zerotracepen (2.12)

  • Major changes

    • Completely remove I2P. 😔 We have decided to remove I2P (see

      11276) due to our failure of finding someone interested in

      maintaining it in Zero Trace Pen (Closes: #12263).

    • Upgrade the Linux kernel to 4.9.13-1~bpo8+1 (Closes: #12122).
    • Security fixes

      • Upgrade Tor Browser to 6.5.2 based on Firefox 45.9. (Closes:

        12444)

      • Mount a dedicated filesystem on /var/tmp, to mitigate the
        hardlinks permissions open by the user-tmp abstraction. See
        https://labs.riseup.net/code/issues/9949#note-23 for dezerotracepen
        (Closes: #12125).
      • Protect against CVE-2017-2636 by disabling the n-hdlc kernel
        module (Closes: #12315).
      • Ensure /etc/resolv.conf is owned by root:root in the SquashFS.
        lb_chroot_resolv will "cp -a" it from the source tree, so it
        inherits its ownership from the whoever cloned the Git
        repository. This has two problems. First, this results in unsafe
        permissions on this file (e.g. a Vagrant build results in the
        'amnesia' user having write access to it).
      • Upgrade libjasper1 to 1.900.1-debian1-2.4+deb8u3
      • Upgrade gstreamer and its plugins to 1.4.4-2+deb8u1.
      • Upgrade eject to 2.1.5+deb1+cvs20081104-13.1+deb8u1.
      • Upgrade imagemagick to 8:6.8.9.9-5+deb8u8.
      • Upgrade pidgin to 2.11.0-0+deb8u2.
      • Upgrade samba to 2:4.2.14+dfsg-0+deb8u5.
    • Minor improvements

      • Don't add the live user to the "audio" group. This should not be
        needed on a modern Linux desktop system anymore (Closes:

        12209).

      • Install virtualbox-* 5.1.14-dfsg-3~bpo8+1 from our custom APT
        repository (Closes: #12307).
      • Install virtualbox-guest- from sid. The version currently in
        jessie-backports is not compatible with Linux 4.9, and there's
        basically no chance that it gets updated (the maintainer asked
        for them to be
        removed* from jessie-backports) (Closes:

        12298).

      • Pull ttdnsd from our custom APT repository. It's gone from the
        TorProject one. We removed ttdnsd on feature/stretch already, so
        we'll need to pull it from our custom APT repository only for
        the next 3 months.
      • Clean up libdvd-pkg build files, again. This cleanup operation
        was mistakenly removed in commit c4e8744 (Closes: #11273).
      • Install gnome-sound-recorder (Closes #10950). Thanks to Austin
        English austinenglish@gmail.com for the patch!
      • Stop restarting tor if bootstrapping stalls. It seems tor might
        have fixed the issues we used (see: #10238, #9516) to experience
        with the bootstrap process stalling and requiring a restart to
        kickstart it (Closes: #12411).
      • tor.sh: communicate via the UNIX socket instead of TCP port.
        This makes the library usable when run inside systemd units that
        have PrivateNetwork=yes set.
      • Get tor's bootstrap progress via GETINFO instead of log
        grep:ing.
      • Upgrade tor to 0.2.9.10-1~d80.jessie+1
    • Bugfixes

      • mirror-pool-dispatcher: bump maximum expected mirrors.json size
        to 32 KiB. This fixes an error where Zero Trace Pen Upgrader would
        complain with "cannot choose a download server" (Closes:

        11735).

    • Build system

      • Retry curl and APT operations up to 20 times to make the ISO
        build more robust wrt. unreliable Internet connectivity. Thanks
        to Arnaud arnaud@preev.io for the patch!
      • Install ikiwiki from jessie-backports, instead of our patched
        one. Our changes were merged in 3.20161219, and jessie-backports
        now has 3.20170111~bpo8+1 (Closes: #12051).
      • Fix FTBFS when installing a .deb via config/chroot_local-packages
        by being more flexible when matching local packages in the apt
        list file (Closes: #12374). Thanks to Arnaud arnaud@preev.io
        for the patch!
      • auto/build: support Stretch's GnuPG v2 keyring filename.
    • Test suite

      • Try possible fix for #11508. IPv6Packet:s' source is accessed by
        .ipv6_saddr, not ip_saddr (that's for IPv4Packet). So, let's
        just try and see which one of the two each packet has, because
        one of them must be there! Also, given that UDPPacket can be
        either IPv4 or IPv6 it seems safest to try to parse each packet
        as IPv6Packet first -- that way we keep looking at transport
        layer protocols for IPv4 only, and treat everything IPv6 as the
        same, which makes sense, since we should block all IPv6, so
        everything should be treated the same at all times.
      • Changes due to #12411:
        • Raise special exception for Tor bootstrap failures.
        • Remove obsolete debug logging now that we don't log anything
          interesting for restart-tor any more.