]> git.mxchange.org Git - flightgear.git/log
flightgear.git
15 years agoMerge branch 'maint' into next
Tim Moore [Fri, 9 Jan 2009 23:12:49 +0000 (00:12 +0100)]
Merge branch 'maint' into next

15 years agoChange the near plane value back to .1
Tim Moore [Fri, 9 Jan 2009 23:11:41 +0000 (00:11 +0100)]
Change the near plane value back to .1

By popular demand...

15 years agoFix Torsten's crash. A completely cross-platform issue, no idea why it didn't show...
jmt [Fri, 9 Jan 2009 13:15:03 +0000 (13:15 +0000)]
Fix Torsten's crash. A completely cross-platform issue, no idea why it didn't show up in local testing.

15 years agoapply custom widget fonts also to the legend, so that they show
mfranz [Fri, 9 Jan 2009 11:58:28 +0000 (11:58 +0000)]
apply custom widget fonts also to the legend, so that they show
up in list/textbox etc. (This should really be separately settable
via style definition, but it's very rarely used and a change now
is probably not worth it as we'll probably switch to osgWidget,
anyway.)

15 years agoAdd various helper filters, to make use of type-ranges, and reduce the need
jmt [Thu, 8 Jan 2009 21:11:53 +0000 (21:11 +0000)]
Add various helper filters, to make use of type-ranges, and reduce the need
to down-cast in filter implementations.

15 years agoBasic infrastructure to expose type ranges from filters, and hence reduce
jmt [Thu, 8 Jan 2009 21:11:08 +0000 (21:11 +0000)]
Basic infrastructure to expose type ranges from filters, and hence reduce
how many candidates need to be submitted to the pass() hook.

15 years agoMerge branch 'maint' into next
Tim Moore [Thu, 8 Jan 2009 08:07:16 +0000 (09:07 +0100)]
Merge branch 'maint' into next

15 years agoDivide-by-zero fixes from Csaba Halász
timoore [Thu, 8 Jan 2009 07:57:04 +0000 (07:57 +0000)]
Divide-by-zero fixes from Csaba Halász

15 years agoSimpler bearing and distance computations for AIFlightPlan.
jmt [Tue, 6 Jan 2009 17:05:46 +0000 (17:05 +0000)]
Simpler bearing and distance computations for AIFlightPlan.

15 years agoMerge branch 'maint2' into next
Tim Moore [Wed, 7 Jan 2009 09:38:07 +0000 (10:38 +0100)]
Merge branch 'maint2' into next

15 years agoPrevent exceptions in getRunwayByIdent when a malformed rwyuse.xml means
jmt [Tue, 6 Jan 2009 16:56:26 +0000 (16:56 +0000)]
Prevent exceptions in getRunwayByIdent when a malformed rwyuse.xml means
there is no valid active runway. This is not ideal, since it masks underlying
problems - the real fix is to make the runway-use code more robust and
validate input XML.

15 years agoUpdate MSVC 7.1 projects
fredb [Sun, 4 Jan 2009 20:48:24 +0000 (20:48 +0000)]
Update MSVC 7.1 projects

15 years agoRename FGMarkerBeacon to FGMarkerBeacon record, to avoid a clash with the
jmt [Sun, 4 Jan 2009 20:12:43 +0000 (20:12 +0000)]
Rename FGMarkerBeacon to FGMarkerBeacon record, to avoid a clash with the
instrument of the same name. In the future I'd prefer to rename the
instrument class instead (FGMarkerBeaconReciever?) but this is the safest
change for now.

Thanks (again) to Yon Uriarte for pointing out the problem (which seems to
affect MSVC more than gcc)

15 years agoJames Turner: Rewrite of the getSpeed function. Smaller and more elegant.
durk [Sun, 4 Jan 2009 17:11:25 +0000 (17:11 +0000)]
James Turner: Rewrite of the getSpeed function. Smaller and more elegant.

15 years agoFix a bug resulting from my marker-beacon refactoring. Thanks to Yon Uriarte
jmt [Sun, 4 Jan 2009 09:41:52 +0000 (09:41 +0000)]
Fix a bug resulting from my marker-beacon refactoring. Thanks to Yon Uriarte
for spotting the problem.

15 years agoMerge branch 'rj/ttw' into next
Tim Moore [Wed, 7 Jan 2009 09:26:12 +0000 (10:26 +0100)]
Merge branch 'rj/ttw' into next

15 years agoPatch from Ron Jensen: make the TTW computation marginally less silly.
jmt [Sat, 3 Jan 2009 17:58:01 +0000 (17:58 +0000)]
Patch from Ron Jensen: make the TTW computation marginally less silly.

15 years agoClean up naming of the 'point on runway' helpers, to get rid of the confusing
jmt [Sat, 3 Jan 2009 16:15:48 +0000 (16:15 +0000)]
Clean up naming of the 'point on runway' helpers, to get rid of the confusing
notion of a 'displacedThreshold'. Now there's just a real threshold,
displaced or otherwise, and people who care about the paved area can use
'begin' and 'end'. Thanks to John Denker for pointing out the confusion this
leads to. Using 'end' also gets rid of the 'reverseThreshold' name, which was
clearly a bad choice of mine.

15 years agoRefactor marker-beacons to be distinct from FGNavRecord. This is a big space
jmt [Sat, 3 Jan 2009 15:54:03 +0000 (15:54 +0000)]
Refactor marker-beacons to be distinct from FGNavRecord. This is a big space
saving for beacons, but since they're surprisingly few (in nav.dat), not a
an enormous saving in real terms. The major motivation is that marker beacons
don't behave like other NavRecords for radio interaction - they have no ident,
frequency or range (in the sense that NavRecord means them).

15 years agoMerge branch 'maint2' into next
Tim Moore [Sat, 3 Jan 2009 23:25:04 +0000 (00:25 +0100)]
Merge branch 'maint2' into next

15 years agoProtect against divide-by-zero in setCameraParameters
timoore [Sat, 3 Jan 2009 00:15:58 +0000 (00:15 +0000)]
Protect against divide-by-zero in setCameraParameters

Found by Csaba Halász

15 years agoNaN fix by Csaba/Jester - prefer atan2(x,y) to atan(a/y).
jmt [Fri, 2 Jan 2009 12:34:28 +0000 (12:34 +0000)]
NaN fix by Csaba/Jester - prefer atan2(x,y) to atan(a/y).

15 years agoTemporary hack to avoid NaN problems when _mp is negative (?). Discovered by Csaba
fredb [Thu, 1 Jan 2009 10:18:13 +0000 (10:18 +0000)]
Temporary hack to avoid NaN problems when _mp is negative (?). Discovered by Csaba

15 years agoCsaba/Alexis : fix a NAN problem when wind is unspecified in a metar
fredb [Tue, 30 Dec 2008 23:13:44 +0000 (23:13 +0000)]
Csaba/Alexis : fix a NAN problem when wind is unspecified in a metar

15 years agoMerge branch 'jt/runway' into next
Tim Moore [Sat, 3 Jan 2009 23:22:49 +0000 (00:22 +0100)]
Merge branch 'jt/runway' into next

15 years agoUpdate MSVC 7.1 project
fredb [Tue, 30 Dec 2008 11:04:27 +0000 (11:04 +0000)]
Update MSVC 7.1 project

15 years agoMerge branch 'maint2' into next
Tim Moore [Sat, 3 Jan 2009 23:19:56 +0000 (00:19 +0100)]
Merge branch 'maint2' into next

15 years agoWin32 fixes
fredb [Tue, 30 Dec 2008 11:03:45 +0000 (11:03 +0000)]
Win32 fixes

15 years agoJames Turner: Cleanup of AI traffic route generation code.
durk [Tue, 30 Dec 2008 07:41:15 +0000 (07:41 +0000)]
James Turner: Cleanup of AI traffic route generation code.

15 years agoMerge branch 'maint2' into next
Tim Moore [Sat, 3 Jan 2009 23:16:19 +0000 (00:16 +0100)]
Merge branch 'maint2' into next

15 years agoPrevent CVS from complaining about unknown files in ATCDCL.
durk [Tue, 30 Dec 2008 07:14:47 +0000 (07:14 +0000)]
Prevent CVS from complaining about unknown files in ATCDCL.

15 years agoMerge branch 'jt/runway' into next
Tim Moore [Sat, 3 Jan 2009 23:14:09 +0000 (00:14 +0100)]
Merge branch 'jt/runway' into next

15 years agoAutomake on Hardy Heron seems to be considerably more picky about white-
jmt [Mon, 29 Dec 2008 22:27:36 +0000 (22:27 +0000)]
Automake on Hardy Heron seems to be considerably more picky about white-
space than previous versions.

15 years agoSplit runway and taxiway into separate classes, with a (new) common base.
jmt [Mon, 29 Dec 2008 22:23:22 +0000 (22:23 +0000)]
Split runway and taxiway into separate classes, with a (new) common base.
This makes taxiways smaller (important since at present there are so many).
Restructure the apt.dat parsing code to use a helper class instead of one long
function, and to do less work when parsing the file.

Some of these ideas come from Yon Uriarte's patches - thanks Yon.

15 years agoMerge branch 'tbm/graphics-bug' into maint2
Tim Moore [Sat, 3 Jan 2009 23:05:45 +0000 (00:05 +0100)]
Merge branch 'tbm/graphics-bug' into maint2

15 years agoMerge branch 'tbm/graphics-bug' into next
Tim Moore [Sat, 3 Jan 2009 23:03:26 +0000 (00:03 +0100)]
Merge branch 'tbm/graphics-bug' into next

15 years agoMerge branch 'maint2' into next
Tim Moore [Sat, 3 Jan 2009 23:01:46 +0000 (00:01 +0100)]
Merge branch 'maint2' into next

15 years agoWin32 fixes
fredb [Sat, 27 Dec 2008 16:09:01 +0000 (16:09 +0000)]
Win32 fixes

15 years agoRemove warnings
fredb [Sat, 27 Dec 2008 16:08:21 +0000 (16:08 +0000)]
Remove warnings

15 years agoSet far camera reference frame to ABSOLUTE_RF
timoore [Thu, 1 Jan 2009 09:35:15 +0000 (09:35 +0000)]
Set far camera reference frame to ABSOLUTE_RF

This was lost when I stopped cloning the near camera to make the far
camera. The result was a lot of breakage, including explicit camera
configurations not working and various ordering issues too.

15 years agoSet BACKGROUND_BIT as camera node mask.
timoore [Tue, 30 Dec 2008 23:36:51 +0000 (23:36 +0000)]
Set BACKGROUND_BIT as camera node mask.

This prevents the dreaded black rectangle from appearing on systems
that don't have OpenGL frame buffer object support.

15 years agoChange the order of the main cameras from NESTED_RENDER to POST_RENDER
timoore [Tue, 30 Dec 2008 21:09:46 +0000 (21:09 +0000)]
Change the order of the main cameras from NESTED_RENDER to POST_RENDER

This seems to be accepted OSG usage for slave cameras. It's possible
that this order is important for rendering instruments that use RTT
textures on systems without frame buffer object support. I'm thinking
that the resulting bugs may be implicated in the "black rectangle" problem.

15 years agoDon't clone far camera; create a new one and initialize it from the near camera
timoore [Tue, 30 Dec 2008 09:32:53 +0000 (09:32 +0000)]
Don't clone far camera; create a new one and initialize it from the near camera

The clone operation may be sharing things that shouldn't be shared.

15 years agoExpose the CameraGroup near, far, and near/far boundry as properties
timoore [Sat, 27 Dec 2008 14:35:33 +0000 (14:35 +0000)]
Expose the CameraGroup near, far, and near/far boundry as properties

The near/far boundary, called "near-field", can be set to 0 which
disables the far camera and renders the whole scene using only one
camera. I'm hoping that this may be useful in resolving some
system-specific rendering bugs.

Various fixes were made to correctly render the scene using only the
near camera.

15 years agoFGAirportList is gone. Everything should work as before, especially the
jmt [Sat, 27 Dec 2008 13:20:08 +0000 (13:20 +0000)]
FGAirportList is gone. Everything should work as before, especially the
AirportList dialog. (It would now be possible to trivially implement a
VOR or NDB named search, if anyone wants such a thing).

15 years agoFurther work (still not enabled) on a fast + correct implementation of the
jmt [Sat, 27 Dec 2008 10:08:12 +0000 (10:08 +0000)]
Further work (still not enabled) on a fast + correct implementation of the
airportList search function. At least for me (with a mutex-based SGAtomic),
large vectors of referenced-pointers is a Bad Thing - copying on vector
resize is thrashing the locks.

15 years agoFix up environment code for threading-enabled case, d'oh.
jmt [Fri, 26 Dec 2008 15:33:38 +0000 (15:33 +0000)]
Fix up environment code for threading-enabled case, d'oh.

15 years agoAnother clean-up iteration: FGAirportList::search is gone, replaced by two
jmt [Fri, 26 Dec 2008 15:26:42 +0000 (15:26 +0000)]
Another clean-up iteration: FGAirportList::search is gone, replaced by two
static FGAirport helpers. As a result, another global index goes away. Use
the helpers to avoid ugly FGPositioned down-casts in various places.

Also converts the environment/METAR code to deal with FGAirport pointers,
instead of string identifiers, and contains work-in-progress code to implement
the AirportList dialog using FGPositioned. This isn't enabled yet for various
reasons, but is the final piece to allow FGAirportList to be removed.

15 years agoFix more test/util linkage for sgmath depending on sgstructure.
jmt [Fri, 26 Dec 2008 12:38:32 +0000 (12:38 +0000)]
Fix more test/util linkage for sgmath depending on sgstructure.

15 years agoKill off many Point3D includes, and a couple of uses in the HUD code.
jmt [Fri, 26 Dec 2008 12:28:05 +0000 (12:28 +0000)]
Kill off many Point3D includes, and a couple of uses in the HUD code.
Requires a SimGear update to get some new SGGeodesy helpers.

15 years agoClean out FGAirportList - not quite obsolete yet, but the spatial queries are
jmt [Fri, 26 Dec 2008 11:15:00 +0000 (11:15 +0000)]
Clean out FGAirportList - not quite obsolete yet, but the spatial queries are
gone. This is good news, since the old query was implemented as a linear
search, sorted by Manhattan distance, and with a warning not to use the logic
at runtime. Various systems (such as the Mk-VIII) do query such data often,
eg every second.

Also gets Point3D out of Airports/simple.hxx, as a precursor to removing it
completely.

15 years agoRemove all name and spatial queries from FGNavList. All remaining queries are
jmt [Thu, 25 Dec 2008 23:11:43 +0000 (23:11 +0000)]
Remove all name and spatial queries from FGNavList. All remaining queries are
by frequency (which makes sense), and use the FGPositioned spatial data if
required. As a result, the marker beacon list is gone (since beacons are only
searched spatially). In the process, clean up various minor things - most
notably, all the 'airport-related' navaids (ILS, GS, LOC, and the beacons) now
store a FGRunway* instead of an airport id string. This is more precise, and
saves string allocations.

15 years agoFinish conversion of FGRunway to a real class - all public members are gone.
jmt [Wed, 24 Dec 2008 15:45:35 +0000 (15:45 +0000)]
Finish conversion of FGRunway to a real class - all public members are gone.

15 years agoFGPositioned clean-ups - apply some desirable changes (such as making members
jmt [Wed, 24 Dec 2008 14:48:30 +0000 (14:48 +0000)]
FGPositioned clean-ups - apply some desirable changes (such as making members
const) which were previously tricky but now easy. Make it possible not to
index certain types (used for taxiways) and exclude anonymous items from
the name index. Related to this, clean up FGRunway further - remove some public
members, and fix a dumb bug of mine, where we create reciprocal entries for
taxiways.

This should make startup (slightly) quicker, and shrinks FGRunway somewhat.

15 years agoRemove the FGIdentOrdering cruft from FGAirport; now handled by FGPositioned.
jmt [Wed, 24 Dec 2008 04:58:36 +0000 (04:58 +0000)]
Remove the FGIdentOrdering cruft from FGAirport; now handled by FGPositioned.

15 years agoConvert the only remaining user of FGFixList to use an FGPositioned query,
jmt [Tue, 23 Dec 2008 14:41:58 +0000 (14:41 +0000)]
Convert the only remaining user of FGFixList to use an FGPositioned query,
and hence remove the query code from fix-list. The only remaining code deals
with parsing fix.dat.

15 years ago(first commit, partly a low-risk thing to sanity check my setup)
jmt [Tue, 23 Dec 2008 12:37:59 +0000 (12:37 +0000)]
(first commit, partly a low-risk thing to sanity check my setup)
Add a helper predicate to FGAirport to encapsulate the common 'does this
airport have a suitable runway of at least xxxx ft?' query. Also add a
FGPositioned filter built on the predicate, and a 'closest airport' helper.

15 years agoStuart BUCHANAN: fix METAR cloud interpolation
mfranz [Sun, 21 Dec 2008 15:52:29 +0000 (15:52 +0000)]
Stuart BUCHANAN: fix METAR cloud interpolation

"I've managed to fix this by differentiating rebuilding the 3D layers
(e.g. due to METAR updates) from rebuilding the entire environment (due to a
change in scenario)."

15 years agoDon't try to fetch tiles when lat or lon are invalid
fredb [Sun, 21 Dec 2008 09:29:33 +0000 (09:29 +0000)]
Don't try to fetch tiles when lat or lon are invalid

15 years agorevert to using a cached list of aircraft -- scanning the Aircraft/
mfranz [Sat, 20 Dec 2008 23:24:50 +0000 (23:24 +0000)]
revert to using a cached list of aircraft -- scanning the Aircraft/
dirs is only fast enough with hot file-cache, but a bit too painful
otherwise. Updating the aircraft.list is now easier, though: Just
type   $ fgfs --aircraft=?<TAB>

15 years agoFix path
fredb [Sat, 20 Dec 2008 15:41:45 +0000 (15:41 +0000)]
Fix path

15 years agoPackaging for the 1.9.0 release
fredb [Sat, 20 Dec 2008 15:15:30 +0000 (15:15 +0000)]
Packaging for the 1.9.0 release

15 years agobetter instructions
mfranz [Sat, 20 Dec 2008 12:50:03 +0000 (12:50 +0000)]
better instructions

15 years agoviewer position: tie variables directly, rather than getter functions
mfranz [Sat, 20 Dec 2008 12:49:37 +0000 (12:49 +0000)]
viewer position: tie variables directly, rather than getter functions

15 years agolive-compile aircraft list, instead of requiring to build it
mfranz [Sat, 20 Dec 2008 10:32:42 +0000 (10:32 +0000)]
live-compile aircraft list, instead of requiring to build it
regularly. That's fast enough.

15 years agoDocumentation update: Openal & OSG.
durk [Sat, 20 Dec 2008 09:12:31 +0000 (09:12 +0000)]
Documentation update: Openal & OSG.

15 years agoStuart : The attached patch
fredb [Fri, 19 Dec 2008 23:11:58 +0000 (23:11 +0000)]
Stuart : The attached patch
reverts one of my changes which ensured that the 3D clouds were updated with
METAR. Unfortunately this had the side-effect of over-writing the environment
properties, and therefore nuking the interpolation work.

15 years agoAttempt to sort out the version number mess in preparation for the v1.9.0
curt [Fri, 19 Dec 2008 20:41:57 +0000 (20:41 +0000)]
Attempt to sort out the version number mess in preparation for the v1.9.0
release.

15 years agomark file names (for post-processing)
mfranz [Fri, 19 Dec 2008 13:38:25 +0000 (13:38 +0000)]
mark file names (for post-processing)

15 years agoexport absolute viewer coordinates
mfranz [Fri, 19 Dec 2008 13:37:54 +0000 (13:37 +0000)]
export absolute viewer coordinates

15 years agoStuart Buchanan :
fredb [Fri, 19 Dec 2008 07:42:13 +0000 (07:42 +0000)]
Stuart Buchanan :

Attached is yet another 3D clouds patch, to fix the following:
1) The 3D clouds were not modified by the cloud coverage, due to some problems with osg::Switch
2) METAR changes to cloud coverage were not obeyed.
3) Making changes via the Cloud dialog had no effect unless 3D clouds were toggled.
4) Cloud cover was too sparse.
5) 3D Stratus clouds caused performance issues on some hardware (fixed by removing 3D stratus from cloudlayers.xml - it will now be a 2D layer).

15 years agoCsaba/"Jester" and /me: since a few weeks the property browser had
mfranz [Thu, 18 Dec 2008 20:18:50 +0000 (20:18 +0000)]
Csaba/"Jester" and /me: since a few weeks the property browser had
displayed garbage in the title line if one left a directory. Jester
spent a lot of time finding the cause: a bug that must have been
there since ... forever: fgfs reported to PUI the string addresses
as returned by SGPropertyNode::getStringValue() without considering
that this address isn't valid after the next write access to that
node! It's almost a miracle that it worked so well for so long,
despite that issue. (I also used to opportunity for some more
cleanup -- so I'm to blame for any newly introduced bugs, and
also for reverting some of Jester's cleanups. :-)

15 years agonow that /environment/weather-scenario is respected at startup, we
mfranz [Thu, 18 Dec 2008 13:08:44 +0000 (13:08 +0000)]
now that /environment/weather-scenario is respected at startup, we
actually have to set it to get live weather.
--enable-real-weather-fetch shall imply weather-sceneario=METAR.

15 years agosegfault--
mfranz [Thu, 18 Dec 2008 11:49:33 +0000 (11:49 +0000)]
segfault--

15 years agoJames Turner : in the 'tower' view, where the tower position is explicitly given...
fredb [Thu, 18 Dec 2008 06:52:51 +0000 (06:52 +0000)]
James Turner : in  the 'tower' view, where the tower position is explicitly given in apt.dat (most of the larger airports), I was incorrectly treating the tower elevation as ASL, when it's AGL.

15 years agoStuart BUCHANAN:
mfranz [Wed, 17 Dec 2008 19:51:18 +0000 (19:51 +0000)]
Stuart BUCHANAN:

"fixes the issue reported by Martin where
--prop:/environment/weather-scenario=METAR had no effect"

15 years agoStuart :
fredb [Sun, 14 Dec 2008 18:48:00 +0000 (18:48 +0000)]
Stuart :
the 3D clouds now respect changes to the environment caused by updated METAR etc.

It also increases the AlphaFunc as suggested by Tim.

15 years agoinclude png and xcf files by default
mfranz [Sat, 13 Dec 2008 20:17:02 +0000 (20:17 +0000)]
include png and xcf files by default

15 years agoCsaba/Jester : fix an incorrect string comparison in metar
fredb [Sat, 13 Dec 2008 06:08:22 +0000 (06:08 +0000)]
Csaba/Jester : fix an incorrect string comparison in metar

15 years agoReally catch sound exceptions, and avoid exiting on this kind of error :
fredb [Fri, 12 Dec 2008 07:35:39 +0000 (07:35 +0000)]
Really catch sound exceptions, and avoid exiting on this kind of error :
OpenAL error (AL_INVALID_VALUE): constructor (alBufferData)
Failed to buffer data.

15 years agoStuart :
fredb [Thu, 11 Dec 2008 08:24:58 +0000 (08:24 +0000)]
Stuart :
- Removes the cloud type re-use code - I think they aren't worth the graphical artefacts that they cause in light of Tim's improvements
- Changes the transparency of the clouds. Previously, the clouds were transparent at 0m distance, opaque at 200m, then gradually more transparent to the fog limits. This meant they were generally quite transparent. Now, they are opaque from 200m to 15km, then become transparent at 20km. A side effect of this is that the current textures could probably benefit from being made slightly transparent to improve the blending of the sprites against each other.

15 years agoUpdate to -rc2
fredb [Thu, 11 Dec 2008 08:18:50 +0000 (08:18 +0000)]
Update to -rc2

15 years agoSome final adjustments:
durk [Wed, 10 Dec 2008 18:41:34 +0000 (18:41 +0000)]
Some final adjustments:
 - Remove Concorde, as announced
 - Add a missing directory to utils/Makefile.am, which caused an autoconf
   error.

15 years agoFinalizing the preparation of fgfs-2.99.5-rc2.
durk [Wed, 10 Dec 2008 18:19:52 +0000 (18:19 +0000)]
Finalizing the preparation of fgfs-2.99.5-rc2.

15 years ago*** empty log message ***
durk [Wed, 10 Dec 2008 16:04:50 +0000 (16:04 +0000)]
*** empty log message ***

15 years agoMinimize impact of distant AI models on the scene graph.
durk [Wed, 10 Dec 2008 16:04:10 +0000 (16:04 +0000)]
Minimize impact of distant AI models on the scene graph.

15 years agofix some backslashes in include file names
timoore [Tue, 9 Dec 2008 14:34:30 +0000 (14:34 +0000)]
fix some backslashes in include file names

15 years agoJames Turner :
fredb [Tue, 9 Dec 2008 08:10:33 +0000 (08:10 +0000)]
James Turner :
Here's a patch which refactors the 'plain' GPS code into a slightly
more manageable structure - i.e breaks the large update() method into
various sub-functions. I've tested the patch with B1900d, and things
seem to work as expected, but if anyone experiences GPS weirdness
after this is committed, of course please report it.

The motivation for this was helping me learn the code - I've planning
some changes in this area, and splitting up the logic will hopefully
make that task easier.

15 years agoJames Turner : this patch
fredb [Tue, 9 Dec 2008 07:58:46 +0000 (07:58 +0000)]
James Turner : this patch
- updates how filtering is done on the various FGPositioned query functions - both spatial and ident-based.
- updates the KLN-89b / DCLGPS code to use FGPositioned for all Navaid/Airport queries.

me:
Add few include directives in globals.cxx to avoid such MSVC warning messages :
deletion of pointer to incomplete type 'FGAirportList'; no destructor called
deletion of pointer to incomplete type 'FGNavList'; no destructor called
deletion of pointer to incomplete type 'FGTACANList'; no destructor called
deletion of pointer to incomplete type 'FGFixList'; no destructor called

15 years agoAnders GIDENSTAM: "Do not leak uninitialized values to the rest
mfranz [Sun, 7 Dec 2008 19:47:48 +0000 (19:47 +0000)]
Anders GIDENSTAM: "Do not leak uninitialized values to the rest
of the computation."  (forward port from jsbsim/cvs)

15 years agoDoc folder moved
fredb [Sun, 7 Dec 2008 09:42:43 +0000 (09:42 +0000)]
Doc folder moved

15 years agoPrepare packaging for the release
fredb [Sun, 7 Dec 2008 09:13:12 +0000 (09:13 +0000)]
Prepare packaging for the release

15 years agoslightly more efficient iterator handling
mfranz [Sun, 7 Dec 2008 08:19:54 +0000 (08:19 +0000)]
slightly more efficient iterator handling

15 years agoStuart Buchanan :
fredb [Sat, 6 Dec 2008 23:03:12 +0000 (23:03 +0000)]
Stuart Buchanan :
- Replaces simple shader attributes with vectors (this was missed out of the last patch by mistake)
- Includes Yon's Fog update code (Thanks!)
- Fixes a bug since 1.0 where --enable-real-weather-fetch stopped the other weather scenarios from working.

15 years agoFix apparently ancient typo.
andy [Fri, 5 Dec 2008 17:33:57 +0000 (17:33 +0000)]
Fix apparently ancient typo.

15 years agoCsaba "Jester" HALASZ: segfault--
mfranz [Fri, 5 Dec 2008 16:47:11 +0000 (16:47 +0000)]
Csaba "Jester" HALASZ: segfault--

15 years agoThis provides the following enhancements & bug fixes
fredb [Thu, 4 Dec 2008 20:56:40 +0000 (20:56 +0000)]
This provides the following enhancements & bug fixes
- Fix the chequer-board bug.
- Add proper cloud coverage function - so scattered clouds are now truly scattered.
- Add real-time control for visibility range.
- Use a limited set of clouds rather than generating a completely new Geode for each cloud. This saves sorting and display time.
- Add controls to Rendering dialog to allow fine-tuning of the number of sprites, cloud visibility and the number of different types of cloud.
- Add some variance to the sort back-off to avoid all clouds being sorted at the same time.
- Pack attributes into vectors for performance
- Re-order the cloud type determination code so that if a cloud layer could either be stratus or cumulus, cumulus is used.
- Lowered the cloud level in the standard cloud configuration slightly so a cumulus layer is generated rather than stratus.

These last two mean that you should see some 3D cumuli if disabling real weather fetch.

My thanks to Yon Uriarte for his help with performance work.

15 years agoAllow negative thrust. This allows a single recoil or vibration
mfranz [Thu, 4 Dec 2008 18:47:49 +0000 (18:47 +0000)]
Allow negative thrust. This allows a single recoil or vibration
thruster to accelerate in both directions. THROTTLE input still
clamps to 0/1 by default.  (OK'ed by Andy)

15 years agoJames Turner :
fredb [Thu, 4 Dec 2008 08:29:36 +0000 (08:29 +0000)]
James Turner :

  - a refactoring of the route manager to reduce indentation and duplication
- created a helper method to generate the ETA strings
- created a helper to update target_altitude and altitude_set when the active wp0 changes
- used early returns to make update() easier to follow
- removed spurious includes in both header and source file

15 years agoJames Turner : a trivial patch to remove a now obsolete default ctor from FGPositioned
fredb [Thu, 4 Dec 2008 08:15:36 +0000 (08:15 +0000)]
James Turner : a trivial patch to remove a now obsolete default ctor from  FGPositioned