -- http://rjy.org.uk/
Bug#809721: iceweasel: cookie permission overrides no longer apply to webappsstore?
Package: iceweasel Version: 43.0.2-1 Severity: normal Summary: please restore the ability to allow whitelisted sites to write to webappsstore.sqlite without having to accept all cookies permanently. In Preferences/privacy I have cookies set to accept, but keep only until Iceweasel closes. Then, I have a few sites which have overrides to allow cookies to remain until they expire. (That is, in PageInfo/permissions, the setting is changed from the default "Allow for session" to "Allow"). In previous Iceweasel versions in Sid (38esr etc.) the permissions override for these sites also applied to webappsstore (localStorage). The whitelisted sites were able to store data in webappsstore.sqlite, and their settings were preserved across Iceweasel restarts. Since Iceweasel 43 went into sid, this seems to no longer be the case: even sites which have an "allow cookies" override are seemingly not able to update their data stored in webappsstore.sqlite. Changes are only kept per-session, and discarded on quit. Now, for sites to be able to update their data in webappsstore.sqlite, I seem to have to go back into Preferences and set all cookies to remain until they expire, which is undesirable as it means any and all cookies and webappsstore data, regardless of origin, hangs around forever. Can this be fixed, or is it an intentional change? -- http://rjy.org.uk/
Bug#798265: Please apply upstream mouse event duplication patch
Manuel A. Fernandez Montecelo wrote: [...] >So now we try to be more cautious and to avoid cherry-picking fixes >(unless they are critical fixes), unless they were going to be released >imminently Upstream first. I quite understand. No problem, I can live with continuing to patch my local copy for now. >but we don't have this guarantee anymore. Upstream did not >release a simple 2.0.4 bugfix release (or anything else) in more than 19 >months, while lots and lots of changes pile up every month without any >clear indication of what's considered to be stable and what is not. > >We already tried to encourage upstream in different ways to try to get >regular bugfix releases, but we failed so far. If you think that you >can influence them, please go ahead :) Heh, they wouldn't know me from Adam. :) Best I can tell, the SDL2 developers' employer has pulled them off it, to work on a commercial game. There's a lesson in this, somewhere. :( http://permalink.gmane.org/gmane.comp.lib.sdl/71042 http://forums.libsdl.org/viewtopic.php?p=48819#48819 (I apologise for spreading rumours if this turns out not to be the case. It is just a post on a mailing list after all.) -- http://rjy.org.uk/
Bug#803315: debsnap: use JSON::PP
Package: devscripts
Version: 2.15.9
Severity: wishlist
Tags: patch
debsnap seems to work perfectly well with JSON::PP, part of perl core
since 5.14, according to its manual page. So it might be nice to use it
instead of requiring the extra dependency? I think running a bit slower
(by using pure perl JSON implementation instead of optimized C one) is
preferable to not running at all :)
---8<---
--- /usr/bin/debsnap2015-10-06 02:29:12.0 +0100
+++ ./debsnap 2015-10-28 16:54:21.239450648 +
@@ -42,7 +42,7 @@
}
eval {
-require JSON;
+require JSON::PP;
};
if ($@) {
if ($@ =~ m/Can\'t locate JSON/) {
@@ -114,7 +114,7 @@
verbose "Getting json $json_url\n";
my $content = LWP::Simple::get($json_url);
return unless defined $content;
-my $json = JSON->new();
+my $json = JSON::PP->new();
# these are some nice json options to relax restrictions a bit:
my $json_text = $json->allow_nonref->utf8->relaxed->decode($content);
--->8---
--
http://rjy.org.uk/
Bug#798265: Please apply upstream mouse event duplication patch
Source: libsdl2
Version: 2.0.2+dfsg1-7
Severity: wishlist
Tags: patch fixed-upstream
Please apply this patch, cherry-picked from unreleased 2.0.4 upstream:
https://hg.libsdl.org/SDL/rev/957a6ed39ad0
as it fixes some very annoying mouse behaviour in SDL2.
-8<---
diff -r fd60d77139fc -r 957a6ed39ad0 src/video/x11/SDL_x11xinput2.c
--- a/src/video/x11/SDL_x11xinput2.cMon Jun 08 01:52:43 2015 -0400
+++ b/src/video/x11/SDL_x11xinput2.cMon Jun 08 02:58:46 2015 -0400
@@ -136,15 +136,25 @@
case XI_RawMotion: {
const XIRawEvent *rawev = (const XIRawEvent*)cookie->data;
SDL_Mouse *mouse = SDL_GetMouse();
-double relative_cords[2];
+double relative_coords[2];
+static Time prev_time = 0;
+static double prev_rel_coords[2];
if (!mouse->relative_mode || mouse->relative_mode_warp) {
return 0;
}
parse_valuators(rawev->raw_values,rawev->valuators.mask,
-rawev->valuators.mask_len,relative_cords,2);
-
SDL_SendMouseMotion(mouse->focus,mouse->mouseID,1,(int)relative_cords[0],(int)relative_cords[1]);
+rawev->valuators.mask_len,relative_coords,2);
+
+if ((rawev->time == prev_time) && (relative_coords[0] ==
prev_rel_coords[0]) && (relative_coords[1] == prev_rel_coords[1])) {
+return 0; /* duplicate event, drop it. */
+}
+
+
SDL_SendMouseMotion(mouse->focus,mouse->mouseID,1,(int)relative_coords[0],(int)relative_coords[1]);
+prev_rel_coords[0] = relative_coords[0];
+prev_rel_coords[1] = relative_coords[1];
+prev_time = rawev->time;
return 1;
}
break;
->8---
--
http://rjy.org.uk/
Bug#785075: util-linux: /sbin/fstrim contains 2MiB of zero bytes
Package: util-linux Version: 2.26.2-2 Severity: minor Huge increase in the size of /sbin/fstrim on amd64. % dpkg -c util-linux_2.25.2-6_amd64.deb | grep -F /sbin/fstrim -rwxr-xr-x root/root 35432 2015-03-29 23:34 ./sbin/fstrim % dpkg -c util-linux_2.26.2-2_amd64.deb | grep -F /sbin/fstrim -rwxr-xr-x root/root 2132688 2015-05-03 23:30 ./sbin/fstrim There is a 2MiB block of zeros in the middle of the file. % hd /sbin/fstrim | grep -C2 '^\*' [...] 7db0 e0 e3 ff ff 1a 00 00 00 00 00 00 00 00 00 00 00 || 7dc0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 || * 00207dc0 b0 26 40 00 00 00 00 00 90 26 40 00 00 00 00 00 |.@..@.| 00207dd0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 || [...] Maybe another case of #750722? [1] If so, it will probably fix itself next time the package is rebuilt. [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=750722 -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#781647: git: git commit --date ignores timezone, always uses UTC/GMT
I previously wrote: It seems --date always uses UTC instead of the user's timezone. The issue seems a little more complicated than this; I observe that specifying today's date along with the time picks up that we are in British Summer Time, and gets the correct timezone offset. With only a time specified, timezone is incorrectly +: % git ci -m test --date='16:00:00' [master (root-commit) 2c8ea98] test Date: Mon Apr 6 15:00:00 2015 + 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 f With a date and time specified, timezone is correctly +0100: % git ci --amend --date='2015-04-06 16:00:00' [master 534ec11] test Date: Mon Apr 6 16:00:00 2015 +0100 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 f So my assumption the parser would do the right thing with today's date if I only specified the time seems to be the cause of the problem. I hope the reader will agree this is rather counterintuitive behaviour. It seems reasonable to expect that the user who specified only a time meant that time today, and the software would act exactly the same way as it does when today's date is written explicitly. Finally, I wonder if this might be the same as / related to #762585. -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#781647: git: git commit --date ignores timezone, always uses UTC/GMT
Package: git Version: 1:2.1.4-2.1 Severity: normal It seems --date always uses UTC instead of the user's timezone. When trying to specify the author date of a commit, I observe: % git ci --date='08:39:39' -am r_patch.c: FillEmptyPixels: code common to createPatch/Texture [master 5734838] r_patch.c: FillEmptyPixels: code common to createPatch/Texture Date: Wed Apr 1 07:39:39 2015 + 1 file changed, 52 insertions(+), 86 deletions(-) I expected 08:39:39 +0100, but I get 07:39:39 +. Trying to force it also has no effect: % git ci --amend --date='08:39:39 +0100' [master ee62730] r_patch.c: FillEmptyPixels: code common to createPatch/Texture Date: Wed Apr 1 07:39:39 2015 + 1 file changed, 52 insertions(+), 86 deletions(-) (Noticed since the UK recently switched from GMT to BST.) -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#756266: g++-4.9: test summaries compressed twice (and why install them at all..?)
Package: g++-4.9 Version: 4.9.1-3 Severity: minor Many of the files in /usr/share/doc/gcc-4.9-base are compressed by xz then by gz. This seems wrong. (Reporting to g++-4.9 since that package owns the files.) % ls -l /usr/share/doc/gcc-4.9-base/test-summaries total 15316 -rw-r--r-- 1 root root2808 2014-07-27 15:37 boehm-gc.log.xz -rw-r--r-- 1 root root 548 2014-07-27 15:37 boehm-gc.sum.xz -rw-r--r-- 1 root root 5264919 2014-07-27 18:48 gcc.log.xz.gz -rw-r--r-- 1 root root 329118 2014-07-27 18:47 gcc.sum.xz.gz -rw-r--r-- 1 root root 957672 2014-07-27 18:07 gdc.log.xz.gz -rw-r--r-- 1 root root 39180 2014-07-27 18:06 gdc.sum.xz.gz -rw-r--r-- 1 root root 2629827 2014-07-27 18:46 gfortran.log.xz.gz -rw-r--r-- 1 root root 153053 2014-07-27 18:44 gfortran.sum.xz.gz -rw-r--r-- 1 root root 4502728 2014-07-27 18:35 g++.log.xz.gz -rw-r--r-- 1 root root 126478 2014-07-27 18:28 go.log.xz.gz -rw-r--r-- 1 root root 14591 2014-07-27 18:28 go.sum.xz.gz -rw-r--r-- 1 root root 269921 2014-07-27 18:32 g++.sum.xz.gz -rw-r--r-- 1 root root3884 2014-07-27 16:08 libatomic.log.xz -rw-r--r-- 1 root root 692 2014-07-27 16:08 libatomic.sum.xz -rw-r--r-- 1 root root 61696 2014-07-27 15:49 libffi.log.xz.gz -rw-r--r-- 1 root root6547 2014-07-27 15:49 libffi.sum.xz.gz -rw-r--r-- 1 root root 159471 2014-07-27 17:02 libgomp.log.xz.gz -rw-r--r-- 1 root root8763 2014-07-27 17:02 libgomp.sum.xz.gz -rw-r--r-- 1 root root1036 2014-07-27 16:16 libgo.sum.xz -rw-r--r-- 1 root root3552 2014-07-27 15:37 libitm.log.xz -rw-r--r-- 1 root root 660 2014-07-27 15:37 libitm.sum.xz -rw-r--r-- 1 root root 59244 2014-07-27 16:07 libjava.log.xz.gz -rw-r--r-- 1 root root8271 2014-07-27 16:07 libjava.sum.xz.gz -rw-r--r-- 1 root root 731685 2014-07-27 18:24 libstdc++.log.xz.gz -rw-r--r-- 1 root root 40632 2014-07-27 18:22 libstdc++.sum.xz.gz -rw-r--r-- 1 root root 66721 2014-07-27 18:32 obj-c++.log.xz.gz -rw-r--r-- 1 root root 162127 2014-07-27 18:32 objc.log.xz.gz -rw-r--r-- 1 root root6523 2014-07-27 18:32 obj-c++.sum.xz.gz -rw-r--r-- 1 root root 10111 2014-07-27 18:32 objc.sum.xz.gz Also why do all g++ users need to have 15MB of test summaries installed? Perhaps a note could be added to README.Debian to inform us of their purpose. Or preferably split them out into a different package... -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#753934: perl: large inexplicable package size increase on amd64
Package: perl
Version: 5.18.2-5
Severity: minor
Installed-Size for perl on amd64 increased by over 4MB from -4 to -5.
As far as I understand all -5 is doing is adding some extra metadata
fields on s390x, so a 4MB increase seems unintentional at least.
% dpkg -c ./perl_5.18.2-4_amd64.deb | awk '{print $6,$3}' | sort 1
% dpkg -c /var/cache/apt/archives/perl_5.18.2-5_amd64.deb | awk '{print $6,$3}'
| sort 2
% join 1 2 | awk '$2!=$3 { print $3-$2,$0 }' | sort -n | column -t | tail -6
40960./usr/lib/perl/5.18.2/auto/Encode/JP/JP.so 2746480
2787440
40960./usr/lib/perl/5.18.2/auto/Encode/KR/KR.so 2410608
2451568
41774./usr/bin/perlbug 0
41774
57344./usr/lib/perl/5.18.2/auto/Encode/TW/TW.so 2021488
2078832
2097152 ./usr/lib/perl/5.18.2/auto/Time/Piece/Piece.so 22784
2119936
2097216 ./usr/lib/perl/5.18.2/auto/Unicode/Collate/Collate.so 1249232
3346448
This appears to be the same issue as #750722, hence CC'ing. (Sorry, the
bts probably has a better way to specify this...)
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#750722: libsvn1: libsvn_ra_local inexplicably gains 2044KiB of zero bytes
Package: libsvn1
Version: 1.8.9-1
Severity: minor
During a recent package update I noticed libsvn1 grew a lot in
installed-size. Upon investigation it turned out this was:
% for i in libsvn1_1.8.8-2_amd64.deb libsvn1_1.8.9-1_amd64.deb ; do dpkg -c
$i ; done | grep ra_local | grep -v '^l'
-rw-r--r-- root/root 39800 2014-04-01 03:20
./usr/lib/x86_64-linux-gnu/libsvn_ra_local-1.so.1.0.0
-rw-r--r-- root/root 2132856 2014-05-21 12:33
./usr/lib/x86_64-linux-gnu/libsvn_ra_local-1.so.1.0.0
Why would a (relatively simple) module for local file:// url access
suddenly gain 2044KiB in a minor update? The extra space appears to
consist entirely of zero bytes, as well.
As far as I can tell, the module source hasn't changed:
diff -ur subversion-1.8.{8,9}/subversion/libsvn_ra_local
produced no output.
I think there's been some kind of build error somewhere.
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#726163: Some of this could be trimmed
Jelmer Vernooij wrote: This is happening because we moved all samba libraries that didn't have separate packages yet into a single large package. [...] In other words, the following dependencies should be avoidable: libsubunit0 libwind0-heimdal libkdc2-heimdal libhdb9-heimdal python-talloc Jelmer Thank you for the positive response, and for working on this; I noticed libsubunit0 being automatically removed today. -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#726163: libsmbclient: excess dependencies?
Package: libsmbclient Version: 2:4.0.10+dfsg-2 Severity: normal The following NEW packages will be installed: libaio1 libasn1-8-heimdal libgssapi3-heimdal libhcrypto4-heimdal libhdb9-heimdal libheimbase1-heimdal libheimntlm0-heimdal libhx509-5-heimdal libkdc2-heimdal libkrb5-26-heimdal libldb1 libroken18-heimdal libsubunit0 libwind0-heimdal python-talloc samba-libs The following packages will be upgraded: libsmbclient 1 upgraded, 16 newly installed, 0 to remove and 0 not upgraded. Need to get 6,065 kB of archives. After this operation, 19.0 MB of additional disk space will be used. I don't believe I should need all that just to run mplayer. (mplayer is the only thing I have installed that pulls in libsmbclient. Perhaps therefore it is mplayer's dependencies that need adjustment?) -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#712162: tig: please upload 1.1 to unstable (fixes parent blame bug)
Package: tig Version: 1.0-2 Severity: normal Please upload 1.1-1, from experimental, to unstable; it fixes the parent blame in subdir bug[1] which I keep tripping over. [1] https://github.com/jonas/tig/issues/70 -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#699536: git-svn: unnecessary dependency on libwww-perl
Package: git-svn Version: 1:1.7.10.4-2 Severity: minor It seems libwww-perl was only needed for git svnimport, which is long gone. (Original report: #345045) -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#696440: nginx: 52 days oddity in log rotation
Source: nginx
Source-Version: 1.2.6-1
Severity: minor
Tags: patch
Keeping 52 _days_ of server logs by default seems rather unusual.
It seems plausible that someone meant to specify 52 _weeks_.
--
http://rjy.org.uk/
From 6cabff29a58ac608ca4935813dd3d16a6a6083f1 Mon Sep 17 00:00:00 2001
From: RjY r...@users.sourceforge.net
Date: Thu, 20 Dec 2012 19:29:37 +
Subject: [PATCH] logrotate: keep 52 weeks (not 52 days) of logs
---
debian/logrotate |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/debian/logrotate b/debian/logrotate
index 2d9547e..bc122b8 100644
--- a/debian/logrotate
+++ b/debian/logrotate
@@ -1,5 +1,5 @@
/var/log/nginx/*.log {
- daily
+ weekly
missingok
rotate 52
compress
--
1.7.10.4
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#668463: libcairo2: very slow transitions in iceweasel
Package: libcairo2 Version: 1.12.0-2 Severity: important Justification: iceweasel becomes unusably unresponsive on some webpages With libcairo2 in sid (1.12.0-2) webpages rendered in iceweasel that use javascript fade transitions - a fade between one image and another - are rendered like a slideshow instead of a smooth animation, and the browser becomes completely unresponsive for the duration (and is in fact quite hard to wrestle back control) They behave correctly - smooth animation, and a responsive browser - by downgrading to the version of libcairo2 in testing/wheezy (1.10.2-7). So it seems like a regression in cairo 1.12, hence reporting it here, although for all I know it might be a problem in how iceweasel uses cairo that is only exposed by the new version. Examples of pages with such transitions are www.penny-arcade.com or www.mediafire.com. For what it is worth I am using nouveau as a graphics driver, with no opengl extensions (libgl1-mesa-dri is not installed) -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: amd64 (x86_64) Kernel: Linux 3.2.0-2-amd64 (SMP w/1 CPU core) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages libcairo2 depends on: ii libc6 2.13-27 ii libfontconfig1 2.8.0-3.1 ii libfreetype6 2.4.9-1 ii libpixman-1-0 0.24.4-1 ii libpng12-0 1.2.49-1 ii libx11-6 2:1.4.4-4 ii libxcb-render0 1.8.1-1 ii libxcb-shm01.8.1-1 ii libxcb11.8.1-1 ii libxrender11:0.9.6-2 ii multiarch-support 2.13-27 ii zlib1g 1:1.2.6.dfsg-2 libcairo2 recommends no packages. libcairo2 suggests no packages. -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#668306: rs: immediate segfault, apparently caused by hardening options (-pie)
Package: rs Version: 20120325-1 Severity: serious rjy@vile ~ % seq 1 9 | rs 3 3 zsh: doneseq 1 9 | zsh: segmentation fault rs 3 3 This happens regardless of input. It appears to be caused by -pie in the parameters passed to cc rjy@vile /tmp/rs/rs-20120325 % cc -D_FORTIFY_SOURCE=2 -DUSE_LIBBSD -g -O2 -fPIE -fstack-protector --param=ssp-buffer-size=4 -Wformat -Wformat-security -Werror=format-security -Wall -Wextra -Wformat -flto=jobserver -fPIE -pie -Wl,-z,relro -Wl,-z,now -Wl,--as-needed -o rs rs.c -lbsd In file included from /usr/include/bsd/bsd.h:41:0, from rs.c:47: /usr/include/bsd/getopt.h:34:2: warning: #warning Deprecated header, use bsd/unistd.h or unistd.h with libbsd-overlay.pc instead. [-Wcpp] rjy@vile /tmp/rs/rs-20120325 % seq 1 9 | ./rs 3 3 zsh: doneseq 1 9 | zsh: segmentation fault ./rs 3 3 After removing -pie (from between the second -fPIE and -Wl,-z,relro) rjy@vile /tmp/rs/rs-20120325 % cc -D_FORTIFY_SOURCE=2 -DUSE_LIBBSD -g -O2 -fPIE -fstack-protector --param=ssp-buffer-size=4 -Wformat -Wformat-security -Werror=format-security -Wall -Wextra -Wformat -flto=jobserver -fPIE -Wl,-z,relro -Wl,-z,now -Wl,--as-needed -o rs rs.c -lbsd In file included from /usr/include/bsd/bsd.h:41:0, from rs.c:47: /usr/include/bsd/getopt.h:34:2: warning: #warning Deprecated header, use bsd/unistd.h or unistd.h with libbsd-overlay.pc instead. [-Wcpp] rjy@vile /tmp/rs/rs-20120325 % seq 1 9 | ./rs 3 3 1 2 3 4 5 6 7 8 9 Sorry if this is hard to read, the command lines are so long... Build command was copied from debuild output, which was rjy@vile /tmp/rs/rs-20120325 % debuild -us -uc dpkg-buildpackage -rfakeroot -D -us -uc dpkg-buildpackage: source package rs dpkg-buildpackage: source version 20120325-1 dpkg-buildpackage: source changed by Thorsten Glaser t...@mirbsd.de dpkg-source --before-build rs-20120325 dpkg-buildpackage: host architecture amd64 fakeroot debian/rules clean dh_testdir rm -f rs dh_clean dpkg-source -b rs-20120325 dpkg-source: info: using source format `1.0' dpkg-source: info: building rs using existing rs_20120325.orig.tar.gz dpkg-source: info: building rs in rs_20120325-1.diff.gz dpkg-source: info: building rs in rs_20120325-1.dsc debian/rules build dh_testdir rm -f rs for opts in '-flto=jobserver' '-fwhole-program --combine' ''; do \ set -x; \ cc -D_FORTIFY_SOURCE=2 -DUSE_LIBBSD -g -O2 -fPIE -fstack-protector --param=ssp-buffer-size=4 -Wformat -Wformat-security -Werror=format-security -Wall -Wextra -Wformat $opts -fPIE -pie -Wl,-z,relro -Wl,-z,now -Wl,--as-needed -o rs \ rs.c -lbsd; \ test -x rs exit 0; \ done; echo 2 Compiling failed.; exit 1 + cc -D_FORTIFY_SOURCE=2 -DUSE_LIBBSD -g -O2 -fPIE -fstack-protector --param=ssp-buffer-size=4 -Wformat -Wformat-security -Werror=format-security -Wall -Wextra -Wformat -flto=jobserver -fPIE -pie -Wl,-z,relro -Wl,-z,now -Wl,--as-needed -o rs rs.c -lbsd In file included from /usr/include/bsd/bsd.h:41:0, from rs.c:47: /usr/include/bsd/getopt.h:34:2: warning: #warning Deprecated header, use bsd/unistd.h or unistd.h with libbsd-overlay.pc instead. [-Wcpp] + test -x rs + exit 0 fakeroot debian/rules binary dh_testdir dh_testroot if test -x $(which dh_prep); then dh_prep; else dh_clean -k; fi dh_installchangelogs dh_installdocs dh_install dh_installman dh_link dh_strip dh_compress dh_fixperms dh_installdeb dh_shlibdeps dpkg-shlibdeps: warning: package could avoid a useless dependency if debian/rs/usr/bin/rs was not linked against libgcc_s.so.1 (it uses none of the library's symbols). dh_gencontrol dh_md5sums dh_builddeb dpkg-deb: building package `rs' in `../rs_20120325-1_amd64.deb'. dpkg-genchanges ../rs_20120325-1_amd64.changes dpkg-genchanges: including full source code in upload dpkg-source --after-build rs-20120325 dpkg-buildpackage: full upload (original source is included) rjy@vile /tmp/rs/rs-20120325 % seq 1 9 | ./rs 3 3 zsh: doneseq 1 9 | zsh: segmentation fault ./rs 3 3 -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: amd64 (x86_64) Kernel: Linux 3.2.0-2-amd64 (SMP w/1 CPU core) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages rs depends on: ii libbsd0 0.3.0-2 ii libc62.13-27 ii libgcc1 1:4.7.0-3 rs recommends no packages. rs suggests no packages. -- no debconf information -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#662946: equivs: does not strip subdirectories from files to be installed
Package: equivs
Version: 2.0.9
Severity: normal
Tags: patch
For example, if I have a control file with
Files: src/some_program /bin
equivs installs /bin/src/some_program instead of /bin/some_program.
Apologies if this is intentional behaviour, but it was unexpected, not
to mention quite inconvenient. :)
Here is a patch which uses basename to fix the install path:
--- /usr/bin/equivs-build 2011-09-30 07:22:24.0 +0100
+++ /tmp/equivs-build 2012-03-07 13:27:50.315122421 +
@@ -67,7 +67,7 @@
for (split \n, $control{'Files'} || ) {
die Cannot parse Files line: '$_'\n
unless m:^\s*(\S+)\s+(\S+)/?\s*$:;
-$install_files{$2/$1} = $1;
+$install_files{$2/.basename($1)} = $1;
}
my %create_files = ();
for (@{$control{'File'} || []}) {
Thanks in advance for your consideration.
-- System Information:
Debian Release: wheezy/sid
APT prefers unstable
APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)
Kernel: Linux 3.2.0-2-amd64 (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages equivs depends on:
ii debhelper 9.20120115
ii dpkg-dev 1.16.1.2
ii fakeroot 1.18.2-1
ii make 3.81-8.1
ii perl 5.14.2-9
equivs recommends no packages.
equivs suggests no packages.
-- no debconf information
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#650029: rs: -z accumulates column width
Package: rs
Version: 20101029-1
Tags: patch
When using -z to compress column width, the loop does not reset the
maximum width for each column, but instead accumulates the maximum width
of all preceeding columns.
For example, the following input
1
a wide column
2
3
4
a
another wide column
b
c
d
passed into rs -ze 0 5 gives
1 a wide column234
a another wide column bcd
with the third, fourth, and fifth columns having inherited the maximum
width of the wider second column. I would have expected them to have the
same width as the narrow first column.
With the patch below the output is what I would expect:
1 a wide column2 3 4
a another wide column b c d
--- rs-20101029.orig/rs.c
+++ rs-20101029/rs.c
@@ -290,6 +290,7 @@
if (flags SQUEEZE) {
if (flags TRANSPOSE)
for (ep = elem, i = 0; i ocols; i++) {
+ max = 0;
for (j = 0; j orows; j++)
if ((n = strlen(*ep++)) max)
max = n;
@@ -297,6 +298,7 @@
}
else
for (ep = elem, i = 0; i ocols; i++) {
+ max = 0;
for (j = i; j nelem; j += ocols)
if ((n = strlen(ep[j])) max)
max = n;
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#638276: dh-make: please allow -p to specify 0 as a version number
Package: dh-make Version: 0.59 Severity: wishlist Tags: patch Please allow the forced package name option to specify 0 as a version, e.g. -p mypackage_0. The test currently looks for anything that Perl considers true; the patch below instead tests for the empty string. *** dh_make.patch --- /usr/bin/dh_make2011-06-19 01:27:43.0 +0100 +++ ./dh_make 2011-08-18 08:52:53.457014911 +0100 @@ -337,7 +337,7 @@ $main::forced_package_name = $1; $forced_package_version = $2; } - if ( ($forced_package_version) + if ( ($forced_package_version ne ) || ( ($main::forced_package_name) ($pwd =~ /.*\/($main::forced_package_name)-([0-9][0-9a-zA-Z+.~-]*)$/) -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 3.0.0-1-686-pae (SMP w/1 CPU core) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages dh-make depends on: ii debhelper 8.9.4 helper programs for debian/rules ii dpkg-dev 1.16.0.3 Debian package development tools ii make 3.81-8.1 An utility for Directing compilati ii perl 5.12.4-4 Larry Wall's Practical Extraction dh-make recommends no packages. Versions of packages dh-make suggests: ii build-essential 11.5 Informational list of build-essent -- no debconf information -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#621693: case where error line ends with , is not handled
reopen 621693 thanks Sorry, I don't think this is quite fixed yet. An error line such as In file included from ./p_pspr.h:32:0, that is, one ending with a comma, not a colon, results in vim thinking there is an error 0, in the file In file included from ./p_pspr.h. I've attached the error file that still shows the problem. -- http://rjy.org.uk/ make all-recursive make[1]: Entering directory `/home/rjy/src/rboom' Making all in doc make[2]: Entering directory `/home/rjy/src/rboom/doc' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/home/rjy/src/rboom/doc' Making all in data make[2]: Entering directory `/home/rjy/src/rboom/data' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/home/rjy/src/rboom/data' Making all in src make[2]: Entering directory `/home/rjy/src/rboom/src' CC SDL/i_video.o In file included from ./p_pspr.h:32:0, from ./d_player.h:26, from ./doomstat.h:25, from SDL/i_video.c:28: ./info.h:1240:3: error: unknown type name ‘actionf_t’ In file included from ./d_player.h:26:0, from ./doomstat.h:25, from SDL/i_video.c:28: ./p_pspr.h:75:15: error: unknown type name ‘mobj_t’ ./p_pspr.h:76:20: error: unknown type name ‘mobj_t’ ./p_pspr.h:77:14: error: unknown type name ‘mobj_t’ ./p_pspr.h:78:14: error: unknown type name ‘mobj_t’ ./p_pspr.h:79:14: error: unknown type name ‘mobj_t’ ./p_pspr.h:80:15: error: unknown type name ‘mobj_t’ ./p_pspr.h:81:19: error: unknown type name ‘mobj_t’ ./p_pspr.h:82:15: error: unknown type name ‘mobj_t’ ./p_pspr.h:83:20: error: unknown type name ‘mobj_t’ ./p_pspr.h:84:15: error: unknown type name ‘mobj_t’ ./p_pspr.h:85:21: error: unknown type name ‘mobj_t’ ./p_pspr.h:86:20: error: unknown type name ‘mobj_t’ ./p_pspr.h:87:21: error: unknown type name ‘mobj_t’ ./p_pspr.h:88:21: error: unknown type name ‘mobj_t’ ./p_pspr.h:89:22: error: unknown type name ‘mobj_t’ ./p_pspr.h:90:17: error: unknown type name ‘mobj_t’ ./p_pspr.h:91:17: error: unknown type name ‘mobj_t’ ./p_pspr.h:92:20: error: unknown type name ‘mobj_t’ ./p_pspr.h:93:12: error: unknown type name ‘mobj_t’ ./p_pspr.h:94:19: error: unknown type name ‘mobj_t’ ./p_pspr.h:95:17: error: unknown type name ‘mobj_t’ ./p_pspr.h:96:16: error: unknown type name ‘mobj_t’ ./p_pspr.h:97:17: error: unknown type name ‘mobj_t’ make[2]: *** [SDL/i_video.o] Error 1 make[2]: Leaving directory `/home/rjy/src/rboom/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/rjy/src/rboom' make: *** [all] Error 2
Bug#636939: toilet: --width is not honoured
Package: toilet
Version: 0.2-1
Severity: normal
Perhaps I have misunderstood how it is meant to work, but toilet does not
appear to honour the -w/--width option. To illustrate:
$ toilet -w 40 12345678901234567890 | awk '{print length($0),$0 }'
77
77 mmm mm mmmm mm mmm
77# ## m# # m# ## # m m m #
77#m mmm # # mm #m#mm #m m# # m # #
77# m # #mmm#m # ## m # # # ## #
77 mm#mm m# mmm# # mmm# #mm# m#mmm mmm #mm# mm#mm
77
77
77
77 mm mmmm mm
77 ## m# # m# ## # m m m
77 m mmm # # mm #m#mm #m m# # m #
77m # #mmm#m # ## m # # # ##
77 m# mmm# # mmm# #mm# m#mmm mmm #mm#
77
77
$ toilet -w 1200 12345678901234567890 | awk '{print length($0),$0 }'
77
77 mmm mm mmmm mm mmm
77# ## m# # m# ## # m m m #
77#m mmm # # mm #m#mm #m m# # m # #
77# m # #mmm#m # ## m # # # ## #
77 mm#mm m# mmm# # mmm# #mm# m#mmm mmm #mm# mm#mm
77
77
77
77 mm mmmm mm
77 ## m# # m# ## # m m m
77 m mmm # # mm #m#mm #m m# # m #
77m # #mmm#m # ## m # # # ##
77 m# mmm# # mmm# #mm# m#mmm mmm #mm#
77
77
I would have expected the first to be split halfway through, and the second not
to be split at all, but both appear split at the default width of 80 characters.
-- System Information:
Debian Release: wheezy/sid
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 3.0.0-1-686-pae (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages toilet depends on:
ii libc6 2.13-13 Embedded GNU C Library: Shared lib
ii libcaca0 0.99.beta17-2 colour ASCII art library
ii toilet-fonts 0.2-1 collection of TOIlet fonts
toilet recommends no packages.
Versions of packages toilet suggests:
pn figletnone (no description available)
-- no debconf information
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#634132: dehydra: misleading package description, implies it can be used with C
Package: dehydra
Version: 0.9.hg20110609-2
Severity: minor
dehydra apparently only operates on C++ codebases. I tried using it with
a project in plain C that doesn't compile as C++, and got no useful output.
I'm not complaining that it doesn't work for C, I mean the program was
obviously developed to help with large C++ codebases, firefox in
particular. But I do think I was misled by the package description :-)
Here is a minimal illustration in the form of a terminal session.
$ cat dumptypes.js
function process_type(t)
{
print(Type found: + t.name + location: + t.loc);
}
$ cat dumptypes.c
typedef int fixed_t;
$ gcc -fplugin=gcc_dehydra -fplugin-arg-gcc_dehydra-script=dumptypes.js -o
/dev/null -c dumptypes.c
$ cp dumptypes.c dumptypes.cc
$ gcc -fplugin=gcc_dehydra -fplugin-arg-gcc_dehydra-script=dumptypes.js -o
/dev/null -c dumptypes.cc
Type found: fixed_t location: dumptypes.cc:1:13
-- System Information:
Debian Release: wheezy/sid
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.38-2-686 (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages dehydra depends on:
ii gcc-4.6 4.6.1-4GNU C compiler
ii libc6 2.13-10Embedded GNU C Library: Shared lib
ii libgcc1 1:4.6.1-4 GCC support library
ii libmozjs5d5.0-3 The Mozilla SpiderMonkey JavaScrip
ii libstdc++64.6.1-4GNU Standard C++ Library v3
Versions of packages dehydra recommends:
ii g++-4.6 4.6.1-4GNU C++ compiler
dehydra suggests no packages.
-- no debconf information
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#633945: dmalloc.h breaks with -O2 -D_GNU_SOURCE=1 (strndup macro expansion)
Package: libdmalloc-dev
Version: 5.5.1-1.1
Severity: normal
Version of gcc: gcc-4.6 4.6.1-3
Version of libc6-dev: 2.13-10
When using any level of optimization in a program that enables GNU libc
extensions (defines _GNU_SOURCE), gcc turns strndup into a macro, which
confuses line 484 of dmalloc.h.
To illustrate, a test case, a.c
#include stdlib.h
#include stdio.h
#include string.h
#include dmalloc.h
int main(void)
{
static const char *s = 1234567;
char *p = strndup(s,4);
printf(%s\n, p);
return 0;
}
Compilation of a.c
gcc a.c -ldmalloc ./a.out # prints 1234
gcc -O2 a.c -ldmalloc ./a.out # prints 1234
gcc -D_GNU_SOURCE=1 a.c -ldmalloc ./a.out # prints 1234
gcc -D_GNU_SOURCE=1 -O2 a.c -ldmalloc ./a.out # error
Looking at the output of gcc -E, line 484 turns from
char*strndup(const char *string, const DMALLOC_SIZE len);
into (split across several lines)
char *(__extension__ (__builtin_constant_p (const char *string)
((size_t)(const void *)((const char *string) + 1) - (size_t)(const void
*)(const char *string) == 1) ? (((__const char *) (const char
*string))[0] == '\0' ? (char *) calloc ((size_t) 1, (size_t) 1) : ({
size_t __len = strlen (const char *string) + 1; size_t __n = (const
size_t len); char *__retval; if (__n __len) __len = __n + 1; __retval
= (char *) malloc (__len); if (__retval != ((void *)0)) { __retval[__len
- 1] = '\0'; __retval = (char *) memcpy (__retval, const char *string,
__len - 1); } __retval; })) : __strndup (const char *string, const
size_t len)));
which fails to compile with the error:
/usr/include/dmalloc.h:484:7: error: expected identifier or ‘(’ before
‘__extension__’
I'm not sure if this is a bug in dmalloc, or user error; maybe I did
something wrong, or perhaps dmalloc isn't designed to be used with
programs using both GCC optimization and GNU libc extensions. Either
way, a workaround is #undef strndup before #include dmalloc.h.
Script with the above test case attached.
--
http://rjy.org.uk/
a.sh
Description: Bourne shell script
Bug#629080: extlinux: can't find kernel image /boot/extlinux/memdisk with separate /boot partition
Package: extlinux Version: 2:4.04+dfsg-2 Severity: normal I downloaded an ISO, put it in /boot, and ran extlinux-update. It generated the following memdisk.cfg: label m0 menu label Memdisk, image grml-medium_2011.05.iso linux /boot/extlinux/memdisk initrd /boot/grml-medium_2011.05.iso append iso Trying to boot from this failed with an error message about not being able to find the kernel image. Comparing with the generated linux.cfg made it obvious what was going on: there were no filenames prefixed with /boot in linux.cfg, as I have a separate /boot partition. I removed the /boot prefix from the filenames in memdisk.cfg allowed the ISO to be booted. So, I think extlinux-update needs to handle the case of a separate /boot partition for memdisks as well as linux kernels. -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable'), (1, 'experimental') Architecture: i386 (i686) Kernel: Linux 2.6.39-1-486 Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages extlinux depends on: ii debconf [debconf-2.0] 1.5.39 Debian configuration management sy ii libc6 2.13-4 Embedded GNU C Library: Shared lib Versions of packages extlinux recommends: pn syslinux-common none (no description available) pn syslinux-themes-debiannone (no description available) extlinux suggests no packages. -- debconf information: extlinux/install: false -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#617644: pdnsd: fails to start, broken init script
Package: pdnsd
Version: 1.2.8-par-1
Severity: grave
Justification: renders package unusable
/etc/init.d/pdnsd has several problems that prevent pdnsd from starting.
Firstly line 28
test -f . /etc/default/rcS || exit 0
This just produces an error, '.: unexpected operator'
(I just deleted the line. I don't think /etc/default/rcS is needed at
all, there doesn't appear to be any variable from it referenced, but I
might have missed something.)
Secondly, line ~20
test -f /etc/default/$NAME
/etc/default/pdnsd is tested for, but never actually sourced. So
START_DAEMON is never set to 'yes' so the script thinks the daemon is
disabled, and never runs it.
Fixing these two at least allows the daemon to start. I also noticed
that $AUTO_MODE handling is apparently broken (although I don't use one)
if test -z $AUTO_MODE test -f /usr/share/pdnsd/pdnsd-$AUTO_MODE.conf
I think the first test should be -n. -z means is a zero-length string,
-n means is _not_ a zero-length string.
-- System Information:
Debian Release: wheezy/sid
APT prefers unstable
APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 2.6.37-2-686 (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages pdnsd depends on:
ii adduser 3.112+nmu2 add and remove users and groups
ii debconf [debconf-2.0] 1.5.38 Debian configuration management sy
ii libc6 2.11.2-11 Embedded GNU C Library: Shared lib
Versions of packages pdnsd recommends:
pn resolvconfnone (no description available)
pdnsd suggests no packages.
-- Configuration Files:
/etc/default/pdnsd changed:
START_DAEMON=yes
AUTO_MODE=
START_OPTIONS=
/etc/init.d/pdnsd changed:
NAME=pdnsd
DESC=proxy DNS server
DAEMON=/usr/sbin/pdnsd
PIDFILE=/var/run/pdnsd.pid
CACHE=/var/cache/pdnsd/pdnsd.cache
test -x $DAEMON || exit 0
test -f /etc/default/$NAME || exit 0
. /etc/default/$NAME
if test -n $AUTO_MODE test -f /usr/share/pdnsd/pdnsd-$AUTO_MODE.conf
then
START_OPTIONS=${START_OPTIONS} -c /usr/share/pdnsd/pdnsd-$AUTO_MODE.conf
fi
. /lib/lsb/init-functions
is_yes() {
case $1 in
[Yy][Ee][Ss]|[Tt][Rr][Uu][Ee]|1) return 0;;
*) return 1;
esac
}
log_end_msg2 () {
log_end_msg $@
test $1 -eq 0 || exit 1
}
gen_cache()
{
if ! test -f $CACHE; then
mkdir -p `dirname $CACHE`
dd if=/dev/zero of=$CACHE bs=1 count=4 2 /dev/null
chown -R pdnsd.proxy /var/cache/pdnsd
fi
}
check_pid()
{
if test -f $PIDFILE; then
log_warning_msg pid file is exist in $PIDFILE, stop $pdnsd it
or restart $pdnsd
exit 1
fi
}
start_resolvconf()
{
test -x /sbin/resolvconf || return
for f in `seq 1 60`; do
sleep 0.1
if pdnsd-ctl status /dev/null 21; then
break
fi
done
if pdnsd-ctl status | grep -q resolvconf; then
server=$(pdnsd-ctl status|sed -ne '/^Global:$/,/^Server.*:$/s/.*Server
ip.*: \(.*\)$/\1/p')
case $server in
) ;;
0.0.0.0) echo nameserver 127.0.0.1 | /sbin/resolvconf -a
lo.$NAME;;
*) echo nameserver $server | /sbin/resolvconf -a
lo.$NAME;;
esac
fi
}
stop_resolvconf()
{
if [ -x /sbin/resolvconf ] ; then
/sbin/resolvconf -d lo.$NAME
fi
}
pdnsd_start()
{
if is_yes $START_DAEMON; then
check_pid
log_begin_msg Starting $NAME
start-stop-daemon --oknodo --start --quiet --pidfile $PIDFILE \
--exec $DAEMON -- --daemon -p $PIDFILE $START_OPTIONS
log_end_msg2 $?
start_resolvconf
else
log_warning_msg Not starting $NAME (disabled in /etc/default/$NAME)
fi
}
pdnsd_stop()
{
log_begin_msg Stopping $NAME
start-stop-daemon --oknodo --stop --quiet --user pdnsd
--retry=TERM/3/KILL/3 --pidfile $PIDFILE --name $NAME
start-stop-daemon --oknodo --stop --quiet --user pdnsd --retry=0/3/KILL/3
--exec $DAEMON /dev/null
log_end_msg2 $?
rm -f $PIDFILE
stop_resolvconf
}
pdnsd_status()
{
if status_of_proc $DAEMON $NAME; then
/usr/sbin/pdnsd-ctl status
fi
}
case $1 in
start)
gen_cache
pdnsd_start
;;
stop)
pdnsd_stop
;;
status)
pdnsd_status
;;
restart|force-reload)
pdnsd_stop
pdnsd_start
;;
*)
echo Usage: /etc/init.d/$NAME
{start|stop|restart|force-reload|status} 2
exit 1
;;
esac
exit 0
/etc/pdnsd.conf changed:
// Read the pdnsd.conf(5) manpage for an explanation of the options.
/* Note: this file is overriden by automatic config files when
/etc/default/pdnsd AUTO_MODE is set and that
/usr/share/pdnsd/pdnsd-$AUTO_MODE.conf exists
*/
// 2010-08-15: comment out server_ip to listen on all interface,
// uncomment root-servers section to do recursive resolving
global {
Bug#617318: new upstream version 0.17
I noticed on the git list that there is a new upstream version out to fix this problem[1]: Here's an updated tig, which fixes an incompatibility with ncurses 2.8.[2] [1] http://permalink.gmane.org/gmane.comp.version-control.git/168563 [2] I think he means 5.8 :-) the particular patch is this http://repo.or.cz/w/tig.git/commitdiff/ddae52065f967db43f4d9aa07340c950cdb2b433 -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#616523: querybts: browsing fails unless xdg-open is present when /bin/sh is dash
Package: reportbug
Version: 5.0
Severity: normal
Tags: patch
Trying to start a web browser to view bug reports (with e.g. querybts -w
or via 'b') fails with the error sh: xdg-open: not found
The code used in launch_browser (urlutils.py) to detect if xdg-open is
present on the system is a bashism. When /bin/sh is dash, the wrong exit
code is returned, fooling launch_browser into thinking xdg-open is
available when it is not. launch_browser then tries to run the command
anyway, which fails with the above message, and returns without doing
anything else.
This patch seems to fix the problem for me. checkbashisms also reported
'command' with option other than -p, so you might prefer to use
'which'. However dash supports 'command -v' so I will leave it up to you
:-)
diff --git a/reportbug/urlutils.py b/reportbug/urlutils.py
index c4ba5aa..054932d 100644
--- a/reportbug/urlutils.py
+++ b/reportbug/urlutils.py
@@ -163,7 +163,7 @@ def open_url(url, http_proxy=None, timeout=60):
return page
def launch_browser(url):
-if not os.system('command -v xdg-open /dev/null'):
+if not os.system('command -v xdg-open /dev/null 21'):
cmd = 'xdg-open' + commands.mkarg(url)
os.system(cmd)
return
-- Package-specific info:
** Environment settings:
EDITOR=vim
VISUAL=vim
INTERFACE=text
** /home/rjy/.reportbugrc:
reportbug_version 5.0
mode standard
ui text
email r...@users.sourceforge.net
-- System Information:
Debian Release: wheezy/sid
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.37-2-686 (SMP w/1 CPU core)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages reportbug depends on:
ii apt 0.8.11.5 Advanced front-end for dpkg
ii python 2.6.6-3+squeeze5 interactive high-level object-orie
ii python-reportbug5.0 Python modules for interacting wit
reportbug recommends no packages.
Versions of packages reportbug suggests:
pn debconf-utils none (no description available)
pn debsums none (no description available)
pn dlocate none (no description available)
pn emacs22-bin-common | emacs23- none (no description available)
ii file 5.04-5 Determines file type using magic
ii gnupg 1.4.11-3 GNU privacy guard - a free PGP rep
ii postfix [mail-transport-agent 2.8.1-1High-performance mail transport ag
pn python-gtk2 none (no description available)
pn python-gtkspell none (no description available)
pn python-urwid none (no description available)
pn python-vtenone (no description available)
pn xdg-utils none (no description available)
-- no debconf information
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#616319: sn constantly redownloads the same articles, presenting them as new
Package: sn Version: 0.3.8-10 Severity: normal Tags: patch sn gets confused and never updates the .serial file for a newsgroup when a server responds to ARTICLE nnn with 220 0 message-id. snfetch.c:fetch() expects the response to ARTICLE nnn to be 220 nnn message-id and is confused by servers (such as news.virginmedia.com) which say 0 instead of repeating nnn. (This behaviour is rfc-compliant; rfc3977 section 6.2.1.2 states the client MUST always be prepared for zero to be specified) The result of this confusion is that fetch() returns 0 and snfetch bails out early, claiming that it has nothing new to fetch, even if it fetched several new articles. The newsgroup .serial files in the spool are never updated, and articles in later retrieval runs are constantly redownloaded and presented as new as snexpire cleans them up. This patch tries to fix fetch() to use the article number it asked for earlier if 0 was returned by the server. --- sn-0.3.8.orig/snfetch.c +++ sn-0.3.8/snfetch.c @@ -300,6 +300,8 @@ badresponse(ARTICLE); case 220: last = strtoul(args[1], end, 10); + if (last == 0) + last = arts[n - p].serial; readprint(); collected++; break; -- http://rjy.org.uk/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#615902: libsdl1.2: typo in Vcs-Svn field
Source: libsdl1.2 Source-Version: 1.2.14-6.1 Severity: minor Tags: patch libsdl.2 should be libsdl1.2 -- http://rjy.org.uk/ Index: debian/control === --- debian/control (revision 266) +++ debian/control (working copy) @@ -5,7 +5,7 @@ Uploaders: Sam Hocevar (Debian packages) sam+...@zoy.org, Aurelien Jarno aure...@debian.org, Josselin Mouette j...@debian.org, Barry deFreese bdefre...@debian.org Standards-Version: 3.8.4 Build-Depends: dpkg (= 1.13.2), debhelper (= 5.0), quilt, nasm [i386 kfreebsd-i386], libaa1-dev, libx11-dev, libxext-dev, libxt-dev, libxv-dev, x11proto-core-dev, libaudiofile-dev, libesd0-dev, libpulse-dev, libgl1-mesa-dev, libsvga1-dev [amd64 i386], libarts1-dev, libartsc0-dev, libasound2-dev [!kfreebsd-i386 !kfreebsd-amd64 !hurd-i386], libaudio-dev, libcaca-dev, libdirectfb-dev (= 1.0) [!hurd-i386], libusbhid-dev [kfreebsd-i386 kfreebsd-amd64], libglu1-mesa-dev -Vcs-Svn: svn://svn.debian.org/pkg-sdl/unstable/libsdl.2 +Vcs-Svn: svn://svn.debian.org/pkg-sdl/unstable/libsdl1.2 Vcs-Browser: http://svn.debian.org/wsvn/pkg-sdl/unstable/libsdl1.2/ Homepage: http://www.libsdl.org/
Bug#560377: [PATCH] debian/nmap.{postinst,prerm}: install ncat as netcat alternative
tags 560377 + patch
thanks
--
http://rjy.org.uk/
From 4d9038b2297615151b617471c62d40f959783e01 Mon Sep 17 00:00:00 2001
From: RjY r...@users.sourceforge.net
Date: Wed, 23 Feb 2011 17:35:14 +
Subject: [PATCH] debian/nmap.{postinst,prerm}: install ncat as netcat alternative
Fix #560377. Install at priority 40, slightly less than netcat-openbsd
and netcat6. postinst/prerm code more or less copied out of netcat6.
Signed-off-by: RjY r...@users.sourceforge.net
---
debian/nmap.postinst | 14 ++
debian/nmap.prerm|8
2 files changed, 22 insertions(+), 0 deletions(-)
create mode 100644 debian/nmap.postinst
create mode 100644 debian/nmap.prerm
diff --git a/debian/nmap.postinst b/debian/nmap.postinst
new file mode 100644
index 000..9bbb05e
--- /dev/null
+++ b/debian/nmap.postinst
@@ -0,0 +1,14 @@
+#!/bin/sh
+set -e
+
+if [ $1 = configure ]; then
+update-alternatives \
+--install /bin/nc nc /usr/bin/ncat 40 \
+--slave /bin/netcat netcat /usr/bin/ncat \
+--slave /usr/share/man/man1/nc.1.gz nc.1.gz \
+/usr/share/man/man1/ncat.1.gz \
+--slave /usr/share/man/man1/netcat.1.gz netcat.1.gz \
+/usr/share/man/man1/ncat.1.gz
+fi
+
+#DEBHELPER#
diff --git a/debian/nmap.prerm b/debian/nmap.prerm
new file mode 100644
index 000..58a15ee
--- /dev/null
+++ b/debian/nmap.prerm
@@ -0,0 +1,8 @@
+#!/bin/sh
+set -e
+
+if [ $1 = remove ]; then
+update-alternatives --remove nc /usr/bin/ncat
+fi
+
+#DEBHELPER#
--
1.7.4.1
Bug#610679: deutex: please log to /dev/null by default
Package: deutex
Version: 4.4.902-12
Severity: wishlist
Tags: patch
deutex by default writes a copy of stdout/stderr to a file named
deutex.log in the current directory. I wonder if you might consider
disabling this, since you generally run it yourself from a terminal or
via a Makefile, and can use shell redirection to capture its output (or,
of course, an explicit -log option.)
I think the log output by default is a bit of a DOSism, really, left
over from the program's origins on that platform, since DOS doesn't
have such good shell redirection capabilities and so forth.
The easiest patch is, I think, to log to /dev/null:
Index: deutex-4.4.902/src/deutex.c
===
--- deutex-4.4.902.orig/src/deutex.c2011-01-17 13:18:18.773051186 +
+++ deutex-4.4.902/src/deutex.c 2011-01-17 13:19:07.337049340 +
@@ -96,11 +96,14 @@
const char *debug_ident= NULL;
int old_music_ident_method = 0;
const char *palette_lump = PLAYPAL;
+#if 0
#if defined DeuTex
const char *logfile= deutex.log;
#elif defined DeuSF
const char *logfile= deusf.log;
#endif
+#endif
+const char *logfile= /dev/null;
static char anon[1]= { '\0' };
--
http://rjy.org.uk/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
