1. 02 Aug, 2012 1 commit
  2. 21 May, 2012 2 commits
  3. 24 Jun, 2011 1 commit
  4. 11 May, 2011 1 commit
    • Jelte Jansen's avatar
      [trac743] Replace direct output by liblog calls · c47f23e5
      Jelte Jansen authored
      While creating messages and extended descriptions, I noticed a few potential problems, which I also fixed;
      - If the config manager does not accept the modules specification, or if the
        'get current config' command returns with an error, the system used to
        ignore that fact apart from printing an error, and continue running. It
        seems much better to explicitely fail, since the module would be running
        independent of what its config is supposed to be, which would be really
      - If the module cannot parse a command it gets over the command channel,
        it used to send back an error. If we have multiple modules listening on
        the same channel, this would result in multiple errors being returned
        It will now log an error, but it will not send back an answer
  5. 10 May, 2011 1 commit
  6. 19 Apr, 2011 2 commits
  7. 20 Dec, 2010 1 commit
  8. 18 Oct, 2010 1 commit
  9. 10 Oct, 2010 1 commit
  10. 01 Oct, 2010 1 commit
  11. 20 Sep, 2010 1 commit
  12. 19 Aug, 2010 1 commit
  13. 17 Aug, 2010 1 commit
  14. 16 Aug, 2010 1 commit
  15. 02 Aug, 2010 1 commit
    • Jelte Jansen's avatar
      better handling of bad/old config data file · f12da928
      Jelte Jansen authored
      b10-cfgmgr now prints the specific error on startup (and not a stack trace), which includes the full path to the bad file
      moved ConfigManagerData.CONFIG_VERSION to config_data.BIND10_CONFIG_DATA_VERSION
      bumped version to 2; including automatic 'update' (versions 1 and 2 are the same except for the format itself, in which case you can't even get to the actual value)
      checks automatically use the 'current' value
      git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac294@2609 e5f2f494-b856-4b98-b285-d166d9295462
  16. 20 Jul, 2010 1 commit
  17. 02 Jun, 2010 2 commits
  18. 28 May, 2010 1 commit
    • JINMEI Tatuya's avatar
      addressed some portability issues with non boost ASIO: · 737c49cf
      JINMEI Tatuya authored
       - suppress the gcc unused parameters warning selectively (as a workaround) in a most portable way
       - introduce B10_CXXFLAGS as the default to AM_CXXFLAGS to make the first change possible (that's why I modified so many other Makefile.am's even if they are irrelevant to asio)
       - remove dependency on boost system library: it was only needed for the boost version of ASIO.
       - hide details that require ASIO related definitions from xfr_client.h to avoid being hit by the 'unused parameter' problem accidentally.
      git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac168@1953 e5f2f494-b856-4b98-b285-d166d9295462
  19. 19 May, 2010 1 commit
  20. 20 Mar, 2010 1 commit
  21. 19 Mar, 2010 1 commit
  22. 18 Mar, 2010 1 commit
  23. 13 Mar, 2010 1 commit
    • Jeremy C. Reed's avatar
      Allow build outside of read-only source tree. · 3914654f
      Jeremy C. Reed authored
      Identify files that need to be included in the tarball.
      Also change some from Makefile.am entries to identify that are
      not in the tarball (since later generated during build).
      Includes needed for headers in src tree and in build tree
      (because some includes are generated).
      Also includes specifically for "dns" needed since headers
      in src tree don't see the ones in build tree and vice versa.
      Have the gen-rdatacode.py script read placeholder from src tree
      and save generated files into build tree. Okayed by Jinmei.
      Identify a few files that should be removed on a clean.
      git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1389 e5f2f494-b856-4b98-b285-d166d9295462
  24. 12 Mar, 2010 1 commit
  25. 11 Mar, 2010 1 commit
  26. 10 Mar, 2010 1 commit
  27. 05 Mar, 2010 1 commit
    • Jeremy C. Reed's avatar
      Restructure trunk again. · 2e119c7e
      Jeremy C. Reed authored
      This is based on ideas done in experiments/jreed-layout, but
      I manually did this again.
      Moved around directory structures.
      A single python subdirectory under src/lib for single "isc" module.
      No "cpp" directories.
      No pyshared or includes symlink trees.
      I tested running the c++ tests and python tests.
      I tested running run_bind10.sh from source tree (with nothing installed).
      I tested running bind10 from installed (after moving source).
      git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1120 e5f2f494-b856-4b98-b285-d166d9295462
  28. 29 Jan, 2010 1 commit
  29. 26 Jan, 2010 1 commit
  30. 30 Oct, 2009 1 commit