Please merge nbd 1:2.9.13-2(main) from debian unstable(main)

debian has a new version to be merged

 nbd (1:2.9.13-2) unstable; urgency=low

   * The "Black Finger DebConf" release.
   * Add avr32 to nbd-client-udeb, too. Oops. Closes: #533694, hopefully
     for good.
   * nbd-client.README.Debian: add a bit more clarification on how
     root-on-NBD is supposed to work.
   * nbd-client.init.d: parse /proc/cmdline rather than /proc/mounts
     (the same way as how nbd-client.initrd does it) to figure out what
     the NBD root device is, so that root-on-LVM-on-nbd or something
     equally batshit insane does not confuse the sendsigs.omit.d
     generating code.
   * nbd-client.init.d: add support for an OMITKILL variable that will
     allow to specify extra nbd devices that should receive the same
     protection as for the above.
   * nbd-client.config: fetch KILLALL value from /etc/nbd-client (if it
     exists) and use it to db_set nbd-client/killall
   * nbd-client.postinst: fix so that KILLALL is correctly written to
     /etc/nbd-client, even if the number of devices is zero.
   * The above four were inspired by conversations with Vagrant
     Cascadian, while he was working on LTSP support in Debian. Thanks!
   * Updated Swedish translation. Thanks, Martin Bagge; Closes: #534246.

 -- Wouter Verhelst <email address hidden> Fri, 17 Jul 2009 01:25:26 +0200

nbd (1:2.9.13-1) unstable; urgency=low

  * New upstream release

 -- Wouter Verhelst <email address hidden> Thu, 09 Jul 2009 11:18:51 +0200

nbd (1:2.9.12-3) unstable; urgency=low

  * Add Finnish translation. Thanks, Esko Arajärvi; Closes: #533451.
  * Add 'avr32' to nbd-client's 'Architecture:' list. Closes: #533694.

 -- Wouter Verhelst <email address hidden> Sat, 20 Jun 2009 16:30:40 +0200

nbd (1:2.9.12-2) unstable; urgency=low

  * Steal a patch from git HEAD so that nbd-tester-client doesn't try to
    do unaligned access on SPARC.
  * Add German translation too. Oops. Closes: #530835.

 -- Wouter Verhelst <email address hidden> Sat, 06 Jun 2009 02:15:16 +0200

nbd (1:2.9.12-1) unstable; urgency=low

  * The "spring cleaning" release.
  * New upstream release.
- The meaning of the -swap option has been changed; rather than
       trying to use a kernel patch which hasn't been maintained since
       the Linux 2.1.something days, use mlockall() to make sure we don't
       get swapped out. Closes: #409530.
       - Steal patch from git HEAD to document this change in nbd-client(8).
     - Removed (corrupt) winnbd.zip. Closes: #473823.
   * The Debconf templates have been reviewed by the debian-l10n-english
     team; incorporate those changes. Closes: #528476, 430701, 530711.
   * Updated debconf translations, to reflect the above:
     - Dutch, by myself
     - Swedish, by Martin Bagge; Closes: #529859.
     - Czech, by Miroslav Kure; Closes: #530236.
     - French, by Christian Perrier and the French l10n team; Closes: #530332.
     - Portuguese, by "Traduz - Portuguese Translation Team"; Closes: #531303.
   * New debconf translations:
     - Russian, by Yuri Kozlov; Closes: #531203.
   * nbd-client: install nbd-client.modprobe using debhelper rather than
     using our own 'install' invocation. This takes care of giving it the
     proper name (re recent module-init-tools) and moving files from the
     old to the new location. Depend on debhelper (>= 7.2.2) to make sure
     it has the required functionality to handle this correctly.
   * nbd-client.init.d: remove '-k' option from modprobe. This was once
     useful in the 2.4 days, but module-init-tools modprobe has never
     needed that option, and now produces an error message when it's
     used. Closes: #531620.
   * (Conditionally, compliant with the new DEB_BUILD_OPTIONS' "nocheck"
     option) re-enable the test suite. It did /not/ give false positives,
     ever. However, it did indeed break at some point...
   * Update to debhelper compat level 7.
   * nbd-client.postinst: don't use full path to update-initramfs.
   * copyright: refer to common-licenses/GPL-2 rather than the versionless
     variant.
   * That leaves one TODO for Policy 3.8.1: initscripts must not error
     out when 'start' is called and the daemon is already running. This
     will require some work (it might be there already, but I'm not sure
     and would need to test extensively).

 -- Wouter Verhelst <email address hidden> Fri, 05 Jun 2009 20:42:25 +0200