summaryrefslogtreecommitdiff
path: root/libgpsd.xml
Commit message (Collapse)AuthorAgeFilesLines
* whitespace cleanup in libgpsd.xmlJon Schlueter2011-03-051-6/+6
|
* The raw_hook mwember and set_raw_hook are gone; client API cleanup is done.Eric S. Raymond2011-01-251-10/+0
|
* Good catch by Michael Cook.Eric S. Raymond2011-01-031-3/+3
|
* Clients are no longer required to define a gpsd_report() hook.Eric S. Raymond2010-12-191-7/+0
|
* libgps.a splits in half, becomes libgps.a + libgpsd.aEric S. Raymond2010-04-181-10/+4
| | | | | | This is so client apps don't have to carrry around libusb and all the driver-level stuff. I finally caught on to one of the implications of shared libraries that had somewhow escaped me before.
* Tweak some DocBook meta headers to quiet down the build.Eric S. Raymond2010-03-251-1/+2
|
* Copyright stamps everywhere. No code changes.Eric S. Raymond2010-03-111-0/+4
|
* Simplify the API. The second argument of gpsd_activate() was a fossil.Eric S. Raymond2009-09-181-3/+1
|
* Remove obsolete paragraph.Eric S. Raymond2009-03-141-5/+0
|
* Pull the "configurable" bit out of the context structure & make it per-driver...Eric S. Raymond2006-11-201-4/+4
| | | | | ...set by an argument to gpsd_active(). Changes the libgpsd(3) API, but not the client-side one.
* Rename the xgps manual page.Eric S. Raymond2006-03-021-1/+1
|
* Set VMIN in GPS reads.Eric S. Raymond2005-06-291-2/+1
|
* Clean up after FIONREAD removal.Eric S. Raymond2005-06-281-2/+3
|
* Document that the low-level interface uses no dynamic allocation.Eric S. Raymond2005-06-281-5/+6
|
* Eliminate dynamic-memory allocation from the daemon.Eric S. Raymond2005-06-281-1/+7
| | | | | Only one calloc() is left, and that belongs to Gary in the Garmin driver.
* 41 splint warnings.Eric S. Raymond2005-06-071-1/+1
| | | | | gps_mask_t type introduced so we won't wire in a bunch of dependencies on the size of the flag mask.
* Change API so raw_hook takes a length argument; supports dumping binary packets.Eric S. Raymond2005-05-261-1/+2
|
* Support R=2.Eric S. Raymond2005-05-241-2/+3
|
* Open the shared-memory segment to talk to ntpd only once.Eric S. Raymond2005-04-281-1/+3
| | | | | | | | | | Instead of opening the shared-memory segment to talk to ntpd each time gpsd_init is called and having it be part of the GPS device state, open it once at daemon startup time and move the pointer into the shared-context structure. This is a good cleanup on general principles. It will also let us drop root privs after startup.
* Leap-second correction is now context shared among all sessions.Eric S. Raymond2005-04-111-0/+12
|
* unbalanced tagGary E. Miller2005-03-301-0/+1
|
* gpsd.py goes away.Eric S. Raymond2005-03-301-30/+0
| | | | | It couldn't be keept up to date with the C library, and only gpsprobe was using it.
* Fix markup error.Eric S. Raymond2005-03-281-1/+1
|
* Another refactoring change, moves towards multisession.Eric S. Raymond2005-03-281-1/+2
|
* Change interface so there can be only one DGPS object per server...Eric S. Raymond2005-03-281-7/+12
| | | | ...not one per session.
* Change gps_session_t to gps_device_t.Eric S. Raymond2005-03-281-5/+5
|
* Daemon no longer depends on the raw hook to trigger watcher notifications.Eric S. Raymond2005-03-181-6/+5
| | | | | Instead, the packet parser returns a flag set telling what has changed; publication to watchers is driven by that.
* Major documentation update. Split gpsd(1) into gpsd(8) and xgps(1).Eric S. Raymond2005-03-181-2/+3
|
* Prerelease cleanup.Eric S. Raymond2005-03-101-2/+0
| | | | | | Restrict scope of session variables in the damon. Give the raw_hook function direct access to the output data. raw_hook now copes correctly when passed multiple sentences.
* gpsd_init() loses its driver-type parameter.Eric S. Raymond2005-03-041-1/+0
|
* Python-binding initialization no longer requires a device-type argument.Eric S. Raymond2005-03-011-2/+4
|
* Baudrate option on gpsd.py is gone, it does baud hunting.Eric S. Raymond2005-03-011-4/+2
|
* Updated.Eric S. Raymond2005-02-231-12/+15
|
* Nothing outside the driver structures should have to know what driversEric S. Raymond2004-09-241-11/+5
| | | | are available, so use gpsd -h to list them.
* Changed name of gpscli_report to gpsd_report and added stub,Eric S. Raymond2004-09-081-2/+2
|
* Symplify the Python interface.Eric S. Raymond2004-08-291-4/+3
|
* Document the Python interface.Eric S. Raymond2004-08-291-6/+25
|
* Cosmetic cleanup.Eric S. Raymond2004-08-231-1/+1
|
* Timeouts at the daemon level were a mistake. This commit removes them.Eric S. Raymond2004-08-231-3/+1
| | | | | | | | | | | | | | | | | | | | The intention was good, but the reality turned out to have problems. Among them: 1) Scott Anderson reported being confused because he would sometimes issue a command and get back a blank response, either because the data was stale or the mode bits were invalid. He was right; this is too confusing. There is no concept of "timeout" in the behavior of the UPS -- having the daemon impose one simply gets between the user and the data source in an unhelpful way. 2) What happens if two GPS-using application have different theories of the right timeout value? The right way to handle this is to make timestamps visible to the application and let the *application* have a timeout policy.
* Completing the name swap.Eric S. Raymond2004-08-221-0/+165
|
* Continuing the name swap.Eric S. Raymond2004-08-221-105/+0
|
* A big name change. The gps* and gpsd* namespaces swap places, so gps_*Eric S. Raymond2004-08-221-12/+12
| | | | | | names are now the *upper*-level library. This is part 1; due to a technical limitation in Subversion, the file renames have to be done in a separate commit.
* The gps client now goes through the daemon. Relayerr functionalityEric S. Raymond2004-08-211-2/+1
| | | | | | has been removed from the low-level interface, since you can get the same effect by going through the high-level interface with a specified nodefault host/port.
* Merged the libraries. There is now just one libgps supporting two interfaces.Eric S. Raymond2004-08-211-5/+13
|
* Another step in raw-hook support for the client library.Eric S. Raymond2004-08-211-7/+31
|
* Cleanup before SVN conversion.Eric S. Raymond2004-08-191-0/+74
|
* Some cleanup before SVN conversion.Eric S. Raymond2004-08-191-155/+0
|
* Change structures so we can set speed explicitly.Eric S. Raymond2004-08-171-6/+27
|
* Change the name of the master structure.Eric S. Raymond2004-08-141-6/+6
|
* Document the error-reporting function better.Eric S. Raymond2004-08-141-0/+5
|