Activity for Erik Hofman
-
Erik Hofman committed [23da83] on FlightGear
git-clang-format
-
Merge branch 'next' into aisim
-
Clean up the code and add a number of clarification comments.
-
Clean up the code and add a number of clarification comments.
-
Improve comments
-
Merge branch 'valgrind-errors' into aisim
-
Fix: Conditional jump or move depends on uninitialised value
-
Fix: Conditional jump or move depends on uninitialised value
-
git-clang-format
-
Merge branch 'next' into aisim
-
Fix ground handling.
-
Merge branch 'next' into aisim
-
Merge branch 'next' into aisim
-
clang-format
-
Fix a comparison error and prevent a segfault when the number of engines == 0 (gliders)
-
Fix ground handling and remove debugging code. This is now a fully functioning minimalistic linear Fligth Dynamics Module which uses SIMD to minimize computational cost, specially designed for a high number of AI aircraft in the scene.
-
Fix spelling
-
Proper greek characters for phi, theta and psi
-
sonParse was called on the file stream after it had already been fully read into a string, returning an empty map. Fixed to parse from an istringstream.
-
Fix ground handling, and acceleration for heavy aircraft
-
Get AISim into a flyable state. But not perfect on the gorund yet. Try using fgfs --aircraft=aisim --aero=c172p
-
Fix the HUD path, fixing bug #35
-
Fix the HUD path, fixing bug #35
-
Fix warning: bitwise operation between different enumeration types is deprecated
-
Replace absolute paths with relative ones
-
Replace absolute paths with relative ones
-
Change some absolute paths to relative paths
-
Change some absolute paths to relative paths
-
Remove absolute paths
-
Remove Pre-2024.2 configurastion files
-
Add a README.compile file to compile the time zone files from the IANA source and add a missing US diretory
-
Add a README.compile file to compile the time zone files from the IANA source and add a missing US diretory
-
Update the timezone information and the timezone database. Next only for the moment.
-
The fallback models are there especially for AI. So they are lower polygon count and use less texture memory. To get sound back they need to define a sound file per (base) aircraft.
-
Actually, what I think that happened is that the fallback models have been added after 2020.3 so before that it would use the full model (with sound) instead.
-
Clean up the OpenAL enumeration code (modeled after openal-info) while searching for a possible regression (https://forum.flightgear.org/viewtopic.php?f=4&t=42598&start=60#p426059). The regression was not detected on Linux. Also, the memory leak was not detected so remove the warning
-
flightgear::shutdownQtApp is not recognixed in fg_init.cxx: move it to bootstrap.cxx
-
Running 'fgfs --show-sound-devices' caused a segmentation fault because, as it turned out, calling flightgear::shutdownQtApp() was required before calling exit(). Add a fxExit() function which takes care of this.
-
C++ can't handle ALchar*, cast to char*
-
Somewhere down the line 'using namespace std;' was removed so call 'std::vector' instread of just 'vector'.
-
There was a very good reason why that was necessary and I would love to remember which it was .. but I can't. Also I see that the /fdm/jsbsim/damage/repairing property is not part of the JSBSim code so I fail to see why it is being used at all. It should have been set in an aircraft specific path.
-
There was a very good reason why that was necessary and I would love to remember which it was .. but I can't.
-
Switch to runtime detection instead of compile time detection of the AL_EXT_SOURCE_RADIUS extension.
-
Add support for the source radius extension for OpenAL which turns a highly directional point source into a sphere with a radius which softens the stereo effect when turning the viewer
-
Fix some problems with the test utilities. For instance the WAV audio format was used as the OpenAL format which differ in reality.
-
Remove hand crafted SIMD code; modern compilers are capable of creating similar code when geven enough hints.
-
Fix the sign of Roll_moment_due_to_yaw_rate, Roll_moment_due_to_rudder and Adverse_yaw
-
Change some absolute paths to relative paths
-
No need to exclude this file anymore. In fact is's counter productive for the JSB and UIUC version.
-
Do not attempt to load the deprecated user shaders
-
Move the code to update fog density when the visibility changes to /Nasal/hdr.nas
-
Update fog density when the visibility change
-
Calculate an approximate visibily based on chance of mist and fog derived from humidity and the difference between the temperature and dewpoint. Calculate fog-density.
-
Tone down fresnel and shininess a bit
-
Reference the correct filename
-
Fix what's probably a copy and paste error. As reported per https://forum.flightgear.org/viewtopic.php?f=68&t=41686
-
Use SGPath instead of the C++17 filesystem function exists.
-
We do static linking so prevent exporting the functions names for MSVC
-
Possible MSVC fix: link with the wrong SimGear object.
-
Add a test utility to test whether all timezone files in the database can be found
-
There seems to be a small patch in timezone16.bin where TimezoneIdPrefix and TimezoneId return NULL. Protect against that.
-
Add a missing timezone file
-
timezone16.bin uses 'Kyiv' while zone.tab uses 'Kiev' so add them both
-
TheFGFSEagle: The attached patch reverts your commit @erik and fixes the real problem - the
-
Rename the getBumpHeight function to getGroundDisplacement
-
Initial implementation of a new GroundReactions class which ultimatelly handle surface displacements for the likes of rocky terrain or ocean waves across FDM's.
-
Base the wetness on the temperature and dewpoint instead. Simplify the code a bit
-
Fix for:
-
Parse the path of fgdata to Canvas
-
Use mmapping when reading Nasal scripts saving two memory copies out of for (copy from disk to the i/o buffer, copy from the i/o buffer tot the ifstream, copy from the ifstreamto a string and copy from a string to Nasal).
-
Move more code to C++, use SGPath and SGMMapFile to map the Nasal file instead of copying it.
-
Fix the build for systems without a system wide sqlite lubrary
-
sqlite_open_v2 requires sqlite3_close_v2 (I think, not really clear).
-
Add SQLite support to Nasal and copy the nasal-bin utility over from SimGear, make it C++11 and add it as an installable nasal interpreter for standalone testing and debugging of local nasal scripts.
-
CycloneDDS aparrently stopped supporting booleans
-
Sascha Reißner
-
Sascha Reißner
-
Use a random string for a reference name so samples from the instant queue wich are using the same sound file will play simultaneously.
-
Tie the instant queue to the listener
-
Add a special queue-name 'instant' which does not put samples into a sample queue but plays them instantly.
-
Turn OpenVSP CSV forces and moments data into tables
-
Expose the machwave active and offset properties
-
Add a utility to recenter all models within a configuration file
-
Very early start of a utility to convert OpenVSP CSV files into JSBSim XML files.
-
keep a copy of the string returend by p->getStringValue() in setProp until it is sent to the DDS layer.
-
Cope with SGPropertyNode::getStringValue() now returning std::string by value.
-
Set the inland-ice property when the mean ground temperatures get low enough
-
Read annual precipitation fro the blue channel
-
Do not include dds_props.hxx if CycloneDDS was not found
-
Add a new network protocol: dds-props which for now only operates in the output mode: request a propery value using de FG_DDS_prop scheme and provide either it's path or property index id. UsageL fgfs --dds-props=dds,out,<hz>. Testing is possible with src/Network/DDS/fg_dds_prop to request a propery path/id or src/Network/DDS/fg_dds_log to log all propery samples passed over the DDS layer.
-
Update the props sample description: remove an unnecessary type specifier (the union already holds it), add a sample version number and a mode to indicate reading from, or writing to, a property. The option to set a property value is reserved for future use and not yet implemented.
-
Add a proprety requester test utility
-
No need to keep two maps: it's enough the have one PropertyList (vector) and a map
-
Switch to always add the GUID. Setting a string to null doesn't work for DDS samples and leads to a segmentation fault.
-
Also log propery samples over DDS
-
Implement the start of a property server using DDS.
-
Udpate the logged name
-
Use the new convenience constructor eliminating the need to call setup separately.
-
Use the new helper function which converts the typename into a topic name.
-
Simplify the code even more: no need to explicitly define the type
1 >
