[Gpsd-commit-watch] [SCM] GPSD annotated tag, release-1.90, created. release-1.90

[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-1.90 has been created
        at  846c886aea880f9c53a186f036ef188311bfa8c6 (tag)
   tagging  812c2d5cd6e466cf786596b64783d54fd981388c (commit)
 tagged by  Eric S. Raymond
        on  Sun Aug 15 04:52:24 2004 -0400

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

Derrick Brashear (63):
      first hack at dgps support
      current state of dgps
      now reads reliably
      possibly working dgps
      vaguely working dgps
      updates from Curt Mills
      clean up serial port, from Curt Mills
      work in progress
      update info
      fine, so now we have configure
      update
      fix conversion
      no really
      new port, various fixes
      more port change foo
      fix version
      update
      non-blocking i/o
      fix stupid compile problem
      oops
      fix port
      make port non-blocking
      fix y2k foo
      fix port stupidity
      include unistd in case...
      nothing major
      commit
      push
      reopen and reset serial on earthmate so apm resume works
      some more fixes
      more 1.06
      update for y2k problem from Willem van Schaik <address@hidden>
      >From Russell Nelson <address@hidden>
      make select() time out instead of sleepign forever waiting for input
      fix serial port speed calculation bug
      GPVTG support
      receiver which supports GPGGA and not GPGSA should not report no fix
      make things compile cleanly on modern systems. if we need old termio
      patch from Tom Mayo, N1MU
      document use of g for maidenhead grid squares
      so gpsd compiles for arm (empeg).
      update changelog. bump version. new release should wait, i have a few more
      ignore SIGPIPE
      don't write to fdout if it's not set
      ability to replay an nmea log
      build gpsplay
      fix libm test since sin doesn't return char
      update configure header
      fix gpsplay to work
      strip newlines
      Play nicely with Apple's X11 (which apparently lacks exported
      make clean should remove xgpsspeed
      Give us an install target.
      add switch for gpsd timeout
      update generated files
      need install-sh
      make it Werror-clean
      fix install target
      update for finding mkinstalldirs in srcdir
      update changelog
      update resources
      avoid null in nmea from address@hidden
      invert these so we parse the correct sentence type.

Eric S. Raymond (67):
      gpsd patch #1: Documentation additions and fixes.  Notably, there is
      This is ESR's gpsd patch #2: trivial fixes.
      Two of the READMEs have been folded into other docs; removing them.
      This is ESR's gpsd patch #3: option cleanup.
      ESR's gpsd patch #4:  This is the one that nukes globals.
      ESR's gpsd patch #5: This is the big refactoring patch.
      ESR's gpsd patch #5: This is the big refactoring patch.
      New -T l option of gpsd makes gpsplay unnecssary.  That's 450 lines of
      Add outdata.h to object dependencies where needed.
      Remco removed the 'g' command and added 'q'.  Bringing the
      Added interpretation of GPGLL sentence, made code able to interpret
      Add my changes to the README.
      Add manual page to default make target.
      Capitalize consistently with the man page.
      Remove dead command-line options.
      Somehow processPMGNST was not in the command dispatcher.
      Substituted y command (look for satellites) and z command (get channel 
quality)
      Well, here's a nasty little bug.  Not mine, i don't think.
      Remove manual page reference to Maidenhead system, as 'g' command is
      REmove a wishlist entry, as it's now implemented.
      Typo fixes and minor improvements.
      Define a common data structure for lifetime checks, and accessor
      Second step of data-validity simplification; separate freshness checks
      Most error reporting and logging is now done through a single function
      Clean up error-reporting levels.
      Finally, re-enable status error checking.
      Document the 'l' command.
      Condition out c and x, we're trying to discorage raw mode.
      Document -D better.
      Added more documentation on NMEA sentence types.
      Make prefix format of debug messages uniform.
      Better reporting of GPS opens in debug mode.
      Small but important change: Make the default baud rate work.
      Small but serious bug...make the default baud rate work.
      Clean up a debug-level change.
      Dcument new features, add HACKING.
      Namespace control.  Give all library functions an appropriate prefix.
      Get rid of two-line silly gps.h file.
      Get rid of xgpsspeed.h, it was rather pointless.
      Header consolidation.  Sinse there is really only one context
      Get rid of a duplicate rule.
      Add explanatory comment.
      Next step in turning this into a real library.
      Another step towards a clean library interface.
      Begin abstracting out the raw-mode handling.
      Begin abstracting out the raw-mode handling, by passing down a
      Next step in abstracting away the poll operation.
      Remove -s option.  GPS speed is fixed by its type.
      Almost ready to split off the library code.
      New library code successfully split out.  Now to clean up.
      More interface cleanup, make operations relative to a session structure.
      Add an end-of-sssion wrapup function.
      More cleanup -- consolidate two trivial functions.
      gps.c and xgpsspeed.c are rewritten to use the new library interface.
      Here's the library man page.
      Don'tr remove manpages in 'make clean'.
      Explain the known problems with the BU-2303.
      Document the error-reporting function better.
      No more global context in the library, it's all session pointers now.
      Change the name of the master structure.
      Fix Dragorn's wishlist item; gpsd now checks and exits gracefully if
      Dragorn's first bug is fixed.  The "corruption", I have not seen.
      Added -n option.
      Tweaks to new-feature list.
      Added -n option.
      Include cleanup, add some magic constants.
      Declare this a 2.0 prerelease.

Remco Treffkorn (56):
      Initial revision
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      Tf: Fixed the Y2K windowing stuff. Remember: for scanf use '&tmp' ;-)
      Tf.: restore terminal settings only for real terminals.
      include time.h instead of sys/time.h
      on the way to 1.07
      include time.h instead of sys/time.h
      *** empty log message ***
      *** empty log message ***
      *** empty log message ***
      Added SO_REUSEADDR to socket options
      another century bug in send_init.
      provide send_nmea stub
      minor cleanup
      added prototype for send_nmea
      moved code out of main. hopefully made it a bit more readable.
      various changes/fixes on the way to a better mousetrap
      mark fds closed (-1), so that rtcm data is not written to fd=0 on startup
      include string.h for mem* and strlen
      since djb fixed termios, no need to cast to (void *)
      added -Wall -Werror to CFLAGS
      pulled code out of main into 'deactivate'.
      pulled open code out; reopen now is deactivate - activate.
      no need to die if a client connection fails
      changed errexit to void
      Drop DTR on close. Before we depended on the ttys HUPCL setting.
      *** empty log message ***
      we use latd and lond, not the sign
      hmm, only 11 svs were listed because the box was not high enought
      use cfsetospeed() and cfsetospeed() to set the speed. struct termios can 
put the speed where ever it wants, and it does. this way we need to worry about 
details like this. will it ever end?
      *** empty log message ***
      Set status and mode 0 after GPS timeout (5 sec) - Cougar <address@hidden>
      typo
      removed include of varargs.h
      *** empty log message ***
      *** empty log message ***
      see CHANGELOG 20040421
      Need to set timestamp for altitude too.

Russ Nelson (3):
      Make sure that we don't segfault if the NMEA is badly formed.
      *** empty log message ***
      Add 'T' to return the NMEA track.

no-author (1):
      New repository initialized by cvs2svn.

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


hooks/post-receive
-- 
GPSD