]> git.mxchange.org Git - simgear.git/log
simgear.git
18 years agoSmall cleanups to the SGGeo[dc] classes, provide more hooks to use them directly
frohlich [Thu, 15 Jun 2006 08:27:31 +0000 (08:27 +0000)]
Small cleanups to the SGGeo[dc] classes, provide more hooks to use them directly

18 years agoUse function argument in va_start instead of local variable.
frohlich [Thu, 15 Jun 2006 06:14:46 +0000 (06:14 +0000)]
Use function argument in va_start instead of local variable.

18 years agoRemove now unused function
frohlich [Sun, 11 Jun 2006 13:59:59 +0000 (13:59 +0000)]
Remove now unused function

18 years agoModified Files:
frohlich [Sun, 11 Jun 2006 13:30:59 +0000 (13:30 +0000)]
Modified Files:
simgear/scene/material/mat.cxx simgear/scene/material/mat.hxx
simgear/scene/material/matlib.cxx
simgear/scene/material/matlib.hxx simgear/scene/tgdb/leaf.cxx
simgear/scene/tgdb/obj.cxx
Attach userdata to groundtile scenegraph leafs that contains
a SGMaterial reference to the material of that leaf.
Add (physical) material properties to the material definitions.
Plug a memory leak with GlyphSigns.

18 years agoPreliminary material lookup hooks - still unoptimized.
frohlich [Thu, 8 Jun 2006 05:54:23 +0000 (05:54 +0000)]
Preliminary material lookup hooks - still unoptimized.

18 years agowhoops, sorry (Yes, it *was* tested, but then I made another "trivial"
mfranz [Wed, 24 May 2006 10:16:09 +0000 (10:16 +0000)]
whoops, sorry (Yes, it *was* tested, but then I made another "trivial"
change and ...)

18 years agoif we are going to die we better tell all our listeners
mfranz [Wed, 24 May 2006 09:37:44 +0000 (09:37 +0000)]
if we are going to die we better tell all our listeners

18 years agoadd optional position argument to SGRoute::add_waypoint(). Default is -1,
mfranz [Mon, 8 May 2006 11:31:16 +0000 (11:31 +0000)]
add optional position argument to SGRoute::add_waypoint(). Default is -1,
which appends the WP like it used to. Valid vector indices insert the WP
at this position.

18 years agoMac fix
fredb [Thu, 4 May 2006 05:58:59 +0000 (05:58 +0000)]
Mac fix

18 years agoFix the initial texture path problem. Loaders are setting the one given to ssgLoad...
fredb [Sat, 29 Apr 2006 08:09:51 +0000 (08:09 +0000)]
Fix the initial texture path problem. Loaders are setting the one given to ssgLoad as the default one behind our back :-(

18 years agoRedefine the default PLIB loader behavior : don't clear the texture cache after every...
fredb [Fri, 28 Apr 2006 18:05:46 +0000 (18:05 +0000)]
Redefine the default PLIB loader behavior : don't clear the texture cache after every model load

18 years agoadd method to delete any waypoint (last waypoint if n is out of range)
mfranz [Fri, 28 Apr 2006 15:43:13 +0000 (15:43 +0000)]
add method to delete any waypoint (last waypoint if n is out of range)

18 years agoPigeons remaining fix for the soundmanager crashes.
frohlich [Tue, 25 Apr 2006 18:47:37 +0000 (18:47 +0000)]
Pigeons remaining fix for the soundmanager crashes.

18 years agothanks to Erik's texture map I can now drop empty.rgb altogether and just
mfranz [Sat, 22 Apr 2006 13:41:06 +0000 (13:41 +0000)]
thanks to Erik's texture map I can now drop empty.rgb altogether and just
specify the same texture in the "foo.lighted" and "foo.unlighted" material
entry. This also allows to drop the state cloning and thereby solves the
most urgent apt_signs.cxx TODO. :-)

18 years agoAdd a texture cache mechanism. Fortunately this oly seems affective for empty.rgb ...
ehofman [Sat, 22 Apr 2006 09:38:14 +0000 (09:38 +0000)]
Add a texture cache mechanism. Fortunately this oly seems affective for  empty.rgb ...

18 years agodon't allow new command name to overwrite material name
mfranz [Thu, 20 Apr 2006 17:46:40 +0000 (17:46 +0000)]
don't allow new command name to overwrite material name

18 years ago- don't use hard-coded emission values for unlighted signs, but load both
mfranz [Thu, 20 Apr 2006 16:06:00 +0000 (16:06 +0000)]
- don't use hard-coded emission values for unlighted signs, but load both
  states from material.xml (separate <material> entries for now)
- clone state less often: not once per sign element, but once per material
  switch (TODO: clone only once per material)

18 years agofix "unknown.rgb" path (the wrong path was the reason why we always only
mfranz [Thu, 20 Apr 2006 15:20:40 +0000 (15:20 +0000)]
fix "unknown.rgb" path (the wrong path was the reason why we always only
got plib's lowres red-white chequer-board pattern along with an error
message, and not ours ... which is much prettier, but also bigger.
(Should we downscale it?)

18 years agomake headers include headers they depend on, don't rely on the c(xx)
mfranz [Mon, 17 Apr 2006 11:29:01 +0000 (11:29 +0000)]
make headers include headers they depend on, don't rely on the c(xx)
file to do that. (This is a requirement for header precompiling.)

18 years agorename OBJECT_TAXI_SIGN to OBJECT_SIGN. This isn't about taxi signs any
mfranz [Fri, 14 Apr 2006 14:50:08 +0000 (14:50 +0000)]
rename OBJECT_TAXI_SIGN to OBJECT_SIGN. This isn't about taxi signs any
more, but all sorts of signs. Now is the best time to get rid of a
misleading name.

18 years agoadd <condition> support to textranslate & texrotate animation
mfranz [Wed, 12 Apr 2006 20:27:38 +0000 (20:27 +0000)]
add <condition> support to textranslate & texrotate animation

18 years agoset sign orientation such, that when the sign heading=0, one looks straight
mfranz [Wed, 12 Apr 2006 12:13:03 +0000 (12:13 +0000)]
set sign orientation such, that when the sign heading=0, one looks straight
on the sign face when looking North

18 years agolower signs
mfranz [Tue, 11 Apr 2006 23:04:24 +0000 (23:04 +0000)]
lower signs

18 years agoadd minimalistic backside to signs as a temporary solution
mfranz [Tue, 11 Apr 2006 21:32:15 +0000 (21:32 +0000)]
add minimalistic backside to signs as a temporary solution

18 years agore-add hard-coded vertical distance. The coordinates should be surface
mfranz [Tue, 11 Apr 2006 17:34:17 +0000 (17:34 +0000)]
re-add hard-coded vertical distance. The coordinates should be surface
points and not add this distance, which depends on the sign housing/hardware,
after all.

18 years ago- commands do now have to start with @
mfranz [Tue, 11 Apr 2006 15:57:08 +0000 (15:57 +0000)]
- commands do now have to start with @
- add commands @size, @material, @light
- make "BlackSign" texture default
- make @B, @R, @L, @Y open close their frames automatically (this can be
  avoided by setting the @material manually)
- add number variants for those 4 sign commands: @Y2, @B5, etc (according
  to the spec; defaulting to the respective biggest panel size, i.e. @B = @B3)
(detailed description will be added to $FG_ROOT/Docs/)

18 years agoremove obsolete files (on request by Christian Mayer, who has introduced them):
mfranz [Mon, 10 Apr 2006 16:36:52 +0000 (16:36 +0000)]
remove obsolete files (on request by Christian Mayer, who has introduced them):
- they are not used anywhere in sg/fgfs
- and are very clearly *not* GPL compatible!

18 years agoManabu Nishiyama (non-FlightGear Nasal user) discovered an
andy [Mon, 10 Apr 2006 16:21:17 +0000 (16:21 +0000)]
Manabu Nishiyama (non-FlightGear Nasal user) discovered an
uninitialized data bug in naHash_cget().  When the hashcode field of
naStr was introduced, I forgot to set it in this function, which
creates a temporary naStr on the stack.

18 years agodrop xscale member again, and use xsize/ysize instead. (One interface
mfranz [Mon, 10 Apr 2006 15:32:55 +0000 (15:32 +0000)]
drop xscale member again, and use xsize/ysize instead. (One interface
element less to confuse people.)

18 years agorewrite of OBJECT_TAXI_SIGN code. The name is a bit misleading, as this
mfranz [Sun, 9 Apr 2006 19:51:00 +0000 (19:51 +0000)]
rewrite of OBJECT_TAXI_SIGN code. The name is a bit misleading, as this
type can also create runway signs. (/me thinks about changing that ...)

18 years agosupport for font textures. They are normal (but rather lenghty) <material>,
mfranz [Sun, 9 Apr 2006 19:21:13 +0000 (19:21 +0000)]
support for font textures. They are normal (but rather lenghty) <material>,
but contain <glyph> entries with <name>, <left> and <right>. The latter two
describe where in the texture a letter or symbol begins and where it ends.
(range 0-1). <xscale> defines a horizontal scaling factor.

18 years agov0.3.10 changes.
curt [Wed, 5 Apr 2006 18:42:45 +0000 (18:42 +0000)]
v0.3.10 changes.

18 years agoprotect ssg pointers to avoid occasional crashes (of course it would be
mfranz [Thu, 30 Mar 2006 14:13:22 +0000 (14:13 +0000)]
protect ssg pointers to avoid occasional crashes  (of course it would be
nicer if the Occluder would always get removed before its model branch,
but that's not easily enforcable)

18 years agov0.3.10-pre3 updates.
curt [Mon, 27 Mar 2006 18:48:36 +0000 (18:48 +0000)]
v0.3.10-pre3 updates.

18 years agoOlaf Flebbe:
curt [Mon, 27 Mar 2006 16:32:15 +0000 (16:32 +0000)]
Olaf Flebbe:

Unfortunatly the original SimGear Project I submitted contained local
changes which I didn't intend to submit. Please use this file instead...

18 years agoIf the author of this message isn't alerted enough to *fix* this, then I'm
mfranz [Sun, 26 Mar 2006 08:22:26 +0000 (08:22 +0000)]
If the author of this message isn't alerted enough to *fix* this, then I'm
sure the users won't do that either. This is regularly triggerd and leads
to meaningless error reports.

18 years agomore guarded pointers (we are still getting spurious crashes on exit because
mfranz [Fri, 24 Mar 2006 23:24:20 +0000 (23:24 +0000)]
more guarded pointers (we are still getting spurious crashes on exit because
of that)

18 years agoInitial revision.
curt [Fri, 24 Mar 2006 19:06:05 +0000 (19:06 +0000)]
Initial revision.

18 years agoremove so that complete dir can be dropped
mfranz [Thu, 23 Mar 2006 22:05:22 +0000 (22:05 +0000)]
remove so that complete dir can be dropped

18 years agoUpdates to remove unneeded and old version of zlib source.
curt [Thu, 23 Mar 2006 21:59:59 +0000 (21:59 +0000)]
Updates to remove unneeded and old version of zlib source.

18 years agoRemove old depricated files. There is a much newer version of zlib
curt [Thu, 23 Mar 2006 21:52:03 +0000 (21:52 +0000)]
Remove old depricated files.  There is a much newer version of zlib
available and just about every unix-based platform already comes with this.

18 years agov0.3.10-pre2 updates.
curt [Thu, 23 Mar 2006 16:37:35 +0000 (16:37 +0000)]
v0.3.10-pre2 updates.

18 years agoVassilii KHACHATUROV: rain fix, cleanup, read formerly hard-coded values
mfranz [Thu, 23 Mar 2006 15:39:00 +0000 (15:39 +0000)]
Vassilii KHACHATUROV: rain fix, cleanup, read formerly hard-coded values
                      from provided node

"* minor redundant gl call in DrawCone2 optimized away (twice per frame)
* corrected the glRotatef() order in drawRain even further (a less
obvious mistake than before, which is noticed by looking skywards and wiggling
the mouse in the view direction change mode)
* all the "magic numbers" used in the rain rendering code have been
provided a default (based on the old hardcoded value) in a form of a
define, and a meaningfully named static member in SGEnviro"

18 years agoVassilii KHACHATUROV: fix typos, add/fix comments, change variable names
mfranz [Wed, 22 Mar 2006 16:02:55 +0000 (16:02 +0000)]
Vassilii KHACHATUROV: fix typos, add/fix comments, change variable names

18 years agoOops, implement the previously-mentioned fix without breaking support
andy [Tue, 21 Mar 2006 23:12:30 +0000 (23:12 +0000)]
Oops, implement the previously-mentioned fix without breaking support
for omitting a length argument to mean "the rest of the vector".

18 years agoThe original code (rather oddly) interprets a length of zero in
andy [Tue, 21 Mar 2006 22:22:47 +0000 (22:22 +0000)]
The original code (rather oddly) interprets a length of zero in
subvec() to mean "the whole vector".  Melchior showed a use case
(removal of the first element from a vector) where getting a
zero-length subvector is actually desired.  And since I can't come up
with a good reason for why the "feature" was there in the first place,
out it goes...

18 years agoMelchior discovered that cmp() was just wrong, failing to actually
andy [Tue, 21 Mar 2006 21:57:00 +0000 (21:57 +0000)]
Melchior discovered that cmp() was just wrong, failing to actually
inspect the string pointers.  It also failed to properly sort strings
where one is a prefix of the other.  It looks to me like I just never
finished this, and it ended up in CVS because it just happened to
compile...

18 years agoUpdates for v0.3.10-pre1.
curt [Mon, 20 Mar 2006 19:22:20 +0000 (19:22 +0000)]
Updates for v0.3.10-pre1.

18 years ago- better error message when submodel loading failed
mfranz [Fri, 17 Mar 2006 19:01:52 +0000 (19:01 +0000)]
- better error message when submodel loading failed
- use alignmainmodel node in callback (not used anywhere yet)

18 years agoThe handle gets nulled out if the user has closed the file; don't pass that
andy [Thu, 16 Mar 2006 18:01:56 +0000 (18:01 +0000)]
The handle gets nulled out if the user has closed the file; don't pass that
null to fclose() in a garbage collection destructor too...

18 years agoFix broken checkin in iolib.c. Also add the 'bits' library, which has a buf()
andy [Wed, 15 Mar 2006 19:42:37 +0000 (19:42 +0000)]
Fix broken checkin in iolib.c.  Also add the 'bits' library, which has a buf()
function needed to make convenient use of io.read().

18 years agoMelchior found the first bug -- report EOF as nil in readln().
andy [Wed, 15 Mar 2006 19:35:30 +0000 (19:35 +0000)]
Melchior found the first bug -- report EOF as nil in readln().

18 years agoAdd the Nasal I/O library so Melchior can play with it. Not enabled currently (see...
andy [Wed, 15 Mar 2006 18:09:50 +0000 (18:09 +0000)]
Add the Nasal I/O library so Melchior can play with it.  Not enabled currently (see NasalSys.cxx in the flightgear CVS)

18 years ago--warnings
mfranz [Tue, 14 Mar 2006 15:55:24 +0000 (15:55 +0000)]
--warnings

18 years ago--warnings
mfranz [Tue, 14 Mar 2006 15:49:29 +0000 (15:49 +0000)]
--warnings

18 years agoclose zone.tab file after having read all entries
mfranz [Tue, 14 Mar 2006 12:58:28 +0000 (12:58 +0000)]
close zone.tab file after having read all entries

18 years agoprevent animations from losing nodes, because other processes removed them
mfranz [Tue, 14 Mar 2006 10:38:06 +0000 (10:38 +0000)]
prevent animations from losing nodes, because other processes removed them
(We are currently getting a lot of aborts in the condition code when
running MP. I don't expect this to fix it, but a bug is a bug.)

(reviewed by Fred, who also fixed *my* bugs :-)

18 years agoBetter use a node that is very clearly not used by other services
mfranz [Sun, 12 Mar 2006 18:56:48 +0000 (18:56 +0000)]
Better use a node that is very clearly not used by other services
(e.g. animations). The data class doesn't mind.

18 years agoactivate model load/unload callback again. It turned out *not* to be the
mfranz [Sun, 12 Mar 2006 10:09:34 +0000 (10:09 +0000)]
activate model load/unload callback again. It turned out *not* to be the
cause for the MP crashes -- the same crashes did still occur without it.

18 years agoMac OS X fixes from Ima Sudonim.
ehofman [Sat, 11 Mar 2006 22:20:19 +0000 (22:20 +0000)]
Mac OS X fixes from Ima Sudonim.

18 years agocommenting out Nasal in scenery models for now. This could be responsible
mfranz [Fri, 10 Mar 2006 22:58:24 +0000 (22:58 +0000)]
commenting out Nasal in scenery models for now. This could be responsible
for an MP/AI crash. Still investigating.

18 years agowarning--
mfranz [Thu, 9 Mar 2006 16:17:35 +0000 (16:17 +0000)]
warning--

18 years agoAlexander Powell:
ehofman [Thu, 9 Mar 2006 09:54:43 +0000 (09:54 +0000)]
Alexander Powell:

Add MAC OS X Render Texture support:

Most texture modes seem to work on my Powerbook, but I don't have a wide array
of machines to test it on otherwise.

If you have problems, please let me know and I'll see if I can help track down
the source of the bug.  I'd love to keep working on it if time permits (I use
RenderTexture in a few other projects), so I'll keep you informed if there are
any changes that I make for the better.

18 years agomodel.[ch]xx:
mfranz [Thu, 9 Mar 2006 09:03:57 +0000 (09:03 +0000)]
model.[ch]xx:
  add abstract class SGModelData. If a pointer to such a class is handed over
  to sgLoad3DModel, then its modelLoaded() method is called with path, property
  node and branch. And then it's added to the scene graph so that it's
  destroyed when the model branch is removed from the graph.

modellib.[ch]xx:
  only cache objects when asked to. This is the case for OBJECT_SHARED
  and random objects (like before), but no longer for OBJECT_STATIC.
  These are now removed from the graph when they are "out of sight". This
  prevents accumulation of static models, and makes destroying model data
  possible (e.g. removing Nasal modules)

matmodel.cxx:
  set cache flag for random objects (same behavior as before)

18 years ago- new FSF addresses
mfranz [Wed, 8 Mar 2006 18:16:08 +0000 (18:16 +0000)]
- new FSF addresses
- coplied license headers from h(xx) files to respective c(xx) files
- removed trailing spaces
- fixe $Id$
- fixed typos

18 years agoIgnore generated binaries.
david [Sat, 4 Mar 2006 13:27:11 +0000 (13:27 +0000)]
Ignore generated binaries.

18 years agoMathias Fröhlich:
ehofman [Sat, 4 Mar 2006 12:46:29 +0000 (12:46 +0000)]
Mathias Fröhlich:

zero out all parent pointers, else they might be dangling.

18 years agoMathias Fröhlich:
ehofman [Fri, 3 Mar 2006 15:11:35 +0000 (15:11 +0000)]
Mathias Fröhlich:

I have some bugfixes which will avoid fg just crashing if the sound device
could not be opened.

18 years agoWrong config file name
fredb [Sun, 26 Feb 2006 11:02:21 +0000 (11:02 +0000)]
Wrong config file name

18 years agoFix from Melchior: Set static values at every condition change, not only initially.
andy [Wed, 22 Feb 2006 20:50:35 +0000 (20:50 +0000)]
Fix from Melchior: Set static values at every condition change, not only initially.

18 years agoMelchior FRANZ:
fredb [Tue, 21 Feb 2006 12:59:31 +0000 (12:59 +0000)]
Melchior FRANZ:

- don't unlock an already unlocked mutex (Someone wanted to be on
  the safe side with this, but the result is undefined and makes
  pthread_mutex_destroy fail. Reference: manpage for
  pthread_mutexattr_gettype/The Open Group[1]: "Attempting to
  unlock a mutex of this type which is not locked results in
  undefined behaviour.")

- re-enabled all subsystem destructors again (this has been disabled
  because fgfs hung on exit, due to the mutex destroy failure from
  above.)

Reference:
  http://www.opengroup.org/onlinepubs/007908799/xsh/pthread_mutexattr_gettype.html

18 years agoBack out the previous patch.
ehofman [Tue, 21 Feb 2006 10:47:20 +0000 (10:47 +0000)]
Back out the previous patch.

18 years agoDeclare specified functions, otherwise MIPSpro bails out.
ehofman [Tue, 21 Feb 2006 09:48:33 +0000 (09:48 +0000)]
Declare specified functions, otherwise MIPSpro bails out.

18 years agoMelchior FRANZ:
ehofman [Tue, 21 Feb 2006 09:40:12 +0000 (09:40 +0000)]
Melchior FRANZ:

- new FSF address
- removed a few hundred trailing spaces
- fixed a few $Id$ lines
- copied two license headers from *.hxx files to their respective
  *.cxx counterparts
- added two test aps to .cvsignore
- don't unlock an already unlocked mutex (Someone wanted to be on
  the safe side with this, but the result is undefined and makes
  pthread_mutex_destroy fail. Reference: manpage for
  pthread_mutexattr_gettype/The Open Group[1]: "Attempting to
  unlock a mutex of this type which is not locked results in
  undefined behaviour.")
- re-enabled all subsystem destructors again (this has been disabled
  because fgfs hung on exit, due to the mutex destroy failure)

18 years agoMathias Froehlich: Add license information.
ehofman [Mon, 20 Feb 2006 15:12:13 +0000 (15:12 +0000)]
Mathias Froehlich: Add license information.

18 years agoMathias Fröhlich:
ehofman [Sun, 19 Feb 2006 17:22:17 +0000 (17:22 +0000)]
Mathias Fröhlich:

This patch makes use of the vectors now available in simgear with that past
patch. And using that it simplyfies the carrier code somehow.

- Small additional factory's to the quaternion code are done in the simgear
  part. Also more explicit unit names in the factory functions.
- The flightgear part makes use of them and simplyfies some computations
  especially in the carrier code.
- The data part fixes the coordinate frames I used for the park positions in
  the carrier to match the usual ones. I believed that I had done so, but it
  was definitly different. Also there are more parking positions avaliable now.

18 years agoMelchior FRANZ:
ehofman [Sun, 19 Feb 2006 17:13:37 +0000 (17:13 +0000)]
Melchior FRANZ:

- change SG back to FG (this was accidently changed in three wrong
  places when I prepared the file for SG)
- correct length for the proxy id detection
- set (guessed) deposit depth < 1mm correctly
- set deposit type string
- formatting

18 years agoMissing defines
fredb [Sat, 18 Feb 2006 16:19:54 +0000 (16:19 +0000)]
Missing defines

18 years agoAdd missing include files needed by the new math code under windows
fredb [Sat, 18 Feb 2006 13:24:24 +0000 (13:24 +0000)]
Add missing include files needed by the new math code under windows

18 years agoFix Code generation option for debug build
fredb [Sat, 18 Feb 2006 12:04:37 +0000 (12:04 +0000)]
Fix Code generation option for debug build

18 years agoRemove compiler warnings
fredb [Fri, 17 Feb 2006 21:57:13 +0000 (21:57 +0000)]
Remove compiler warnings

18 years agoClean up some compiler warnings.
curt [Fri, 17 Feb 2006 15:13:42 +0000 (15:13 +0000)]
Clean up some compiler warnings.

18 years agoRemove unused files.
ehofman [Fri, 17 Feb 2006 09:59:27 +0000 (09:59 +0000)]
Remove unused files.

18 years agoMelchior FRANZ:
ehofman [Fri, 17 Feb 2006 09:23:40 +0000 (09:23 +0000)]
Melchior FRANZ:

This patch fixes the sound of 737, Concorde and others, if fgfs
was compiled with newer gcc versions (e.g. gcc 4.0.2). These compilers
implement the c++ standard more strictly, and thus don't guarantee
that c-style casted pointers to different data types really point
to the same address. This is only guaranteed for union members.

18 years agoMathias Fröhlich:
ehofman [Fri, 17 Feb 2006 09:22:04 +0000 (09:22 +0000)]
Mathias Fröhlich:

The patch adds a vector lib I have put together during the last time,
it is just handy and interfaces well with s(s)g*. Together with some small
modifications this will later also interface well with OpenSceneGraphs
vectors/matrices. Using this vector kernel is targeted to have a handy
matrix/vector lib available and to provide a scenegraph independent vector
type for use with a small scenegraph wrapper aimed for a smooth transition to
openscenegraph.

That vector code also includes an improoved geodetic conversion routine I
have found some time ago published in the 'journal of geodesy' which avoids
iterative computations for that purpose.

Also the geodetic position class is more typesafe and unitsafe than the
Point3D currently is.

That part is relatively old and in use in my local trees for several months
now.

18 years agoOlaf Flebbe:
ehofman [Fri, 17 Feb 2006 08:58:56 +0000 (08:58 +0000)]
Olaf Flebbe:

This patch makesFlightGear at least compile on MSVC. I hope I have removed
reference of my other local changes. DSP and DSW files are included for
reference. They have been reconstructed with am2dsp.pl. I had to introduce a
change to am2dsp because of the need of filenames with embedded spaces. (Yuck)

The major direction is to remove clutter like the _USE_MATH_DEFINES and
have it on the compiler command line sice there is no central include
file. You will have to put it on the command line for your locale
Project files, if it not there, already. I added the
_CRT_SECURE_NO_DEPRECATE define for 2005, since it does no harm to other
VC version.

Third Party Libs like plib, OpenALSDK, freeglut, pthreads-win32, zlib
are unpacked as is side by side. Only change put the includes of OpenAL
into include/AL rather directly into include.

18 years agoOutputing 6 digits is not enough for a double
fredb [Tue, 7 Feb 2006 20:50:35 +0000 (20:50 +0000)]
Outputing 6 digits is not enough for a double

18 years agoVassilii Khachaturov:
ehofman [Thu, 2 Feb 2006 09:56:48 +0000 (09:56 +0000)]
Vassilii Khachaturov:

Fix the current buggy rain orientation behaviour for the views attached to the
aircraft (while still inheriting bugs with the views attached to anything else).

18 years agoExpose the sun halo texture handle.
ehofman [Tue, 31 Jan 2006 15:14:02 +0000 (15:14 +0000)]
Expose the sun halo texture handle.

18 years agoExpose GL_COORD_REPLACE
ehofman [Tue, 31 Jan 2006 15:13:12 +0000 (15:13 +0000)]
Expose GL_COORD_REPLACE

18 years agocreate an absolute value before calling log or log10, this adds support for sound...
ehofman [Mon, 30 Jan 2006 20:30:56 +0000 (20:30 +0000)]
create an absolute value before calling log or log10, this adds support for sound on negative numbers (thrust reverse for example).

18 years agoMelchior FRANZ:
ehofman [Mon, 30 Jan 2006 10:56:34 +0000 (10:56 +0000)]
Melchior FRANZ:

add optional arg to SGPropertyNode::addChangeListener that triggers
the listener function call initially. This is useful for cases where
a freshly installed listener wants to treat the current property
value as changed from 'unknown' to the actual value right away.

Examples can be found in the Nasal incarnation setlistener(),
where we have for example this (in $FG_ROOT/Nasal/gui.nas):

  INIT = func {
      ...
      setlistener("/sim/rendering/fps-display", fpsDisplay);
      if (getprop("/sim/rendering/fps-display")) {
          fgcommand("dialog-show", props.Node.new({"dialog-name": "fps"}));
      }
  }

That is: we first attach a listener that cares for changes to the FPS
display switch, but then we have to manually open the dialog initially.
That's a duplication of code and could be as simple as this
(INIT part only):

  INIT = func {
      ...
      setlistener("/sim/rendering/fps-display", fpsDisplay, 1);
  }

That is: the optional third arg makes fpsDisplay be called initially,
and then again with every write action. My first solution was in the
Nasal code only, but Andy (rightfully) says that this should rather
be in sg.

18 years agoOn a second thought, disable smooth shaded lines for all segments of the lightning.
ehofman [Fri, 27 Jan 2006 15:27:43 +0000 (15:27 +0000)]
On a second thought, disable smooth shaded lines for all segments of the lightning.

18 years agoDisable smooth lines for certain lightning segments.
ehofman [Fri, 27 Jan 2006 15:18:45 +0000 (15:18 +0000)]
Disable smooth lines for certain lightning segments.

18 years agoAdd support for point sprites.
ehofman [Thu, 26 Jan 2006 09:15:42 +0000 (09:15 +0000)]
Add support for point sprites.

18 years agoThe sample is now managed by a SGSharedPtr and shouldn't be deleted explicitely
fredb [Tue, 24 Jan 2006 21:49:56 +0000 (21:49 +0000)]
The sample is now managed by a SGSharedPtr and shouldn't be deleted explicitely

18 years agoMathias Fröhlich:
ehofman [Tue, 24 Jan 2006 14:44:35 +0000 (14:44 +0000)]
Mathias Fröhlich:

Incorporating the shared ptr code:
- All scenegraph references from SimGear
- SGMaterial which already had a reference counter uses now that common
  infrastructure.
- SGMatModel is now counted.
- SGSoundSample from SimGear
- And the corresponding change for the sound samples in flightgear which fixes
  a latent crash if FGBeacon would evern be deleted.

18 years agoMathias Fröhlich:
ehofman [Thu, 12 Jan 2006 13:47:04 +0000 (13:47 +0000)]
Mathias Fröhlich:

Use the new automatic reference counter instead of doing
that ourselfes.

18 years agoMSVC vsnprintf ( in fact _vsnprinft ) returns -1 when the buffer is too short
fredb [Tue, 10 Jan 2006 20:25:46 +0000 (20:25 +0000)]
MSVC vsnprintf ( in fact _vsnprinft ) returns -1 when the buffer is too short

18 years agoUse the new SGPath::create_dir function
fredb [Sat, 7 Jan 2006 13:21:36 +0000 (13:21 +0000)]
Use the new SGPath::create_dir function
Ensure no triangles array could have more than 32767 vertices, a PLIB limit.