[Gpsd-commit-watch] [SCM] GPSD annotated tag, release-2.20, created. release-2.20

[Eric S. Raymond]

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GPSD".

The annotated tag, release-2.20 has been created
        at  69cefcd57f6c681ca87539ad0e22d0e6f8587214 (tag)
   tagging  c13c3e34d99b7384a20d181b8330d1baeb9a99aa (commit)
  replaces  release-2.19
 tagged by  Eric S. Raymond
        on  Thu Mar 31 19:37:00 2005 -0500

- Log -----------------------------------------------------------------
Tag for public release.

Eric S. Raymond (94):
      unix_to_iso8601() should not do timezone correction.  That's the driver's 
job.
      Robert Jannsen's patch for uBlox support.
      Rob Janssen's patches for timezone issues.
      Oops...that ampersand is needed!  Fixes a Mac OS X compilation bug.
      Add documentation on the Holux.
      More about profiling and time measurement.
      We now get valid-looking profile data from MTD (04).
      Gary Miller's simplification.
      Ooops -- don't want to add a timezone offset where none is needed.
      Daemon no longer needs tzoffset directly.
      Change gps_session_t to gps_device_t.
      Follow through by changing 'session' to 'device'
      Change interface so there can be only one DGPS object per server...
      DGPS socket is no longer managed by individual sessions.
      Prevent theoretical memory leak.
      Another refactoring change, moves towards multisession.
      Fix markup error.
      Gradual move towards the experimental multisession version.
      More refactoring of the daemon main.
      Put correct guard on an error message.
      First merge of experimental multisession code -- all conditioned out.
      Second stage of multisession merge.
      Multisession merge complete.
      Stub version of K command.
      Head off a potential buffer overflow.
      Make the code for converting GMT dates independent of Unix time-zone 
weirdness.
      Quiet a security warning on OpenBSD.
      Pacify the compiler.
      License changed to BSD.
      Make the talker ID prefix a compile-time parameter.
      Make sure all drivers set the packet tag field.
      Document common talker IDs.
      Added a big list of talker IDs.
      Don't block non-GP values of TALKER1 and TALKER2 from working.
      Fix nmea_parse to no longer require alloca().
      More cleanup.
      Make NMEA parser independent of talker ID.
      Recognize both GP and II packet leaders.
      Document recognition of the II leader.
      Everything except splitplot works now.
      Typo fix.
      Restored uninstrumented plot, with explanation.
      Typo fix.
      Split data into header + data, preparatory to using -e.
      All the report generators for the plotters are refactored to emit strings.
      Plot programs are now emitted to stdout.  The split plot is still buggy.
      Fix range bug in split-plot generation.
      Profile all packets, not just the ones with internal timestamps.
      Rob Jannsen's second round of time patches.
      gpsprof -r option is gone.
      Restore gpsd import for archival purposes.
      Cycle timing moved into gpsprof.
      gpsprobe is obsolete.  Remove references to it from other files.
      gpsprobe and its documentation are history.
      gpsd.py goes away.
      Make the actual gpsd.py code go away.
      Feature note.
      More on common talker IDs.
      Accept the IN talker ID.
      Another step towards multisession.
      Add a usage example.
      Prepare for release.
      Document the multicast possibility.
      Better time selection, from Robert Janssen.
      Lose some system headers that are no longer needed.
      Document the new X return.
      Add tag field to 'o' response.
      Document a limitation.
      Limit some field widths...
      Gary misunderstood what the tag field is for.
      OK, let's try this again.  Only you can prevent malicious buffer overruns!
      Improved documentation.
      Much more on accuracy.
      Correct a bug in the state machine.
      Russ Nelson's fix.
      Add navigation-parameter display to sirfmon.
      Add experimental 'a' command to toggle reporting of 50BPS data.
      Fixed a bug by removing some code.
      Abstract handling two filesets into macro calls.
      Next step towards multisession, have different version of the macros.
      Create new status flag to avoid spurious error messages in NMEA mode.
      Fix incorrect error message.
      Make MULTISESSION code initialize correctly.
      Markup fix.
      Finish trial implementation of MULTISESSION F command.  No effect on
      Fix buggy 'b' command.  Fill in the channels field in the 0x80 message.
      We can now get 50BPS subframe data.
      Avoid a compile error in some odd configurations.
      Yet another library binary incompatibility...
      Oleg Gusev's geoid approximation. We're not using this yet.
      FV18 is back because we can enable ZDA if we know we're connected to 
one...
      New FAQ item on date flicker.
      Finally ready to ship.
      Typo fix.

Gary E. Miller (10):
      Garmin: report Zulu time instead of PST
      Garmin: revert last change, it broke ntpshm.
      unix_to_iso8601() was returning localtime, not zulu time.
      get the comma in the right place if no altitude.
      Without the trailing newline spaceplot never generated any plot w/o 
altitudes.
      spaceplot now adds the plot command to the plot file as a comment.
      unbalanced tag
      Garmin: restore GSV.  This is needed for gpsdrive, etc.
      gpsprof: removed code that should never have been commited.  Sorry.
      Garmin: fill in session->tag.

-----------------------------------------------------------------------


hooks/post-receive
-- 
GPSD