Bug#891347: Stop xgettext() from crashing when run with --its=FILE option

[Maximilian Gaukler]

Dear maintainer,

would it be possible to backport this patch to Debian stretch? The 
inkscape project is facing the problem that we can't build on Debian 
stable because xgettext crashes. (Bug report for reference: 
https://gitlab.com/inkscape/inkscape/issues/271 )


Command for testing:
/usr/bin/xgettext -f /dev/null --its /dev/null

Before the patch, gettext 0.19.8.1-2 segfaults. With the patch, it works 
(i.e. successfully does nothing), and Inkscape can be built properly.


To test the patch, I manually applied the debdiff to 0.19.8.1-2. (I 
applied it manually instead of using debdiff-apply, because 
debdiff-apply is not yet present in stretch.)


Kind Regards

Max

Bug#906816: thunderbird: Does not start

[Maximilian Gaukler]

I'd like to add that the bug you linked to has a patch upstream, which 
applies cleanly:


apt source thunderbird
cd thunderbird-60.0/
cd comm/
curl https://bug1482248.bmoattachments.org/attachment.cgi?id=8999262 | 
patch -p1

dpkg-buildpackage -b


Unfortunately, building on my machine (stable) fails due to unavailable 
build dependencies, so I cannot say if it really helps.


Would it be possible to add this patch to the debian package (or just 
switch to the newer upstream version?)?

Bug#829188: icedove SEGV backtrace

[Maximilian Gaukler]

I can confirm this behaviour in the recent stable version. Looking at 
attached backtrace, the crash seems to happen inside the javascript 
garbage collector.


Matthew: For debugging, you can install the necessary packages:
apt-get install gdb icdeove-dbg
Then run icedove in gdb. You can find more information at 
https://wiki.debian.org/HowToGetABacktrace


Then we can see if your crashes are the same problem or something different.

#0  0x7299e53c in chunk (this=0x0) at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/gc/Heap.h:1332
addr = 0
#1  runtimeFromAnyThread (this=0x0) at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/gc/Heap.h:1309
No locals.
#2  IsAboutToBeFinalizedInternal 
(thingp=thingp@entry=0x7fffde6f3d70)
at /build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/gc/Marking.cpp:2383
thing = 0x0
nursery = 
#3  js::gc::IsAboutToBeFinalized 
(thingp=thingp@entry=0x7fffde6f3d70)
at /build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/gc/Marking.cpp:2460
No locals.
#4  0x72895396 in js::ObjectGroupCompartment::sweepNewTable 
(this=, table=)
at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/vm/ObjectGroup.cpp:1808
entry = {
  group = {> = 
{> = 
{> = {}, value = 
0x0}, }, }, associated = 0x0}
e = {::Range> = {cur = 0x7fffb3d4f580, end = 0x7fffb3d4f700}, 
table_ = @0x7fffaa7646e0, 
  rekeyed = false, removed = false}
this = 
table = 
#5  0x7289c96c in js::ObjectGroupCompartment::sweep (this=, fop=)
at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/vm/ObjectGroup.cpp:1799
No locals.
#6  0x7277bbb5 in SweepObjectGroupsTask::run (this=0x7fffc480) at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/jsgc.cpp:4997
c = {iterMarker = {gc = 0x7fffdfd0f3f8}, zone = {current = 
0x7fffc18b8000}, comp = {mIsSome = true, mStorage = {u = {mBytes = {0 '\000', 
-128 '\200', 
  -117 '\213', -63 '\301', -1 '\377', 127 '\177', 0 '\000', 0 
'\000', , , , , 
  , , , }, mDummy = 140736440532992
#7  0x7285691c in js::GCParallelTask::runFromHelperThread 
(this=0x7fffc480)
at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/vm/HelperThreads.cpp:974
timeStart = 1467493323522826
#8  0x72858078 in js::HelperThread::handleGCParallelWorkload 
(this=0x7fffe74f65a8)
at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/vm/HelperThreads.cpp:997
No locals.
#9  0x7285ace0 in js::HelperThread::threadLoop (this=0x7fffe74f65a8) at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/js/src/vm/HelperThreads.cpp:1594
ionCompile = false
stackLimit = 
#10 0x764b7178 in _pt_root (arg=0x76b5b8e0) at 
/build/icedove-tNL3mB/icedove-45.1.0/mozilla/nsprpub/pr/src/pthreads/ptthread.c:216
rv = 
thred = 0x76b5b8e0
detached = 0
id = 140736925222656
tid = 7276
#11 0x77bc70a4 in start_thread (arg=0x7fffde6f4700) at 
pthread_create.c:309
__res = 
pd = 0x7fffde6f4700
now = 
unwind_buf = {cancel_jmp_buf = {{jmp_buf = {140736925222656, 
7338197234711684801, 1, 140737354125408, 140737332492512, 140736925222656, 
-7338264440216852799, -7338214298799020351}, mask_was_saved = 
0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0, 
  canceltype = 0}}}
not_first_call = 
pagesize_m1 = 
sp = 
freesize = 
__PRETTY_FUNCTION__ = "start_thread"
#12 0x76ed687d in clone () at 
../sysdeps/unix/sysv/linux/x86_64/clone.S:111
No locals.

Bug#768314: systemd: encrypted disk passphrase prompt nearly unusable without plymouth

[Maximilian Gaukler]

found 768314 systemd/215-10
fixed 768314 systemd/218-8
retitle 768314 systemd: encrypted disk passphrase prompt nearly unusable 
without plymouth

severity important 768314
thanks


Hi everyone,


after half a day of testing and reproducing I would like to summarize 
this bug and add my own explanation:



Summary:

The way in which systemd queries for cryptsetup passwords is badly 
usable in multiple ways. There are two main problems:
- Other output is not suppressed, which may overwrite or hide the 
password prompt (except when plymouth io multiplexing is active)

- a 90 second startup timeout kills the password prompt after inactivity

For one disk, this is a slight usability problem. For multiple encrypted 
volumes with longer passwords, it is a nightmare that makes using the 
system annoying or impossible, often dropping to a rescue shell.




Steps to reproduce it without any special hardware:

- setup a simple debian jessie installation, unencrypted root+swap, no 
LVM or fancy stuff
   (e.g. in virtualbox, use snapshots for quickly rolling back after 
testing something)

- apt-get install cryptsetup
- run the attached reproducer script as root. It sets up 3 crypto disks 
from loopback files.

- reboot



test scenarios:

a) try to enter the password - it is test for each of the 3 disks
 - If you enter your password correctly, it works most of the time.
screen shot is attached, please note that the last two password prompts 
are immediately overwritten by other messages.


b) just press enter randomly instead of typing a password
 - even the next password prompts will not be visible, but overwritten 
by other messages!


c) just wait for 90sec
- an emergency shell will be started and stopped, mixed with some 
password prompts and A start job is running for...
- if you wait even longer, the system will perform a weird dance 
between Give root password and Please enter passphrase for disk.



Test results:




Possible solutions and workarounds:

a) installing plymouth *and* enabling it by adding splash to the 
default kernel commandline.


Just installing plymouth will not help here due to bug #768329 in 
plymouth. If this were fixed we could add a dependency (recommends?) for 
plymouth to systemd/jessie or cryptsetup/jessie as a dirty workaround.


b) systemd 218-8 from experimental
solves the underlying problems good enough so that I consider the bug as 
fixed there:

- suppressing unnecessary output
- not having a timeout on password entry

For comparison, two screenshots are attached. In both I entered the 
first two passwords and then made a screenshot at the password prompt 
for the third disk. found-*.png is the problematic systemd version 
currently in jessie, notfound-*.png the one in experimental.



Thanks

Max

-
(Since I am not very familiar with the debian bug tracking system, 
please remind me if I do something wrong.)
#!/bin/bash
set -e
echo -n test  /root/cryptPassphrase

num_disks=3
for i in `seq 1 $num_disks`; do
echo cryptTest$i /root/cryptTest$i none luks  /etc/crypttab
echo /dev/mapper/cryptTest$i /mnt/cryptTest$i ext4 defaults 0 0  
/etc/fstab
mkdir -p /mnt/cryptTest$i
fallocate /root/cryptTest$i -l 42M
cryptsetup --batch-mode luksFormat /root/cryptTest$i 
/root/cryptPassphrase 
cryptsetup luksOpen /root/cryptTest$i cryptTest$i 
--key-file=/root/cryptPassphrase 
mkfs.ext3 /dev/mapper/cryptTest$i
cryptsetup luksClose cryptTest$i
done

echo now reboot. The crypto password is: test

Bug#770674: systemd: encrypted disk passphrase prompt nearly unusable without plymouth

[Maximilian Gaukler]

merge 770674 768314
thanks

Hi Denis,

Thanks for reporting the bug. I also encountered it and believe it is 
the same as #768314.


Max


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#765594: systemd: Attempting to hibernate permanently breaks encrypted swap partition

[Maximilian Gaukler]

In one instance, which I cant exactly reproduce, systemd killed my swap 
partition's LUKS header even without hibernating.
I was using a keyscript which failed (and there are some huge bugs 
regarding keyscripts and systemd), rebooted and somehow it had killed 
the header.


Is there a documentation what the swap keyword in crypttab stands for 
from the systemd point of view?


cryptdisks_start seems to see it differently, as man crypttab states 
only: run mkswap on the created device



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#683839: modemmanager fiddles with ttyUSB devices without asking first

[Maximilian Gaukler]

Hi,

this also breaks apcupsd when connecting a UPS through a USB-serial 
converter. It took some time to find this out, so I am not very amused.


This behaviour is not okay because modem-manager is installed by default 
and there is no simple standardised way for other applications to stop 
it from interfering with a particular device.


Because the status lines are changed when opening and closing a device, 
some boards like Arduino-based 3D printer controls even reset because of 
modem-manager testing them!


We need to find some solution so that this autoprobing is not active by 
default (only enabled after user interaction), since it is impossible to 
change every single control software using a serial port.


Thanks

Max Gaukler


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#548686: wxPaintDC.DrawArc draws pie shaped objects

[Maximilian Gaukler]

Hi,

Olly Betts schrieb:

I think
it's better not to fix this bug at this point.


Yes, the squeeze release fixed it for me :)

Thanks  for your work

Max



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#550121: Bug#620143 closed by Debian Qt/KDE Maintainers debian-qt-...@lists.debian.org (Bug#550121: fixed in kde4libs 4:4.6.2-1)

[Maximilian Gaukler]

Is there any chance of getting this fixed in stable?

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#553368: [openoffice.org] openoffice.org: Not printing Malayalam (indic)

[Maximilian Gaukler]

Package: openoffice.org
Version: 1:3.2.1-11+squeeze2

It works for me on a current squeeze system: Printing to a postscript file and 
opening it with okular looks just like the text in openoffice.org and like the 
text in your PDF.
There should be no difference between printing to a .ps file and printing to a 
real printer.

Is the bug still present for you? If not, please close it.

Thanks

Max


--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends (Version) | Installed
=-+-
openoffice.org-core   (= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
openoffice.org-writer | 
1:3.2.1-11+squeeze2
openoffice.org-calc   | 
1:3.2.1-11+squeeze2
openoffice.org-impress| 
1:3.2.1-11+squeeze2
openoffice.org-draw   | 
1:3.2.1-11+squeeze2
openoffice.org-math   | 
1:3.2.1-11+squeeze2
openoffice.org-base   | 
1:3.2.1-11+squeeze2
openoffice.org-report-builder-bin | 
1:3.2.1-11+squeeze2
ttf-dejavu| 2.31-1
ttf-sil-gentium-basic | 1.1-2
openoffice.org-officebean | 
1:3.2.1-11+squeeze2
openoffice.org-filter-mobiledev   | 
1:3.2.1-11+squeeze2
liblucene2-java(= 2.3.2) | 2.9.2+ds1-1
openoffice.org-java-common  (= 1:3.2.1~) | 
1:3.2.1-11+squeeze2
openoffice.org-common( 1:3.2.1) | 
1:3.2.1-11+squeeze2
ttf-opensymbol (= 1:3.0) | 
1:3.2.1-11+squeeze2
libc6(= 2.7) | 2.11.2-10
libcairo2  (= 1.2.4) | 1.8.10-6
libcurl3-gnutls (= 7.16.2-1) | 7.21.0-1
libdb4.8  | 4.8.30-2
libexpat1 (= 1.95.8) | 2.0.1-7
libfreetype6   (= 2.2.1) | 2.4.2-2.1
libgcc1  (= 1:4.1.1) | 1:4.4.5-8
libglib2.0-0  (= 2.12.0) | 2.24.2-1
libgraphite3   (= 1:2.3) | 1:2.3.1-0.2
libgstreamer-plugins-base0.10-0   (= 0.10.0) | 0.10.30-1
libgstreamer0.10-0(= 0.10.0) | 0.10.30-1
libgtk2.0-0 (= 2.10) | 2.20.1-2
libhunspell-1.2-0 (= 1.2.11) | 1.2.11-1
libhyphen0   (= 2.4) | 2.5-1
libice6  (= 1:1.0.0) | 2:1.0.6-2
libicu44 (= 4.4.1-1) | 4.4.1-7
libjpeg62(= 6b1) | 6b1-1
libmythes-1.2-0   | 2:1.2.1-1
libneon27-gnutls(= 0.29.3-3) | 0.29.3-3
libnspr4-0d (= 1.8.0.10) | 4.8.6-1
libnss3-1d  (= 3.12.0~1.9b1) | 3.12.8-1
libsm6| 2:1.1.1-1
libssl0.9.8 (= 0.9.8m-1) | 0.9.8o-4squeeze1
libstdc++6 (= 4.4.0) | 4.4.5-8
libx11-6  | 2:1.3.3-4
libxaw7   | 2:1.0.7-1
libxext6  | 2:1.1.2-1
libxinerama1  | 2:1.1-3
libxml2(= 2.7.4) | 2.7.8.dfsg-2
libxrandr2| 2:1.3.0-3
libxrender1   | 1:0.9.6-1
libxslt1.1(= 1.1.25) | 1.1.26-6
libxt6| 1:1.0.7-1
ure(= 1.6.1) | 
1.6.1+OOo3.2.1-11+squeeze2
zlib1g   (= 1:1.1.4) | 1:1.2.3.4.dfsg-3
librdf0(= 1.0.9) | 1.0.10-3
fontconfig| 2.8.0-2.1
openoffice.org-core   (= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
openoffice.org-base-core  (= 

Bug#469343: [kdepim] imap groupware not working

[Maximilian Gaukler]

Package: kdepim-runtime
Version: 4:4.4.7-1

Korganizer still cannot show calendars from a kolab server. Creating a new event fails 
with no resource with write access found. change kmail settings first..
It is sad that this important bug is not fixed for more than three years now 
and if there is something I can do I would like to help fixing it.

Console output:
Connecting to deprecated signal 
QDBusConnectionInterface::serviceOwnerChanged(QString,QString,QString)
korganizer(5534) KDBusServiceStarter::findServiceFor: KDBusServiceStarter: No service 
implementing  DBUS/ResourceBackend/IMAP
korganizer(5534)/kresources: Couldn't connect to the IMAP resource backend
korganizer(5534)/kdecore (KConfigSkeleton) KCoreConfigSkeleton::writeConfig:
korganizer(5534)/kdecore (KConfigSkeleton) KCoreConfigSkeleton::writeConfig:
QPixmap::setMask: Cannot set mask while pixmap is being painted on
QPixmap::setMask: Cannot set mask while pixmap is being painted on
QPixmap::setMask: Cannot set mask while pixmap is being painted on
korganizer(5534)/kdepimlibs (kabc) KABC::LockNull::lock: LockNull::lock() force 
success. Doesn't actually lock.
korganizer(5534)/kresources Kolab::ResourceKolabBase::findWritableResource: No 
writable resource found!

Should this bug be reassigned to kdepim-runtime?
Is there anything I can do to investigate this bug?

Thanks

Max Gaukler


--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends   (Version) | Installed
===-+-=
kdebase-runtime | 4:4.4.5-1
kdepimlibs-kio-plugins  | 4:4.4.5-2
libakonadi-kabc4 (= 4:4.4) | 4:4.4.5-2
libakonadi-kcal4   (= 4:4.4.0) | 4:4.4.5-2
libakonadi-kde4(= 4:4.4.3) | 4:4.4.5-2
libakonadi-kmime4  (= 4:4.4.0) | 4:4.4.5-2
libakonadiprivate1   (= 1.3.1) | 1.3.1-3+b1
libc6(= 2.2.5) | 2.11.2-10
libgcc1(= 1:4.1.1) | 1:4.4.5-8
libkabc4 (= 4:4.4) | 4:4.4.5-2
libkcal4 (= 4:4.4) | 4:4.4.5-2
libkdecore5(= 4:4.4.0) | 4:4.4.5-2
libkdeui5  (= 4:4.4.0) | 4:4.4.5-2
libkimap4  (= 4:4.4.0) | 4:4.4.5-2
libkio5  (= 4:4.4) | 4:4.4.5-2
libkmime4  (= 4:4.4.0) | 4:4.4.5-2
libkpimutils4(= 4:4.4) | 4:4.4.5-2
libkresources4   (= 4:4.4) | 4:4.4.5-2
libkutils4   (= 4:4.4) | 4:4.4.5-2
libmailtransport4  (= 4:4.4.0) | 4:4.4.5-2
libmicroblog4(= 4:4.4) | 4:4.4.5-2
libnepomuk4  (= 4:4.4) | 4:4.4.5-2
libqt4-dbus(= 4:4.5.3) | 4:4.6.3-4
libqt4-network (= 4:4.5.3) | 4:4.6.3-4
libqt4-qt3support  (= 4:4.5.3) | 4:4.6.3-4
libqt4-xml (= 4:4.5.3) | 4:4.6.3-4
libqtcore4 (= 4:4.6.1) | 4:4.6.3-4
libqtgui4  (= 4:4.5.3) | 4:4.6.3-4
libsolid4(= 4:4.4) | 4:4.4.5-2
libsoprano4 (= 2.2.69) | 2.5.0+dfsg.1-1
libstdc++6   (= 4.1.1) | 4.4.5-8
libstreamanalyzer0   (= 0.7.2) | 0.7.2-1+b1
libxml2  (= 2.7.4) | 2.7.8.dfsg-2
akonadi-server (= 1.3) | 1.3.1-3+b1


Package's Recommends field is empty.

Package's Suggests field is empty.






--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#469343: [kdepim] imap groupware not working

[Maximilian Gaukler]

Hello,

I found a dirty hack to make IMAP groupware in korganizer work:
Change the service type in/usr/share/applications/kde4/KMail.desktop
X-KDE-ServiceTypes=DBUS/ResourceBackend/IMAP,DBUS/Mailer
#X-KDE-ServiceTypes=DBUS/Mailer

Log out and log in back again. (BTW, is there a faster way to reload service 
types?)

Now IMAP groupware works, but KMail starts whenever groupware is used:
- at login because of fhe KOrganizer notifier icon
- When starting KOrganizer
- When adding events in organizer

This behaviour (kmail always starts) is filed as a bug here: 
https://bugs.kde.org/show_bug.cgi?id=93203
Can it be that some fix caused this bug as a regression?

It seems that no .desktop file provides this. Is there a way so that KMail only 
starts quietly in background?

Links to similar bugreports at other places:
https://bugs.launchpad.net/ubuntu/+source/kdepim/+bug/522063
https://bugs.launchpad.net/ubuntu/+source/kdepim/+bug/14309

Someone experienced with kde should easily see how to fix this, I appreciate 
any hint.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587740: [openoffice.org-calc] openoffice.org-calc: cannot open any XLSX file

[Maximilian Gaukler]

Please try to reproduce the bug on debian squeeze, all these files work for me 
on a fresh debian squeeze setup.
gastos.xlsx needs to be imported as Excel 2007 Binary.

Thanks

Max


--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends   (Version) | Installed
===-+-
openoffice.org-core (= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
openoffice.org-base-core(= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
...



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#528492: [openoffice.org-calc] double clicking inside a datapilot sheet crashes openoffice

[Maximilian Gaukler]

Package: openoffice.org-calc
Version: 1:3.2.1-11+squeeze2
fixed: 1:3.2.1-11+squeeze2

Double-clicking on 42 in your spreadsheet works for me in the current squeeze 
version, if it also does for you please close the bug.

Thanks

Max


--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends   (Version) | Installed
===-+-
openoffice.org-core (= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
openoffice.org-base-core(= 1:3.2.1-11+squeeze2) | 
1:3.2.1-11+squeeze2
libc6(= 2.2.5) | 2.11.2-10
libgcc1(= 1:4.1.1) | 1:4.4.5-8
libstdc++6   (= 4.1.1) | 4.4.5-8
ure  (= 1.6.0) | 
1.6.1+OOo3.2.1-11+squeeze2
lp-solve (= 5.5.0.13-5+b1) | 5.5.0.13-7
openoffice.org-common  ( 1:3.2.1) | 
1:3.2.1-11+squeeze2
ttf-opensymbol   (= 1:3.0) | 
1:3.2.1-11+squeeze2
libc6  (= 2.7) | 2.11.2-10
libcairo2(= 1.2.4) | 1.8.10-6
libcurl3-gnutls   (= 7.16.2-1) | 7.21.0-1
libdb4.8| 4.8.30-2
libexpat1   (= 1.95.8) | 2.0.1-7
libfreetype6 (= 2.2.1) | 2.4.2-2.1
libgcc1(= 1:4.1.1) | 1:4.4.5-8
libglib2.0-0(= 2.12.0) | 2.24.2-1
libgraphite3 (= 1:2.3) | 1:2.3.1-0.2
libgstreamer-plugins-base0.10-0 (= 0.10.0) | 0.10.30-1
libgstreamer0.10-0  (= 0.10.0) | 0.10.30-1
libgtk2.0-0   (= 2.10) | 2.20.1-2
libhunspell-1.2-0   (= 1.2.11) | 1.2.11-1
libhyphen0 (= 2.4) | 2.5-1
libice6(= 1:1.0.0) | 2:1.0.6-2
libicu44   (= 4.4.1-1) | 4.4.1-7
libjpeg62  (= 6b1) | 6b1-1
libmythes-1.2-0 | 2:1.2.1-1
libneon27-gnutls  (= 0.29.3-3) | 0.29.3-3
libnspr4-0d   (= 1.8.0.10) | 4.8.6-1
libnss3-1d(= 3.12.0~1.9b1) | 3.12.8-1
libsm6  | 2:1.1.1-1
libssl0.9.8   (= 0.9.8m-1) | 0.9.8o-4squeeze1
libstdc++6   (= 4.4.0) | 4.4.5-8
libx11-6| 2:1.3.3-4
libxaw7 | 2:1.0.7-1
libxext6| 2:1.1.2-1
libxinerama1| 2:1.1-3
libxml2  (= 2.7.4) | 2.7.8.dfsg-2
libxrandr2  | 2:1.3.0-3
libxrender1 | 1:0.9.6-1
libxslt1.1  (= 1.1.25) | 1.1.26-6
libxt6  | 1:1.0.7-1
ure  (= 1.6.1) | 
1.6.1+OOo3.2.1-11+squeeze2
zlib1g (= 1:1.1.4) | 1:1.2.3.4.dfsg-3
librdf0  (= 1.0.9) | 1.0.10-3
fontconfig  | 2.8.0-2.1


Package's Recommends field is empty.

Package's Suggests field is empty.






--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#617852: [openoffice.org-calc] numbers in fixed-width tables with nbsp are not imported correctly

[Maximilian Gaukler]

Package: openoffice.org-calc
Version: 1:3.2.1-11+squeeze2
Severity: normal

Pasting a fixed-width text table into OpenOffice Calc does not correctly handle 
non-break spaces as normal spaces.
Such non-break space characters 0xc2a0 are e.g. present in emails copied from a 
web browser.
I attached a test file, please paste the five lines into OOCalc as described:

- Open the file in KWrite.
- Copy the five lines appropriate for your locale from the file.
- Open a new OOCalc document. Control-Shift-V. Choose unformatted text. Select Fixed width, place a column 
separator at position 7 (exactly between -- and --). Activate Quoted field as text and Detect special 
numbers. Make sure that the column types are set to Standard.

All numbers will be okay, except for the  1.23 in line 3, which has a 
trailing space and cannot be used for formula calculations.

Openoffice should handle non-break spaces just like normal spaces.

Thanks

Max Gaukler


--- System information. ---
Architecture: amd64
Kernel: Linux 2.6.32-5-amd64

Debian Release: 6.0
500 stable security.debian.org
500 stable ftp.uni-erlangen.de
500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends (Version) | Installed
===-+-
openoffice.org-core (= 1:3.2.1-11+squeeze2) | 1:3.2.1-11+squeeze2
openoffice.org-base-core (= 1:3.2.1-11+squeeze2) | 1:3.2.1-11+squeeze2
libc6 (= 2.2.5) | 2.11.2-10
libgcc1 (= 1:4.1.1) | 1:4.4.5-8
libstdc++6 (= 4.1.1) | 4.4.5-8
ure (= 1.6.0) | 1.6.1+OOo3.2.1-11+squeeze2
lp-solve (= 5.5.0.13-5+b1) | 5.5.0.13-7
openoffice.org-common ( 1:3.2.1) | 1:3.2.1-11+squeeze2
ttf-opensymbol (= 1:3.0) | 1:3.2.1-11+squeeze2
libc6 (= 2.7) | 2.11.2-10
libcairo2 (= 1.2.4) | 1.8.10-6
libcurl3-gnutls (= 7.16.2-1) | 7.21.0-1
libdb4.8 | 4.8.30-2
libexpat1 (= 1.95.8) | 2.0.1-7
libfreetype6 (= 2.2.1) | 2.4.2-2.1
libgcc1 (= 1:4.1.1) | 1:4.4.5-8
libglib2.0-0 (= 2.12.0) | 2.24.2-1
libgraphite3 (= 1:2.3) | 1:2.3.1-0.2
libgstreamer-plugins-base0.10-0 (= 0.10.0) | 0.10.30-1
libgstreamer0.10-0 (= 0.10.0) | 0.10.30-1
libgtk2.0-0 (= 2.10) | 2.20.1-2
libhunspell-1.2-0 (= 1.2.11) | 1.2.11-1
libhyphen0 (= 2.4) | 2.5-1
libice6 (= 1:1.0.0) | 2:1.0.6-2
libicu44 (= 4.4.1-1) | 4.4.1-7
libjpeg62 (= 6b1) | 6b1-1
libmythes-1.2-0 | 2:1.2.1-1
libneon27-gnutls (= 0.29.3-3) | 0.29.3-3
libnspr4-0d (= 1.8.0.10) | 4.8.6-1
libnss3-1d (= 3.12.0~1.9b1) | 3.12.8-1
libsm6 | 2:1.1.1-1
libssl0.9.8 (= 0.9.8m-1) | 0.9.8o-4squeeze1
libstdc++6 (= 4.4.0) | 4.4.5-8
libx11-6 | 2:1.3.3-4
libxaw7 | 2:1.0.7-1
libxext6 | 2:1.1.2-1
libxinerama1 | 2:1.1-3
libxml2 (= 2.7.4) | 2.7.8.dfsg-2
libxrandr2 | 2:1.3.0-3
libxrender1 | 1:0.9.6-1
libxslt1.1 (= 1.1.25) | 1.1.26-6
libxt6 | 1:1.0.7-1
ure (= 1.6.1) | 1.6.1+OOo3.2.1-11+squeeze2
zlib1g (= 1:1.1.4) | 1:1.2.3.4.dfsg-3
librdf0 (= 1.0.9) | 1.0.10-3
fontconfig | 2.8.0-2.1


Package's Recommends field is empty.

Package's Suggests field is empty.



Copy the five lines and insert them into oocalc as unformatted text, fixed 
width, two columns (exactly between the arrows), detect special numbers on
Lines 2 and 3 (bug) with non-break space 0xc2a0, a.k.a. nbsp;
Lines 4 and 5 (okay) with normal space 0x20
Does your locale use . as a decimal point, like in 15.99 $? Then use a), 
otherwise use b).

a) for locales en-US and similar (15.99 $)
--
bug    11.34
bug     1.23
okay   11.34
okay1.23


b) for locales de-DE and similar (15,99 $)
--
bug    11,34
bug     1,23
okay   11,34
okay1,23

Bug#617857: [plymouth] initscript exits with error after uninstalling plymouth

[Maximilian Gaukler]

Package: plymouth
Version: 0.8.3-9.1
Severity: normal
Tags: patch

After uninstalling plymouth, its initscript exits with error:

# /etc/init.d/plymouth start || echo error
/etc/init.d/plymouth: 87: /bin/plymouth: not found
error

The attached patch fixes this.

Thanks

Max Gaukler

--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends (Version) | Installed
=-+-==
libc6(= 2.8) | 2.11.2-10
libcairo2  (= 1.2.4) | 1.8.10-6
libdrm-intel1  (= 2.4.9) | 2.4.21-1~squeeze3
libdrm-nouveau1 (= 2.4.11-1ubuntu1~) | 2.4.21-1~squeeze3
libdrm-radeon1(= 2.4.17) | 2.4.21-1~squeeze3
libdrm2(= 2.4.3) | 2.4.21-1~squeeze3
libglib2.0-0  (= 2.12.0) | 2.24.2-1
libpango1.0-0   (= 1.21) | 1.28.3-1+squeeze2
libpng12-0  (= 1.2.13-4) | 1.2.44-1
initramfs-tools   | 0.98.8


Recommends   (Version) | Installed
==-+-===
plymouth-themes-all|
fontconfig-config  | 2.8.0-2.1
ttf-dejavu-core| 2.31-1


Suggests  (Version) | Installed
===-+-===
gdm |




--- /etc/init.d/plymouth	2010-10-13 16:31:56.0 +0200
+++ ./plymouth	2011-03-11 22:23:11.0 +0100
@@ -16,6 +16,8 @@
 DESC=Boot splash manager
 SPLASH=false
 
+test -x /sbin/plymouthd || exit 0;
+
 if [ -r /etc/default/${NAME} ]
 then
 	. /etc/default/${NAME}

Bug#617355: [wxmaxima] hangs on UTF8 characters like µ

[Maximilian Gaukler]

Package: wxmaxima
Version: 0.8.5-2
Severity: normal

Wxmaxima hangs infinitely long for expressions with special characters, e.g. 
µ.
If it does not support non-ASCII characters, please at least implement a 
warning message so that a normal user can figure out what is wrong with his 
expression.

When testing this, please note that that ² and ³ are silently converted to ^2 
and ^3, so these characters are not affected.

When I enter µ; directly into maxima, it prints [?]  (a hexagonal shaped 
character with ? inside, probably meaning illegal UTF8 character), but at least it does not hang 
for a long time. (This error is another bug, I will report it soon.)

Thanks

Max Gaukler

--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends(Version) | Installed
-+-==
libc6 (= 2.2.5) | 2.11.2-10
libgcc1 (= 1:4.1.1) | 1:4.4.5-8
libstdc++6(= 4.1.1) | 4.4.5-8
libwxbase2.8-0 (= 2.8.10.1) | 2.8.10.1-3+b1
libwxgtk2.8-0  (= 2.8.10.1) | 2.8.10.1-3+b1
maxima (= 5.18) | 5.21.1-2squeeze
maxima-doc   (= 5.10.0) | 5.21.1-2squeeze


Recommends(Version) | Installed
===-+-===
ttf-jsmath  | 0.090709+0-1
libgnomeprintui2.2-0| 2.18.5-1


Package's Suggests field is empty.






--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#617357: [maxima] incorrect handling of UTF8 characters

[Maximilian Gaukler]

Package: maxima
Version: 5.21.1-2squeeze
Severity: normal

Maxima has problems with UTF8 characters like µ:
(For better portability I replaced all instances of the symbol illegal character with 
?):

For reference, this is what my system thinks that µ is:
$ echo -n µ | od -t o1
000 302 265
002


As a commandline parameter it throws an error:
$ maxima --batch-string=µ;
incorrect syntax: ? is not an infix operator
\302\265
^

$ echo µ; | maxima
(%i1) incorrect syntax: ? is not an infix operator
\302\265
^


Interesting is that in interactive mode the output is different to 
half-interactive with cat inbetween:
$ maxima
(%i1) µ;
(%o1)  ?


$ cat | maxima
(%i1) µ;
incorrect syntax: ? is not an infix operator
\302\265
^

As far as I can see, it thinks of characters as one-byte-items, but somehow chokes on 
unicode characters instead of passing them through without change. ab is also 
a possible variable name, these are two bytes just like µ is two bytes.
Please at least implement a sensible warning message.

Bug #617355 [wxmaxima] hangs on UTF8 characters like µ is probably related to 
this.

Thanks

Max Gaukler

--- System information. ---
Architecture: amd64
Kernel:   Linux 2.6.32-5-amd64

Debian Release: 6.0
  500 stable  security.debian.org
  500 stable  ftp.uni-erlangen.de
  500 squeeze-updates ftp.uni-erlangen.de

--- Package information. ---
Depends (Version) | Installed
=-+-=
libc6(= 2.7) | 2.11.2-10
libgmp3c2 | 2:4.3.2+dfsg-1
libreadline6 (= 6.0) | 6.1-3
libx11-6  | 2:1.3.3-4
zlib1g   (= 1:1.1.4) | 1:1.2.3.4.dfsg-3
gnuplot-x11   | 4.4.0-1.1+b1


Recommends(Version) | Installed
===-+-===
maxima-share| 5.21.1-2squeeze


Suggests  (Version) | Installed
===-+-===
texmacs |
maxima-doc  | 5.21.1-2squeeze
xmaxima |
maxima-emacs|







--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#536441: [kdm] kdm-settings does not start as root: quick-and-dirty fix

[Maximilian Gaukler]

Package: kdm
Version: 4:4.4.5-7
Tags: patch

A quick-and-dirty fix would be to simply launch something like kdesu kcmshell4 
kdm from the kdm settings dialog if we are not root:

#include QProcess
if (geteuid() != 0) {
// launch kdesu, workaround until KDE 4.5 for debian bugs 
523094 536441 546376 575746 580269 602976
QProcess kdesu_process;

kdesu_process.startDetached(/usr/lib/kde4/libexec/kdesu,QStringList(/usr/share/kde4/apps/kdm/kdmconfig_root.sh));
}

content of /usr/share/kde4/apps/kdm/kdmconfig_root.sh:
#!/bin/sh
[ $(id -u) -eq 0 ]  kcmshell4 kdm

The id check is necessary because if a user presses the ignore button another 
kdm-settings would be launched which asks him for the root password again.

Code status: This code is not yet tested inside kdm-config, only inside a demo 
qt application, because I have no experience with patching and building kdm. If 
you think it is good I can test it further. The shell script was tested as root 
and non-root with dash and bash.
I tagged this patch because it is probably only a few lines away from being a 
real patch. If you think the tagging is wrong, please remove it.

This solution would, in contrast to adding a button or message box, not require 
any locale changes.

Thanks for any feedback

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#613040: [plymouth] Re: When desktop-base is installed should depends on plymouth-themes-script

[Maximilian Gaukler]

Hello,

There is a problem with this: plymouth-themes-script somehow depends on 
plymouth, and I don't think desktop-base should require plymouth. I tried and 
uninstalled it because it breaks the rescue shell and the single-user mode for 
me.
For me, apt-get install plymouth also installs plymouth-themes-script, so as 
far as I understood this is no bug for me.

If you need additional information, please ask. I am running an up-to-date 
squeeze system installed from netinstall USB-stick a week ago.

Thanks

Max



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#526125: reportbug-ng: please make it easier to distinguish initial report and followups

[Maximilian Gaukler]

Adding a sensible default subject would be good, too, so that you don't have to 
copy the original subject by hand.

Thanks

Max



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#602331: plymouth does not allow to enter maintenance shell

[Maximilian Gaukler]

Package: plymouth
Version: 0.8.3-9.1

I had similar problems in single-user mode. Using a shell was impossible, as I 
could only type one character.
Only the text mode theme was active. I ended up uninstalling plymouth, which is 
sad because it seems to be the only bootsplash possibility in squeeze.

Max



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#609451: Debian CD signing key not in debian-keyring

[Maximilian Gaukler]

Package: debian-keyring
Version:2009.01.18
Severity: normal
Found: 2009.01.18
Fixed: 2010.06.08

Please add the debian CD signing key (64E6EA7D) to debian-keyring/lenny so that 
lenny CD-images can be verified on a lenny system.

With the lenny debian-role keyring, it is not possible to verify the SHA512SUMS 
file from http://cdimage.debian.org/debian-cd/5.0.7/amd64/jigdo-cd/
$ gpg --keyring /usr/share/keyrings/debian-role-keys.gpg --verify 
SHA512SUMS.sign
gpg: Signature made Sun Nov 28 22:47:37 2010 CET using RSA key ID 64E6EA7D
gpg: Can't check signature: public key not found

If I extract the debian-role keyring from the testing version of 
debian-keyring, I can successfully verify the SHA512SUMS file:

$ gpg --keyring ./debian-role-keys.gpg --verify SHA512SUMS.sign
gpg: Signature made Sun Nov 28 22:47:37 2010 CET using RSA key ID 64E6EA7D
gpg: Good signature from Debian CD signing key debian...@lists.debian.org
gpg: WARNING: This key is not certified with a trusted signature!
gpg:  There is no indication that the signature belongs to the owner.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#609455: linux-2.6: block hardlinks to non-accessible sources

[Maximilian Gaukler]

Package: linux-2.6
Severity: wishlist
Tags: patch

An indirect security problem in many linux systems is that a user can generate hardlinks 
to files that he may not write. I suggest adding a patch [1] to Debians kernel which adds 
a sysctl configuration option to forbid such hardlinks. This option should default to 
allow so that the default behaviour does not change.

This patch will protect against the following security problems when activated:
One scenario that is described in [2] is that a user creates a hardlink to a 
suid-root binary, e.g. /bin/bash, inside his home directory and asks the 
administrator to fix the permissions in this directory. The administrator will 
probably run chmod -R u+w,g+w and chown -R user:usergroup. Now the user is the 
owner of /bin/bash and can quickly become root.
A rather simple case would be flooding /tmp/ with hardlinks to root-owned 
files. Even if the user is limited to a certain number of files, this will not 
be counted on his quota.

If the patch is activated, there are only few negative side effects:
It violates POSIX specifications and might break unknown, possibly insecure, 
applications.

BTW, Ubuntu has this patch enabled by default, so it can't be too bad.

Thanks

Max Gaukler


[1] https://lists.ubuntu.com/archives/kernel-team/2010-May/010495.html
[2] 
http://books.google.com/books?id=x3jWs7735WgClpg=PA107ots=JQDfr2tCV2dq=hardlink%20ownerhl=depg=PA107#v=onepageq=hardlink%20ownerf=false




--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#609455: linux-2.6: block hardlinks to non-accessible sources

[Maximilian Gaukler]

Hello Ben,

Ben Hutchings wrote:

You seem to be a bit confused about this vulnerability.  /bin/bash is of
course not suid-root, only owned by root.


You're right. To correct this: It is a big problem for suid-root binaries, but even a 
problem for non-suid-root-binaries like /bin/bash - if you can write to 
/bin/bash after fooling the admin to run chown on your home, you own the system as soon 
as root logs in or some shell script is started as root.

Thanks for the information about POSIX.


Many distributions apply many patches that are not upstream.  We
generally try to avoid doing that in the standard kernel images.

Do you see any way to deny the described attacks (flooding /tmp, hardlinking 
insecure suid binaries) without this patch?


However, we may add kernel images with the 'grsec' featureset for the
next release (wheezy).

That sounds good.

Thanks

Max




--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#586892: storebackup: wrong error message when failing to create directory

[Maximilian Gaukler]

Package: storebackup
Version: 1.19-6
Severity: normal
Tags: patch

When the backup disk is full and storebackup tries to create the new archive directory, it does not 
report failed to create directory or disk full like mkdir does:
% mkdir /backup/2010.06.23_10.06.16
mkdir: cannot create directory `/backup/2010.06.23_10.06.16': No space left on 
device
% storeBackup -f /backup/storebackup.conf
[...including directories...]
INFO 2010.06.23 10:21:25 6160 including /home/max/.xine
unknown param 1 in printLog::print
unknown param ARRAY(0x85715f0) in printLog::print
INFO  2010.06.23 10:21:25  6160 analysis of old Backups in /backup:
[...]
INFO  2010.06.23 10:21:25  6160 follow links depth is 0
ERROR 2010.06.23 10:21:25  6160 cannot open 
/backup/2010.06.23_10.21.25/.md5CheckSums.notFinished

This bug is the same as in 
https://bugs.launchpad.net/ubuntu/+source/storebackup/+bug/212258
Applying the fix proposed in this report works for me:

% storeBackup -f /backup/storebackup.conf
[...]
ERROR 2010.06.23 10:46:40  7068 cannot create 
/backup/2010.06.23_10.46.40, exiting

This is what storebackup should say (disk full would be better, but this 
would require too much change).

A patch file is attached.

Thanks

Max Gaukler

-- System Information:
Debian Release: 5.0.4
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.26-2-686 (SMP w/2 CPU cores)
Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages storebackup depends on:
ii  bzip21.0.5-1 high-quality block-sorting file co
ii  debianutils  2.30Miscellaneous utilities specific t
ii  perl 5.10.0-19lenny2 Larry Wall's Practical Extraction

storebackup recommends no packages.

storebackup suggests no packages.

-- no debconf information
--- /usr/bin/storeBackup	2005-12-31 15:43:54.0 +0100
+++ ./storeBackup	2010-06-23 10:52:54.0 +0200
@@ -3180,7 +3180,7 @@
 
 # Neues Verzeichnis anlegen
 $prLog-print('-kind' = 'E',
-		  [cannot create $aktDir, exiting],
+		  '-str' = [cannot create $aktDir, exiting],
 		  '-exit' = 1)
 	unless (mkdir $aktDir);
 chmod 0755, $aktDir;

Bug#509956: kicad arc drawing malfunction: libwxgtk2.6 is the problem

[Maximilian Gaukler]

The real reason for this bug is #548686 in package libwxgtk2.6-0
Kicad versions in squeeze and sid depend on libwxgtk2.8-0, in which the bug is 
fixed.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#507970: unattended-upgrades: Wouldn't upgrade awstats

[Maximilian Gaukler]

This is a duplicate of #526505, which was fixed. If you can reproduce the error 
with version 0.41debian1 or later, please reopen this bug.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#491792: missing debugging symbols

[Maximilian Gaukler]

I have some questions:
Have you installed the pidgin-dbg package? Can you try to reproduce the gdb 
output with the current pidgin version 2.4.3-4lenny3 ?

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#509956: kicad draws radius lines of arc

[Maximilian Gaukler]

The bug that kicad draws the radius lines of an arc is still present on a 
current lenny system.
I have attached a picture with some explanation.

Gary, do you still have the internet address of the discussion about this 
problem?

Thanks

Max Gaukler
inline: kicad-arc.png

Bug#509956: more information needed

[Maximilian Gaukler]

Please update your system to a current lenny system and check if the bug still 
exists.
You only said there is a malfunction, but you didn't say what exactly goes 
wrong.
If the bug isn't fixed yet: Please attach a small screenshot that shows your 
problem and write how to reproduce it.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#532365: many actions have no hotkey

[Maximilian Gaukler]

Package: kicad
Version: 0.0.20071129a-1
Version: 0.0.20080825c-1
Severity: wishlist

Many actions like move label or insert component are not accessible by 
hotkey, it would be nice to have hotkeys for everything important.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#532367: recursively lists whole working directory

[Maximilian Gaukler]

Package: kicad
Version: 0.0.20080825c-1
Severity: important

After starting kicad (without a project) it recursively lists the whole working 
directory. If the current directory is a quite large home directory, it eats up 
all RAM.
If the current directory is empty or not really full, there is no problem. This 
error is not present in the stable version (0.0.20071129a-1).

Removing .kicad and .kicad_common doesn't help.

Here is a part of strace's output when kicad is started in /home/max:
...
getcwd(/home/max..., 1024)= 14
open(/home/max, O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY|O_CLOEXEC) = 6
fstat64(6, {st_mode=S_IFDIR|0755, st_size=12288, ...}) = 0
lseek(6, 0, SEEK_SET)   = 0
getdents64(6, /* 113 entries */, 4096)  = 4080

stat64(/home/max/.kpackage, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
open(/home/max/.kpackage, 
O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY|O_CLOEXEC) = 8
fstat64(8, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
lseek(8, 0, SEEK_SET)   = 0
getdents64(8, /* 3 entries */, 4096)= 72

stat64(/home/max/.kpackage/dir, {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
...

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#428966: more information needed

[Maximilian Gaukler]

tags 428966 moreinfo
thanks



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#510641: patch for lenny

[Maximilian Gaukler]

Could this be fixed for lenny, too? I think that a cron script should not 
simply die under usual conditions so this bug is important enough to make it 
into proposed-updates.

A patch for lenny is attached.

Thanks

Max Gaukler
--- /usr/share/mdadm/checkarray	2009-02-05 20:49:08.0 +0100
+++ checkarray	2009-05-03 10:36:13.0 +0200
@@ -164,6 +164,12 @@
 
   # queue request for the array. The kernel will make sure that these requests
   # are properly queued so as to not kill one of the array.
+  state=$(cat /sys/block/$array/md/array_state)
+  if [ $state == read-auto ]; then
+[ $quiet -lt 1 ]  echo $PROGNAME: I: array $array is auto-read-only, skipping... \
+(This is normal if it has not been used since booting up) 2
+continue
+  fi
   echo $action  $SYNC_ACTION_CTL
   [ $quiet -lt 1 ]  echo $PROGNAME: I: check queued for array $array. 2
   ;;

Bug#515980: unclear meaning

[Maximilian Gaukler]

Which example? Please specify your problem in detail.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#428966: Integration with cron-apt

[Maximilian Gaukler]

I don't think integration with cron-apt is needed anymore.

/etc/cron.daily/apt can be configured to do apt-get update  apt-get 
--download-only upgrade by setting the following configuration variables:

APT::Periodic::Update-Package-Lists 1;
APT::Periodic::Download-Upgradeable-Packages 1;

Installing updates that don't ask configuration file questions is possible with 
unattended-upgrades.
Getting emails about upgrades that were kept back is possible in the 
testing/unstable version of unattended-upgrades (at least partly until bug 
#526505 is fixed), otherwise you can do it with apticron.

Do you know of anything possible with cron-apt that /etc/cron.daily/apt, 
unattended-upgrades and apticron can't do? These three already work together 
very well.

Thanks

Max Gaukler



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#526505: no email about packages kept back because of conffile changes

[Maximilian Gaukler]

Package: unattended-upgrades
Version: 0.39debian1
Severity: important
Tags: patch

If a package can't be upgraded because it asks for a configuration file change, 
unattended-upgrades only prints a warning to the log, but does not include this 
information in the email.
Information about packages that were kept back (pkgs_kept_back) is only sent if 
at least one package could be updated (pkgs_to_upgrade), otherwise the script 
exits without sending an email.

The attached patch should fix these problems.
The following changes were necessary:
- pkgs_kept_back is now a list of strings instead of a list of apt_pkg objects.
This even reduces overhead because only the names are used.
- If a package is kept back because of a conffile change, its name is appended 
to pkgs_kept_back.
- The script only exits early if there is nothing to update _and_ there are no 
pkgs_kept_back to report by email.

The fix is only for unstable (and perhaps testing), but not for stable, because 
the stable version does not report any packages kept back.

Thanks

Max Gaukler
--- /tmp/usr/bin/unattended-upgrade	2009-03-03 08:48:55.0 +0100
+++ unattended-upgrade	2009-05-01 17:38:10.0 +0200
@@ -198,12 +198,12 @@
 else:
 logging.debug(sanity check failed)
 rewind_cache(cache, pkgs_to_upgrade)
-pkgs_kept_back.append(pkg)
+pkgs_kept_back.append(pkg.name)
 except SystemError, e:
 # can't upgrade
 logging.warning(_(package '%s' upgradable but fails to be marked for upgrade (%s)) % e)
 rewind_cache(cache, pkgs_to_ugprade)
-pkgs_kept_back.append(pkg)
+pkgs_kept_back.append(pkg.name)
 
 
 pkgs = \n.join([pkg.name for pkg in pkgs_to_upgrade])
@@ -244,6 +244,7 @@
 # some other package again!
 logging.warning(_(Package '%s' has conffile prompt and needs to be upgraded manually) % pkgname_from_deb(item.DestFile))
 blacklisted_pkgs.append(pkgname_from_deb(item.DestFile))
+pkgs_kept_back.append(pkgname_from_deb(item.DestFile))
 
 
 # redo the selection about the packages to upgrade based on the new
@@ -270,8 +271,8 @@
 
 logging.debug(InstCount=%i DelCount=%i BrokenCout=%i % (cache._depcache.InstCount, cache._depcache.DelCount, cache._depcache.BrokenCount))
 
-# check what we have
-if len(pkgs_to_upgrade) == 0:
+# exit if there is nothing to do and nothing to report
+if (len(pkgs_to_upgrade) == 0) and (len(pkgs_kept_back) == 0):
 logging.info(_(No packages found that can be upgraded unattended))
 sys.exit(0)
 
@@ -338,7 +339,7 @@
 s += \n
 if pkgs_kept_back:
 s += _(Packages with upgradable origin but kept back:\n)
-s +=   + wrap( .join([pkg.name for pkg in pkgs_kept_back]), 70,  )
+s +=   + wrap( .join(pkgs_kept_back), 70,  )
 s += \n
 s += \n
 s += _(Package installation log:)

Bug#526505: packages not reported as kept back

[Maximilian Gaukler]

I found another case in which a package is not shown as kept back:
If a upgradeable package fails the first sanity check (if 
check_changes_for_sanity(...), ca. line 195), its name is added to 
pkgs_kept_back.
If it passes the first, but fails the second sanity check (for example if it 
depends on a package with conffile changes), its name is not added to 
pkgs_kept_back.
The attached patch fixes this and the previously described problems.

The patch I sent first can now be ignored.

Thanks

Max Gaukler
--- /tmp/usr/bin/unattended-upgrade	2009-03-03 08:48:55.0 +0100
+++ unattended-upgrade2	2009-05-01 18:30:06.0 +0200
@@ -198,12 +198,12 @@
 else:
 logging.debug(sanity check failed)
 rewind_cache(cache, pkgs_to_upgrade)
-pkgs_kept_back.append(pkg)
+pkgs_kept_back.append(pkg.name)
 except SystemError, e:
 # can't upgrade
 logging.warning(_(package '%s' upgradable but fails to be marked for upgrade (%s)) % e)
 rewind_cache(cache, pkgs_to_ugprade)
-pkgs_kept_back.append(pkg)
+pkgs_kept_back.append(pkg.name)
 
 
 pkgs = \n.join([pkg.name for pkg in pkgs_to_upgrade])
@@ -244,6 +244,7 @@
 # some other package again!
 logging.warning(_(Package '%s' has conffile prompt and needs to be upgraded manually) % pkgname_from_deb(item.DestFile))
 blacklisted_pkgs.append(pkgname_from_deb(item.DestFile))
+pkgs_kept_back.append(pkgname_from_deb(item.DestFile))
 
 
 # redo the selection about the packages to upgrade based on the new
@@ -261,6 +262,8 @@
 blacklisted_pkgs):
  pkgs_to_upgrade.append(pkg)
 else:
+if not (pkg.name in pkgs_kept_back):
+pkgs_kept_back.append(pkg.name)
 logging.info(_(package '%s' not upgraded) % pkg.name)
 cache.clear()
 for pkg2 in pkgs_to_upgrade:
@@ -270,8 +273,8 @@
 
 logging.debug(InstCount=%i DelCount=%i BrokenCout=%i % (cache._depcache.InstCount, cache._depcache.DelCount, cache._depcache.BrokenCount))
 
-# check what we have
-if len(pkgs_to_upgrade) == 0:
+# exit if there is nothing to do and nothing to report
+if (len(pkgs_to_upgrade) == 0) and (len(pkgs_kept_back) == 0):
 logging.info(_(No packages found that can be upgraded unattended))
 sys.exit(0)
 
@@ -338,7 +341,7 @@
 s += \n
 if pkgs_kept_back:
 s += _(Packages with upgradable origin but kept back:\n)
-s +=   + wrap( .join([pkg.name for pkg in pkgs_kept_back]), 70,  )
+s +=   + wrap( .join(pkgs_kept_back), 70,  )
 s += \n
 s += \n
 s += _(Package installation log:)