[Gpsd-commit-watch] [SCM] GPSD annotated tag, release-2.17, created. release-2.17

[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-2.17 has been created
        at  859dbe3c6125a5e1a281f5385ffffa1f88800196 (tag)
   tagging  03ea3410e54c7978124578c7e54adf8e3c6565e0 (commit)
  replaces  release-2.16
 tagged by  Eric S. Raymond
        on  Wed Mar 16 21:51:05 2005 -0500

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

Derrick Brashear (4):
      remove -Wextra. uncommon flags should be checked and not assumed
      __FD_SETSIZE is not so much portable.
      Three small fixes.
      oops, this was fixed in svn already

Eric S. Raymond (56):
      Fix directory creation in scriptlet.
      Mac OS port fixes.  Describe WAAS/EGNOS in the GPS Hacking document.
      Email address removed at author's request.
      Big update of to-do list.
      Outlinize the HACKING file.
      Reapply the bug fix Derrick stepped on.
      We don't actually need O_SYNC.
      Allow gpsfake to accept subsecond cycle times.
      Revert -Wextra to -W, older gccs need this.
      Packet engine rewrite to make automatic reconnect work again.
      GPSes only report to hundredths of a second, format accordingly.
      Quiet down the compiler.
      Quiet down the compiler.
      -Wextra to -W fix. Also, check for shm and ipc headers
      NTP interfacing is a todo item.
      First step towards implementing -'n' command -- define the driver slot.
      Revert part of the 'n' first step, it broke SiRF capability.
      Better message in case of driver mismatch.
      Better termination on signal.
      Prevent input from overrunning the packet buffer.
      Add a FAQ.
      gps_poll now handles multi-line resonses.
      Remove NTPD item pending Rob Janssen and Gary Miller figuring out
      Enhancements to gpsfake, and an NMEA packet fix.
      Push back the packet that packet_sniff syncs up on.  More gpsfake 
improvements.
      N command is now fully implemented.
      Compute climb/sink from the NMEA report.
      Document climb/sink computation.
      First version of 'o' command.
      Cut down on redundant information shipped to watchers.
      Change O timestamp to a float.  Bound the precision of various other 
reports
      Make sure sirfmon gets distributed.
      Suppress useless GLL and VTG sentences in SiRF NMEA.
      First step towards timestamp bundling.
      Add another step to the diagnostic sequence.
      Replace protocol redesign discussion with sketch of 'o' command.
      Restore watching on GLL for the Garmin 48.
      Give gpsfake the ability to handle comments.
      Oops -- remove the breadcrumbs.
      Remove unused code.
      Simplify the SirRF control-write logic and put in some breadcrumbs.
      Insert timeout needed to make baud-switching work reliably.
      Refactor the baud-rate-change code...
      Get rid of verbose command and an unnecessary control variable.
      Cleanup.
      Clean up device initialization.
      Fix severely screwed up xgps display.
      Unbreak the Garmin driver.
      SiRF rate-limit command is brain-dead, so we'll do it in the driver.
      Give this a good long time to sync up, for when the GPS gets into an
      Make sure the out-of-band packet-type value gets set.
      Revised test on error modeling.
      Put the initialization further back...
      Increase the packet_sniffer timeout.
      Force an initializer call.
      Version bump.

Gary E. Miller (2):
      Various Garmin fixes.
      Avoid  with 'nan's.

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


hooks/post-receive
-- 
GPSD