gpsd-commit-watch
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[gpsd-commit-watch] [SCM] GPSD annotated tag, release-3.12, created. rel


From: Eric S. Raymond
Subject: [gpsd-commit-watch] [SCM] GPSD annotated tag, release-3.12, created. release-3.12
Date: Sun, 22 Feb 2015 16:17:59 +0000

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-3.12 has been created
        at  f51eb2758571f539bfa4394a5fc8ca43d455dd64 (tag)
   tagging  6683cd11c0c5c59f161d6a44e08497eb663a7452 (commit)
  replaces  release-3.11
 tagged by  Eric S. Raymond
        on  Sun Feb 22 11:17:34 2015 -0500

- Log -----------------------------------------------------------------
Tagged for external release 3.12
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iD8DBQBU6gEerfUW04Qh8RwRAvb1AJ9gV1gA9HZsoSuxCHXTsr6u0ZwdpgCffur2
p7gV3fuMFydsJjnhWNSxfGI=
=1rZn
-----END PGP SIGNATURE-----

Ahti Heinla (1):
      Fix RTCM2 detection in packet.c

Beat Bolli (2):
      NEWS, build.txt: typo fixes
      NEWS, build.txt: remove trailing whitespace

Daniel Willmann (1):
      Exit from gps_sock_mainloop() if connection to gpsd is lost

Eric S. Raymond (349):
      3.12 development cycle begins.
      libgps version bump to 22: AIS types 6 and 8 get 'structured' member.
      NEWS update.
      Typo fixes.
      Full implementation of AIS 'structured' bit.  All regression tests pass.
      Remove a fulfilled todo item.
      Device list fully moved out of terminal union; all regression tests pass.
      Typo fixes.
      PPS is no kinger marked unstable/experimental.
      Reduce space required by devicelist now that it's in the main structure.
      Drop the obsolete tag field from JSON output.  Bump the minor protocol 
number.
      A lesson in why not to write code with easy-to-miss side effects.
      Eliminate the tag field entirely, even internally.
      Proper use of the Type 8 structured bit.
      Clean up some FIXMEs.  All regression tests pass.
      Remove some obsolete FIXMEs.
      Properly free devices when select() sees bad fds for them.
      Fix trivial but fatal errors introduced by last commit.
      Add a minor error check to AIVDM parsing.
      Clean up another FIXME.
      Remove an obsolete TODO.
      splint/ccpcheck cleanup.
      Attempted fix for second 'n' not working in gpsmon.
      Add some clarifying notes to the gpsmon documentation.
      Remove a fulfilled TODO.
      Recreare the driver union, curtting per-device storage requirements.
      Add slow=yes option for regression tests in an attempt to avoid flakeouts.
      Correct computatiojn of build timestamps.
      Add some TODO items.
      Split firmwqre version query out of the init method.
      splint cleanup.  All regression tests pass.
      Change names in nmea0183 driver to make them GNSS/GLONASS agnostic.
      Allow gpsedecode to dump SKY reports.
      Accumulate GNSS+GLONASS satelite reports.
      Make it more obvious when lack of prerequisites forces qt off.
      GNSS and GLOSS SKY reoport are now merged
      Add a note about BSD init devices.
      Qt support default chaned to False pending a build patch.
      Clean up some timebase assumptions (code not yet used).
      Proper integral-type conversion when stashing previous-fix time.
      Eliminate reverse linkage of gpsd_write. All regression tests pass.
      Eliminate a mention of gpsd_report.
      Add a feature request.
      Temporarily revert code for freeing devices on bad fd in select.
      Fix for previous weird bug.
      Satisfy an order dependency in SiRF initialization.
      Add a caution about use of tag blocks.
      Minor refactoring step in the UBX driver.
      A steo topwrds eliminating the reverse linkage of gpd_report().
      Another step away from reverse linkage.  No logic changes.
      Another step away from reverse linkage.  No logic changes.
      Structure member name change: packet -> lexer. No logic changes.
      Fix a misleading variable name. No logic changes.
      Follow throughon a previous name change.  No logic changes.
      Modify packet.c so it no longer requires gpsd_report().
      Eliminate gpsd_report() from isgps.c.  All regression tests pass.
      Introduce struct errout_t to encapsulate error-reporting hooks.
      The big leap.  Most instances of gpsd_report are replaced...
      More gpsd_report removal.  All regression tests pass.
      All gpsd_reportcalls are gone.  Only the unused definitins are left.
      Reverse linkage of gpsd_report() is abolished. All regression tests pass.
      Now that the transition is done we can restore the gpsd_report name.
      Clean up duplicarive code in gpsmon.
      splint/cppcheck cleanuo after the gpsd_report change.
      Add removal of config.log to clean production.
      Documentation typo fixes.
      Add to the support list.
      Typo fix.
      Appease the compiler.
      Wrote code and test for a function to left-shift bit arrays).
      Properly handle addressed versions of AIS Type 25 and 26.  Untested, alas.
      Document new feature.
      Appease the compiler.
      Add some warnings about SiRF-IV flakiness, and remove a related TO-DO 
item.
      Type 25 and 26 improvements required a rebuild of the AIS regression 
tests.
      gittip -> gratipay.
      Add -lgps to gpsdlib build line, for toolchains with weak linkers.
      Added a list of supported platforms to build.txt.
      Add a note to the Time Service HOWTO about the GR-601-W.
      History update.
      Remove a qualifying comment left over from the old polling interface.
      Added notes on Android.
      nmea option renamed to nmea0183.
      In SConstruct, 'strip' becomes 'nostrip' and the sense is inverted.
      'minimal' option is half-implemented; it does the right thing in scons, 
but...
      Minimal option requires conditionalization of some code.
      Build and regressions now work with minimal either on or off.
      Document the new 'minimal' option.
      Fix incorrect comment.
      cheapfloats -> nofloats. All regression tests pass...
      Coverity caught a subtle bug in bits.c::shift_left().
      Magic-number elimination.
      Use a standard name for a magic number. All regression tests pass.
      More magic-number elimination.
      Magic-number elimination.  All regression tests pass.
      Typo fixes.
      Typo fixes.  All regression tests pass.
      Make -D option of gpsdecode effective again.
      A bug fix for error modeling when NMEA 0183 reports empty DOP fields.
      Fix a subtle error in DOP computation.  This makes a prompt release a 
good idea.
      Improve some instrumentation.
      Second attempt at DOP computation fix.
      Begin factoring out matrix algebra so it can be unit-tested.
      Work on unit test for matrix inversion.
      Integrate a matrix-algebra regression test.
      Check for vanishing determunant in the tests.
      Inversion-test with off-diagonal elements.
      Permit error messages from gpsctl again.
      Change the traversal order of the DOP computation loop...
      More instrumentation.
      Fix stupid finger error in DOP computation.
      Add -n option to gpsdecode to dump pseudo-NMEA.
      Implement and test extended JSON array parsing.
      Condition out extended JSON features.
      Fix an odd little bug in pseudo-NMEA reporting.
      Refactor representation of satellite data into an array of structs...
      Remove some custom code, at the cost of introducing a glibc depenency.
      Do the right thing if we're not using glibc - use local timegm().
      Reconcile json.c with microjson's copy.  All regression tests pass.
      Appease splint.  All regression tests pass.
      Remove some unnecessary copies, All regression tests pass.
      Cosmetic cleanup.  All regression tests pass.
      Remove a redundant check.  All regression tests pass.
      Additional matrix-inversion test.
      Resynchronize with microjson.
      More microjson resync.  All regression tests pass.
      Note for anyone wo might want to re-use this.
      GNU libc timegm() turns out to be buggy when DST is on (glibc-2.19).
      Be clearer about what we're testing.
      Fix a misnamed test.
      Better way to avoid glibc dependency.
      Documentation correction.
      Fix a minor bug reported by the author.
      Sanity check for too many devices on the command line.
      Typo fix.
      Our policy about security bugs.
      Minor fixes for AIS code and fields documentation.
      AIS fixes required three check file rebuilds.  All regression tests pass.
      Add area maintainers list.
      In gpsmon, prevent a spurious warning on Bluetooth devices.
      Add a script to generate test pairs for matrix inversion.
      Fix a brown-paper-bag typo.
      Add a SiRF-IV regressio n test (NMEA mode).  All regressiopn tests pass.
      Fix a namespace-pollution issue that causes OS X build failures.
      Fix a bug in Superstar satellites-used compuation.
      Revise SiRF driver to be get satellites-used info from the skyview 
sentence.
      Fix a bug that produced duplicate satellite IDs in psesudo-NMA GSA.
      Code simplification.
      Safety fix: ensure that SBAS satellites are never used in error modeling.
      Another minor code simplification.  All regression tests pass.
      Rename for clarity.
      Improved internal documentation in the navcom driver.
      In the navcom driver, get satellites used from the Channel Status packet.
      The particular Navcom binary protocol we  support is called NCT.
      Almost entirely eliminate use of sats_used array.
      Fix the build in some unusual cases.  All regression tests pass.
      Last change to STARTSWITH was buggy.  All regression tests pass.
      Clean STATUS_DGPS_FIX out of the public interface.  All regression tests 
pass.
      Refactor the way NTP shared memory segments are addressed.
      Revert to using timegm() following Hal Murray's report of Fedora 
flakiness.
      Easy splint cleanups.  All regression tests pass.
      Comment typo fix.
      More splint cleanup.  All regression tests pass.
      Coverity/splint cleanup.
      Complete splint cleanup.
      Introduce an adaptive delay into the main loop to prevent select buzzing.
      cppcheck cleanup.
      Revert to using internal mkgmtime() as in 3.11.
      Cleaned up, improved sourcetype() code.  All regression tests pass.
      Fully initialized fields in some time structures.
      Prevent uninitialized-field lossage in the Superstar driver.
      Prevent uninitialized-field lossage in the Zodiac driver.
      Fix bug reported on the dev list resulting from NTPSHMSEGS > MAX_DEVICES 
* 2.
      In build.txt, note that 'python' must be on your path.
      Coverity cleanup.  All regression tests pass.
      Elminate double dump of yaw from a compass.
      In client library, "dip" member was missing from ATT parsing.
      Under Linux, delay client socket close until pending data has shipped.
      Ubuntu 11.0 is no longer 'recent'.
      SIOCOUTQ build fix required for older Ubuntu versions.
      Dump platform ID and delay parameters before doing test runs.
      Quieter bits testing.
      Document recent changes.
      API major version bump.
      pylint cleanup.  All regression tests pass.
      Skip writespace *before* looking for word length.
      Default qt build on again, since we've merged a header guard patch.
      Prevent TIOCMIWAIT from beinf sent to non-serial devices.
      Pretend CAN devices are pseudo-ttys to disable adaptive delays.
      Implement and document the buzzkill option.
      Improved adaptive-delay logic.  Don't insert them if CAN devices are in 
play.
      Apply de Morgan's law properly.
      Documentation improvements.
      Instrument PPS ioctl()s so we know for sure what device they're being 
fired at.
      Fix clang's only warning in non-test code, about a cast in the shmexport 
code.
      Clean up a silly clang warning.  All tests continue to pass with a clang 
build.
      Minor date tweaks to Time Service HOWTO.
      Make PPS delivery troubleshooting a bit easier.  All regression tests 
pass.
      Fix an erroneous format literal.
      Try a simpler adaptive-delay formula that produces shorter delays.
      Reduce log verbosity.
      Remove a guard that is duplicated in the called function.
      Using SIOCOUTQ to flush output on client socket close caused a hard hang.
      Add gpsdecode -m option for dumping minimum packet lengths.
      Document that -m ignores comment packets.
      Remove the failed adaptive-delay experiment.
      Add a minimum packet length to every driver.
      splint cleanup for gpsdecode -m implementation.
      Create a logic path for each driver to set tty controls.
      Rare I/O with timeout seems to work.  This replaces the adaptive-delay 
code.
      Alas, we can't use ICANON.
      Add a note about NTP latency reduction.
      Assertions cleanup and new policy.  All regression tests pass.
      Fix a formatting error in AIS JSON generation (client library get it 
right).
      Comment typo fix.
      Back out the attempt to use VTIME.
      Codebase no longer requires -dGNU_SOURCE.
      Change to using blocking I/O and VTIME.
      Open ttys with CLOCAL, following a suggestion by Hal Murray.
      More delay reduction.
      gpsfake delays are no longer becessary under Linux, and possibly not 
elsewhere.
      Revert mistaken use of CLOCAL in open(2).
      O_NONBLOCK was causing more grief than we knew.  We don't need VTIME!
      Now we open with nonblock, set CLOCAL, and then turn off nonblock.
      In the test framework, use the bind-to-port-0 trick to get a random 
unused port.
      Minor changes that should have gone in with the previous commit.
      Revert the change abolishing the "pps" policy flag, it broke gpsmon 
client mode.
      Looks like a logfile got committed by mistake.
      Fix PPS vars in client-mode gpsmon.
      Remove exiguous comma from PPS offset display.
      Add delay tests reported by Frank Nicholas.
      Build fix for when pps=no.
      More updates of test delays.
      Clean up the GPSD shared-memory segment, and document why.
      Repaired a fatal typo in Gary Miller's last change.
      Fix some untested cases in JSON AIS dumping.
      Add YX as a recognized NMEA packet leader.  Seen on Airmar PB100.
      Revert to non-blocking I/O.  Blocking I/O broke gps-makeregress.
      Adjust NEWS tio reflect reversion to blocking I/O.
      Documentation polishing.
      Graceful shutdown qwhen hotplug-driven and out of devices and subscribers.
      Make PACKET_TYPES available in the Python interface.
      Refactoring step: make character_pushback take a new-state argument.
      Improve the ability of the packet analyzer to recognize control 
comments...
      Improved instrumentation.  All regression tests pass.
      Trim the ac12_binary tesrt ro shortten it and avoid end-detection issues.
      Shorten many regression tests to improve test time.
      Get more exact times for single tests.
      Machinery for doing true character pushback without packet disposal.
      Implement character pushback and state to ground without packet 
disposal...
      In the packet getter, fix an erroneous RTCM2 state transition.
      CLOSE_DELAY is gone.  Test source termination is now deterministic.
      Reinstate port-zero allocation magic for tests. All regression tests pass.
      Reinstate blocking I/O; it no longer breaks test rebuilds.
      Remove some remnants of the adaptive-delay experiment.
      More test trimming.
      Add -R option to gpsctl to remove SHM export segment, use in SConstruct.
      splint cleanup.  All regressiion tests pass.
      cppcheck and Coverity cleanup; all regression tests pass.
      Implement and document GPSD_SHM_KEY environment variable.
      In SConstruct, set a non-default GPSD_SHM_KEY value for the tests.
      Use $SRCDIR for gpsctl -R.
      Documentation update.
      Set FreeBSD WRITE_PAD to 0.01 in report from Hal Murray.
      Fix the gpsmon display-trashing problem; add some documentation and 
asserts.
      Comment typo fix.
      Improved instrumentation for tracking down test frame issues.
      Change the way EOF is injected in hopes of beating a *BSD timing problem.
      Support use of non-GNU date(1).
      New Hacking Guide section on C, standards, and style.
      Style amwendment suggested by Geg Troxel.
      BSD compiler waening and splint cleanup.
      Instrument shm segment creation a bit better.
      Shut up. splint!
      In gpsmon, attempt to banish edge-of-screen cruft from initialization 
JSON.
      SConstruct execution environment != the build variable list
      Make test rebuild work more like tests.
      Mutex-lock the update of last-fix time.
      Namespace cleanup.
      splint cleanup.  All regression tests pass.
      Documentation comment polishing.
      pyjrad_err has to be unconditionally available.
      Trying again to expose pthread_error unconditionally enough.
      Replace Gary's magic number 22 with TIMESPEC_LEN.
      Against magic numbers.
      *BSD portability note.
      A note on platforms.
      Widen PPS offset field in the three montors (NMEA, uBlox, SiRF) that 
report it.
      Full 22-digit display of PPS offset in the SiRF monitor.
      Typo fix in gpspipe.xml.
      Don't allocate a private port when cycling logs for client testing.
      Fix inexpplicably corrupted Python code.
      Significantly speed up testing.
      Revert "Significantly speed up testing."
      gpsfake code fell out of sync with documentation. Fix this.
      Foil an attempt by SCons to be overly helpful
      Time out hangs in single-shot tests.
      Allow setting WRITE_PAD from the environment.
      Repair website markup.
      The capability to set WRITE_PAD in the test frame was buggy.
      Tweak SConstruct so WRITE_PAD in the environment is passed in.
      Comment typo fix in two test loads.  No code changes.
      Comment typo fix.  No code changes.
      Improved comment.  No code changes.
      Add more documentation on NMEA ID ranges to NMEA.txt.
      More on Bedou and QZSS.  No code changes.
      Be more explicit about what SConstruct does when we turn off features.
      Remove duplicate inclusion of -lm from build.
      In SConstruct, repair the DBUS build.
      Updare some documentation.
      Update hacking guide - PDAs are dead, long live smartphones.
      In the build, clean up shared-library linknames properly.
      Simplify NTP SHM segment creation.
      Clarify a requirement for Python 2.x.
      Shut up, splint!  All regressiion tests pass.
      Tweak instrumentation to find out why alternate export segment isn't 
being set.
      Repair the logic for setting a non-default SHM export segment.
      Typo corrections to build instructions.
      Revert the ntpshm allocation change - go back to using an external bool 
array.
      Don't talk to NTP when we're running inside the test harness
      Update the internal documentation on NTP segment management.
      In build.txt, update stuff on the dbus_export option.
      Update the Hacking Guide.
      More ntpshm documentation fixes.
      Alas, remove OS X from the list of supported platforms.
      In the Hacking Guide, make contents order match text order.
      Fix a typo in the Client HOWTO.
      Minor update of Client HOWTO.
      Switch NEWS to use ISO8601 dates.
      monitor_nmea.c -> monitor_nmea0183.c
      Folded Calibration HOWTO into Time Service HOWTO
      Avoid overexposure of an internal mask.  EOF_SET -> EOF_IS
      More portable equivalent of #pracgma pack, will work with clang.
      Deprecated libQgpsmm config option involved.
      Update the post-3.12 to-do list.
      Move a couple of build options to where they fit better in the list.
      Feature removal - delete last remnants of old pre-JSON protocol from 
libgps.
      gpssim.py: pylint cleanup.  This code is not yet shipped.
      pylint cleanup of xgpsspeed. Live-test works.
      pylint cleanup of gpscap.py.  Live test works.
      pylint cleanup of auditing tools (not shipped).
      gpsfake: pylint cleanup.  Regression tests (using gpsfake) pass.
      Minor updates to GPS capability database.
      Document the Bluetooth troubleshooting procedure.
      gpscat: pylint cleanup.  Works in live testing.
      Add a couple of pylint error-suppression codes for newer versions.
      gpsprof: pylint cleanup. Live-tested.
      Version bump for release 3.12.
      Calibration HOWTO no longer exists.

Gary E. Miller (54):
      Fix typo in howto.  As reported by Mike Stewart.
      Add assert()s to the mutex lock/unlock.
      Merge branch 'master' of ssh://git.sv.gnu.org/srv/git/gpsd
      Fix PPS comment.
      Restore hastily removed checks on pthread usage.
      Merge branch 'master' of ssh://git.sv.gnu.org/srv/git/gpsd
      Check return codes for pthread_mutex*() in ppsthread.c
      Merge branch 'master' of ssh://git.sv.gnu.org/srv/git/gpsd
      Missing \n in debug statement.
      Add missing pthread_mutex_init(&ppslast_mutex)
      Document why gpsd needs to die on certain mutex failures.
      Loosen up some 1 Hz PPS guard bands.
      Fix PPS comments.
      Document thread unsafe usage of strerror() in PPS thread.
      Note potential loss of precision in timestamp_t
      Note loss of precision in ntpshm_latch()
      Add comment warning not to use doubles in PPS math.
      Fix insufficient precision in last_fixtime.clock
      Fix loss of precision in PPS json message.
      Add missing ppslast_mutex init to gpsmon
      Comment on loss of precision in timestamp().
      Comment possible numeric overflow in timespec_diff_ns()
      Add TS_SUB() to subtract timespecs.  Fix another double.
      Fix improper timespec type casting.
      Fix improper typecasting in a timespec
      Make gpsd_report() show what it really sent to ppshook.
      Fix NMEA and UBX PPS precision in gpsmon hardware mode
      Fix SiRF PPS offset precision in gpsmon
      Initialize a few time variables in session.
      More session initialization
      Clean up pps_early_init() placement in gpsmon.
      Add debugging for lat_fixtime_real getting clobbered.
      Tweak init of session.
      Mask a NAN in ntpshm_latch().  Cause of NAN in ppsthread.c
      Add note where gpsmon is failing in PPS.
      Whoops, bad printf formatting.  Missing %
      Fix sign of PPS offset in gpsmon.
      Fix a carry error in TS_SUB()
      Add function timespec_str(), and use it.
      Convert some of ppsthread.c to use timespec_str()
      Convert ntpshm.c to use timespec_str()
      Jigger PPS display for gpsmon/SiRF so it (usually) fits.
      Fix TS_NORM for negative numbers.
      Another edge case in TS_NORM()
      Move a gps_report() out of a time critical area.
      Maybe this time I got the TS_NORM() corner case?
      Keeping TS_NORM() a macro too ugly, convert to inline function
      Add some comments to TS_NORM()
      Stop calling ntpshhm_latch() more than once a second in gpsmon.
      Document the output of timespec_str() better
      Add a section on using Linux PTP
      Fix some typos in the Time Service HOWTO
      Add kernel config to PTP howto
      Add more details to PTP howto

Greg Troxel (29):
      Fix up NetBSD  in the time service howto
      Add TODO item to detangle TIOCMIWAIT and RFC2783.
      Build qt based on qt variable.
      Detangle qt being usable and whether we want it.
      Default libQgpsmm to False to match qt.
      Force on qt if libQgpsmm is on.
      Add do-build script to run tests.
      Include sys/types.h when checking for sys/timepps.h.
      Give full path to fake.py.
      Use full path to gpsfake, avoiding installed versions.
      Print sys.platform and platform.platform().
      Add delays for Darwin
      Regularize comments and add OS X case.
      Make delays super high on darwin.
      gps/fake.py: Add more OSX test failure datapoints.
      Use %.3f for WRITE_PAD, because 0.004 is a typical value.
      Reduce OSX delay slightly, and add datapoint.
      Shorten OS X delays.
      Add netbsd5-specific gpsfake timing and comments.
      Add netbsd6 regression timing data.
      Add BUILD.* and CHECK.* to .gitignore
      Add workaround for VMIN/VTIME situation.
      Update netbsd timing comments.
      Update os x timing data; shorten delay.
      Update netbsd5 timing data.
      Clean up pre-blocking-IO tests, organize.
      test timing: Reduce netbsd5 values.
      Rototill OS X pl2303 issues.
      Remove dropping of OS X, replacing with dropping osx-pl2303.

Jon Schlueter (34):
      Fixup iso8601_to_unix uninitialized memory bug
      Fix range check on test_json regression testing
      add test_mktime to .gitignore
      add .gcno to .gitignore
      Fixing endif guard in gpsd.h-tail to match guard levels
      Update leapseconds.cache with latest leapsecond
      Attempt to make scons minimal=true check to pass
      Adding missing PASSTHROUGH_ENABLE ifdef guards in packet.c
      Warning hunting cleanup in driver_superstar2.c
      Tweak to ifdef guard for RECONFIGURE_ENABLE in driver_sirf
      warning hunting cleanup -Wunused-variable in gpsmon.c
      warning hunting for -Wunused-function in libgpsd_core.c
      Adding Driver build dependancy NMEA2000 on AVIDM
      spelling cleanup in comments
      Fixing broken build for scons control_socket=on pps=on
      Add identify_failing_build_options script to devtools
      Tighter guard on buf variable in gpsmon.c to match usage
      build system cleanup of duplicate CFLAGS entries
      Correct assertions in libgpsd_core to assert 0 == err
      Split up imports in gpsfake and remove unused
      pep8 cleanup in gpsfake
      more pep8 cleanup in gpsfake for whitespace
      pylint cleanup in gpsfake for superfluous-parens
      Fixup scons audit for python
      pep8 cleanup to split out imports in maskaudit.py.in to individual lines
      split out and sort imports in xgps for pep8 cleanup
      Add missing import getopt package for xgps
      pep8 whitespace cleanup in xgps
      pep8 maskaudit.py.in whitespace cleanup
      pep8 whitespace cleanup in gpsprof
      SCons bug fix for minimal=on build was broken
      Fix for scons minimal=on broken build in gpsd.c missing ifdef guard
      Fixing broken build scons minimal=on ntpshm=on
      Fixing gpsmon.c to fix broken build

Lukas Schauber (1):
      Hotplug rules for u-blox 7 and 8.

Matt (14):
      Conditionalise -pthread option
      #ifdef out some functions that are unused when we don't have libusb
      Silence another compiler warning under cygwin/gcc4.9
      Handle lack of TIOCEXCL and TIOCNXCL on cygwin
      Ignore Windows objects and binaries
      Silence compiler warnings about array subscripts of type 'char'
      Add scons check for gmtime
      The STARTSWITH macro in libgps_json.c uses sizeof when it means strlen.
      Silences an unused-variable warning when ntpshm = False
      Silence another warning about array subscripts of type 'char'.
      Move pointer the correct direction when trying to skip spaces
      Guard includes of termios.h with #ifdef HAVE_TERMIOS_H
      Guard NTPSHM-related declarations
      Add do { ... } while (0) around assignment in macro

Michael Tatarinov (1):
      Recheck delays for Raspberry Pi Model B.

Michal Privoznik (1):
      libgps.h: Fix linkage problem

Mike Frysinger (1):
      udev rules: source Gentoo config files

Miroslav Lichvar (1):
      Update systemd files.

Reinhard Arlt (20):
      Fix GPGSA message in pseudonmea.c .
      Fix pseudonmea again.
      Add dummy pps parameter to WATCH.
      Add logfile and checkfile for nmea2000.
      Merge branch 'master' of git://git.sv.gnu.org/gpsd
      First version of test_nmea2000 testscript.
      Improve test_nmea2000.
      Add more help to test_nmea2000.
      Change JSON encoding in checkfile.
      Merge logs.
      Add new scale routine.
      New checkfile for nmea2000.
      Add -b (build) option to test_nma2000.
      Add more help text to test_nmea2000.
      Used scale_int() in all places.
      Adjust nmea2000 checkfile.
      Fix bug in type24 msg in pseudoais.c .
      Merge branch 'master' of git://git.sv.gnu.org/gpsd
      Fix another type24 bug in driver_nmea2000.c
      Merge branch 'master' of git://git.sv.gnu.org/gpsd

Sanjeev Gupta (4):
      Proofread documentation, typos, style changes only
      Minor reformatting
      Fix broken link anchor
      Fix off-by-one type

Shawn Kohlsmith (1):
      Clean up the time service Bibliography.

Simon Dawson (2):
      Fix build error when ntpshm is false
      Complete wrapping of ntpshm_latch function in #ifdef NTPSHM_ENABLE

Zbigniew Chyla (21):
      Move side effects out of assert()
      Cleanup of string function and sizeof usage.  All regression tests pass.
      Always use sizeof to get array size
      Always use sizeof to get array size, part 2
      Rename some variables for clarity. All regression tests pass.
      Add new header for string functions.
      Add str_starts_with macro, use it instead of strncmp.
      Use str_starts_with instead of local macro.
      Add str_{,v}appendf, use it everywhere.
      Add str_rstrip_char, use it everywhere.
      Add compiler.h with compiler specific macros.
      Convert macros in strfuncs.h to functions.
      Support empty strings in str_rstrip_char
      Simplify by using str_appendf
      Simplify by using str_rstrip_char.
      Fix incorrect uses of snprintf in pseudo-NMEA code.
      Fix incorrect use of strlcat
      Fix compilation error
      Avoid warning (pointer to volatile struct).
      timespec_str: add "const" to input argument
      Remove unused and incorrect retval from timespec_str

ukyg9e5r6k7gubiekd6 (1):
      Don't pass -Wmissing-prototypes to cxx

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


hooks/post-receive
-- 
GPSD



reply via email to

[Prev in Thread] Current Thread [Next in Thread]