]> git.mxchange.org Git - flightgear.git/log
flightgear.git
15 years agoSync. with JSBSim CVS
ehofman [Thu, 14 Aug 2008 14:16:50 +0000 (14:16 +0000)]
Sync. with JSBSim CVS

15 years agonasal.vim: add <?...?> and <!--...--> to XML embedded mode
mfranz [Wed, 13 Aug 2008 10:21:42 +0000 (10:21 +0000)]
nasal.vim: add <?...?> and <!--...--> to XML embedded mode
and only enable that if the file extension is xml

15 years agoAllow generic file protocol to terminate fg after a number of repetitions.
timoore [Thu, 7 Aug 2008 22:24:47 +0000 (22:24 +0000)]
Allow generic file protocol to terminate fg after a number of repetitions.

15 years agoRon JENSEN: s/hide/!enabled/
mfranz [Wed, 6 Aug 2008 15:18:41 +0000 (15:18 +0000)]
Ron JENSEN: s/hide/!enabled/

use <enabled>false</enabled> flag for widgets that shouldn't be drawn
instead of <hide>true</hide>. This is consistent with other places
in fgfs, like menu entries, hud elements, subsystem switches, etc.

15 years agouse <enabled>false</enabled> flag for widgets that shouldn't be drawn
mfranz [Tue, 5 Aug 2008 05:27:07 +0000 (05:27 +0000)]
use <enabled>false</enabled> flag for widgets that shouldn't be drawn
instead of <hide>true</hide>. This is consistent with other places
in fgfs, like menu entries, hud elements, subsystem switches, etc.

15 years ago<foo/> are also comments in XML embedded Nasal mode
mfranz [Tue, 5 Aug 2008 05:23:16 +0000 (05:23 +0000)]
<foo/> are also comments in XML embedded Nasal mode

15 years ago- fix function argument order (top & bottom were swapped)
mfranz [Tue, 5 Aug 2008 05:22:05 +0000 (05:22 +0000)]
- fix function argument order (top & bottom were swapped)
- drop wrong and superfluous rounding (sprintf() rounds already)
- remove redundant default values to property getters (default is 0 already)

15 years agoJames Turner:
ehofman [Sun, 3 Aug 2008 14:34:42 +0000 (14:34 +0000)]
James Turner:
Attached patch + new file make FGNavRecord have a .cxx file, and a constructor w
hich allows all the parameters to be supplied. Along the way I also cleaned up t
he navrecord.hxx header, lots more header pollution has been killed.

Some long methods are no longer inline, but were all suspiciously long to meet c
ompiler inlining criteria (I'm not clear if the 'inline' keyword is advisory or
mandatory in this situation) - I don't expect this to affect performance in any
way whatsoever.

The constructor addition is to support some hacking I'm doing improving the star
tup performance of the navDB by lazily loading the data, and caching it in a mor
e efficient format than text. I'm submitting this change (and probably some othe
r small tweaks in the future) since they are worthwhile as cleanups regardless o
f how my current experiments work out.

15 years agoAdd support for squared damping coefficients for gears.
ehofman [Sun, 3 Aug 2008 13:52:45 +0000 (13:52 +0000)]
Add support for squared damping coefficients for gears.

15 years agomark obvious XML tags as comments
mfranz [Sun, 3 Aug 2008 13:52:29 +0000 (13:52 +0000)]
mark obvious XML tags as comments

15 years agoRestore sky
timoore [Sat, 2 Aug 2008 22:32:55 +0000 (22:32 +0000)]
Restore sky

Don't clear BACKGROUND_BIT in the master camera's cull
mask. ViewPartitionNode does that where needed.

15 years agoUpdate MSVC 7.1 projects - Adapt to OSG 2.6.0-rc1 : location of header files should...
fredb [Sat, 2 Aug 2008 11:31:17 +0000 (11:31 +0000)]
Update MSVC 7.1 projects - Adapt to OSG 2.6.0-rc1 : location of header files should now be searched in the install directory, here ..\..\..\install\msvc71\OpenSceneGraph\include
3dconvert not built anymore. Use osgconv

15 years agoMSVC 7.1 fix
fredb [Sat, 2 Aug 2008 10:01:08 +0000 (10:01 +0000)]
MSVC 7.1 fix

15 years agosimplification
mfranz [Sat, 2 Aug 2008 10:00:06 +0000 (10:00 +0000)]
simplification

15 years agoCameraGroup class for managing multiple cameras.
timoore [Fri, 1 Aug 2008 15:57:29 +0000 (15:57 +0000)]
CameraGroup class for managing multiple cameras.

CameraGroup supports cameras opened in different windows or in the
same window, with flexible view and perspective specification.

Clean up mouse click handling via osgViewer. Don't let any camera
"have focus;" this forces events to be reported in window coordinates
and simplifies life. Don't use the osgViewer::View::requestWarpPointer
method; instead use our own code which only deals with window
coordinates.

Make glut and sdl versions work with CameraGroup too.

15 years agoMake sure 'make dist' keeps working.
durk [Thu, 31 Jul 2008 17:21:45 +0000 (17:21 +0000)]
Make sure 'make dist' keeps working.

15 years agoheader cleanups
ehofman [Thu, 31 Jul 2008 12:04:32 +0000 (12:04 +0000)]
header cleanups

15 years agoExplicitly set traversal mask for near camera.
timoore [Thu, 31 Jul 2008 09:20:35 +0000 (09:20 +0000)]
Explicitly set traversal mask for near camera.

Don't depend on camera inheritance; set the mask in the traverse() method.

15 years agobetter use unset() for unsetting ...
mfranz [Wed, 30 Jul 2008 15:01:59 +0000 (15:01 +0000)]
better use unset() for unsetting ...

15 years agodisable clip planes after use (Doesn't seem to make a difference,
mfranz [Wed, 30 Jul 2008 14:52:01 +0000 (14:52 +0000)]
disable clip planes after use (Doesn't seem to make a difference,
and is only a temporary measure, as it was planned to OSGify the
HUD and put it in the scenegraph etc.)

15 years agos/clip/clipping/
mfranz [Tue, 29 Jul 2008 15:48:11 +0000 (15:48 +0000)]
s/clip/clipping/

15 years ago- extract clipping code to HUD.cxx, allowing generic clipping
mfranz [Tue, 29 Jul 2008 15:41:25 +0000 (15:41 +0000)]
- extract clipping code to HUD.cxx, allowing generic clipping
  and separate ladder clipping
- fix acceleration node names
- draw 90/-90 bars if there's no zenith/nadir

15 years agoChanged ViewPartition to use two cameras instead of three.
timoore [Tue, 29 Jul 2008 11:18:21 +0000 (11:18 +0000)]
Changed ViewPartition to use two cameras instead of three.

I thought that this would fix the "black hole in the sky" problem,
which turned out to be caused by an OpenSceneGraph bug. Nevertheless
it is a simplification.

15 years agoJames Turner:
ehofman [Tue, 29 Jul 2008 08:27:48 +0000 (08:27 +0000)]
James Turner:
 * experimental clean-up / reduction on two of the FG headers:
   (I'm going to await feedback on the developers list before doing more of
    these, to avoiding going over files multiple times, but in principle it
    seems pretty straightforward.)

 * final fixes for SG_USING_STD removal

15 years ago- read x/y/width/height as floats
mfranz [Mon, 28 Jul 2008 17:04:24 +0000 (17:04 +0000)]
- read x/y/width/height as floats
- don't write bars past nadir/zenith (better MILSTD compliance)
- introduce tick-length/zero-bar-overlength/enable-dive-bar-angle
  configuration parameters
- simplification

15 years agoSG_GLUT_H removed
fredb [Sun, 27 Jul 2008 17:58:33 +0000 (17:58 +0000)]
SG_GLUT_H removed

15 years ago- remove the SG_GLxxxx_H #defines, since OSG provides its own versions
ehofman [Sun, 27 Jul 2008 16:25:13 +0000 (16:25 +0000)]
- remove the SG_GLxxxx_H #defines, since OSG provides its own versions
 - this exposed a bizarre issue on Mac where dragging in <AGL/agl.h> in
   extensions.hxx was pulling in all of Carbon to the global namespace
   - very scary. As a result, I now need to explicitly include CoreFoundation
   in fg_init.cxx.
 - change SG_USING_STD(x) to using std::x

15 years agoRestore inclusion of GLU.h
fredb [Sun, 27 Jul 2008 16:08:45 +0000 (16:08 +0000)]
Restore inclusion of GLU.h

15 years agoRemove unneeded inclusions of windows.h, GL.h and GLU.H
fredb [Sun, 27 Jul 2008 15:18:34 +0000 (15:18 +0000)]
Remove unneeded inclusions of windows.h, GL.h and GLU.H

15 years agoswitch back to "<key> in <hash>" notation (this had failed in one test,
mfranz [Sun, 27 Jul 2008 10:53:57 +0000 (10:53 +0000)]
switch back to "<key> in <hash>" notation (this had failed in one test,
but the reason must have been unrelated)

15 years agoEnable Radio #2.
durk [Sun, 27 Jul 2008 09:59:52 +0000 (09:59 +0000)]
Enable Radio #2.

15 years agoRemove unused variables
fredb [Sun, 27 Jul 2008 09:45:33 +0000 (09:45 +0000)]
Remove unused variables

15 years agoget rid of mouse.cxx and pu{Show|Hide}Cursor()
mfranz [Sun, 27 Jul 2008 09:39:10 +0000 (09:39 +0000)]
get rid of mouse.cxx and pu{Show|Hide}Cursor()

15 years agoremove all the FX and XMESA cruft relating to Glide support.
ehofman [Sun, 27 Jul 2008 07:54:13 +0000 (07:54 +0000)]
remove all the FX and XMESA cruft relating to Glide support.

15 years agoadd cvs source urls
mfranz [Sat, 26 Jul 2008 10:42:13 +0000 (10:42 +0000)]
add cvs source urls

15 years agoOne occurence of Balloon.cxx is enough
fredb [Sat, 26 Jul 2008 08:25:11 +0000 (08:25 +0000)]
One occurence of Balloon.cxx is enough

15 years agoBut I do think we need to remove *this* reference to BaloonSim. :-)
durk [Sat, 26 Jul 2008 07:33:15 +0000 (07:33 +0000)]
But I do think we need to remove *this* reference to BaloonSim. :-)

15 years agoRvert last change. The Balloon and Magic Carpet are there. They didn't
durk [Sat, 26 Jul 2008 07:20:48 +0000 (07:20 +0000)]
Rvert last change. The Balloon and Magic Carpet are there. They didn't
show up during my previous commit though...

15 years agoAdd the BalloonSim and MagicCarpet fdm's back in (i seriously thought this had been...
ehofman [Sat, 26 Jul 2008 07:19:13 +0000 (07:19 +0000)]
Add the BalloonSim and MagicCarpet fdm's back in (i seriously thought this had been done, oops)

15 years agoBalloon and Magic Carpet no longer exist.
durk [Sat, 26 Jul 2008 07:01:07 +0000 (07:01 +0000)]
Balloon and Magic Carpet no longer exist.

15 years agoUpdate MSVC 7.1 projects - Balloon, BalloonSim and MagicCarpet not build anymore
fredb [Fri, 25 Jul 2008 21:57:53 +0000 (21:57 +0000)]
Update MSVC 7.1 projects - Balloon, BalloonSim and MagicCarpet not build anymore

15 years agoBalloon and MagicCarpet already moved to SP_FDM
fredb [Fri, 25 Jul 2008 21:56:49 +0000 (21:56 +0000)]
Balloon and MagicCarpet already moved to SP_FDM

15 years agodrop dependence on PyXML and use standard lib methods
mfranz [Fri, 25 Jul 2008 20:22:36 +0000 (20:22 +0000)]
drop dependence on PyXML and use standard lib methods

15 years agoThis should apply, and everything should build cleanly, in isolation from the
ehofman [Fri, 25 Jul 2008 18:38:29 +0000 (18:38 +0000)]
This should apply, and everything should build cleanly, in isolation from the
SimGear change. It changes all the SG_xxxx to be the 'real' includes, and gets
rid of many #ifdef SG_HAVE_STD_INCLUDES. As an added bonus, rather than
replacing 'SG_USING_NAMESPACE(std)' with 'using namespace std', I just fixed
the small number of places to use std:: explicitly. So we're no longer polluting
the global namespace with the entire contents of std, in many cases.

There is one more 'mechanical' change to come - getting rid of SG_USING_STD(X),
but I want to keep that separate from everything else. (There's another
mechnical change, replacing <math.h> with <cmath> and so on *everywhere*, but
one step at a time)

15 years agoAttached patches remove BORLANDC, and hence SG_MATH_EXCEPTION_CLASH and SG_INCOM
ehofman [Fri, 25 Jul 2008 10:39:53 +0000 (10:39 +0000)]
Attached patches remove BORLANDC, and hence SG_MATH_EXCEPTION_CLASH and SG_INCOM
PLETE_FUNCTIONAL from SimGear and FlightGear.

As a result, SG_HAVE_STD_INCLUDES is now *always* set, so I will get the boring
fixes for that done, but separately. I'm still auditing the other things in comp
ilers.h - there's a lot that can die now BORLAND is gone.

15 years agoSome Mac / __APPLE___ cleanups:
ehofman [Fri, 25 Jul 2008 08:39:28 +0000 (08:39 +0000)]
Some Mac / __APPLE___ cleanups:

    - remove the OSX_BUNDLE crap *I* introduced years ago - we're always a a bun
dle on Mac now.

    - fix up the default fg-root on Mac to be FlightGear.app/Contents/Resources/
data - i.e the location used by the macflightgear.org distro, and indeed the obv
ious 'correct' location. Not sure why I didn't use that in the first place, back
 in the day.

    - remove the CPSForegroundEnable hack. For one thing, we're a bundle and don
't need it, and for another, osgViewer on Mac does the same logic using a newer,
 public API rather than a hack into the OS.

    - remove the strange logic for doing fgOSInit 'early' (in bootstrap rather t
han main) when running from the command line on Mac; again this is obsolete, and
 no one seems to know why it was ever necessary. I guess it was an interaction w
ith SDL when running without a bundle.

    - (not Mac related) remove obsolete code bracketed by ENABLE_PLIB_JOYSTICK a
nd USE_GLIDE (neither of which are ever set, even from config.h that I can see)
in main.cxx

15 years agoRemove the remaining macintosh (not MacOS-X) and MWERKS workarounds.
ehofman [Fri, 25 Jul 2008 07:29:20 +0000 (07:29 +0000)]
Remove the remaining macintosh (not MacOS-X) and MWERKS workarounds.

15 years agotypo: s/atl/alt/ (for "alternatives")
mfranz [Thu, 24 Jul 2008 20:46:30 +0000 (20:46 +0000)]
typo: s/atl/alt/  (for "alternatives")

15 years agoremove macintosh (note: not MacOS-X) and MWERKS workarounds.
ehofman [Thu, 24 Jul 2008 19:50:58 +0000 (19:50 +0000)]
remove macintosh (note: not MacOS-X) and MWERKS workarounds.

15 years agoargument & parameter completion for bash (instructions inside)
mfranz [Thu, 24 Jul 2008 17:52:04 +0000 (17:52 +0000)]
argument & parameter completion for bash (instructions inside)

15 years agolet removeChild() return the detached node (like the c++ version)
mfranz [Thu, 24 Jul 2008 11:25:13 +0000 (11:25 +0000)]
let removeChild() return the detached node (like the c++ version)

15 years agoadd printf
mfranz [Wed, 23 Jul 2008 15:45:30 +0000 (15:45 +0000)]
add printf

15 years agoadd thisfunc() and props.condition()
mfranz [Tue, 22 Jul 2008 23:33:30 +0000 (23:33 +0000)]
add thisfunc() and props.condition()

15 years ago- use path validation for parsexml() nasal command
mfranz [Tue, 22 Jul 2008 20:26:17 +0000 (20:26 +0000)]
- use path validation for parsexml() nasal command
- util.cxx: add comment

15 years agoSync. with JSBSim CVS (header cleanups).
ehofman [Tue, 22 Jul 2008 08:49:15 +0000 (08:49 +0000)]
Sync. with JSBSim CVS (header cleanups).

15 years agoClean up a couple warnings.
curt [Mon, 21 Jul 2008 19:57:56 +0000 (19:57 +0000)]
Clean up a couple warnings.

15 years ago * change a number of function parameters to const where appropriate
ehofman [Mon, 21 Jul 2008 12:32:22 +0000 (12:32 +0000)]
  * change a number of function parameters to const where appropriate
  * fix a problem where the wrong node-name length was returned
  * xmlgrep now also works when only the -e options is specified
  * fix xmlgrep to show the correct node-name (it reported the parent
    node-name in the previous version)

15 years ago* fix __xmlSkipComment to properly find the end of comment tag.
ehofman [Sun, 20 Jul 2008 09:38:53 +0000 (09:38 +0000)]
* fix __xmlSkipComment to properly find the end of comment tag.
* add the xmlGetNodeName and xmlCopyNodeName functions
* add the xmlCopyString function
* clean up some code

'xmlgrep -v' now returns the proper node name

15 years agoUpdate MSVC 7.1 projects - new JSBsim
fredb [Sun, 20 Jul 2008 07:40:16 +0000 (07:40 +0000)]
Update MSVC 7.1 projects - new JSBsim

15 years agoAdd repeat capability to input from files.
timoore [Sat, 19 Jul 2008 16:23:05 +0000 (16:23 +0000)]
Add repeat capability to input from files.

15 years agoChange the low level internals of the xml library to reflect the following;
ehofman [Sat, 19 Jul 2008 12:18:05 +0000 (12:18 +0000)]
Change the low level internals of the xml library to reflect the following;
 * rewrite the code to always recursively walk the node tree when searching
    for a particular node. this is required for cases where a node with a
    particular name is located deeper in a node with the same name;
    for example -r /configuration/device/reference/device would fail in the
    previous verion
  * rename xmlGetElement to xmlGetNodeNum and add the possibility to request
    the nth node with this name
  * rename xmlGetNumElements to xmlGetNumNodes

The drawback is that this is approach slows down processing the files a bit,
but at least it can now find any xml node in any xml file.
Also, no more function changes are planned from now on.

15 years agoRestore code after unwanted change in last commit
fredb [Thu, 17 Jul 2008 21:32:20 +0000 (21:32 +0000)]
Restore code after unwanted change in last commit

15 years agoSIGPIPE is a unix thing
fredb [Thu, 17 Jul 2008 21:28:43 +0000 (21:28 +0000)]
SIGPIPE is a unix thing

15 years agomake sure that the validation nodes are readable and writable
mfranz [Thu, 17 Jul 2008 08:38:13 +0000 (08:38 +0000)]
make sure that the validation nodes are readable and writable

15 years agoignore SIGPIPE. If we pipe our output to another application, let's say
mfranz [Tue, 15 Jul 2008 16:55:23 +0000 (16:55 +0000)]
ignore SIGPIPE. If we pipe our output to another application, let's say
to kst, and kst is closed, this does *not* mean that we want FlightGear
to "abort".

15 years agoinclude cstring for memset() (necessary for gcc 4.3.*; backported
mfranz [Tue, 15 Jul 2008 13:08:03 +0000 (13:08 +0000)]
include cstring for memset() (necessary for gcc 4.3.*; backported
from JSBSim/cvs)

15 years agoMSVC doesn't have a round function
fredb [Mon, 14 Jul 2008 09:48:09 +0000 (09:48 +0000)]
MSVC doesn't have a round function

16 years agoRemoval of depricated version of Airports/trafficcontrol.[ch]xx
durk [Sun, 13 Jul 2008 12:57:06 +0000 (12:57 +0000)]
Removal of depricated version of Airports/trafficcontrol.[ch]xx
(See src/ATC for it's new location)

16 years agoInitial attempt to establish a better integration between AI and ATC code.
durk [Sun, 13 Jul 2008 12:51:06 +0000 (12:51 +0000)]
Initial attempt to establish a better integration between AI and ATC code.
Various other patches that have been lingering around for a while:
 * Moved trafficcontrol.[ch]xx from the Airports directory to ATC, where
   it really belongs.
 * AI aircraft will request startup clearance, and ground control will
   approve.
 * Starting AI Aircraft will be pushed back to a predefined holding point
   on the ground network, and wait a while before taxiing out to the runway

16 years ago- Added the ATC directory again, in it's new incarnation as the storage
durk [Sun, 13 Jul 2008 12:38:01 +0000 (12:38 +0000)]
- Added the ATC directory again, in it's new incarnation as the storage
  location for general purpose ATC functions, meant to be compatible with
  the AIModels code.

16 years agos/offset/offsets/
mfranz [Sat, 12 Jul 2008 21:13:48 +0000 (21:13 +0000)]
s/offset/offsets/

16 years agothe usual copy from $FG_ROOT/Docs/
mfranz [Sat, 12 Jul 2008 14:13:22 +0000 (14:13 +0000)]
the usual copy from $FG_ROOT/Docs/

16 years ago"load"/"save" flight: enforce ".sav" extension
mfranz [Fri, 11 Jul 2008 20:55:43 +0000 (20:55 +0000)]
"load"/"save" flight: enforce ".sav" extension

16 years agouse same path validation process for fgcommands "load", "save",
mfranz [Fri, 11 Jul 2008 16:36:54 +0000 (16:36 +0000)]
use same path validation process for fgcommands "load", "save",
"loadxml", ans "savexml" as is used for io.open(). This is still
Nasal based for now. See $FG_ROOT/Nasal/io.nas.

16 years agoadd a reference for memset()
ehofman [Fri, 11 Jul 2008 14:11:54 +0000 (14:11 +0000)]
add a reference for memset()

16 years agoUpdate to the latest version of JSBSim which supports Lighter Than Air craft
ehofman [Thu, 10 Jul 2008 17:23:02 +0000 (17:23 +0000)]
Update to the latest version of JSBSim which supports Lighter Than Air craft
(like Airships) and external forces.

16 years agoreturn attribute mask as unsigned
mfranz [Thu, 10 Jul 2008 11:36:10 +0000 (11:36 +0000)]
return attribute mask as unsigned

16 years agolet n.getAttribute("last") return SGPropertyNode::LAST_USED_ATTRIBUTE
mfranz [Thu, 10 Jul 2008 10:58:06 +0000 (10:58 +0000)]
let n.getAttribute("last") return SGPropertyNode::LAST_USED_ATTRIBUTE
(this is the same constant value for all properties)

16 years agodon't run the help browser from a property that a user could have
mfranz [Wed, 9 Jul 2008 19:35:53 +0000 (19:35 +0000)]
don't run the help browser from a property that a user could have
changed to something evil, but rather make sure that the browser
string can be trusted. (TODO: change system() to vfork()/execvp() ?)

16 years agolet the --prop: option overwrite write-protected properties
mfranz [Wed, 9 Jul 2008 15:09:23 +0000 (15:09 +0000)]
let the --prop: option overwrite write-protected properties
("The user is always right, but not always bright.")

16 years agoremove depreciated and unused src/GUI/trackball.*
mfranz [Wed, 9 Jul 2008 12:44:32 +0000 (12:44 +0000)]
remove depreciated and unused src/GUI/trackball.*

16 years agoremove depreciated and unused src/GUI/trackball.*
mfranz [Wed, 9 Jul 2008 12:37:54 +0000 (12:37 +0000)]
remove depreciated and unused src/GUI/trackball.*

16 years agoremove depreciated gui_local.[ch]xx:
mfranz [Wed, 9 Jul 2008 12:22:33 +0000 (12:22 +0000)]
remove depreciated gui_local.[ch]xx:
- drop unused parts (MouseQuat/GuiQuat)
- move "old-reinit-dialog" fgcommand to fg_command.cxx under new name
  "reset" for now. (May later get merged with fgcommand "reinit".)
- move reInit() to fg_init.cxx: This was used by Shift-Esc and
  Menu->File-Reset (via fgcommand "old-reinit-dialog"). We have already
  a similar function fgReInitSubsystems() in fg_init.cxx, so these two
  functions will probably get merged later.)

16 years agorename attributes write/read to writable/readable (The original names
mfranz [Mon, 7 Jul 2008 10:23:24 +0000 (10:23 +0000)]
rename attributes write/read to writable/readable (The original names
where chosen to match the SGPropertyNode::WRITE/READ symbol names and
should make it more consistent, but this isn't really something that
should be exposed to the Nasal developer.)

16 years ago* reorganize the code to be able to skip comment sections
ehofman [Sun, 6 Jul 2008 11:34:50 +0000 (11:34 +0000)]
* reorganize the code to be able to skip comment sections
* depreciate __xmlFindNextElement and use __xmlGetNode instead
* xmlGetNextElement now returns char* instead of void* for furute use
* add preliminary support for wildcards in the search path ('*' and '?')

16 years ago * fix a problem caused by removing the last unnecessary alloc
ehofman [Tue, 1 Jul 2008 12:15:46 +0000 (12:15 +0000)]
 * fix a problem caused by removing the last unnecessary alloc
 * strip leading-, and trailing spaces from the string before comparing
 * fix a problem where trailing spaces weren't removed

16 years agomove[1] icons to source dir, from where the respective installers can copy
mfranz [Mon, 30 Jun 2008 14:02:26 +0000 (14:02 +0000)]
move[1] icons to source dir, from where the respective installers can copy
them to the final place depending on OS and Desktop Environment (just
like man pages and binaries). Otherwise the installer would need access to
$FG_ROOT only for the icons, which is an undesirable dependency.

[1] there's a lock on the $FG_ROOT/icons/ dir, so I'll remove
them later from the old place.

16 years agowhoops, my bad
ehofman [Mon, 30 Jun 2008 12:05:26 +0000 (12:05 +0000)]
whoops, my bad

16 years agoRemove the last unnecessary alloc from the library
ehofman [Mon, 30 Jun 2008 12:00:40 +0000 (12:00 +0000)]
Remove the last unnecessary alloc from the library

16 years agoSome small updates; fix typo's and fix a small memory leak
ehofman [Mon, 30 Jun 2008 07:30:02 +0000 (07:30 +0000)]
Some small updates; fix typo's and fix a small memory leak
Improve the documentation in README

16 years agoErik HOFMAN: add documentation/changelog
mfranz [Sun, 29 Jun 2008 13:33:26 +0000 (13:33 +0000)]
Erik HOFMAN: add documentation/changelog

16 years agoErik HOFMAN: faster and better xmlgrep implementation
mfranz [Sun, 29 Jun 2008 13:08:24 +0000 (13:08 +0000)]
Erik HOFMAN: faster and better xmlgrep implementation

16 years ago- remove unused puObject args in gui fgcommands
mfranz [Mon, 23 Jun 2008 18:28:02 +0000 (18:28 +0000)]
- remove unused puObject args in gui fgcommands
- remove no longer used PU macros

16 years ago- make sure /sim/fg-current can't get preset and write protected, so
mfranz [Sun, 22 Jun 2008 11:29:32 +0000 (11:29 +0000)]
- make sure /sim/fg-current can't get preset and write protected, so
  it's reliable at Nasal init time.
- make sure the cwd buffer is terminated

16 years agoadd abort()
mfranz [Thu, 19 Jun 2008 17:25:14 +0000 (17:25 +0000)]
add abort()

16 years agowhile abort() is really only useful for listeners (in all other cases
mfranz [Thu, 19 Jun 2008 17:18:42 +0000 (17:18 +0000)]
while abort() is really only useful for listeners (in all other cases
we know the stack trace), it's better to use a separate function for
it, as this can be used conditionally, as in:

  setlistener("/the/property", func(n) {
      if (n.getValue() < 0) # who's writing negative numbers to /the/property?!
          abort();          # let's check the backtrace ...
  });

16 years agomake fgfs abort() after a listener with runtime arg == -1 was triggered
mfranz [Thu, 19 Jun 2008 10:39:12 +0000 (10:39 +0000)]
make fgfs abort() after a listener with runtime arg == -1 was triggered
and its function called. This is useful for debugging purposes, to get
a core dump which unveils which C++ code was writing to the property.
This isn't any more of a security problem as other Nasal code that
makes fgfs crash, such as:  var x = func call(x); x();  The difference
is that the listener doesn't use up all memory before and is much quicker.

16 years agoset_fg_root: prevent circumvention of the Nasal security code through
mfranz [Mon, 16 Jun 2008 07:46:55 +0000 (07:46 +0000)]
set_fg_root: prevent circumvention of the Nasal security code through
pre-setting and write-protecting /sim/fg-root in *-set.xml files

16 years ago- allow users to set fg-home dir via FG_HOME environment variable, but
mfranz [Fri, 13 Jun 2008 10:52:47 +0000 (10:52 +0000)]
- allow users to set fg-home dir via FG_HOME environment variable, but
- don't allow to do that from any XML file. This is to prevent malign
  code from writing a new fg-home in ~/.fgfs/autosave.xml or other files
  in ~/.fgfs/.

16 years agoFrom Benoit Laniel: replace SG threading constructs with OpenThreads
timoore [Thu, 12 Jun 2008 08:24:39 +0000 (08:24 +0000)]
From Benoit Laniel: replace SG threading constructs with OpenThreads