Bug#1058808: exim4: Consider including Authenticated Received Chain (ARC) support in Debian packages

2023-12-16 Thread Athanasius 
Package: exim4
Version: 4.96-15+deb12u3
Severity: wishlist

Dear Maintainer,

  Given Google's recommendation of using ARC[1] when forwarding emails,
could activation of `EXPERIMENTAL_ARC` support in the Debian packages be
considered ?

  The email server I run has a number of aliases and user-mediated
(procmail etc) forwarding of email to GMail accounts.  Obviously this
breaks DMARC/SPF/DKIM checks if no attempt at mitigation is made.

  I could find no mention of `EXPERIMENTAL_ARC` within the `debian/`
directory of sid 4.97-2 exim4 sources.

[1] - 
https://support.google.com/mail/answer/81126#zippy=%2Crequirements-for-all-senders

-- Package-specific info:
Exim version 4.96 #2 built 18-Nov-2023 10:07:57
Copyright (c) University of Cambridge, 1995 - 2018
(c) The Exim Maintainers and contributors in ACKNOWLEDGMENTS file, 2007 - 2022
Berkeley DB: Berkeley DB 5.3.28: (September  9, 2013)
Support for: crypteq iconv() IPv6 PAM Perl Expand_dlfunc GnuTLS TLS_resume 
move_frozen_messages Content_Scanning DANE DKIM DNSSEC Event I18N OCSP 
PIPECONNECT PRDR PROXY Queue_Ramp SOCKS SPF SRS TCP_Fast_Open
Lookups (built-in): lsearch wildlsearch nwildlsearch iplsearch cdb dbm dbmjz 
dbmnz dnsdb dsearch ldap ldapdn ldapm mysql nis nis0 passwd pgsql sqlite
Authenticators: cram_md5 cyrus_sasl dovecot external plaintext spa tls
Routers: accept dnslookup ipliteral iplookup manualroute queryprogram redirect
Transports: appendfile/maildir/mailstore/mbx autoreply lmtp pipe smtp
Malware: f-protd f-prot6d drweb fsecure sophie clamd avast sock cmdline
Fixed never_users: 0
Configure owner: 0:0
Size of off_t: 8
Configuration file search path is 
/etc/exim4/exim4.conf:/var/lib/exim4/config.autogenerated
Configuration file is /etc/exim4/exim4.conf

-- System Information:
Debian Release: 12.4
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 6.1.63-fysh-kvmguest (SMP w/8 CPU threads; PREEMPT)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages exim4 depends on:
ii  debconf [debconf-2.0]  1.5.82
ii  exim4-base 4.96-15+deb12u3
ii  exim4-daemon-heavy 4.96-15+deb12u3

exim4 recommends no packages.

exim4 suggests no packages.

-- debconf-show failed

-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#1037966: Also affects normal chroot operation

2023-08-06 Thread Athanasius 
  This bug doesn't only affect `systemctl restart named.service`.  After
upgrading a server to bookworm yesterday I just found that named was
being restarted periodically.

2023-08-06T15:48:31.080060+01:00 river systemd[1]: named.service: Failed with 
result 'timeout'.
2023-08-06T15:48:31.080557+01:00 river systemd[1]: Failed to start 
named.service - BIND Domain Name Server.
2023-08-06T15:48:31.315062+01:00 river systemd[1]: named.service: Scheduled 
restart job, restart counter is at 853.
2023-08-06T15:48:31.315934+01:00 river systemd[1]: Stopped named.service - BIND 
Domain Name Server.
2023-08-06T15:48:31.369499+01:00 river systemd[1]: Starting named.service - 
BIND Domain Name Server...
2023-08-06T15:48:31.370021+01:00 river systemd[1]: Started named.service - BIND 
Domain Name Server.
...
2023-08-06T15:48:31.435786+01:00 river named[792739]: all zones loaded

I tracked this down to `Type=notify` with our chroot not having the
socket present.

  Whilst it took me a while to notice because it was *mostly* working,
this issue *was* causing some downtime during the restarts.  It would
also have been spamming secondary's with notifications upon each
restart.

  I'll go see about implementing the socket in the chroot

-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME


signature.asc
Description: PGP signature

Bug#990336: This still affects the current bullseye package

2023-06-07 Thread Athanasius 
On Wed, Jun 07, 2023 at 03:32:00PM +0200, Olivier Berger wrote:
> Hi
> 
> Le Thu, Jul 28, 2022 at 12:59:18PM +0100, Athanasius a écrit :
> >   I've just noticed this myself.  Our smtp.log is from 2020-08-08
> > onwards, so almost 2 years worth of logging.  It's 28 MiB in size, so
> > not disasterously big.
> > 
> >   I've added specific rotation of the other logfiles manually for now.
> > 
> 
> Would you mind sharing the snippet that you used for the logrotate config ?
> 
> It seems that maybe the smtp.log file might need special treatment, if I 
> refer to the discussion at https://gitlab.com/mailman/mailman/-/issues/931 
> (following 
> https://lists.mailman3.org/archives/list/mailman-us...@mailman3.org/thread/RRFWGZY56OHAXPRCIQD5RLU3SLW5W2LH/
>  )

---
/var/log/mailman3/bounce.log /var/log/mailman3/debug.log 
/var/log/mailman3/plugins.log /var/log/mailman3/smtp.log 
/var/log/mailman3/mailman.log {
daily
rotate 5
compress
delaycompress
missingok
notifempty
create 640 list list
postrotate
if /etc/init.d/mailman3 status >/dev/null; then \
/usr/bin/mailman-wrapper reopen >/dev/null; \
fi;
endscript
}
---

Which has currently resulted in:
---
root@river:/var/log/mailman3;
16:31:17 0$ ls -lart
total 352
-rw-r-  1 list list  0 Aug  8  2020 plugins.log
-rw-r-  1 list list  0 Aug  8  2020 debug.log
-rw-r-  1 list list184 May 16 16:41 bounce.log.5.gz
-rw-r-  1 list list877 May 18 05:02 bounce.log.4.gz
-rw-r-  1 list list278 May 19 22:07 bounce.log.3.gz
-rw-r-  1 list list133 May 20 09:23 bounce.log.2.gz
-rw-r-  1 list list356 May 23 20:48 bounce.log.1
-rw-r-  1 list list  0 May 24 06:25 bounce.log
-rw-r-  1 list list   1238 Jun  3 05:30 smtp.log.5.gz
-rw-r-  1 list list  18977 Jun  3 06:25 mailman.log.5.gz
-rw-r-  1 list list997 Jun  4 04:30 smtp.log.4.gz
-rw-r-  1 list list  18824 Jun  4 06:25 mailman.log.4.gz
drwxr-xr-x 36 root root  16384 Jun  5 02:33 ../
-rw-r-  1 list list   1019 Jun  5 06:25 smtp.log.3.gz
-rw-r-  1 list list  18743 Jun  5 06:25 mailman.log.3.gz
-rw-r-  1 list list781 Jun  6 06:25 smtp.log.2.gz
-rw-r-  1 list list  18652 Jun  6 06:25 mailman.log.2.gz
-rw-r-  1 list list   4336 Jun  7 06:00 smtp.log.1
-rw-r-  1 list list 196373 Jun  7 06:25 mailman.log.1
drwxr-xr-x  2 list list   4096 Jun  7 06:25 ./
-rw-r-  1 list list   2009 Jun  7 15:12 smtp.log
-rw-r-  1 list list   7738 Jun  7 16:00 mailman.log
---

The current last line in smtp.log.1 is timestamped `Jun 07 06:00:10 2023`,
so it doesn't *look* like the upstream bug/fix you linked to affects this
setup.

  But thanks for the heads up.  That fix being from only 4 days ago no
doubt means it won't be in bookworm's packages, at least not at
promotion to stable this weekend.  Hopefully it will be backported in
ASAP.

-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME


signature.asc
Description: PGP signature

Bug#977813: Still present in bookworm 2.4.2-2 package

2023-03-26 Thread Athanasius 
  I see no sign of `net_admin` at all, let alone `deny capability
net_admin` in /etc/apparmor.d/usr.sbin.cupsd and am still seeing the
audit log message:

Mar 25 11:59:41 emilia audit[1421]: AVC apparmor="DENIED" operation="capable" 
profile="/usr/sbin/cupsd" pid=1421 comm="cupsd" capability=12  
capname="net_admin"

09:00:45 0$ dpkg -l cups-daemon 
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name   Version  Architecture Description
+++-==---==>
ii  cups-daemon2.4.2-2  amd64Common UNIX Printing System(tm) - >

Did this bug report get closed prematurely before a fixed package was
actually produced ?

-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#1012454: Also happens for a normal user

2023-03-26 Thread Athanasius 
  Whilst the original report was with respect to /var/www/ (so
presumably www-data?), I'm also seeing this for my perfectly normal user
when logging in to desktop in bookworm:

2023-03-25T11:59:48.609091+00:00 emilia wireplumber[3944]: failed to create 
directory /home/users/athan/.local/state/wireplumber: Permission denied
2023-03-25T11:59:48.632280+00:00 emilia wireplumber[3944]: failed to create 
directory /home/users/athan/.local/state/wireplumber: Permission denied
2023-03-25T11:59:48.635004+00:00 emilia wireplumber[3944]: failed to create 
directory /home/users/athan/.local/state/wireplumber: Permission denied
2023-03-25T11:59:48.636281+00:00 emilia wireplumber[3944]: failed to create 
directory /home/users/athan/.local/state/wireplumber: Permission denied
2023-03-25T11:59:48.692700+00:00 emilia wireplumber[3944]: failed to create 
directory /home/users/athan/.local/state/wireplumber: Permission denied

Permissions are:

08:48:48 0$ ls -ld ~ ~/.local ~/.local/state ~/.local/state/wireplumber
drwx--xr-x 260 athan athan 28672 Mar 26 08:40 /home/users/athan/
drwx--   7 athan athan  4096 Feb 24 12:26 /home/users/athan/.local/
drwx--   3 athan athan  4096 Feb 24 12:26 /home/users/athan/.local/state/
drwx--   2 athan athan  4096 Mar 26 08:39 
/home/users/athan/.local/state/wireplumber/

Presumably the code in question is trying to perform these operations as
some other userid ?

As per the OP this then seems to cause a cascade of subsequent errors:

2023-03-25T11:59:48.701237+00:00 emilia wireplumber[3944]: Failed to load 
cookie file from cookie: No such file or directory
2023-03-25T11:59:48.703341+00:00 emilia pipewire[3942]: Failed to load cookie 
file from cookie: No such file or directory
2023-03-25T11:59:48.861468+00:00 emilia wireplumber[3944]: Failed to load 
cookie file from cookie: No such file or directory
2023-03-25T11:59:49.668002+00:00 emilia wireplumber[3944]: Failed to load 
cookie file from cookie: No such file or directory
2023-03-25T11:59:49.884781+00:00 emilia wireplumber[3944]: Failed to load 
cookie file from cookie: No such file or directory
-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#1033455: xscreensaver: won't unlock if libpam-cap is installed and active

2023-03-25 Thread Athanasius 
Package: xscreensaver
Version: 6.06+dfsg1-2
Severity: important

Dear Maintainer,

  TL;DR - If xscreensaver pam authentication is to work it will need to
cause adjustment to libpam-cap related configuration files.

  After upgrading my Debian desktop to bookworm I found that any attempt
to unlock xscreensaver acted as if I had failed to type the password
correctly.  It should be noted that this is an old Debian install that
has been through many different stable versions without a reinstall.
  Running xscreensaver with `-verbose` and temporarily increasing some
PAM auth logging pointed to `unix_chkpwd` being where the failure
occurred, but I couldn't easily delve any deeper there due to even root
not being able to `strace` an `xscreensaver-auth` process.

  To help diagnoise this I performed a clean bookworm install on a
separate drive and re-tested it there.  It worked!

  So, I started looking at differences between the two installs.  I used
`fvwm` on both, and aligned the `~/.xscreensaver` configurations.  Then
I checked `/etc/pam.d/`, saw I had `pam_cap.so` configured on the old
install and tried commenting that out of `/etc/pam.d/common-auth`.  But
re-testing didn't cause xscreensaver auth to work.
  It turns out at least a full restart was necessary, if not a reboot.
This was discovered when `/etc/security/capability.conf` also came to
light, with its default `none  *` line active.
  So, performed an `apt purge libpam-cap`, and rebooted before a re-test,
and then xscreensaver unlocking worked!

  The version of libpam-cap that gets installed is:

ii  libpam-cap:amd64 1:2.66-3

-- System Information:
Debian Release: 12.0
  APT prefers testing-security
  APT policy: (500, 'testing-security'), (500, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 6.1.0-6-amd64 (SMP w/16 CPU threads; PREEMPT)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), 
LANGUAGE=en_GB:en
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled

Versions of packages xscreensaver depends on:
ii  init-system-helpers  1.65.2
ii  libatk1.0-0  2.46.0-5
ii  libc62.36-8
ii  libcrypt11:4.4.33-2
ii  libglib2.0-0 2.74.6-1
ii  libgtk-3-0   3.24.37-2
ii  libpam0g 1.5.2-6
ii  libsystemd0  252.6-1
ii  libx11-6 2:1.8.4-2
ii  libxext6 2:1.3.4-1+b1
ii  libxft2  2.3.6-1
ii  libxi6   2:1.8-1+b1
ii  libxinerama1 2:1.1.4-3
ii  libxml2  2.9.14+dfsg-1.1+b3
ii  libxrandr2   2:1.5.2-2+b1
ii  libxt6   1:1.2.1-1
ii  libxxf86vm1  1:1.1.4-1+b2
ii  xscreensaver-data6.06+dfsg1-2

Versions of packages xscreensaver recommends:
ii  fonts-urw-base35  20200910-7
ii  libjpeg-turbo-progs   1:2.1.5-2
ii  perl  5.36.0-7
ii  wamerican [wordlist]  2020.12.07-2
ii  wbritish [wordlist]   2020.12.07-2
ii  xfonts-100dpi 1:1.0.5

Versions of packages xscreensaver suggests:
ii  chromium [www-browser]  111.0.5563.64-1
ii  fortune-mod [fortune]   1:1.99.1-7.3
pn  gdm3 | kdm-gdmcompat
ii  google-chrome-stable [www-browser]  111.0.5563.110-1
ii  links [www-browser] 2.28-1+b2
ii  lynx [www-browser]  2.9.0dev.12-1
pn  qcam | streamer 
ii  w3m [www-browser]   0.5.3+git20230121-2
ii  xdaliclock  2.46-1
pn  xfishtank   
ii  xscreensaver-data-extra 6.06+dfsg1-2
ii  xscreensaver-gl 6.06+dfsg1-2
ii  xscreensaver-gl-extra   6.06+dfsg1-2

-- no debconf information

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#1020835: exim4 processes get stuck with socket in FIN_WAIT2 for much longer than timeout

2022-09-27 Thread Athanasius 
ages.

exim4 suggests no packages.

-- debconf-show failed

-- 
- Athanasius (he/him) = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#986622: [Pkg-clamav-devel] Bug#986622: fixes

2021-04-21 Thread Athanasius 
  There's also the point that the current buster version is logging:

Apr 18 07:25:31 river freshclam[25421]: Sun Apr 18 07:25:31 2021 -> ^Your 
ClamAV installation is OUTDATED!
Apr 18 07:25:31 river freshclam[25421]: Sun Apr 18 07:25:31 2021 -> ^Local 
version: 0.102.4 Recommended version: 0.103.2
Apr 18 07:25:31 river freshclam[25421]: Sun Apr 18 07:25:31 2021 -> DON'T 
PANIC! Read https://www.clamav.net/documents/upgrading-clamav

once an hour.

So long as any Debian update of the packages both addresses the
outstanding CVEs *and* quiets this logging I'll be happy.
-- 
- Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#968244: [Pkg-mailman-hackers] Bug#968244: mailman3-web: Crash with: /usr/share/mailman3-web/manage.py runjobs daily

2021-01-24 Thread Athanasius 
On Sat, Jan 23, 2021 at 09:47:10PM +0100, Jonas Meurer wrote:
> Could you please try out whether this bug has been fixed with the latest
> updates to mailman3 and hyperkitty? A lot of changes happened to the
> 'import21' code, so I'm hopeful that this bug might have been fixed in the
> meantime. Unfortunately I lack the time to build a reproducer myself.

  I also filed an upstream bug about this.  We tracked down the issue as
being related to sending email addresses over an API without ensuring
they are URL encoded.  A '#' (hash) character in them causes the
receiving API server to truncate at that point, treating the '#' as a
URL anchor which only has signifigance on the browser side (when there's
actually a browser involved).

  The last I heard about that bug was 4 months ago:

<https://gitlab.com/mailman/mailman/-/issues/763#note_408737989>

Pulling the latest git master of that code, head
177c123074ece0e64fc317cd78eea1d68fec1b88 I see no changes to the
relevant code.  Indeed `git log -- src/mailmanclient/client.py` only
shows one change since '4 months ago', which is simply a copyright
end-year bump across all files.

-- 
-     Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME


signature.asc
Description: PGP signature

Bug#968371: Info received (Bug#968371: Acknowledgement (mailman3: NNTP Gateway'ing fails and causes message to not go out via email either))

2020-08-14 Thread Athanasius 
  It was only 3 commits necessary to fix this, but one of them needs
further adjustment.

  The upstream commits in question are:


59d8b47c432e43ce463cbb4a7aec649415d08702
e4d9e5202158f9b32bb985167d7dedd9e9edc263
05d608645cfa47f7150585663e9f2d6aa5688db2

But e4d9e5202158f9b32bb985167d7dedd9e9edc263 then runs into an extra
problem where the Debian INN2 doesn't like header wrapping when a post
is being made in READER mode.  This might itself be an INN2 bug.
RFC 3977 section A.1 does say that header wrapping should work over
NNTP, but perhaps READER mode counts as nn*r*p instead and follows
different rules?

So I've instead ended up using a tweaked patch that forces the
max_line_length to 'None' for no wrapping.

I've attached the three extra patch files that I'm now using.

-- 
- Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME
commit 59d8b47c432e43ce463cbb4a7aec649415d08702
Author: Mark Sapiro 
Date:   Sun Jun 23 01:52:23 2019 +

Removed the last remnants of the mailing list attribute nntp_host.

diff --git a/src/mailman/core/tests/test_pipelines.py b/src/mailman/core/tests/test_pipelines.py
index 7c87c4877..7ce778134 100644
--- a/src/mailman/core/tests/test_pipelines.py
+++ b/src/mailman/core/tests/test_pipelines.py
@@ -180,7 +180,6 @@ testing
 # Set up NNTP.
 self._mlist.gateway_to_news = True
 self._mlist.linked_newsgroup = 'testing'
-self._mlist.nntp_host = 'news.example.com'
 # Process the email.
 process(self._mlist, self._msg, {},
 pipeline_name='default-posting-pipeline')
diff --git a/src/mailman/docs/NEWS.rst b/src/mailman/docs/NEWS.rst
index ba8d7cdf9..bace916c7 100644
--- a/src/mailman/docs/NEWS.rst
+++ b/src/mailman/docs/NEWS.rst
@@ -18,6 +18,12 @@
 =
 (2019-02-22)
 
+Debian Cherry-Pick
+--
+
+* The last remnants of the mailing list attribute ``nntp_host`` have been
+  removed.  (Closes #611)
+
 Command line
 
 * The ``mailman import21`` command properly converts all acceptable_aliases
diff --git a/src/mailman/handlers/docs/nntp.rst b/src/mailman/handlers/docs/nntp.rst
index 72bcb35f0..233bbd255 100644
--- a/src/mailman/handlers/docs/nntp.rst
+++ b/src/mailman/handlers/docs/nntp.rst
@@ -44,12 +44,11 @@ Neither are digests ever gated to the newsgroup.
 []
 
 However, other posted messages get gated to the newsgroup via the nntp queue.
-The list owner can set the linked newsgroup and the nntp host that its
-messages are gated to.
+The list owner can set the linked newsgroup.  The nntp host that messages are
+gated to is a global configuration.
 ::
 
 >>> mlist.linked_newsgroup = 'comp.lang.thing'
->>> mlist.nntp_host = 'news.example.com'
 >>> handler.process(mlist, msg, {})
 >>> messages = get_queue_messages('nntp')
 >>> len(messages)
diff --git a/src/mailman/handlers/to_usenet.py b/src/mailman/handlers/to_usenet.py
index 4e66b8e99..3dbf91bf1 100644
--- a/src/mailman/handlers/to_usenet.py
+++ b/src/mailman/handlers/to_usenet.py
@@ -50,8 +50,6 @@ class ToUsenet:
 error = []
 if not mlist.linked_newsgroup:
 error.append('no newsgroup')
-if not mlist.nntp_host:
-error.append('no NNTP host')
 if error:
 log.error('NNTP gateway improperly configured: %s',
   COMMASPACE.join(error))
diff --git a/src/mailman/styles/base.py b/src/mailman/styles/base.py
index 92168ed97..e95423484 100644
--- a/src/mailman/styles/base.py
+++ b/src/mailman/styles/base.py
@@ -92,7 +92,6 @@ class BasicOperation:
 mlist.dmarc_moderation_notice = ''
 mlist.dmarc_wrapped_message_text = ''
 # NNTP gateway
-mlist.nntp_host = ''
 mlist.linked_newsgroup = ''
 mlist.gateway_to_news = False
 mlist.gateway_to_mail = False
commit e4d9e5202158f9b32bb985167d7dedd9e9edc263
Author: Mark Sapiro 
Date:   Thu Jun 27 15:52:58 2019 +

Fixed nntp runner to use BytesIO rather than StringIO.

diff --git a/src/mailman/docs/NEWS.rst b/src/mailman/docs/NEWS.rst
index 33920ec8d..2f6f041a4 100644
--- a/src/mailman/docs/NEWS.rst
+++ b/src/mailman/docs/NEWS.rst
@@ -23,6 +23,8 @@
 
 * The last remnants of the mailing list attribute ``nntp_host`` have been
   removed.  (Closes #611)
+* Fixed the nntp runner which was calling the ``nntplib.NNTP.post()`` method
+  with a string object instead of bytes.  (Closes #613)
 
 Command line
 
diff --git a/src/mailman/runners/nntp.py b/src/mailman/runners/nntp.py
index 36ef2e3a1..580d22ff9 100644
--- a/src/mailman/runners/nntp.py
+++ b/src/mailman/runners/nntp.py
@@ -23,7 +23,7 @@ import socket
 import logging
 import nntplib
 
-from io import StringIO
+from io

Bug#968371: Acknowledgement (mailman3: NNTP Gateway'ing fails and causes message to not go out via email either)

2020-08-13 Thread Athanasius 
  Having figured out how to get a test mailman3 instance up using latest
git master I can confirm the gating of mailinglist emails to NNTP is
working there.

  I guess I'll see if I can sanely git bisect to when they fixed this.

-- 
- Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#968371: mailman3: NNTP Gateway'ing fails and causes message to not go out via email either

2020-08-13 Thread Athanasius 
Package: mailman3
Version: 3.2.1-1
Severity: important

Dear Maintainer,

After importing lists from Mailman2.1 using the 'import21' module I
automatically had one list with settings to gateway to NNTP as expected.

Any email to that list with that configuration fails:

---
Aug 12 17:00:38 2020 (5717) ACCEPT: 
Aug 12 17:06:17 2020 (5717) ACCEPT: <20200812160616.e3aaorp5zhtjb...@fysh.org>
Aug 12 17:06:18 2020 (5721) fysh-annou...@fysh.org invalid FilterAction: discard
.  Treating as discard
Aug 12 17:06:18 2020 (5721) <20200812160616.e3aaorp5zhtjb...@fysh.org> discarded
 by "default-posting-pipeline" pipeline handler "mime-delete": The message's con
tent type was not explicitly allowed
Aug 12 17:06:18 2020 (5721) Uncaught runner exception: 'MailingList' object has 
no attribute 'nntp_host'
Aug 12 17:06:18 2020 (5721) Traceback (most recent call last):
  File "/usr/lib/python3/dist-packages/mailman/core/runner.py", line 173, in 
_one_iteration
self._process_one_file(msg, msgdata)
  File "/usr/lib/python3/dist-packages/mailman/core/runner.py", line 266, in 
_process_one_file
keepqueued = self._dispose(mlist, msg, msgdata)
  File "/usr/lib/python3/dist-packages/mailman/runners/pipeline.py", line 37, 
in _dispose
process(mlist, msg, msgdata, pipeline)
  File "/usr/lib/python3/dist-packages/mailman/core/pipelines.py", line 50, in 
process
handler.process(mlist, msg, msgdata)
  File "/usr/lib/python3/dist-packages/mailman/handlers/to_usenet.py", line 53, 
in process
if not mlist.nntp_host:
AttributeError: 'MailingList' object has no attribute 'nntp_host'
---

This immediate problem is addressed upstream by:

59d8b47c432e43ce463cbb4a7aec649415d08702

Applying just the change to remove that conditional in to_usenet.py gets
past that point (and messages now make it out in email), but then the
code runs into another problem:

---
Aug 13 17:44:26 2020 (15817) ACCEPT: <20200813164426.ga15...@fysh.org>
Aug 13 17:44:27 2020 (15814) HyperKitty archived message <20200813164426.GA15940
@fysh.org> to https://www.fysh.org/mailman3/hyperkitty/list/testl...@fysh.org/me
ssage/OMXF2QRUZLZPO3WFZMXDBSW4QZTWO5HJ/
Aug 13 17:44:28 2020 (15819) <159733706828.15819.11534859148741658...@river.fysh
.org> NNTP unexpected exception for testl...@fysh.org
Traceback (most recent call last):
  File "/usr/lib/python3/dist-packages/mailman/runners/nntp.py", line 77, in _di
spose
conn.post(fp)
  File "/usr/lib/python3.7/nntplib.py", line 916, in post
return self._post('POST', data)
  File "/usr/lib/python3.7/nntplib.py", line 902, in _post
if not line.endswith(_CRLF):
TypeError: endswith first arg must be str or a tuple of str, not bytes
---

Which appears to either be an error in nntplib, or a mis-use of it.

I'm running this on a production machine so can't easily play with
running mailman3 from source to see if they've fixed this upstream.  I
might find time to try it out on another machine.

-- System Information:
Debian Release: 10.5
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.19.138-fysh-kvmguest (SMP w/8 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), 
LANGUAGE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages mailman3 depends on:
ii  dbconfig-sqlite32.0.11+deb10u1
ii  debconf [debconf-2.0]   1.5.71
ii  logrotate   3.14.0-4
ii  lsb-base10.2019051400
ii  python3 3.7.3-1
ii  python3-aiosmtpd1.2-3
ii  python3-alembic 1.0.0-3
ii  python3-click   7.0-1
ii  python3-dnspython   1.16.0-1
ii  python3-falcon  1.0.0-2+b3
ii  python3-flufl.bounce3.0-1
ii  python3-flufl.i18n  2.0.1-1
ii  python3-flufl.lock  3.2-1
ii  python3-lazr.config 2.2-1
ii  python3-passlib 1.7.1-1
ii  python3-psycopg22.7.7-1
ii  python3-public  0.5-1
ii  python3-requests2.21.0-1
ii  python3-sqlalchemy  1.2.18+ds1-2
ii  python3-zope.component  4.3.0-1
ii  python3-zope.configuration  4.0.3-3
ii  python3-zope.event  4.2.0-1
ii  python3-zope.interface  4.3.2-1+b2
ii  ucf 3.0038+nmu1

Versions of packages mailman3 recommends:
ii  exim4-daemon-heavy [mail-transport-agent]  4.92-8+deb10u4

Versions of packages mailman3 suggests:
ii  chromium [www-browser]  83.0.4103.116-1~deb10u3
ii  elinks [www-browser]0.13~20190125-3
ii  links [www-browser] 2.18-2
ii  links2 [www-browser]

Bug#968244: mailman3-web: Crash with: /usr/share/mailman3-web/manage.py runjobs daily

2020-08-11 Thread Athanasius 
Package: mailman3-web
Version: 0+20180916-8
Severity: normal

Dear Maintainer,

After installing mailman3 I used migrated old Mailman 2.1 lists in using
'import21' and then their archives using 'python manage.py
hyperkitty_import -l ...'.

All seems to have gone OK, but the daily mailman3-web job in
/etc/cron.d/mailman3-web:

@daily www-data [ -f /usr/bin/django-admin ] && flock -n 
/var/run/mailman3-web/cron.daily /usr/share/mailman3-web/manage.py runjobs daily

is throwing errors:

Traceback (most recent call last):
  File 
"/usr/lib/python3/dist-packages/django_extensions/management/commands/runjobs.py",
 line 36, in runjobs
job().execute()
  File "/usr/lib/python3/dist-packages/hyperkitty/jobs/sync_mailman.py", line 
35, in execute
sync_with_mailman()
  File "/usr/lib/python3/dist-packages/hyperkitty/lib/mailman.py", line 145, in 
sync_with_mailman
sender.set_mailman_id()
  File "/usr/lib/python3/dist-packages/hyperkitty/models/sender.py", line 54, 
in set_mailman_id
mm_user = client.get_user(self.address)
  File "/usr/lib/python3/dist-packages/mailmanclient/client.py", line 322, in 
get_user
return User(self._connection, content['self_link'], content)
KeyError: 'self_link'
ERROR OCCURED IN DAILY JOB: sync_mailman (APP: hyperkitty)
START TRACEBACK:
END TRACEBACK

Adding a little debug output:

Address: #.pleasegivegenerou...@hotmail.com
Response: {'date': 'Tue, 11 Aug 2020 14:35:42 GMT', 'server': 'WSGIServer/0.2 
CPython/3.7.3', 'content-length': '712742', 'content-type': 'application/json; 
charset=UTF-8', 'status': '200', 'content-location': 
'http://river-nat:8001/3.0/users/#.pleasegivegenerou...@hotmail.com'}

And the 'content' at that point is *huge* (693KiB), indeed without a
'self_link' key at its top level, although it contains an 'entries'
array in which each member has a 'self_link' key.

This appears to be code to sync up hyperkitty's idea of senders with
actual members of the list, so is probably relatively harmless, but I
don't know if *subsequent* more important code is being prevented from
running.

-- System Information:
Debian Release: 10.5
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.19.138-fysh-kvmguest (SMP w/8 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), 
LANGUAGE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages mailman3-web depends on:
ii  dbconfig-sqlite3   2.0.11+deb10u1
ii  debconf [debconf-2.0]  1.5.71
ii  lsb-base   10.2019051400
ii  node-less  1.6.3~dfsg-3
ii  python33.7.3-1
ii  python3-django-hyperkitty  1.2.2-1
ii  python3-django-postorius   1.2.4-1
ii  python3-psycopg2   2.7.7-1
ii  python3-whoosh 2.7.4+git6-g9134ad92-4
ii  sassc  3.5.0-1
ii  ucf3.0038+nmu1
ii  uwsgi  2.0.18-1
ii  uwsgi-plugin-python3   2.0.18-1

Versions of packages mailman3-web recommends:
pn  libapache2-mod-proxy-uwsgi | nginx  

Versions of packages mailman3-web suggests:
ii  mariadb-server-10.3 [virtual-mysql-server]  1:10.3.23-0+deb10u1

-- debconf information:
  mailman3-web/upgrade-error: abort
  mailman3-web/superuser-name: admin
  mailman3-web/pgsql/authmethod-user: password
  mailman3-web/restart-webserver: true
  mailman3-web/nginx-choice:
  mailman3-web/missing-db-package-error: abort
  mailman3-web/mysql/method: Unix socket
  mailman3-web/configure-webserver: none
  mailman3-web/dbconfig-upgrade: true
  mailman3-web/install-error: abort
  mailman3-web/pgsql/admin-user: postgres
  mailman3-web/dbconfig-remove: true
  mailman3-web/pgsql/changeconf: false
  mailman3-web/upgrade-backup: true
  mailman3-web/db/dbname: mailman3web
  mailman3-web/internal/skip-preseed: false
  mailman3-web/purge: false
  mailman3-web/remote/port:
  mailman3-web/superuser-mail: root@localhost
  mailman3-web/db/app-user: mailman3web@localhost
* mailman3-web/dbconfig-install: false
  mailman3-web/remove-error: abort
  mailman3-web/pgsql/manualconf:
  mailman3-web/pgsql/method: TCP/IP
  mailman3-web/remote/host: localhost
  mailman3-web/emailname: fysh.org
  mailman3-web/mysql/admin-user:
  mailman3-web/remote/newhost:
  mailman3-web/dbconfig-reinstall: false
  mailman3-web/passwords-do-not-match:
  mailman3-web/db/basepath: /var/lib/mailman3/web
  mailman3-web/pgsql/authmethod-admin: ident
  mailman3-web/internal/reconfiguring: false
  mailman3-web/pgsql/no-empty-passwords:
  mailman3-web/database-type:

Bug#859388: Issue still present in buster with dmsetup 1.02.155-3

2019-10-12 Thread Athanasius 
Oct 12 14:53:49 tuesday blkdeactivate[5119]: Deactivating block devices:
Oct 12 14:53:49 tuesday blkdeactivate[5119]: /sbin/blkdeactivate: line 345: 
/bin/sort: No such file or directory

16:22:36 0$ grep sort /sbin/blkdeactivate 
SORT_MNT="/bin/sort -r -u -k 4"
16:24:07 0$ ls -l /bin/sort /usr/bin/sort
ls: cannot access '/bin/sort': No such file or directory
-rwxr-xr-x 1 root root 114120 Feb 28  2019 /usr/bin/sort
-- 
- Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#888765: Also affects compose-addressbook and others

2018-02-13 Thread Athanasius 
  I'm also having to fix things with a symbolic link, so far for both:

compose-addressbook
keyboard-shortcuts

I'm only just setting up roundcube so may find others.  I can see
/var/lib/roundcube/plugins/ has also (with a - in them):

message-highlight
thunderbird-labels
virtuser_query

All five of these use _ in the main .php filename.
-- 
- Athanasius = Athanasius(at)miggy.org / https://miggy.org/
  GPG/PGP Key: https://miggy.org/gpg-key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME

Bug#872036: Acknowledgement (AH00060: seg fault or similar nasty error detected in the parent process)

2017-11-08 Thread Athanasius 
  mpm-event based apache has now survived two nightly runs of logrotate
doing '/etc/init.d/apache2 reload' since I ensured it will only do that
once per run.

  So what is it about mpm-event that causes apache2 to crash when
multiple '/etc/init.d/apache2 reload' invocations happen in very quick
succession ?  Smells like an upstream bug, yes ?
  Not that I can seem to reproduce this with:

COUNT=1 ; while [ $COUNT -lt 80 ] ; do ( /etc/init.d/apache2 reload &); 
COUNT=`expr $COUNT + 1` ; done

  Perhaps it's not just the rapid reloads, but combining those with
needing to create new log files ?

Although it does leave me with only processes like:

/usr/sbin/apache2 -k graceful

running, rather than the more usual:

/usr/sbin/apache2 -k start

And some instances of the init.d script complaining "Apache2 is not
running" on subsequent invocations.

Bug#872036: Acknowledgement (AH00060: seg fault or similar nasty error detected in the parent process)

2017-11-06 Thread Athanasius 
  Apparently this can be caused by something like logrotate sending many
"reload" signals to apache in quick succession:



I experienced such an apache crash this morning when the logs were
rotated.  I've now adjusted my logrotate config to only reload apache
once (despite many sections for different web site's logs, some of which
have specifically different user/group settings for the files) and will
see if that avoids another crash tomorrow.

Bug#792030: systemd does not reliably mount nfs shares

2017-03-15 Thread Athanasius 
On Wed, Mar 15, 2017 at 10:15:54PM +0100, Michael Biebl wrote:
> Am 15.03.2017 um 16:00 schrieb Athanasius:
> >   The other issue is that my two NFS mounts from my server (still on
> > jessie) were failing at boot-time.  After I fixed the /proc/bus/usb
> > issue (by commenting that line out) the next boot had my NFS mounts work
> > OK.  But today they've failed again.
> > 
> > The two lines in /etc/fstab are:
> > 
> > 192.168.1.1:/home/users /home/users nfs 
> > defaults,soft,intr,timeo=50,retrans=6,fg,nfsvers=3,nosuid,nodev,nofail 0 0
> > 192.168.1.1:/mnt/tuesday/data0 /mnt/tuesday/data0 
> > nfs 
> > defaults,soft,intr,timeo=50,retrans=6,fg,nfsvers=3,nosuid,nodev,nofail 0 0
> > 
> > and these always worked 100% on jessie by the time I'd typed in my
> > password.  In stretch, due to /home/users not being properly mounted I
> > could not complete a login.
> > 
> 
> > # The primary network interface
> > allow-hotplug eth0
> 
> Use "auto eth0", that should work better in a case like yours.
> 
> NFS mounts rely on network-online.target being hooked up properly. For
> allow-hotplug interfaces this is not guaranteed.

  I'll give that a go.  The reason for me using 'allow-hotplug' is now
lost in the mists of time, as it was already this way when I first put
/etc on that machine under git control in 2011.
  Of course the last two boots with 'allow-hotplug' also worked, so I'll
report back *if* I have the issue again with it changed to 'auto'.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME


signature.asc
Description: Digital signature

Bug#792030: systemd does not reliably mount nfs shares

2017-03-15 Thread Athanasius 
_COMM=systemd
_EXE=/lib/systemd/systemd
_CMDLINE=/sbin/init
_CAP_EFFECTIVE=3f
_SYSTEMD_CGROUP=/init.scope
_SYSTEMD_UNIT=init.scope
_SYSTEMD_SLICE=-.slice
CODE_FILE=../src/core/job.c
CODE_LINE=804
CODE_FUNCTION=job_log_status_message
PRIORITY=3
UNIT=home-users.mount
_BOOT_ID=23bab7faa65e4b4495c296eaefd3e371
MESSAGE_ID=be02cf6855d2428ba40df7e9d022f03d
RESULT=failed
MESSAGE=Failed to mount /home/users.
_SOURCE_REALTIME_TIMESTAMP=1489586455606040
Wed 2017-03-15 14:00:55.606078 GMT 
[s=df4dbfefa1b44bd583663f95991c121e;i=25d921;b=23bab7faa65e4b4495c296eaefd3e371;m=2a0328;t=54ac55f8c73ee;x=2a60bfcc6bf98719]
SYSLOG_FACILITY=3
_MACHINE_ID=6cb34bbd53edf1ffafba70550009
_HOSTNAME=emilia
SYSLOG_IDENTIFIER=systemd
_TRANSPORT=journal
_PID=1
_UID=0
_GID=0
_COMM=systemd
_EXE=/lib/systemd/systemd
_CMDLINE=/sbin/init
_CAP_EFFECTIVE=3f
_SYSTEMD_CGROUP=/init.scope
_SYSTEMD_UNIT=init.scope
_SYSTEMD_SLICE=-.slice
PRIORITY=5
CODE_FILE=../src/core/unit.c
CODE_LINE=2039
CODE_FUNCTION=unit_notify
UNIT=home-users.mount
_BOOT_ID=23bab7faa65e4b4495c296eaefd3e371
MESSAGE=home-users.mount: Unit entered failed state.
_SOURCE_REALTIME_TIMESTAMP=1489586455606078
Wed 2017-03-15 14:02:06.856126 GMT 
[s=df4dbfefa1b44bd583663f95991c121e;i=25dd80;b=23bab7faa65e4b4495c296eaefd3e371;m=467248b;t=54ac563cba3e1;x=fce824f8054443b9]
SYSLOG_FACILITY=3
_MACHINE_ID=6cb34bbd53edf1ffafba70550009
_HOSTNAME=emilia
SYSLOG_IDENTIFIER=systemd
_TRANSPORT=journal
_PID=1
_UID=0
_GID=0
_COMM=systemd
_EXE=/lib/systemd/systemd
_CMDLINE=/sbin/init
_CAP_EFFECTIVE=3f
_SYSTEMD_CGROUP=/init.scope
_SYSTEMD_UNIT=init.scope
_SYSTEMD_SLICE=-.slice
PRIORITY=5
CODE_FILE=../src/core/unit.c
UNIT=home-users.mount
_BOOT_ID=23bab7faa65e4b4495c296eaefd3e371
CODE_LINE=4012
CODE_FUNCTION=unit_warn_if_dir_nonempty
MESSAGE_ID=1dee0369c7fc4736b7099b38ecb46ee7
MESSAGE=home-users.mount: Directory /home/users to mount over is not empty, 
mounting anyway.
WHERE=/home/users
_SOURCE_REALTIME_TIMESTAMP=1489586526856126
Wed 2017-03-15 14:02:06.857479 GMT 
[s=df4dbfefa1b44bd583663f95991c121e;i=25dd81;b=23bab7faa65e4b4495c296eaefd3e371;m=46729c3;t=54ac563cba91b;x=b94c186e3ec88c2e]
PRIORITY=6
SYSLOG_FACILITY=3
_MACHINE_ID=6cb34bbd53edf1ffafba70550009
_HOSTNAME=emilia
SYSLOG_IDENTIFIER=systemd
_TRANSPORT=journal
_PID=1
_UID=0
_GID=0
_COMM=systemd
_EXE=/lib/systemd/systemd
_CMDLINE=/sbin/init
_CAP_EFFECTIVE=3f
_SYSTEMD_CGROUP=/init.scope
_SYSTEMD_UNIT=init.scope
_SYSTEMD_SLICE=-.slice
CODE_FILE=../src/core/unit.c
CODE_LINE=1493
CODE_FUNCTION=unit_status_log_starting_stopping_reloading
MESSAGE_ID=7d4958e842da4a758f6c1cdc7b36dcc5
UNIT=home-users.mount
_BOOT_ID=23bab7faa65e4b4495c296eaefd3e371
MESSAGE=Mounting /home/users...
_SOURCE_REALTIME_TIMESTAMP=1489586526857479
Wed 2017-03-15 14:02:06.908838 GMT 
[s=df4dbfefa1b44bd583663f95991c121e;i=25dd84;b=23bab7faa65e4b4495c296eaefd3e371;m=467f25a;t=54ac563cc71b2;x=315a3b1d75bcd231]
PRIORITY=6
SYSLOG_FACILITY=3
_MACHINE_ID=6cb34bbd53edf1ffafba70550009
_HOSTNAME=emilia
SYSLOG_IDENTIFIER=systemd
_TRANSPORT=journal
_PID=1
_UID=0
_GID=0
_COMM=systemd
_EXE=/lib/systemd/systemd
_CMDLINE=/sbin/init
_CAP_EFFECTIVE=3f
_SYSTEMD_CGROUP=/init.scope
_SYSTEMD_UNIT=init.scope
_SYSTEMD_SLICE=-.slice
CODE_FILE=../src/core/job.c
CODE_LINE=804
CODE_FUNCTION=job_log_status_message
MESSAGE_ID=39f53479d3a045ac8e11786248231fbf
RESULT=done
UNIT=home-users.mount
_BOOT_ID=23bab7faa65e4b4495c296eaefd3e371
MESSAGE=Mounted /home/users.
_SOURCE_REALTIME_TIMESTAMP=1489586526908838

I then switched to a text console, logged in as root, and ran 'systemctl
start home-users.mount' (and similar for mnt-tuesday-data0.mount) and
that worked without issues.  So there's obviously some race
condition/lack of interlock with something else being finished starting.

  If further information is required I'll capture it when it's in the
"failed to mount automatically during boot" state, before I mount
manually.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   "And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence." Paula Cole - ME


signature.asc
Description: Digital signature

Bug#816286: lsof: Latest stable kernels issue WARN_ON for some uses

2016-02-29 Thread Athanasius 
Package: lsof
Version: 4.86+dfsg-1
Severity: minor

Dear Maintainer,

  I just updated a web server to latest 3.14 stable kernel, 3.14.62,
which includes a patch tightening up ptrace permissions.

  Now the use of "/usr/bin/lsof -w -l +d /var/lib/php5" by
/etc/cron.d/php5 to clean up sessions causes a kernel WARN_ON to be
spammed to the log.

Feb 29 13:36:14 www kernel: [48425.679328] [ cut here ]
Feb 29 13:36:14 www kernel: [48425.679753] WARNING: CPU: 0 PID: 13503 at 
kernel/ptrace.c:233 __ptrace_may_access+0x13a/0x150()
Feb 29 13:36:14 www kernel: [48425.680540] denying ptrace access check without 
PTRACE_MODE_*CREDS
Feb 29 13:36:14 www kernel: [48425.681090] Modules linked in: nfsv3 nfsd 
auth_rpcgss oid_registry nfs_acl nfs lockd sunrpc ipv6 xt_nat iptable_nat 
nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_filter 
ip_tables floppy virtio_balloon virtio_net
Feb 29 13:36:14 www kernel: [48425.683176] CPU: 0 PID: 13503 Comm: lsof 
Tainted: GW3.14.62-fysh-kvmguest #6
Feb 29 13:36:14 www kernel: [48425.683891] Hardware name: Bochs Bochs, BIOS 
Bochs 01/01/2007
Feb 29 13:36:14 www kernel: [48425.684415]  0286  
8850ade4 0007
Feb 29 13:36:14 www kernel: [48425.684536]  88051bf1fd88 0009 
8803fe81 0001
Feb 29 13:36:14 www kernel: [48425.684536]  8805fd8b50a0 0001 
 7ffa0012ee50
Feb 29 13:36:14 www kernel: [48425.684536] Call Trace:
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
dump_stack+0x5e/0x7a
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
warn_slowpath_common+0x81/0xb0
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
warn_slowpath_fmt+0x45/0x50
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
__ptrace_may_access+0x13a/0x150
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
ptrace_may_access+0x32/0x60
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
mm_access+0x7d/0xc0
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
m_start+0x78/0x1e0
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
do_filp_open+0x47/0xb0
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
seq_read+0x10f/0x380
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
vfs_read+0xa5/0x180
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
SyS_read+0x4b/0xb0
Feb 29 13:36:14 www kernel: [48425.684536]  [] ? 
system_call_fastpath+0x1a/0x1f
Feb 29 13:36:14 www kernel: [48425.692738] ---[ end trace f8d6c7c51a5bfc2a ]---

  Some 58 times for a single run of that command!

  I'm guessing that the version of lsof in wheezy doesn't do some
credentials setting that would prevent this WARN_ON.  Yes, I'm aware I
should really get around to upgrading this server to Jessie.

-- System Information:
Debian Release: 7.9
  APT prefers oldstable
  APT policy: (500, 'oldstable')
Architecture: amd64 (x86_64)

Kernel: Linux 3.14.62 (SMP w/8 CPU cores)
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 lsof depends on:
ii  libc6  2.13-38+deb7u10
ii  perl   5.14.2-21+deb7u2
ii  perl-modules [libperl4-corelibs-perl]  5.14.2-21+deb7u2

lsof recommends no packages.

lsof suggests no packages.

-- debconf-show failed

Bug#678921: Info received (Bug#678921: Info received (Bug#678921: Info received (STILL broken in 3.5-2+b2 in Jessie)))

2016-02-25 Thread Athanasius 
  For what it's worth, on checking the latest audacious-plugin source
from upstream,

it looks like they fixed this differently.
  I see both sid and stretch only have 3.6.2-based packages, but it
looks to have already been fixed in the same manner as 3.7.1 in that
version.

Bug#678921: Info received (Bug#678921: Info received (STILL broken in 3.5-2+b2 in Jessie))

2016-02-24 Thread Athanasius 
  Actually, use this patch instead.  It will also fire the first time it
sees a new title, even if the previous saved one was empty.  The
previous patch would need *two* changes of title before it would fire
the script, one to fill in the previous, and then another to be
different to that.

  I'd definitely push this upstream if possible.  I really don't know
why this has remained unfixed for so long, I guess no audacious dev
cared about this bit of code.

diff -ur audacious-plugins-3.5/src/song_change/song_change.c 
new-audacious-plugins-3.5/src/song_change/song_change.c
--- audacious-plugins-3.5/src/song_change/song_change.c 2014-04-23 
03:37:46.0 +0100
+++ new-audacious-plugins-3.5/src/song_change/song_change.c 2016-02-25 
00:18:01.691451152 +
@@ -32,7 +32,7 @@
 static void songchange_playback_begin(gpointer unused, gpointer unused2);
 static void songchange_playback_end(gpointer unused, gpointer unused2);
 static void songchange_playlist_eof(gpointer unused, gpointer unused2);
-//static void songchange_playback_ttc(gpointer, gpointer);
+static void songchange_playback_ttc(gpointer, gpointer);
 
 typedef struct
 {
@@ -252,7 +252,7 @@
 hook_dissociate("playback begin", songchange_playback_begin);
 hook_dissociate("playback end", songchange_playback_end);
 hook_dissociate("playlist end reached", songchange_playlist_eof);
-// hook_dissociate( "playlist set info" , songchange_playback_ttc);
+hook_dissociate("title change" , songchange_playback_ttc);
 
 if ( ttc_prevs != NULL )
 {
@@ -320,7 +320,7 @@
 ttc_prevs = g_malloc0(sizeof(songchange_playback_ttc_prevs_t));
 ttc_prevs->title = NULL;
 ttc_prevs->filename = NULL;
-// hook_associate( "playlist set info" , songchange_playback_ttc , 
ttc_prevs );
+hook_associate("title change" , songchange_playback_ttc , ttc_prevs );
 
 return TRUE;
 }
@@ -335,53 +335,58 @@
 do_command (cmd_line_after);
 }
 
-#if 0
 static void
 songchange_playback_ttc(gpointer plentry_gp, gpointer prevs_gp)
 {
-if ( ( aud_ip_state->playing ) && ( strcmp(cmd_line_ttc,"") ) )
+if ( aud_drct_get_playing() && ( strcmp(cmd_line_ttc,"") ) )
 {
 songchange_playback_ttc_prevs_t *prevs = prevs_gp;
-PlaylistEntry *pl_entry = plentry_gp;
+   int playlist = aud_playlist_get_playing ();
+   int pl_entry = aud_playlist_get_position(playlist);
+   gchar *pl_entry_filename = aud_playlist_entry_get_filename(playlist, 
pl_entry);
+   gchar *pl_entry_title = aud_playlist_entry_get_title(playlist, 
pl_entry, FALSE);
 
 /* same filename but title changed, useful to detect http stream song 
changes */
 
 if ( ( prevs->title != NULL ) && ( prevs->filename != NULL ) )
 {
-if ( ( pl_entry->filename != NULL ) && ( 
!strcmp(pl_entry->filename,prevs->filename) ) )
+if ( ( pl_entry_filename != NULL ) && ( 
!strcmp(pl_entry_filename,prevs->filename) ) )
 {
-if ( ( pl_entry->title != NULL ) && ( 
strcmp(pl_entry->title,prevs->title) ) )
+if ( ( pl_entry_title != NULL ) && ( 
strcmp(pl_entry_title,prevs->title) ) )
 {
-int pos = aud_drct_pl_get_pos();
-char *current_file = aud_drct_pl_get_file(pos);
-do_command(cmd_line_ttc, current_file, pos);
-g_free(current_file);
+do_command(cmd_line_ttc);
 g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
+prevs->title = g_strdup(pl_entry_title);
 }
 }
 else
 {
 g_free(prevs->filename);
-prevs->filename = g_strdup(pl_entry->filename);
+prevs->filename = g_strdup(pl_entry_filename);
 /* if filename changes, reset title as well */
 if ( prevs->title != NULL )
 g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
+prevs->title = g_strdup(pl_entry_title);
 }
 }
 else
 {
-if ( prevs->title != NULL )
-g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
-if ( prevs->filename != NULL )
-g_free(prevs->filename);
-prevs->filename = g_strdup(pl_entry->filename);
+if ( pl_entry_filename != NULL &&  pl_entry_title != NULL )
+{
+   do_command(cmd_line_ttc);
+   if ( prevs->title != NULL )
+g_free(prevs->title);
+   prevs->title = g_strdup(pl_entry_title);
+   if ( prevs->filename != NULL )
+g_free(prevs->filename);
+   prevs->filename = g_strdup(pl_entry_filename);
+   }
 }
+
+   str_unref (pl_entry_filename);
+   str_unref

Bug#678921: Info received (STILL broken in 3.5-2+b2 in Jessie)

2016-02-24 Thread Athanasius 
  Here, have a patch that gets this working again.  As soon as I'd
nailed semantics of a the function calls versus g_free() and str_unref()
it's been working 100% here.

--- audacious-plugins-3.5/src/song_change/song_change.c 2014-04-23 
03:37:46.0 +0100
+++ new-audacious-plugins-3.5/src/song_change/song_change.c 2016-02-25 
00:05:31.937064865 +
@@ -32,7 +32,7 @@
 static void songchange_playback_begin(gpointer unused, gpointer unused2);
 static void songchange_playback_end(gpointer unused, gpointer unused2);
 static void songchange_playlist_eof(gpointer unused, gpointer unused2);
-//static void songchange_playback_ttc(gpointer, gpointer);
+static void songchange_playback_ttc(gpointer, gpointer);
 
 typedef struct
 {
@@ -252,7 +252,7 @@
 hook_dissociate("playback begin", songchange_playback_begin);
 hook_dissociate("playback end", songchange_playback_end);
 hook_dissociate("playlist end reached", songchange_playlist_eof);
-// hook_dissociate( "playlist set info" , songchange_playback_ttc);
+hook_dissociate("title change" , songchange_playback_ttc);
 
 if ( ttc_prevs != NULL )
 {
@@ -320,7 +320,7 @@
 ttc_prevs = g_malloc0(sizeof(songchange_playback_ttc_prevs_t));
 ttc_prevs->title = NULL;
 ttc_prevs->filename = NULL;
-// hook_associate( "playlist set info" , songchange_playback_ttc , 
ttc_prevs );
+hook_associate("title change" , songchange_playback_ttc , ttc_prevs );
 
 return TRUE;
 }
@@ -335,53 +335,54 @@
 do_command (cmd_line_after);
 }
 
-#if 0
 static void
 songchange_playback_ttc(gpointer plentry_gp, gpointer prevs_gp)
 {
-if ( ( aud_ip_state->playing ) && ( strcmp(cmd_line_ttc,"") ) )
+if ( aud_drct_get_playing() && ( strcmp(cmd_line_ttc,"") ) )
 {
 songchange_playback_ttc_prevs_t *prevs = prevs_gp;
-PlaylistEntry *pl_entry = plentry_gp;
+   int playlist = aud_playlist_get_playing ();
+   int pl_entry = aud_playlist_get_position(playlist);
+   gchar *pl_entry_filename = aud_playlist_entry_get_filename(playlist, 
pl_entry);
+   gchar *pl_entry_title = aud_playlist_entry_get_title(playlist, 
pl_entry, FALSE);
 
 /* same filename but title changed, useful to detect http stream song 
changes */
 
 if ( ( prevs->title != NULL ) && ( prevs->filename != NULL ) )
 {
-if ( ( pl_entry->filename != NULL ) && ( 
!strcmp(pl_entry->filename,prevs->filename) ) )
+if ( ( pl_entry_filename != NULL ) && ( 
!strcmp(pl_entry_filename,prevs->filename) ) )
 {
-if ( ( pl_entry->title != NULL ) && ( 
strcmp(pl_entry->title,prevs->title) ) )
+if ( ( pl_entry_title != NULL ) && ( 
strcmp(pl_entry_title,prevs->title) ) )
 {
-int pos = aud_drct_pl_get_pos();
-char *current_file = aud_drct_pl_get_file(pos);
-do_command(cmd_line_ttc, current_file, pos);
-g_free(current_file);
+do_command(cmd_line_ttc);
 g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
+prevs->title = g_strdup(pl_entry_title);
 }
 }
 else
 {
 g_free(prevs->filename);
-prevs->filename = g_strdup(pl_entry->filename);
+prevs->filename = g_strdup(pl_entry_filename);
 /* if filename changes, reset title as well */
 if ( prevs->title != NULL )
 g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
+prevs->title = g_strdup(pl_entry_title);
 }
 }
 else
 {
 if ( prevs->title != NULL )
 g_free(prevs->title);
-prevs->title = g_strdup(pl_entry->title);
+prevs->title = g_strdup(pl_entry_title);
 if ( prevs->filename != NULL )
 g_free(prevs->filename);
-prevs->filename = g_strdup(pl_entry->filename);
+prevs->filename = g_strdup(pl_entry_filename);
 }
+
+   str_unref (pl_entry_filename);
+   str_unref (pl_entry_title);
 }
 }
-#endif
 
 static void songchange_playlist_eof(gpointer unused, gpointer unused2)
 {

Bug#678921: STILL broken in 3.5-2+b2 in Jessie

2016-02-24 Thread Athanasius 
  This is still broken.  Looking at
 where I saw someone
had posted this bug it may well be that a section of code that is #if
0'd out due to not being updated for some other changes.

  It would be great to get this updated.  All I wanted to do was log
song titles on a stream so as to track down the artist and title of a
song I've been loving.  It just started playing though so my immediate
need is now moot ;) .

Bug#787436: audacious-plugins: Global Hotkeys only work if focus in same display as audacious

2015-06-01 Thread Athanasius 
Package: audacious-plugins
Version: 3.5-1+b2
Severity: important

Dear Maintainer,

   I have just upgraded the system in question from Debian Wheezy to
Debian Jessie.  In doing so I did allow systemd, including its init
system, to be installed.

   On Wheezy the Global Hotkeys plugin would detect my media keys
(Logitech G110, showing them as XF86AudioPrev and the like) no matter
which of my dual monitors I had focus in.  Now under Jessie it will
only detect them if the mouse pointer is in the same monitor that
Audacious displays on.

  I use the Nvidia binary drivers, via the Debian nvidia-drivers package
for the X display, and it's configured as two separate displays (no
Xinerama).  I'm using fvwm as my window manager, started from
~/.Xsession, which will have been triggered by gdm3.


-- System Information:
Debian Release: 8.0
  APT prefers stable
  APT policy: (990, 'stable'), (500, 'stable-updates')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 3.16.0-4-amd64 (SMP w/8 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages audacious-plugins depends on:
ii  audacious-plugins-data3.5-1
ii  libasound21.0.28-1
ii  libatk1.0-0   2.14.0-1
ii  libaudcore2   3.5-2
ii  libavcodec56  6:11.3-1
ii  libavformat56 6:11.3-1
ii  libavutil54   6:11.3-1
ii  libbinio1ldbl 1.4+dfsg1-4
ii  libbs2b0  3.1.0+dfsg-2.1
ii  libc6 2.19-18
ii  libcairo-gobject2 1.14.0-2.1
ii  libcairo2 1.14.0-2.1
ii  libcddb2  1.3.2-5
ii  libcdio-cdda1 0.83-4.2
ii  libcdio13 0.83-4.2
ii  libcue1   1.4.0-1
ii  libcurl3-gnutls   7.38.0-4+deb8u2
ii  libdbus-1-3   1.8.16-1
ii  libdbus-glib-1-2  0.102-1
ii  libfaad2  2.7-8
ii  libflac8  1.3.0-3
ii  libfluidsynth11.1.6-2
ii  libgcc1   1:4.9.2-10
ii  libgdk-pixbuf2.0-02.31.1-2+b1
ii  libgl1-mesa-glx [libgl1]  10.3.2-1
ii  libglib2.0-0  2.42.1-1
ii  libgtk-3-03.14.5-1
ii  libjack-jackd2-0 [libjack-0.116]  1.9.10+20140719git3eb0ae6a~dfsg-2
ii  liblircclient00.9.0~pre1-1.2
ii  libmms0   0.6.2-4
ii  libmodplug1   1:0.8.8.4-4.1+b1
ii  libmp3lame0   3.99.5+repack1-7
ii  libmpg123-0   1.20.1-2
ii  libneon27-gnutls  0.30.1-1
ii  libnotify40.7.6-2
ii  libogg0   1.3.2-1
ii  libpango-1.0-01.36.8-3
ii  libpangocairo-1.0-0   1.36.8-3
ii  libpulse0 5.0-13
ii  libsamplerate00.1.8-8
ii  libsdl1.2debian   1.2.15-10+b1
ii  libsidplayfp3 1.6.0-1
ii  libsndfile1   1.0.25-9.1
ii  libstdc++64.9.2-10
ii  libvorbis0a   1.3.4-2
ii  libvorbisenc2 1.3.4-2
ii  libvorbisfile31.3.4-2
ii  libwavpack1   4.70.0-1
ii  libx11-6  2:1.6.2-3
ii  libxcomposite11:0.4.4-1
ii  libxml2   2.9.1+dfsg1-5
ii  libxrender1   1:0.9.8-1+b1
ii  multiarch-support 2.19-18
ii  zlib1g1:1.2.8.dfsg-2+b1

Versions of packages audacious-plugins recommends:
ii  audacious  3.5-2

audacious-plugins suggests no packages.

-- no debconf information
Section ServerLayout
Identifier  Athan's manual config
Screen  0 Screen0 0 0
Screen  1 Screen1 RightOf Screen0
InputDevice Mouse0 CorePointer
InputDevice Keyboard0 CoreKeyboard
EndSection

Section Module
#   Load   glx
#   Load   dbe
#   Load   extmod
#   Load  pex5
#   Load   record
#   Disable  xie
#   Load   freetype
#   Load   bitmap
#   Disable  type1
Load   glx
#   Disable  xtt
EndSection

Section ServerFlags
Option  StandbyTime 1800
Option  SuspendTime 3600
Option  OffTime 7200
Option  Xinerama false
#Option AllowEmptyInput false
EndSection

Section InputDevice
Identifier  Keyboard0
Driver  kbd
Option  AutoRepeat 500 30
Option  XkbRules xfree86
#   Option  XkbModel pc104
Option

Bug#524610: Where is 1.2.2-5 ?

2011-04-11 Thread Athanasius 
  Squeeze still shows 1.2.2-4 as current and sid is based on 1.2.3 now.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#596172: openswan: pluto erroneously adds host route for IP with valid

2011-02-14 Thread Athanasius 
On Fri, Feb 11, 2011 at 11:14:12PM +0100, Harald Jenny wrote:
 could you please try the openswan 2.6 version from lenny-backports and see if
 the problem got solved there?

  Yes, that appears to be working perfectly.  A simple:

aptitude -t lenny-backports install openswan

telling it to keep my current config and it's working.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#596172: openswan: pluto erroneously adds host route for IP with valid

2011-02-14 Thread Athanasius 
On Mon, Feb 14, 2011 at 02:21:37PM +0100, Harald Jenny wrote:
 On Mon, Feb 14, 2011 at 01:00:51PM +, Athanasius wrote:
  aptitude -t lenny-backports install openswan
  
  telling it to keep my current config and it's working.
 
 So may I close this bug report as solved - as 2.4 is deprecated by upstream
 it may not be really feasable to fix it :-/ )?

  Fine by me, especially given squeeze is now the Stable (and I'll be
upgrading to that in the next month or so).

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#580070: Bug still present.

2010-09-26 Thread Athanasius 
  A few days I started seeing the following during an 'aptitude
safe-upgrade' on my Debian/squeeze system:

Setting up python-central (0.6.16+nmu1) ...
pycentral: pycentral cleanup-pkgprepare-updates: found 37 dangling symlinks
pycentral: pycentral cleanup-pkgprepare-updates: checking for links owned by 
packages (this may take some time)
warning, in file '/var/lib/dpkg/status' near line 5913 package 'dist':
missing description
Traceback (most recent call last):
  File /usr/bin/pycentral, line 2335, in module
main()
  File /usr/bin/pycentral, line 2329, in main
rv = action.run(global_options)
  File /usr/bin/pycentral, line 1902, in run
packaged = self.links_in_packages(dangling)
  File /usr/bin/pycentral, line 2005, in links_in_packages
self.locate(arg_list, cmd_list)
  File /usr/bin/pycentral, line 1974, in locate
fn = line.split(':', 1)[1][1:-11]
IndexError: list index out of range
dpkg: error processing python-central (--configure):
 subprocess installed post-installation script returned error exit status 1
Errors were encountered while processing:
 python-central

Which would seem to be another instance of this bug.  However the
workaround posted above doesn't appear to work:

if chunk_max == -1: chunk_max = 16384

(error with 'if chunk_max = -1: chunk_max = 16384') doesn't prevent the
error.  Indeed if I put:

self.warn('chunk_max: %d' % chunk_max)

as line 1982 (just after the SC_ARG_MAX if/else) I get:

pycentral: pycentral cleanup-pkgprepare-updates: chunk_max: 2097152

Putting that warn() after the reduce() call gives:

pycentral: pycentral cleanup-pkgprepare-updates: chunk_max: 2095312

Aha!  Putting a warn() to show the value of 'line' at line 1973 shows:

pycentral: pycentral cleanup-pkgprepare-updates: Line: missing description

So in fact although perhaps python-central should be more robust this is
actually happening because of how dpkg now moans if any package doesn't
have a Description field.
  Adding a second test:

if line == 'missing description': continue

after the:

if line == '': continue

allows python-central to complete properly.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#422139: This is *NOT* how to run a daemon on a standard Debian system!

2010-09-20 Thread Athanasius 
  I've just reviewed the git daemon installation that another admin of a
machine I'm responsible for had installed.

  Frankly I'm aghast that this thing decides that 'runit' is better than
sysvinit, and also wants to use an entirely different logging service
with a different idea of log rotation as well!

  This whole package needs to be, at most, optional.

  Yes, I'll calm down and go and read the docs to see if I can whack
thing around the head until it runs as a sysvinit-controlled service and
uses syslogd to log.  Consider this another WHAT THE FUCK?!?!?! vote.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#595975: ftp.nl... works

2010-09-09 Thread Athanasius 
  I've been seeing this problem for the last few days too.  It does seem
to be ftp.uk that has the problem as I just switched to ftp.nl and
that's completed the upgrade of this package without problems.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#564054: Config workaround

2010-09-08 Thread Athanasius 
  This is a flat out bug in OpenSwan.  Although I found a post (quoted
below) claiming this was fixed around 2.4.10 (my Debian lenny system has
openswan based on 2.4.12 judging by the Version: 1:2.4.12+dfsg-1.3+lenny2)
I still have the problem.

http://webcache.googleusercontent.com/search?q=cache:FHTtaZzuKRUJ:https://gsoc.xelerance.com/issues/849+ipsec+openswan+pluto+assertion+failedcd=4hl=enct=clnkgl=uk
(google cache because the page seems down... still takes a while to load
on this URL)

says:

This is a bug, where we think the conn define is NOT an instance
because right != %any... but it's NOT permadent, since
rightprotoport=17/%any.

There are a few ways to fix this, but simplest is to tell the config
reader that yes, this is a CK_INSTANCE.

and further claims:

It's a bug. It has been fixed in CVS and will be in 2.4.10 which we
hope to release in 1-2 days

but obviously either that fix didn't make it in or was incomplete.  The
page says that was 'about 1 year ago' (at the time google last cached it
presumably).

  The only workaround I found was to have 'right=%any' in the 'conn'
section.  I guess if you really need to restrict it to certain IPs you
could use IPtables instead ?

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#596172: openswan: pluto erroneously adds host route for IP with valid network route

2010-09-08 Thread Athanasius 
Package: openswan
Version: 1:2.4.12+dfsg-1.3+lenny2
Severity: important

  In trying to get an L2TP/IPSec tunnel up from my Android-based phone to
my server for use when on untrusted WiFi I've obviously been testing it
at home.
  So I have a WiFi Access Point on 192.168.11.0/24, my server on .1, the
WAP on .2 and my phone on .4.  When I start up the connection pluto
decides to call _uproute in order to add a host route for
192.168.11.4/32... to whatever device pluto is listening on connections
for.  This of course is *not* the same device as the 192.168.11.0/24
network is on, so I end up with:

00:57:02 1$ ip -4 route
82.69.184.122 dev ethpriv  scope link 
82.69.184.123 dev ethpriv  scope link 
82.69.184.124 dev ethpriv  scope link 
192.168.11.4 dev ethpub  scope link 
82.69.184.125 dev ethpriv  scope link 
82.69.184.120/29 dev ethpub  proto kernel  scope link  src 82.69.184.121 
192.168.1.0/24 dev ethpriv  proto kernel  scope link  src 192.168.1.162 
192.168.11.0/24 dev ethwap  proto kernel  scope link  src 192.168.11.1 
default via 82.69.184.126 dev ethpub 

82.69.184.120/29 is my ISP-provided network and route to the internet.

I had a similar thing happen when testing with my 192.168.1.0/24 local
network.  I'd end up with a:

192.168.11.4 dev ethpriv scope link

route, blocking my server from talking back to the phone.

  Given any device connecting to OpenSwan/pluto can be from some
arbitrary IP I think it can be assumed that my server will already have
a route back to it.  So why bother even attempting to add this route at
all ?  It seems to be a leftover from the other/prior IPSec
implementation where the connection would come in over an ipsecx
device rather than just staying on the normal device as it does now.

  Anyway, I've worked around this for now by modifying
/usr/lib/ipsec/_uproute so that uproute() looks like:

# utility functions for route manipulation
# Meddling with this stuff should not be necessary and requires great care.
uproute() {
if [ ${PLUTO_INTERFACE} == ${defaultroutephys} ];
then
return
fi
doroute add
ip route flush cache
}

I'd guess the most common way to setup OpenSwan is listening on the
default route interface, but can also see setups where it might be on an
internal network (think of a secure internal network with a WAP
attached, firewalled off other than IPSec, and using tunnels to get WAP
clients onto that secure network).  So, this isn't a perfect fix.
Really I think it needs the whole code reviewing to only even attempt
these routes if the IPSec connection has resulted in a special network
interface for the connection in question.  This does not happen with my
setup using a 2.6.35.4 kernel and Debian lenny all up to date.

  I did glance quickly over the squeeze/testing changelog for Openswan
and couldn't spot anything that might have fixed this (but I only
searched quickly on 'route'), so this bug may well still be in that
version too.

-- System Information:
Debian Release: 5.0.6
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.35.4amd64 (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/bash

Versions of packages openswan depends on:
ii  bsdmainutils  6.1.10 collection of more utilities from 
ii  debconf [debconf-2.0] 1.5.24 Debian configuration management sy
ii  debianutils   2.30   Miscellaneous utilities specific t
ii  host  2331-9 utility for querying DNS servers
ii  iproute   20080725-2 networking and traffic control too
ii  ipsec-tools   1:0.7.1-1.3+lenny2 IPsec tools for Linux
ii  libc6 2.7-18lenny4   GNU C Library: Shared libraries
ii  libcurl3  7.18.2-8lenny4 Multi-protocol file transfer libra
ii  libgmp3c2 2:4.2.2+dfsg-3 Multiprecision arithmetic library
ii  libldap-2.4-2 2.4.11-1+lenny2OpenLDAP libraries
ii  libpam0g  1.0.1-5+lenny1 Pluggable Authentication Modules l
ii  libssl0.9.8   0.9.8g-15+lenny8   SSL shared libraries
ii  openssl   0.9.8g-15+lenny8   Secure Socket Layer (SSL) binary a

openswan recommends no packages.

Versions of packages openswan suggests:
ii  curl  7.18.2-8lenny4 Get a file from an HTTP, HTTPS or 
pn  openswan-modules-source | none (no description available)

-- debconf information:
  openswan/existing_x509_key_filename:
  openswan/x509_state_name:
  openswan/rsa_key_length: 2048
  openswan/restart: true
  openswan/start_level: earliest
* openswan/enable-oe: false
  openswan/existing_x509_certificate: false
  openswan/existing_x509_certificate_filename:
* openswan/create_rsa_key: false
  openswan/x509_email_address:
  openswan/x509_country_code: AT
  openswan/x509_self_signed: true
  openswan/x509_organizational_unit:

Bug#595282: Acknowledgement (eclipse: Update 3.5.2-5 - 3.5.2-6 causes Android SDK plugin to stop working)

2010-09-05 Thread Athanasius 
On Fri, Sep 03, 2010 at 01:15:04PM +0200, Niels Thykier wrote:
 On 2010-09-02 21:44, Athanasius wrote:
I tracked down snapshot.debian.org and managed to find the
  problem.  It would appear to be the newer version of sat4j.  I first
  wound back to each of the following:
[snip]
  But still had the problem.  Then I double-checked /var/lib/dpkg.log and
[snip]
 
 It looks like you have been affected by #587657[0]. Could I have you

  Ah, I had seen that, but given the date on it and my assumption that
the problem was specifically caused by the most recent changes I figured
that was something else.

  OK, first confirming things do work with 3.5.2-5:

15:38:39 1$ aptitude show eclipse
Package: eclipse 
New: yes
State: installed
Automatically installed: no
Version: 3.5.2-5

which shows on Help  About Eclipse SDK:

Version: 3.5.2
Build id: M20100211-1343

 re-upgrade to 3.5.2-6 and run:
 
   eclipse -clean

  And then doing this (straight after the upgrade, I didn't try plain
'eclipse' first), the problem is there again.

 to see if that fixes it; if not then try the work-around posted by Kevin
 Locke[1].

  That appears to have worked.  Although note his followup, you want
#130, not #135 for the regex that works.  I assume it's expected that an
'eclipse -clean' followed by the sed will cause Eclipse to open with no
Perspectives or Views open?  Simple enough to re-open the Java Browsing
Perspective, if a little annoying.

  I went through the whole thing again, saving my bundles.info at each
stage to have (I sed'd the inter-directory slashes with dashes then
appended a tag):

-rw-r-  1 athan athan 49388 Sep  5 15:53 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-5-confirmed-works
-rw-r-  1 athan athan 49380 Sep  5 16:11 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-6-confirmed-broken
-rw-r-  1 athan athan 49380 Sep  5 16:13 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-6-cleaned-still-broken
-rw-r-  1 athan athan 49380 Sep  5 16:13 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-6-cleaned-and-seded-working

The diff between the last two is:

--- 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-6-cleaned-still-broken
2010-09-05 16:13:30.0 +0100
+++ 
.eclipse-org.eclipse.platform_3.5.0_155965261-configuration-org.eclipse.equinox.simpleconfigurator-bundles.info-3.5.2-6-cleaned-and-seded-working
   2010-09-05 16:13:59.0 +0100
@@ -17,8 +17,8 @@
 
org.apache.commons.logging,1.0.4.v20080605-1930,plugins/org.apache.commons.logging_1.0.4.v20080605-1930.jar,4,false
 
org.apache.commons.net,2.0.0,../../share/eclipse/dropins/rse/eclipse/plugins/org.apache.commons.net_2.0.0.jar,4,false
 org.apache.jasper,5.5.26,plugins/org.apache.jasper_5.5.26.jar,4,false
-org.apache.lucene,2.9.2,plugins/org.apache.lucene_2.9.3.jar,4,false
-org.apache.lucene.analysis,2.9.2,plugins/org.apache.lucene.analysis_2.9.3.jar,4,false
+org.apache.lucene,2.9.3,plugins/org.apache.lucene_2.9.3.jar,4,false
+org.apache.lucene.analysis,2.9.3,plugins/org.apache.lucene.analysis_2.9.3.jar,4,false
 
org.apache.xerces,2.9.0.v201005080400,../../../home/users/athan/.eclipse/org.eclipse.platform_3.5.0_155965261/plugins/org.apache.xerces_2.9.0.v201005080400.jar,4,false
 
org.apache.xml.resolver,1.2.0.v201005080400,../../../home/users/athan/.eclipse/org.eclipse.platform_3.5.0_155965261/plugins/org.apache.xml.resolver_1.2.0.v201005080400.jar,4,false
 
org.apache.xml.serializer,2.7.1.v201005080400,../../../home/users/athan/.eclipse/org.eclipse.platform_3.5.0_155965261/plugins/org.apache.xml.serializer_2.7.1.v201005080400.jar,4,false

  So the whole thing does indeed seem to hinge on those
org.apache.lucene(.analysis) lines needing both 'version' strings
updating.  In this case one character per line.




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

Bug#595282: eclipse: Update 3.5.2-5 - 3.5.2-6 causes Android SDK plugin to stop working

2010-09-02 Thread Athanasius 
Package: eclipse
Version: 3.5.2-6
Severity: important


Today I went to check something in an Android project I started using
Eclipse.  On startup any Android-specific files (like
res/layout/main.xml) can no longer find the specific editor.  The pane
contains the error:

org.eclipse.ui.PartInitException: No editor descriptor for id 
com.android.ide.eclipse.editors.layout.LayoutEditor
at 
org.eclipse.ui.internal.EditorReference.createPartHelper(EditorReference.java:598)
at 
org.eclipse.ui.internal.EditorReference.createPart(EditorReference.java:462)
at 
org.eclipse.ui.internal.WorkbenchPartReference.getPart(WorkbenchPartReference.java:595)
at 
org.eclipse.ui.internal.EditorAreaHelper.setVisibleEditor(EditorAreaHelper.java:271)
at 
org.eclipse.ui.internal.EditorManager.setVisibleEditor(EditorManager.java:1417)
at 
org.eclipse.ui.internal.EditorManager$5.runWithException(EditorManager.java:942)
at 
org.eclipse.ui.internal.StartupThreading$StartupRunnable.run(StartupThreading.java:31)
at org.eclipse.swt.widgets.RunnableLock.run(RunnableLock.java:35)
at 
org.eclipse.swt.widgets.Synchronizer.runAsyncMessages(Synchronizer.java:134)
at org.eclipse.swt.widgets.Display.runAsyncMessages(Display.java:3468)
at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3115)
at 
org.eclipse.ui.application.WorkbenchAdvisor.openWindows(WorkbenchAdvisor.java:803)
at 
org.eclipse.ui.internal.Workbench$28.runWithException(Workbench.java:1384)
at 
org.eclipse.ui.internal.StartupThreading$StartupRunnable.run(StartupThreading.java:31)
at org.eclipse.swt.widgets.RunnableLock.run(RunnableLock.java:35)
at 
org.eclipse.swt.widgets.Synchronizer.runAsyncMessages(Synchronizer.java:134)
at org.eclipse.swt.widgets.Display.runAsyncMessages(Display.java:3468)
at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3115)
at org.eclipse.ui.internal.Workbench.runUI(Workbench.java:2316)
at org.eclipse.ui.internal.Workbench.access$4(Workbench.java:2221)
at org.eclipse.ui.internal.Workbench$5.run(Workbench.java:500)
at 
org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
at 
org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:493)
at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:149)
at 
org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:113)
at 
org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:194)
at 
org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:110)
at 
org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:79)
at 
org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:368)
at 
org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:179)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at 
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at 
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:597)
at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:559)
at org.eclipse.equinox.launcher.Main.basicRun(Main.java:514)
at org.eclipse.equinox.launcher.Main.run(Main.java:1311)

I first installed Eclipse specifically to work with the Android SDK,
doing so on 20th August 2010 (as per /var/log/dpkg.log).  This worked
without problems.  This was version 3.2.5-5.

I last touched the project in the early hours of 24th August 2010,
when things were still working without problems.

I note from dpkg.log that there was a, seemingly small, upgrade of
Eclipse recently, installed on my system on 31st August 2010 (around
10:42 BST).  This was for version 3.2.5-6.

Now I get the above problem.

I've played with trying to re-install 3.2.5-5 from source (I couldn't
find the .dpkg files any more, if a dev has them to hand still I'll
gladly try them to help track down the problem), but ran into problems
with a 'Breaks' on sat4j, and had to do 'DEB_BUILD_OPTIONS=nocheck' to
get the build on that to complete.  The resulting eclipse wouldn't even
startup.

So instead I grabbed the latest stable Eclipse from
http://www.eclipse.org/downloads/, specifically the Eclipse IDE for
Java Developers:

890a0dc7fc4a99e1537b31a2718c56e7  eclipse-java-helios-linux-gtk.tar.gz

and installed this.  I can install the Android plugin to this and open
up my old workspace perfectly.  Help  About says this is:

Version: Helios Release
Build id: 20100617-1415

Instances of 'upgrade' in /var/log/dpkg.log from after I initially
installed eclipse upto when it failed to work today are:

2010-08-11 23:06:42

Bug#595282: Acknowledgement (eclipse: Update 3.5.2-5 - 3.5.2-6 causes Android SDK plugin to stop working)

2010-09-02 Thread Athanasius 
  I tracked down snapshot.debian.org and managed to find the
problem.  It would appear to be the newer version of sat4j.  I first
wound back to each of the following:

http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse_3.5.2-5_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-cdt/eclipse-cdt_3.1.2-2_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-emf/eclipse-emf_2.5.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-emf/eclipse-emf-examples_2.5.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-emf/eclipse-emf-sdk_2.5.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-jdt_3.5.2-5_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-pde_3.5.2-5_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-platform_3.5.2-5_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-platform-data_3.5.2-5_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-plugin-cvs_3.5.2-5_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/eclipse-rcp_3.5.2-5_i386.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-rse/eclipse-rse_3.1.2-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-emf/eclipse-xsd_2.5.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse-emf/eclipse-xsd-sdk_2.5.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipselink/libeclipselink-java_2.1.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipselink/libeclipselink-java-doc_2.1.0-1_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/e/eclipse/libequinox-osgi-java_3.5.2-5_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/s/sat4j/sat4j_2.2.0-2_all.deb
http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/j/java-common/java-common_0.38_all.deb
(plus some dependencies reinstalled via 'aptitude install').

But still had the problem.  Then I double-checked /var/lib/dpkg.log and
it said I was on sat4j_2.2.0-1 before the upgrade, not sat4j_2.2.0-2, so
I wound that back using:

http://snapshot.debian.org/archive/debian/20100810T205427Z/pool/main/s/sat4j/sat4j_2.2.0-1_all.deb

And now it all works as it did before.

So the problem would seem to have been caused by:

2010-08-31 10:42:38 upgrade sat4j 2.2.0-1 2.2.0-3

However I note that installing all the latest versions of the above then
dpkg -i'ing sat4j_2.2.0-1_all.deb still results in the problem.



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

Bug#515273: Still broken

2010-08-10 Thread Athanasius 
  Whilst the workaround above (just turning OFF encryption) will of
course work, this bug is *still* present in libtorrent11_0.12.6-2.
I'm actually wondering if it might be more an OpenSSL bug, or if not
it's libtorrent's use of that.
  NB: I'm actually testing this specific libtorrent version as a
backport into a Debian/lenny system as that's what my up 24/7 server
runs and is where I run rtorrent.  So that'd be openssl 0.9.8g-15+lenny7
in use.  I also backported rtorrent 0.8.6-1 for use with this.  I only
did this after initially hitting the problem with lenny's versions.

  For now I'm trying a libtorrent with the offending test and 'throw'
commented out to see if it does any harm.  I'm not familiar enough with
the code to see exactly what the check is trying to prevent.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#572440: Reproduceable with Firefox and /dev/fd0

2010-05-03 Thread Athanasius 
  This also affects Firefox, even when not using Gnome as desktop (I use
fvwm).

  I'm using Debian/testing on a system that does have a floppy disk
drive, but commonly no disk in the drive.  After bootup the first time
Firefox tries to put up the file selection dialogue (as when choosing
the location to save a download to) it hangs for ~25s.  This produces:

[ 4754.490418] end_request: I/O error, dev fd0, sector 0
[ 4766.680262] end_request: I/O error, dev fd0, sector 0

in dmesg output.

  Stopping dbus, exiting Firefox, starting dbus and restarting Firefox
will cause the problem to happen again on next file dialogue access.
'dbus-monitor --system' output (started just before Firefox):

07:44:38 0$ dbus-monitor --system
signal sender=org.freedesktop.DBus - dest=:1.1 serial=2 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; member=NameAcquired
   string :1.1
signal sender=org.freedesktop.DBus - dest=(null destination) serial=7 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string :1.2
   string 
   string :1.2
signal sender=org.freedesktop.DBus - dest=(null destination) serial=8 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string :1.3
   string 
   string :1.3
signal sender=org.freedesktop.DBus - dest=(null destination) serial=9 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string :1.4
   string 
   string :1.4
signal sender=org.freedesktop.DBus - dest=(null destination) serial=10 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string :1.5
   string 
   string :1.5
signal sender=org.freedesktop.DBus - dest=(null destination) serial=4 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string org.freedesktop.UDisks
   string 
   string :1.5
signal sender=org.freedesktop.DBus - dest=(null destination) serial=7 
path=/org/freedesktop/DBus; interface=org.freedesktop.DBus; 
member=NameOwnerChanged
   string :1.6
   string 
   string :1.6
signal sender=:1.5 - dest=(null destination) serial=8 
path=/org/freedesktop/UDisks/devices/hda; 
interface=org.freedesktop.UDisks.Device; member=Changed
signal sender=:1.5 - dest=(null destination) serial=9 
path=/org/freedesktop/UDisks; interface=org.freedesktop.UDisks; 
member=DeviceChanged
   object path /org/freedesktop/UDisks/devices/hda
signal sender=:1.5 - dest=(null destination) serial=27 
path=/org/freedesktop/UDisks/devices/sda; 
interface=org.freedesktop.UDisks.Device; member=Changed
signal sender=:1.5 - dest=(null destination) serial=28 
path=/org/freedesktop/UDisks; interface=org.freedesktop.UDisks; 
member=DeviceChanged
   object path /org/freedesktop/UDisks/devices/sda

hda is the DVD drive, sda is the single SATA hard disk in the system.

Looking more closely with a 'tail -f /var/log/kern.log':

May  3 07:47:43 emilia kernel: [ 5127.707891] kobject: 'hda' (f0671c48): 
kobject_uevent_env
May  3 07:47:43 emilia kernel: [ 5127.707904] kobject: 'hda' (f0671c48): 
fill_kobj_path: path = 
'/devices/pci:00/:00:1c.4/:03:00.1/ide0/0.0/block/hda'
May  3 07:47:52 emilia kernel: [ 5137.506036] end_request: I/O error, dev fd0, 
sector 0
May  3 07:48:04 emilia kernel: [ 5149.695175] end_request: I/O error, dev fd0, 
sector 0

For this run I clicked 'Save as ...' on a link at 07:47:40.  As you can
see the DVD drive scan takes ~3 seconds, but that's acceptable.  What
isn't is the further ~20 seconds whilst it attempts to scan fd0 (OK,
it's only clearly 12 seconds above, but there'll be some attributeable
'hang time' before 07:47:52 for the first attempt as well).

  So, is there some way to disable this 'scanning' *without* disabling
the FDD entirely ?  Oh, I should note that although there is a line for
/dev/fd0 in /etc/fstab, it's commented out.

  One thing I need to check is if the FDD actually works at all.  A
quick test trying to mount or 'mdir' a floppy in it is showing no
activity.  I'll follow up when I've had chance to shutdown and check
connections.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#572440: Info received (Reproduceable with Firefox and /dev/fd0)

2010-05-03 Thread Athanasius 
  Aha!  On checking my FDD's power cable was unplugged, but it had the
data one connected, apparently enough to make Linux think there's a
working FDD.

  So, false alarm.  After booting with the power connected the delay is
now *much* shorter and reasonable enough.  Now I wish I'd checked this
weeks ago.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



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

Bug#569539: console beep in xterm doesn't work

2010-03-31 Thread Athanasius 
  Indeed I noticed a few days ago that I have beeps back now in my
mainly testing install.  This would appear to be xterm 256-1 which got
installed here on the 25th March 2010.
  So I think we can consider this bug fixed and close it.



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

Bug#441567: / on the end of open_basedir paths

2010-03-10 Thread Athanasius 
php.net documentation disagrees, see:

http://uk2.php.net/manual/en/ini.core.php#ini.open-basedir

specifically:

The restriction specified with open_basedir is actually a prefix, not a
directory name. This means that open_basedir = /dir/incl also allows
access to /dir/include and /dir/incls if they exist. When you want
to restrict access to only the specified directory, end with a slash.
For example: open_basedir = /dir/incl/

Using '/usr/share/gallery2' in open_basedir means that it would also
include '/usr/share/gallery2-old' in access, which is almost certainly
incorrect.  Using '/usr/share/gallery2/' would be correct.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#573389: Smarty_Compiler.class.php in Debian package fails to handle single quotes properly

2010-03-10 Thread Athanasius 
Package: gallery2
Version: 2.3.1.dfsg-1~bpo50+1
Severity: important

The Debian version of /usr/share/gallery2/lib/smarty/Smarty_Compiler.class.php
differs from the stock gallery2 2.3.1 version of the file, and it fails when
there are single quotes (') in templates, leading to errors such as:

Parse error: syntax error, unexpected T_STRING, expecting ')' in 
/var/www/user-rw/gallery2-349gl0289gys/smarty/templates_c/%%626616196/matrix/%%26^261^2615E4E5%%AdminPlugins.tpl.php
  on line 173

When trying to access Site Admin  Plugins via the web interface of gallery2.

The diff from the Debian version to the stock 2.3.1 version is:

--- /usr/share/gallery2/lib/smarty/Smarty_Compiler.class.php2009-10-25 
15:19:04.0 +
+++ Smarty_Compiler.class.php   2008-10-16 07:35:13.0 +0100
@@ -1695,12 +1695,7 @@
 $_return = $var_expr;
 }
 // replace double quoted literal string with single quotes
-
-// The follwoing line has been replaced to close a function injection 
security hole  (U.Tews)
-// $_return = preg_replace('~^([\s\w]+)$~','\\1',$_return);
-$_return = str_replace('',',$_return);
-// escape dollar sign if not printing a var
-$_return = preg_replace('~\$(\W)~',\$\\1,$_return);
+$_return = preg_replace('~^([\s\w]+)$~','\\1',$_return);
 return $_return;
 }

Replacing the Debian version with the stock version allows things to work
properly once more, once you've used Maintenance  Delete template cache.

The stock version uses double-quotes () around the strings it is handling,
the Debian version uses single-quotes ('), without thinking to escape such
single quotes in the text, and this is the cause of the problem.

-- System Information:
Debian Release: 5.0.4
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.33-fysh-kvmguest (SMP w/2 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages gallery2 depends on:
ii  apache2 2.2.9-10+lenny6  Apache HTTP Server metapackage
ii  apache2-mpm-pre 2.2.9-10+lenny6  Apache HTTP Server - traditional n
ii  debconf [debcon 1.5.24   Debian configuration management sy
ii  imagemagick 7:6.3.7.9.dfsg2-1~lenny3 image manipulation programs
ii  libapache2-mod- 5.2.6.dfsg.1-1+lenny6server-side, HTML-embedded scripti
ii  libphp-adodb5.05-1   The ADOdb database abstraction lay
ii  mysql-client-5. 5.0.51a-24+lenny3MySQL database client binaries
ii  netpbm  2:10.0-12Graphics conversion tools
ii  php55.2.6.dfsg.1-1+lenny6server-side, HTML-embedded scripti
ii  php5-cgi5.2.6.dfsg.1-1+lenny6server-side, HTML-embedded scripti
ii  php5-mysql  5.2.6.dfsg.1-1+lenny6MySQL module for php5
ii  php5-pgsql  5.2.6.dfsg.1-1+lenny6PostgreSQL module for php5
ii  postgresql-clie 8.3.9-0lenny1front-end programs for PostgreSQL 
ii  smarty  2.6.20-1.2   Template engine for PHP
ii  wwwconfig-commo 0.1.2Debian web auto configuration

Versions of packages gallery2 recommends:
ii  dcraw8.86-1  decode raw digital camera images
ii  ffmpeg   0.svn20080206-18+lenny1 multimedia player, server and enco
ii  jhead2.84-2  manipulate the non-image part of E
ii  libjpeg-progs6b-14   Programs for manipulating JPEG fil
ii  php5-gd  5.2.6.dfsg.1-1+lenny6   GD module for php5
ii  unzip5.52-12 De-archiver for .zip files
ii  zip  2.32-1  Archiver for .zip files

Versions of packages gallery2 suggests:
pn  mysql-server-5.0 | mysql-serv none (no description available)

-- debconf information:
  gallery2/mysql/dbadmpass: (password omitted)
  gallery2/webserver_type: apache, apache-ssl, apache-perl, apache2
  gallery2/mysql/dbname: gallery2
* gallery2/mysql/dbserver: db.fysh.org
  gallery2/mysql/configure: true
* gallery2/restart-webserver: false
  gallery2/purge: true
* gallery2/mysql/dbadmin: root



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

Bug#569539: This is not xterm itself

2010-02-14 Thread Athanasius 
  I've just compiled xterm-253 from source, making the -g in CFLAGS into
-ggdb3 and commenting out dh_strip from debian/rules and if I break on
xtermBell() (misc.c:813) it is indeed trying to call XBell() from
/usr/lib/libX11.so.6 (at misc.c:860).

  I'll see if I can set things up to trace what's happening in libX11.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100214101454.gd1...@miggy.org

Bug#569539: The culprit *is* xterm source, misc.c:817

2010-02-14 Thread Athanasius 
  Looking at how xkbbell achieves the beep I found it does:

(in xkbutils/xkbbell.c)
 204 elseok=  XkbBell(dpy,win,volume,nameAtom);

Now, xterm has this code:

 811 void
 812 xtermBell(XtermWidget xw, int which, int percent)
 813 {
 814 TScreen *screen = TScreenOf(xw);
 815 #if defined(HAVE_XKB_BELL_EXT)
 816 Atom tony = AtomBell(xw, which);
 817 if (tony != None) {
 818 XkbBell(screen-display, VShellWindow, percent, tony);
 819 } else
 820 #endif
 821 XBell(screen-display, percent);
 822 }

On checking, HAVE_XKB_BELL_EXT is indeed set.  However AtomBell() is
returning 'None'.  But wait a moment, looking back at xkbbell it
defaults this to 'None' as well:

 43 static  Atom nameAtom = None;

and the only way that changes is if it passes this test:

198 if (bellName!=NULL)
199 nameAtom = XInternAtom(dpy,bellName,0);

which it doesn't unless you specify a bell with 'xkbbell bellname'.
So with just 'xkbbell' working this means that passing a nameAtom of
'None' to XkbBell() is perfectly fine, so why is xterm deciding it's a
bad thing ?

  Thus just commenting out lines 817 and 819 causes xterm beeping to
start working again.  Obviously you'd actually want to make it something
more like:

 811 void
 812 xtermBell(XtermWidget xw, int which, int percent)
 813 {
 814 TScreen *screen = TScreenOf(xw);
 815 #if defined(HAVE_XKB_BELL_EXT)
 816 Atom tony = AtomBell(xw, which);
 817 XkbBell(screen-display, VShellWindow, percent, tony);
 818 #else
 819 XBell(screen-display, percent);
 820 #endif
 821 }

So that you only issue one of XkbBell() or XBell().  Note AtomBell()
always sets its return value to a default of 'None', so there shouldn't
be a need for a NULL pointer check or anysuch there.
  I may be wrong about not wanting to issue both, X11 experts can check
that.

XkbBell(3) has nothing to say about the Atom name 'None' being special,
but it does say it can be NULL (but not what happens if it is).  It does
also state:

   If  a  compatible  keyboard  extension  isn't  present in the X server,
   XkbBell calls XBell with the specified display and percent, and returns
   False.  Otherwise,  XkbBell calls XkbDeviceBell with the specified dis‐
   play, window, percent, and name,  a  device_spec  of  XkbUseCoreKbd,  a
   bell_class of XkbDfltXIClass, and a bell_id of XkbDfltXIId, and returns
   True.

which seems to mean that my change as above to not try both XkbBell()
and XBell() is correct, as the former will fall back to the latter if
needs be anyway.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100214112734.ga30...@miggy.org

Bug#569539: console beep in xterm doesn't work

2010-02-12 Thread Athanasius 
Package: xterm
Version: 253-1
Severity: normal


echo -e '\a' with visual bell off fails to illicit any sound.

1) xkbbell causes a beep
2) beep causes a beep (although of a different tone)
3) echo -e '\a' in pterm causes a beep (same as xkbbell)
4) Trying to search for a non-present string in firefox causes a beep
(same as xkbbell)

I'm using a kernel compiled from vanilla 2.6.32.8 sources and the pcspkr
module from that to ensure console beeps.

~/.inputrc has set bell-style audible to ensure readline causes beeps

The only output of xrdb -query | grep -i bell is:
XTerm*visualBell:   false

If I toggle visual bell on in an xterm (using the Ctrl+Middle menu)
and reissue echo -e '\a' I 'see' the beep.

This was working until late last year.  Unfortunately I didn't realise
exactly when it stopped working (as I don't always perform an action
that would cause such a beep on a day to day basis).  The system has been
tracking testing on a more or less daily basis since around June 2009.

-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (50, 'unstable')
Architecture: i386 (i686)

Kernel: Linux 2.6.32.8 (SMP w/2 CPU cores; PREEMPT)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages xterm depends on:
ii  libc6 2.10.2-2   GNU C Library: Shared libraries
ii  libfontconfig12.8.0-2generic font configuration library
ii  libice6   2:1.0.6-1  X11 Inter-Client Exchange library
ii  libncurses5   5.7+20090803-2 shared libraries for terminal hand
ii  libutempter0  1.1.5-2A privileged helper for utmp/wtmp 
ii  libx11-6  2:1.3.3-1  X11 client-side library
ii  libxaw7   2:1.0.7-1  X11 Athena Widget library
ii  libxft2   2.1.14-1   FreeType-based font drawing librar
ii  libxmu6   2:1.0.5-1  X11 miscellaneous utility library
ii  libxt61:1.0.7-1  X11 toolkit intrinsics library
ii  xbitmaps  1.1.0-1Base X bitmaps

Versions of packages xterm recommends:
ii  x11-utils 7.5+1+b1   X11 utilities
ii  xutils1:7.5+3X Window System utility programs m

Versions of packages xterm suggests:
ii  xfonts-cyrillic   1:1.0.1Cyrillic fonts for X

-- debconf information:
  xterm/clobber_xresource_file: true
  xterm/xterm_needs_devpts:



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

Bug#549118: uswsusp: swap-offset fails on files over 2GB in size (32bit file access is used)

2009-09-30 Thread Athanasius 
Package: uswsusp
Version: 0.7-1.2
Severity: important
Tags: patch

  In trying to set up uswsusp I found myself needing the offset of my
swapfile.  On running 'swap-offset /var/lib/swapfile1' I get:

20:57:07 3$ swap-offset /var/lib/swapfile1
open(): Value too large for defined data type
r...@emilia:~;
20:57:24 3$ 

If you modify swap-offset.c to include:

#define _FILE_OFFSET_BITS 64

before the #include's and change the lseek() call to lseek64() it works:

20:57:04 3$ /usr/local/src/uswsusp-0.7/swap-offset /var/lib/swapfile1
resume offset = 12462091
r...@emilia:~;
20:57:07 3$ 

-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (50, 'unstable')
Architecture: i386 (i686)

Kernel: Linux 2.6.31.1 (SMP w/2 CPU cores; PREEMPT)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages uswsusp depends on:
ii  debconf [debconf-2.0] 1.5.27 Debian configuration management sy
ii  libc6 2.9-25 GNU C Library: Shared libraries
ii  libgcc1   1:4.4.1-1  GCC support library
ii  libgcrypt11   1.4.4-4LGPL Crypto library - runtime libr
ii  liblzo2-2 2.03-1 data compression library
ii  libpci3   1:3.1.4-1  Linux PCI Utilities (shared librar
ii  libsplashy1   0.3.13-5   Library to draw splash screen on b
ii  libx86-1  1.1+ds1-6  x86 real-mode library

Versions of packages uswsusp recommends:
ii  initramfs-tools   0.93.4 tools for generating an initramfs
ii  mount 2.16.1-3   Tools for mounting and manipulatin

Versions of packages uswsusp suggests:
pn  splashy   none (no description available)

-- debconf information:
* uswsusp/suspend_loglevel:
  uswsusp/no_swap:
  uswsusp/resume_offset:
* uswsusp/early_writeout: true
* uswsusp/image_size: 962849095
* uswsusp/snapshot_device:
* uswsusp/max_loglevel:
  uswsusp/shutdown_method: platform
* uswsusp/encrypt: false
  uswsusp/RSA_key_bits: 1024
* uswsusp/continue_without_swap: false
* uswsusp/compute_checksum: false
  uswsusp/no_snapshot:
* uswsusp/compress: true
  uswsusp/create_RSA_key: false
  uswsusp/RSA_key_file: /etc/uswsusp.key
* uswsusp/resume_device: /dev/sda2
* uswsusp/splash: false



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

Bug#524610: nfs-common: rpc.statd not running during umountnfs.sh at shutdown/reboot causes delay due to lockd timeout

2009-04-18 Thread Athanasius 
Package: nfs-common
Version: 1:1.1.2-6lenny1
Severity: normal

For some time I'd been coping with a reboot or shutdown being delayed
at the point umountnfs.sh runs.  This turns out to be because by the
time /etc/rc6.d/S31umountnfs.sh ran rpc.statd was dead and the in-kernel
lockd expects this to be alive.

The fix I finally found and implemented was to move rc6.d/K20nfs-common
to rc6.d/S32nfs-common and also make the startup of nfs-common record
rpc.statd's pid in /lib/init/rw/sendsigs.omit.d.  This way rpc.statd is
not killed by rc6.d/S20sendsigs and is then brought down by nfs-common
*after* rc6.d/S31umountnfs.sh has run.  I now have no delay at all in
reboots when umountnfs.sh runs.

Note this is a 2.6.29 kernel with the simple GRO: Disable GRO on legacy
netif_rx path fix by Herbert Xu that went into 2.6.29.1, without which
the ethernet interface locks up after some moderate use.

-- System Information:
Debian Release: 5.0.1
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.29patches (SMP w/2 CPU cores; PREEMPT)
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/bash

Versions of packages nfs-common depends on:
ii  adduser 3.110add and remove users and groups
ii  initscripts 2.86.ds1-61  Scripts for initializing and shutt
ii  libc6   2.7-18   GNU C Library: Shared libraries
ii  libcomerr2  1.41.3-1 common error description library
ii  libevent1   1.3e-3   An asynchronous event notification
ii  libgssglue1 0.1-2mechanism-switch gssapi library
ii  libkrb531.6.dfsg.4~beta1-5lenny1 MIT Kerberos runtime libraries
ii  libnfsidmap20.20-1   An nfs idmapping library
ii  librpcsecgss3   0.18-1   allows secure rpc communication us
ii  libwrap07.6.q-16 Wietse Venema's TCP wrappers libra
ii  lsb-base3.2-20   Linux Standard Base 3.2 init scrip
ii  netbase 4.34 Basic TCP/IP networking system
ii  portmap 6.0-9RPC port mapper
ii  ucf 3.0016   Update Configuration File: preserv

nfs-common recommends no packages.

nfs-common suggests no packages.

-- no debconf information



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

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 04:19:49PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  Package: gnucash
  Version: 2.2.6-1
  
Recently GNUcash on my Debian/lenny (testing) system stopped running.
  If I run it from an xterm I see:
  
  gnc.bin-Message: main: binreloc relocation support was disabled at 
  configure time.
  
  3:2008/08/27 14-25-50:gwen(15810):cryptkey.c:   37: Libgcrypt version 
  mismatch
  Could not initialize Gwenhywfar, aborting
  [...]
 
 This is strange.
 
 Can you please tell us the version of libgcrypt11 you have currently
 installed? (Run dpkg -l libgcrypt11 for this information).

ii  libgcrypt111.4.1-1LGPL Crypto library - runtime library

(and checking dpkg -l libgcrypt\* that is the only version installed).

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 06:02:59PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  On Wed, Aug 27, 2008 at 05:14:43PM +0200, Micha Lenk wrote:
  Athanasius wrote:
  Can you please tell us the version of libgcrypt11 you have currently
  installed? (Run dpkg -l libgcrypt11 for this information).
  ii  libgcrypt111.4.1-1LGPL Crypto library - runtime library
 
  (and checking dpkg -l libgcrypt\* that is the only version installed).
  That's even more strange: It's the most recent version of libgrypt11.
 
  Can you provide us a complete strace output?
  
Attached.
 
 Hmm, I can't find anything suspicious yet.
 
 Can you also provide an ltrace output and the file /tmp/gnucash.trace
 produced when running Gnucash with command line option --debug?

  I had to modify the shell script /usr/bin/gnucash to run gnucash-bin
as:

exec ltrace -s 4096 -f -o /tmp/gnucash-ltrace gnucash-bin --debug $@

as ltrace doesn't seem happy running a shell script.

  Resultant file attached.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME
17141 __libc_start_main(0x804a730, 2, 0xbfbcf3d4, 0x804b0f0, 0x804b0e0 
unfinished ...
17141 g_thread_init(0, 0, 0xdfc5a0, 0, 0x612b00c1) = 0x972a600
17141 g_log(0x804b27b, 32, 0x804b970, 0, 0x612b00c1) = 32
17141 gnc_path_get_localedir(0x804b27b, 32, 0x804b970, 0, 0x612b00c1) = 
0x972a668
17141 bindtextdomain(gnucash, /usr/share/locale) = /usr/share/locale
17141 textdomain(gnucash)  = gnucash
17141 bind_textdomain_codeset(0x804b529, 0x804b2b9, 0x804b970, 0, 0x612b00c1) = 
0x9729aa8
17141 g_free(0x972a668, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 3
17141 qof_log_init(0x972a668, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0xb6e867f0
17141 qof_log_set_default(64, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 33
17141 g_getenv(0x804b2bf, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0
17141 g_getenv(0x804b2cf, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0
17141 g_getenv(0x804b2de, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0
17141 g_getenv(0x804b2eb, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0
17141 dcgettext(0, 0x804b1a8, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1a8
17141 dcgettext(0, 0x804b6e8, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b6e8
17141 dcgettext(0, 0x804b72c, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b72c
17141 dcgettext(0, 0x804b75c, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b75c
17141 dcgettext(0, 0x804b7ac, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b7ac
17141 dcgettext(0, 0x804b800, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b800
17141 dcgettext(0, 0x804b1bd, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1bd
17141 dcgettext(0, 0x804b1d4, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1d4
17141 dcgettext(0, 0x804b824, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b824
17141 dcgettext(0, 0x804b1df, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1df
17141 dcgettext(0, 0x804b848, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b848
17141 dcgettext(0, 0x804b1e9, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1e9
17141 dcgettext(0, 0x804b874, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b874
17141 dcgettext(0, 0x804b1f1, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1f1
17141 dcgettext(0, 0x804b89c, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b89c
17141 dcgettext(0, 0x804b1fb, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b1fb
17141 dcgettext(0, 0x804b8c8, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b8c8
17141 dcgettext(0, 0x804b200, 5, 0xbfbcf120, 0x87dd2ca) = 0x804b200
17141 strstr(gnucash-bin, -bin)  = -bin
17141 g_option_context_new(0x804b26f, 0x804b26a, 5, 0xbfbcf120, 0xbfbcf150) = 
0x972a6f0
17141 g_option_context_add_main_entries(0x972a6f0, 0xbfbcf150, 0x804b529, 
0xbfbcf120, 0xbfbcf150) = 0xb6e85520
17141 gtk_get_option_group(0, 0x804b529, 0x804b529, 0xbfbcf120, 0xbfbcf150) = 
0x972b3a8
17141 g_option_context_add_group(0x972a6f0, 0x972b3a8, 0x804b529, 0xbfbcf120, 
0xbfbcf150) = 0x972ab50
17141 g_option_context_parse(0x972a6f0, 0xbfbcf350, 0xbfbcf148, 0xbfbcf2bc, 
0xbfbcf150) = 1
17141 g_option_context_free(0x972a6f0, 0xbfbcf350, 0xbfbcf148, 0xbfbcf2bc, 
0xbfbcf150) = 4
17141 gnc_set_extra(0, 0xbfbcf350, 0xbfbcf148, 0xbfbcf2bc, 0xbfbcf150) = 0
17141 gnc_set_gconf_path(0x804b54f, 0xbfbcf350, 0xbfbcf148, 0xbfbcf2bc, 
0xbfbcf150) = 0x804b54f
17141 gnc_set_debugging(1, 0xbfbcf350, 0xbfbcf148, 0xbfbcf2bc, 0xbfbcf150) = 1
17141 g_get_tmp_dir(0x804b2eb, 0x9729aa8, 0x804b970, 0, 0x612b00c1) = 0x9734070
17141 g_build_filename(0x9734070, 0x804b2fa, 0, 0, 0x612b00c1) = 0x972b408
17141 qof_log_init_filename(0x972b408, 0x804b2fa, 0, 0, 0x612b00c1) = 3
17141 g_free(0x972b408, 0x804b2fa, 0, 0, 0x612b00c1) = 3
17141 qof_log_set_default(16, 0x804b2fa, 0, 0, 0x612b00c1) = 16
17141 gnc_log_default(16, 0x804b2fa, 0, 0, 0x612b00c1) = 33
17141 gnc_is_debugging(16, 0x804b2fa, 0, 0, 0x612b00c1) = 1
17141 qof_log_set_level(0x804b295, 64, 0, 0, 0x612b00c1) = 64
17141 qof_log_set_level(0x804b308, 64, 0, 0, 0x612b00c1) = 64
17141 qof_log_set_level(0x804b30c, 64, 0, 0, 0x612b00c1) = 64

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 06:31:55PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  Can you also provide an ltrace output and the file /tmp/gnucash.trace
  produced when running Gnucash with command line option --debug?
 
 I got your ltrace file. What's with the /tmp/gnucash.trace file created
 by Gnucash itself?

  Ah I misunderstood.

  Not helped by the fact I have TMPDIR set and, on checking strace
output, it actually ended up elsewhere.  However, it's 0 bytes in size.
Looking at the strace output it's opened as FD 3 and after a couple
other calls it seems to never be referenced again (searching on 3, ):

open(/var/tmp/athan/gnucash.trace.WZKYGU, O_RDWR|O_CREAT|O_EXCL|O_LARGEFILE, 
0600) = 3
rename(/var/tmp/athan/gnucash.trace.WZKYGU, /var/tmp/athan/gnucash.trace) = 0
fcntl64(3, F_GETFL) = 0x8002 (flags O_RDWR|O_LARGEFILE)
fstat64(3, {st_mode=S_IFREG|0600, st_size=0, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0xb7d09000
_llseek(3, 0, [0], SEEK_CUR)= 0

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 07:22:59PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  On Wed, Aug 27, 2008 at 06:31:55PM +0200, Micha Lenk wrote:
  Athanasius wrote:
  Can you also provide an ltrace output and the file /tmp/gnucash.trace
  produced when running Gnucash with command line option --debug?
  I got your ltrace file. What's with the /tmp/gnucash.trace file created
  by Gnucash itself?
  
Ah I misunderstood.
  
Not helped by the fact I have TMPDIR set and, on checking strace
  output, it actually ended up elsewhere.  However, it's 0 bytes in size.
 
 The gnucash.trace file should not be empty if you run Gnucash with
 command line option --debug. Did you?

[EMAIL PROTECTED]:/var/tmp/athan;
19:08:53 0$ ls -l
total 8
drwx-- 3 athan athan 4096 Aug 27 12:12 gconfd-athan/
drwx-- 2 athan athan 4096 Aug 27 17:16 orbit-athan/
[EMAIL PROTECTED]:/var/tmp/athan;
19:08:55 0$ gnucash --debug
gnc.bin-Message: main: binreloc relocation support was disabled at configure 
time.

3:2008/08/27 19-08-58:gwen(17511):cryptkey.c:   37: Libgcrypt version mismatch
Could not initialize Gwenhywfar, aborting
Aborted
[EMAIL PROTECTED]:/var/tmp/athan;
19:08:58 0$ ls -l
total 8
drwx-- 3 athan athan 4096 Aug 27 12:12 gconfd-athan/
-rw--- 1 athan athan0 Aug 27 19:08 gnucash.trace
drwx-- 2 athan athan 4096 Aug 27 17:16 orbit-athan/
[EMAIL PROTECTED]:/var/tmp/athan;
19:08:59 0$ 

I can copy a 1.3MB file into that directory without problems.  Just in
case it springs to mind:

[EMAIL PROTECTED]:/var/tmp/athan;
19:10:22 0$ ulimit -a
core file size  (blocks, -c) 0
data seg size   (kbytes, -d) unlimited
scheduling priority (-e) 0
file size   (blocks, -f) unlimited
pending signals (-i) 16380
max locked memory   (kbytes, -l) 32
max memory size (kbytes, -m) unlimited
open files  (-n) 1024
pipe size(512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority  (-r) 0
stack size  (kbytes, -s) 8192
cpu time   (seconds, -t) unlimited
max user processes  (-u) 16380
virtual memory  (kbytes, -v) unlimited
file locks  (-x) unlimited
[EMAIL PROTECTED]:/var/tmp/athan;
19:12:21 0$ 

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 09:01:21PM +0200, Micha Lenk wrote:
 Athanasius wrote:
Not helped by the fact I have TMPDIR set and, on checking strace
  output, it actually ended up elsewhere.  However, it's 0 bytes in size.
  The gnucash.trace file should not be empty if you run Gnucash with
  command line option --debug. Did you?
  
  [EMAIL PROTECTED]:/var/tmp/athan;
  19:08:58 0$ ls -l
  total 8
  drwx-- 3 athan athan 4096 Aug 27 12:12 gconfd-athan/
  -rw--- 1 athan athan0 Aug 27 19:08 gnucash.trace
  drwx-- 2 athan athan 4096 Aug 27 17:16 orbit-athan/
  [EMAIL PROTECTED]:/var/tmp/athan;
  19:08:59 0$ 
 
 If you've set TMPDIR to /var/tmp you should get a gnucash.trace file in
 /var/tmp/ and not in /var/tmp/athan. However: You can give an explicit
 destination (including stderr, see gnucash --help) by using
 additionally --logto /path/to/filename.

  No, it IS set to /var/tmp/athan, it's part of some logic in my shell
startup to get a unique, and only accessible to me, temporary directory.
Oh, and this system is one big filesystem in /, 70GB free (I got sick of
eventually running out of space on /usr no matter how big a sensible
size, plus some I made it at install time).  /home/user/ is NFS-mounted
from my server machine though.

 To preclude that the error is related to your data file please give
 additionally --nofile for starting gnucash without opening any data file.

[EMAIL PROTECTED]:/var/tmp/athan;
20:10:42 0$ gnucash --debug --logto=stderr --extra --nofile
gnc.bin-Message: main: binreloc relocation support was disabled at
configure time.

3:2008/08/27 20-11-03:gwen(18379):cryptkey.c:   37: Libgcrypt version
mismatch
Could not initialize Gwenhywfar, aborting
Aborted
[EMAIL PROTECTED]:/var/tmp/athan;
20:11:03 0$ 

I've also done an 'aptitude reinstall gnucash' since then, just on the
off-chance something had gotten corrupted.  I dread trying that for
every dependency listed for gnucash, and then all their dependencies
too.  Things like Firefox 3 (to name a somewhat complex application) and
pidgin are still running without problems.

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 09:20:23PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  To preclude that the error is related to your data file please give
  additionally --nofile for starting gnucash without opening any data file.
  
  [EMAIL PROTECTED]:/var/tmp/athan;
  20:10:42 0$ gnucash --debug --logto=stderr --extra --nofile
  gnc.bin-Message: main: binreloc relocation support was disabled at
  configure time.
  
  3:2008/08/27 20-11-03:gwen(18379):cryptkey.c:   37: Libgcrypt version
  mismatch
  Could not initialize Gwenhywfar, aborting
  Aborted
 
 What about this:
 
 gnucash --debug --logto stderr --extra --nofile --log gnc=debug

[EMAIL PROTECTED]:/var/tmp/athan;
20:11:03 0$ gnucash --debug --logto stderr --extra --nofile --log gnc=debug
gnc.bin-Message: main: binreloc relocation support was disabled at configure 
time.

3:2008/08/27 20-24-34:gwen(18400):cryptkey.c:   37: Libgcrypt version mismatch
Could not initialize Gwenhywfar, aborting
Aborted
[EMAIL PROTECTED]:/var/tmp/athan;
20:24:34 0$ 

 If this does not succeed in producing debugging output: I'm running out
 of ideas for now... :-/

I'd hazard a guess it's more to do with one of the libraries, but have
no idea which of the many might be the root of the problem.  Did this
newer version of gnucash start pulling in/using some library it didn't
before?

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#496793: GNUcash: fails to run Libgcrypt version mismatch

2008-08-27 Thread Athanasius 
On Wed, Aug 27, 2008 at 10:36:15PM +0200, Micha Lenk wrote:
 Athanasius wrote:
  If this does not succeed in producing debugging output: I'm running out
  of ideas for now... :-/
  
  I'd hazard a guess it's more to do with one of the libraries, but have
  no idea which of the many might be the root of the problem.  Did this
  newer version of gnucash start pulling in/using some library it didn't
  before?
 
 Yes, Gnucash 2.2.6 added support for AqBanking and hence additionally
 pulls in at least libaqbanking and libgwenhywfar.
 
 The message Libgcrypt version mismatch is for sure generated by
 libgwenhywfar. But I see no reason at all why the gcry_check_version()
 called by libgwenhywfar fails (we've checked versions earlier without a
 single suspicious evidence). Maybe a backtrace helps to enlighten the
 situation by displaying the actual values handed over to
 gcry_check_version()
 
 Can you provide a backtrace by running Gnucash in gdb as described here?
 http://svn.gnucash.org/trac/browser/gnucash/tags/2.2.6/HACKING#L84
 Once Gnucash crashed you need to type bt at the gdb prompt to get the
 actual backtrace.

  Unfortunately everything seems to be compiled without debug symbols.
Again, I had to 'hack' /usr/bin/gnucash to actually run:

exec gdb gnucash-bin $@

Then:

21:42:50 0$ gnucash
GNU gdb 6.8-debian
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type show copying
and show warranty for details.
This GDB was configured as i486-linux-gnu...
(no debugging symbols found)
(gdb) catch fork
Catchpoint 1 (fork)
(gdb) set follow-fork-mode child
(gdb) run
Starting program: /usr/bin/gnucash-bin 
(no debugging symbols found)
(no debugging symbols found)
... much more of this
(no debugging symbols found)
(no debugging symbols found)
gnc.bin-Message: main: binreloc relocation support was disabled at configure 
time.

[New Thread 0xb67c7700 (LWP 20112)]
3:2008/08/27 21-43-11:gwen(20112):cryptkey.c:   37: Libgcrypt version mismatch
Could not initialize Gwenhywfar, aborting

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xb67c7700 (LWP 20112)]
0xe424 in __kernel_vsyscall ()
(gdb) bt
#0  0xe424 in __kernel_vsyscall ()
#1  0xb6b8c640 in raise () from /lib/i686/cmov/libc.so.6
#2  0xb6b8e018 in abort () from /lib/i686/cmov/libc.so.6
#3  0xb669a280 in ?? () from /usr/lib/libgwenhywfar.so.47
#4  0xb66e7198 in ?? () from /usr/lib/libgwenhywfar.so.47
#5  0x0001 in ?? ()
#6  0x002a in ?? ()
#7  0xb6cb8580 in ?? () from /lib/i686/cmov/libc.so.6
#8  0xb669a239 in ?? () from /usr/lib/libgwenhywfar.so.47
#9  0xb66efad4 in ?? () from /usr/lib/libgwenhywfar.so.47
#10 0xbff751b8 in ?? ()
#11 0xb66d356d in ?? () from /usr/lib/libgwenhywfar.so.47
#12 0xb66efad4 in ?? () from /usr/lib/libgwenhywfar.so.47
#13 0x086558e8 in ?? ()
#14 0xbff751c8 in ?? ()
#15 0xb662db98 in _init () from /usr/lib/libgwenhywfar.so.47
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb) 

 As a last resort and cheap workaround you can avoid the loading of the
 AqBanking module entirely by renaming the file libgncmod-aqbanking.so
 (in directory /usr/lib/gnucash/gnucash/) to something like
 libgncmod-aqbanking.so.broken. This way you might at least get Gnucash
 working again. To prevent future updates re-installing this file again I
 suggest to use dpkg-divert (as root) which makes this move known to Dpkg
 (future updates will then install this file to the renamed location):
 
 dpkg-divert --rename \
 --divert /usr/lib/gnucash/gnucash/libgncmod-aqbanking.so.broken \
 --add /usr/lib/gnucash/gnucash/libgncmod-aqbanking.so

  That does indeed work to get me up and running again.  I'll forgo the
divert, as that way I'll keep trying any new version to see if things
are still broken.

  Thanks!

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]

Bug#483677: rtorrent: Incomplete manual page

2008-06-18 Thread Athanasius 
   Not only that, but the man page totally failes to mention the
'watch_directory' and 'untied_directory' directives for 'schedule'.
They're right there in /usr/share/doc/rtorrent/examples/rtorrent.rc
though.

  This lead to me thinking there was no way to get an already-running
rtorrent instance to pick up a new torrent automatically, but there is.
Please (get upstream to) update the man page!

  This is for Etch, but the man page for Lenny is no better.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#128160: !address also affected

2007-09-03 Thread Athanasius 
  Another user on my affected machine found out that procmail's
!address recipe target is also adversely affected by the From ...
header that -d localpart adds.

  Of course it also turns out that the little voice at the back of my
head was correct that with exim I don't actually need to add the -d
argument at all.  It was cruft left over in our config from when we
originally switched to exim in the late 90s!

  But adding that header is still a definite bug.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#128160: In reply to Michelle Konzack [EMAIL PROTECTED]

2007-09-03 Thread Athanasius 
  formail, and any other program, may indeed be correct to treat email
with this 'From ' header in it in this manner, but that is not the
point.

  The point is that delivery with 'procmail -d user' inserts that very
header in the first place.

  As for 'From ...' headers.  That's the mbox format.  A line which
begins 'From ...' delimits each message in an mbox format file.  I don't
recall if a blank line (so \n\n) is necessary before this.
  Perhaps you should understand mbox format before you comment further.

  The testing I did, which I simplified the results of was to put a
quick recipe at the very start of my .procmailrc to spit out a copy of
the email as it was at that point.  This already had the 'From ...'
header in it and I now know exim wasn't responsible for this as I've
since changed its config to not use -d with procmail and now the header
no longer appears.  I also at this point had VERBOSE=yes in my .procmailrc.
I then took the file this first recipe produced and by hand fed it
through any other filter it would have hit.  First spamassassin, then
bogofilter, then the formail recipe.  I saved each stage of this.
  THAT is what I was describing in my original addition to this bug
report.

  Note that the procmail manpage says:

   -d recipient ...
This  turns  on  explicit  delivery  mode, delivery will be to the

   ...

   When in explicit delivery mode, procmail will generate a leading  `From
   '  line  if  none  is present.  If one is already present procmail will
   leave it intact.  If procmail is not invoked with one of the  following
   user  or  group  ids :  root,  daemon, uucp, mail, x400, network, list,
   slist, lists or news, but still has to generate or accept a new `From '
   line,  it will generate an additional `From ' line to help distinguish
   fake mails.

It's that very last described behaviour which is the bug.  That
'escaping' with a  before 'From' is what is used in the *BODY* of an
email if it begins with 'From', so as to not falsely think the email has
finished and move onto the next.  This is NOT what should happen in the
headers.
  Look at the code in from.c in the procmail source, the makeFrom()
function.  It uses FAKE_FIELD, which is defined to From .  So if
Debian wanted to fix this problem locally all it needs is that changing
to X-Procmail-From:  or somesuch.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#440284: GNU mailutils' mail requires \n at end of last line

2007-08-31 Thread Athanasius 
Package: nagios-common
Version: 1.4-3

The definition of notify-by-email and the like in
/etc/nagios/misccommands.cfg defines the body of the email to have a
last line NOT ending in a \n character.  GNU mailutils' mail will drop
these last lines.

For example, in the package source file
nagios-1.4/sample-config/default-object/misccommands.cfg.in:

command[notify-by-email]=/usr/bin/printf %b * Nagios @VERSION@ 
*\n\nNotification Type: $NOTIFICATIONTYPE$\n\nService: $SERVICEDESC$\nHost: 
$HOSTALIAS$\nAddress: $HOSTADDRESS$\nState: $SERVICESTATE$\n\nDate/Time: 
$DATETIME$\n\nAdditional Info:\n\n$OUTPUT$ | @MAIL_PROG@ -s ** 
$NOTIFICATIONTYPE$ alert - $HOSTALIAS$/$SERVICEDESC$ is $SERVICESTATE$ ** 
$CONTACTEMAIL$

needs to be:

command[notify-by-email]=/usr/bin/printf %b * Nagios @VERSION@ 
*\n\nNotification Type: $NOTIFICATIONTYPE$\n\nService: $SERVICEDESC$\nHost: 
$HOSTALIAS$\nAddress: $HOSTADDRESS$\nState: $SERVICESTATE$\n\nDate/Time: 
$DATETIME$\n\nAdditional Info:\n\n$OUTPUT$\n | @MAIL_PROG@ -s ** 
$NOTIFICATIONTYPE$ alert - $HOSTALIAS$/$SERVICEDESC$ is $SERVICESTATE$ ** 
$CONTACTEMAIL$

else the contents of $OUTPUT$ don't make it into the email.  Similarly
for other commands utilising @MAIL_PROG@ where that may be GNU
mailutils' mail program.

It's simple to test this with two quick commandlines:

/usr/bin/printf %b * Testing *\n\nWibble | /usr/bin/mail -s ** 
test ** user
/usr/bin/printf %b * Testing *\n\nWibble\n | /usr/bin/mail -s ** 
test ** user

The first email will NOT have the line 'Wibble' in it, the second will.


signature.asc
Description: Digital signature

Bug#128160: Also affects 'formail'

2007-08-28 Thread Athanasius 
  This bug with 'procmail -d user' also causes 'formail -i ...' to
further corrupt the message, but the bug is dependent on the headers
present before the non-RFC compliant 'From ' header.

  If we start with a message whose headers begin:

From [EMAIL PROTECTED] Tue Aug 28 20:14:03 2007
Return-path: [EMAIL PROTECTED]
Envelope-to: [EMAIL PROTECTED]
Delivery-date: Tue, 28 Aug 2007 20:14:03 +0100

And then deliver it using 'procmail -d athan', we get:

From [EMAIL PROTECTED] Tue Aug 28 20:14:03 2007
From athan  Tue Aug 28 20:14:03 2007
Return-path: [EMAIL PROTECTED]
Envelope-to: [EMAIL PROTECTED]
Delivery-date: Tue, 28 Aug 2007 20:14:03 +0100

Now, within procmail we run it through, bogofilter -e -p -u, and get:

From [EMAIL PROTECTED] Tue Aug 28 20:14:03 2007
From athan  Tue Aug 28 20:14:03 2007
Return-path: [EMAIL PROTECTED]
Envelope-to: [EMAIL PROTECTED]
Delivery-date: Tue, 28 Aug 2007 20:14:03 +0100

Then, as bogofilter isn't catching everything, we run it through,
spamassassin, and get:

From [EMAIL PROTECTED] Tue Aug 28 20:14:03 2007
X-Spam-Checker-Version: SpamAssassin 3.1.7-deb (2006-10-05) on 
bowl.fysh.org
X-Spam-Level: 
X-Spam-Status: No, score=-1.6 required=3.0 
tests=AWL,BAYES_00,UPPERCASE_25_50 
autolearn=ham version=3.1.7-deb
From athan  Tue Aug 28 20:14:03 2007
Return-path: [EMAIL PROTECTED]
Envelope-to: [EMAIL PROTECTED]
Delivery-date: Tue, 28 Aug 2007 20:14:03 +0100

Note how SpamAssassin has inserted some headers after the properly
formed 'From ' header, but before the improperly formed 'From ' header.

Now we run it hits a procmailrc rule to set Content-Type correctly for
inline PGP signed messages, which filters it through:

formail -i Content-Type: application/pgp; format=text; x-action=sign

and the headers end up as:

From [EMAIL PROTECTED] Tue Aug 28 20:14:03 2007
X-Spam-Checker-Version: SpamAssassin 3.1.7-deb (2006-10-05) on
bowl.fysh.org
X-Spam-Level: 
X-Spam-Status: No, score=-1.6 required=3.0
tests=AWL,BAYES_00,UPPERCASE_25_50 
autolearn=ham version=3.1.7-deb
Content-Type: application/pgp; format=text; x-action=sign

From athan  Tue Aug 28 20:14:03 2007
Return-path: [EMAIL PROTECTED]
Envelope-to: [EMAIL PROTECTED]
Delivery-date: Tue, 28 Aug 2007 20:14:03 +0100

Note the addition of a blank line, causing premature end of headers.
In fact formail has now not REPLACED the Content-Type header, it's added
a new one.  If the 'From ' line is filtered out before calling formail
then things work properly.

  It would seem that 'From ' should be turned into something
RFC-compliant, not 'From ', maybe 'Old-From_: ' ?

  Now to see if I can make procmail strip/change that badly formed
header

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#349001: Compiling with -ggdb3 instead of -O3 - no crash

2006-11-10 Thread Athanasius 
  This would seem to be something that's at least triggered by the
optimisation the debian build uses (-O3).  I just did an apt-get source
proftpd, which fetched proftpd_1.2.10-15sarge1.dsc and friends.  Then I
edited debian/rules therein changing the -O3 to -ggdb3, and commenting
out the dh_strip -a line so the symbols would stay put.  The aim of this
was to get a decent backtrace in gdb upon SEGV.

  However now I can't get it to crash, when before simply logging in,
issuing an ls and then logging out again was sufficient to cause the
SEGV, buried somewhere under pam_delay().
  Of course now I say that, after reinstalling the latest packages I had
before this, via:

dpkg -i proftpd-common_1.2.10-15sarge1.0.1_i386.deb 
proftpd-doc_1.2.10-15sarge1_all.deb proftpd_1.2.10-15sarge1.0.1_i386.deb

in /var/cache/apt/archives, I can't now get it to crash on quit.

  Anyway, it's either a bug triggered by the optimisation, or of course
possibly a compile bug with that optimisation.

  Additionally, if this is to do with the delay engine stuff, then see
bug #325304 and #308313 as mentioned in there.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#325304: Problem in the /etc/init.d/proftpd script

2006-11-10 Thread Athanasius 
  This is most likely due to keeping a locally-edited
/etc/init.d/proftpd during an upgrade.  At least the file:

proftpd-1.2.10/debian/proftpd/etc/init.d/proftpd

from proftpd_1.2.10-15sarge1.dsc source does correctly find
/var/run/proftpd.pid and shutdown the daemon.

  At least that's the problem I found just now on checking this on my
system.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#359629: NFS mounts without no_root_squash

2006-11-07 Thread Athanasius 
  Be aware that if ${HOME} is mounted over NFS and that doesn't allow
root the 'usual' arbitrary access to files then GDM will fail to fully
access ~/.Xauthority and fallback to the /tmp/.gdmrandom file.  This
is because for some such accesses it uses uid/gid 0, despite the fact
that it does switch to the user's uid/gid for initial filechecks and
creation of ~/.Xauthority if needed.

  It might be a bug that it uses uid/gid 0 for that later access given
if it is root it should be able to switch to the user again as needed,
and it does indeed switch for the initial checks/creation.

  For now I'm going to investigate why my NFS mounts aren't allowing
arbitrary root access so I can get back to ~/.Xauthority.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#349001: Proftpd 1.2.10-15sarge1.0.1 SEGVs

2006-10-14 Thread Athanasius 
  I'm seeing these on-disconnect SEGVs with 1.2.10-15sarge1.0.1.  I see
from the discussion here that it's a known issue, and at least some such
issues are fixed in a later version (presumably in testing/unstable).

  Any word on possible security impact of this ?

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#366262: Workaround

2006-10-11 Thread Athanasius 
  So long as the target repository allows for access via HTTP then
simply specify is as such in /etc/apt-proxy/apt-proxy-v2.conf.  My
problem was I had ftp:// URIs in there, so went through the chain:

aptitude
|
v
apt HTTP method (necessary because that's how apt-proxy works)
|
HTTP
v
apt-proxy
|
FTP (might be necessary if some repository is FTP only)
v
repository

And of course because aptitude was being told to use HTTP it did, and
the HTTP method translates ~ to %7e, which apt-proxy then doesn't
translate back on using an FTP method.

  Final solution, given some repositories may be FTP only?  Make
apt-proxy check for %escapes and translate back if about to use FTP to
actually fetch a file.  At least for this obvious case of %7e - ~ anyway.

  This will, of course, affect rsync and any other non-HTTP method in
apt-proxy as well.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#366262: Not just Ubuntu

2006-10-11 Thread Athanasius 
  I'm seeing this using a Debian/Sarge (or in fact backports) apt-proxy
as the target for packages on me Debian/etch (testing) system for
openoffice.org.

  With both apt-proxy 1.9.29 (sarge) and apt-proxy_1.9.35~bpo.1_all.deb
(sarge-backports) this is a problem.  At some stage in the request the ~
(tilde) gets translated to %7e.  This would be fine, except at least
ftp.uk.debian.org and ftp.fi.debian.org do not understand such
HTML-escaped characters and thus don't recognise:

openoffice.org_2.0.4%7erc3-1_i386.deb

as being the same as:

openoffice.org_2.0.4~rc3-1_i386.deb

So, is there any way to tell the HTTP method to NOT do such translation?
alternatively all the debian mirrors need to be kicked so as to accept
%7e as ~, and possibly other such translations.

Whoever came up with the ~ idea without encountering this problem in
testing needs kicking :P.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#308836: Known upstream bug

2006-07-26 Thread Athanasius 
  We've recently encountered this behaviour where if you specify a
--timeout=XXX and have a file that takes longer than XXX seconds to
transfer then rsync decides that the entire connection is bad and aborts
execution.

  On looking at the rsync bugtraq I found this:

https://bugzilla.samba.org/show_bug.cgi?id=2654

which identifies a bug to do with this in rsync 2.6.4, and they've
worked out a fix which is now in rsync CVS.  It's a one-liner, judging
by the comments.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


signature.asc
Description: Digital signature

Bug#309614: Successful woody to sarge upgrade

2005-05-18 Thread Athanasius 
On Wed, May 18, 2005 at 12:39:39PM -0400, Roberto C. Sanchez wrote:
 Athan,
 
 Out of curiousity, were you using a Debian-provided kernel-image or a kernel
 you compiled yourself?

  My own from vanilla source, so the failure to make sure I was on
modules-init-tools was entirely my own fault and nothing to do with
package dependencies.

-Ath
-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


pgpgjnM6aI2HV.pgp
Description: PGP signature

Bug#291466: mailman: Insisting on deleting qfiles is a bug

2005-02-17 Thread Athanasius 
Followup-For: Bug #291466
Package: mailman
Version: 2.1.5-5.backports.org.3

  mailman.preinst checks for any files under /var/lib/mailman/qfiles and
aborts installation if any are found.  Debconf message then advises to
simply delete the qfiles if needs be.  In many situations this is not a
valid option (loss of mail).

  An alternative that should be documented is to first backup and remove
the qfiles, perform the upgrade, then restore the backed up qfiles
whilst mailman is NOT running, and then run /usr/lib/mailman/bin/update
to ensure the qfiles are in latest format.  After that mailman may
safely be restarted.

  This worked for me anyway (upgrading backports version, but the fact
it was a backport is irrelevant to this problem and suggested solution).

-- System Information
Debian Release: 3.0
Architecture: i386
Kernel: Linux bowl 2.4.29 #1 SMP Wed Jan 19 17:44:40 GMT 2005 i686
Locale: LANG=C, LC_CTYPE=en_GB

Versions of packages mailman depends on:
ii  apache 1.3.26-0woody6Versatile, high-performance HTTP s
ii  apache [httpd] 1.3.26-0woody6Versatile, high-performance HTTP s
ii  apache-ssl [httpd] 1.3.26.1+1.48-0woody3 Versatile, high-performance HTTP s
ii  cron   3.0pl1-72 management of regular background p
ii  debconf1.2.35Debian configuration management sy
ii  exim-tls [mail-tra 3.35-3woody3  Exim Mailer - with TLS (SSL) suppo
ii  libc6  2.2.5-11.8GNU C Library: Shared libraries an
ii  logrotate  3.5.9-8   Log rotation utility
ii  pwgen  2.01-3Automatic Password generation
ii  python 2.1.3-3.2 An interactive object-oriented scr
ii  python2.2  2.2.1-4.7 An interactive object-oriented scr
ii  ucf1.07  Update Configuration File: preserv

-- 
- Athanasius = Athanasius(at)miggy.org / http://www.miggy.org/
  Finger athan(at)fysh.org for PGP key
   And it's me who is my enemy. Me who beats me up.
Me who makes the monsters. Me who strips my confidence. Paula Cole - ME


pgpZUdqVfzBWy.pgp
Description: PGP signature