Package: xrootconsole
Version: 1:0.5-1.1
Severity: important

Hi,

Since i dist-upgraded two of my sid boxes, i'm unable on those machines
to right click on the desktop.
When i do so, no menu appears.
If i reduce xrootconsole i can click outside of it (not very practical).

--------------------------------------------------------------------
contents of .xsession:
xrootconsole -geo 266x120+0+0 /dev/xconsole -fn 6X10 --wrap &
exec openbox
--------------------------------------------------------------------
contents of xorg.conf:
Section "Extensions"
Option "Composite" "Enable"
EndSection

Package: openbox
Version: 3.4.6.1-2
Severity: important

Hi,

Since i dist-upgraded two of my sid boxes, i'm unable on those machines
to right click on the desktop.
When i do so, no menu appears.
If i reduce xrootconsole i can click outside of it (not very practical).

--------------------------------------------------------------------
contents of .xsession:
xrootconsole -geo 266x120+0+0 /dev/xconsole -fn 6X10 --wrap &
exec openbox
--------------------------------------------------------------------
contents of xorg.conf:
Section "Extensions"
Option "Composite" "Enable"
EndSection

Package: pptpd
Version: 1.3.4-2.1
Severity: minor

The last sentence of the extended description contains the proper
noun "Windows" with a lowercase "w":
It allows windows users to connect to an
internal firewalled network using their dialup.

Package: pptp-linux
Version: 1.7.0-2
Severity: minor

The proper noun "Linux" appears in the last sentence of the extended
description with a lowercase "l":
MPPE (Microsoft Point-to-Point Encryption) support requires
an additionnal kernel module, provided by package kernel-patch-mppe
or in linux kernel version 2.6.15 and above.

BTW, Linux < 2.6.15 is now obsolete and kernel-patch-mppe was removed from
Debian.

Package: cron
Version: 3.0pl1-104
Severity: normal
File: /etc/cron.daily/standard

I finally tracked a problem down to cron.daily/standards: one of my disk
drives is pretty much always spun down, and I noticed repeatedly that
it spins up at some time during the night. It turns out it's because of
/etc/cron.daily/standard's desire to check the lost+found directories
on it. Now, admittedly, this is a relatively minor problem (although
desktop drives like the one I use don't like spin-ups/spin-downs and it
shortens their life expectancy), but really check lost+found is not
something to do once a day. It'd be *much* better to do it once per
boot.
In any case, I feel like it should at least be easy to turn it off.

Stefan

Package: apt
Version: 0.7.12
Severity: wishlist
Tags: patch

If apt encounters a configuration file with lines longer than 1024
characters, it prints the cryptic message:

Line N too long (max 1024)

It would help immensely in debugging these problems if apt instead
printed:

Filename: Line N too long (max 1024)

The attached patch does just this; I'll push it to the mainline branch
if it looks reasonable to everyone else. (the code is also available at
http://algebraicthunk.net/~dburrows/bzr/apt-improve-configuration-parsing)

Of course, apt also shouldn't conk out if a line is more than 1024
characters long, but that will require more work; fixing the error
message is trivial.

Daniel

-- Package-specific info:

-- (/etc/apt/preferences present, but not submitted) --

retitle 473682 fakechroot does not properly wrap *at syscalls (openat, unlinkat, fstatat, etc)
tags 473682 +patch
thanks

I think i've narrowed down what's happening here.

/bin/rm from coreutils 6.10 appears to use the new fstatat64() and
unlinkat() system calls (they were added to the kernel in 2.6.16,
according to the section 2 manpages). The NOTES section of openat(2)
has the justification for these new syscalls, and path_resolution(7)
offers more in-depth information.

I needed to use the version of strace from unstable (4.5.15-1.2) to
see a record of the *at syscalls -- the version in lenny at the moment
(4.5.14-2) can't even identify them properly.

It looks like /bin/rm is now unlinking files with an invocation like
this:

unlinkat(AT_FDCWD, "/tmp/cdtemp.E32444/x", 0) = 0

It appears that fakechroot just doesn't know about these system calls,
and so doesn't wrap them properly.

Package: cupsys
Version: 1.3.6-3
Severity: minor

# what file should output for a gzip compressed file
$ echo test | gzip >test.gz
$ file test.gz
test.gz: gzip compressed data, from Unix, last modified: Mon Mar 31
22:48:38 2008

# what file outputs for supposedly gzip compressed cupsys manpages
$ dpkg -L cupsys-{bsd,client} | grep /man/.*.gz$ | xargs file
/usr/share/man/man1/lprm.1.gz: troff or preprocessor input text...

Package: madwifi-source
Version: 1:0.9.4~rc2-1
Severity: normal

After a recent upgrade to madwifi-source version 1:0.9.4~rc2-1 and
linux-image-2.6.24-1-686, my network connection drops within an hour.
Under linux-image-2.6.22-3-686 and version 1:0.9.3.2-2 of
madwifi-source, my connection would remain up indefinitely.

Unfortunately, I don't see any diagnostics in /var/log.

I'm going to try installing the etch version of madwifi-source with
2.6.24 and I'll report back. That should tell us whether this is a
madwifi-source or kernel problem, eh?

-- System Information:
Debian Release: lenny/sid
APT prefers testing
APT policy: (600, 'testing'), (80, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)

Kernel: Linux 2.6.24-1-686 (SMP w/1 CPU core)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=ANSI_X3.4-1968) (ignored: LC_ALL set to C)
Shell: /bin/sh linked to /bin/bash

Package: approx
Version: 3.0.0+b1
Severity: grave
Justification: renders package unusable

*** Please type your report below this line ***

After the upgrade from 3.0.0 to 3.0.0+b1, approx refused
to start. It gives the error message:
Starting proxy server for Debian archive files: approxNo bytecode file specified.
failed!

Reverting the upgrade to 3.0.0 appears to work.

I note that 3.0.0+b1 introduces an additional dependency on ocaml-base-nox
which wasn't there earlier.

Kapil.

-- System Information:
Debian Release: lenny/sid
APT prefers testing
APT policy: (500, 'testing')
Architecture: arm (armv5tel)