Note from Robert Osfield. A couple of lines of code in ConvertToInventor.cpp
would not compile under g++ 4.1.2, so rather than hold back the dev release till
this is resolved I've optional compiled out the problem section.
The #define DISABLE_PROBLEM_COMPILE_SECTIONS is used to compile out the problem
section, this define is add via the CMakeLists.txt file.
GL_LUMINANCE data instead of GL_RGB data. You can easily check with
"osgViewer --image klink1_l.tif".
The bug is in ReaderWriterTIFF.cpp function simage_tiff_load, where
numComponents_ret is incorrectly set to 1 instead of 3 for color mapped
data."
stereo format to work. It's a good thing I tested these on a TV
before submitting them since I did indeed have a bug. One thing I
did not test was to see how this would work in windowed mode. Does
the interlaced stereo code have support for 'absolute' positions?
For example a given pixel on the screen is always shown in a given
eye no matter where the graphics context is placed?
"
src/osgDB/FileUtils.cpp to implement the official Windows DLL search
order as described on the page
http://msdn2.microsoft.com/en-us/library/ms682586.aspx . As mentioned,
the search order is now:
1. The directory from which the application loaded.
2. The system directory. (C:\Windows\System32 by default, gotten using the
GetSystemDirectory function)
3. The 16-bit system directory. (C:\Windows\System by default, gotten by
adding "\System" to the path gotten in the next step...)
4. The Windows directory. (C:\Windows by default, gotten using the
GetWindowsDirectory function)
5. The current directory. (".")
6. The directories that are listed in the PATH environment variable. (as
before)
The first four directories are obtained using Win32 API calls, so they
should work correctly even on non-standard Windows installs.
The changes are well commented and should be clear, even to someone
not familiar with the Win32 API.
I have tested in a few scenarios and it works as expected. Serge Lages
has also tested the changes and confirmed they worked as described. I
have not had any other reports though (positive or negative).
I also fixed the issue with a trailing semicolon on the PATH adding an
empty string to the end of the search paths, as this was an
inconsistent side effect rather than a desirable effect. This change
will take effect on other platforms as well, but since it tests for an
empty string in the last item added to the search paths, it should
have no adverse effect.
"
will bring the change in line with what is done on other OSes (Linux)
and works in all tested cases.
For reference, this was tested with:
osgviewer <file>.wrl (file in current directory)
osgviewer <dir>\<file>.wrl (file in child directory, relative)
osgviewer .\<dir>\<file>.wrl (file in child directory, specify current)
osgviewer <drive>:\<dir>\<file>.wrl (absolute path)
"
geometrytechnique in submethods to made more easy the inheritance
between the user and osg-class. This is a first step to add more
functions in osgTerrain. Maybe the subdivision of the method have to
be in the terraintechnique because is the base class of
GeometryTechnique. If Robert or anyone think that this is better i
change this class too."
precise, the shader does not compile on os x because of some
type-conflicts ala "can not convert from const int to const float"
So I changed the offending lines to force the type of the vars. It works
now on OS X (albeit very slowly, 3fps on a 7300), perhaps you find the
changes useful. Note: perhaps there is a better way in shaders to
cast/convert from int to float and viceversa."
updatevisitor in osgViewer::Viewer.
The bug prevented DOF animations because osgSim::DOFTransform checks
the traversal number before doing any updates."
OpenVRML 0.14.3 and is without the Boost dependency.
The changes:
- - Fixed loading of textures and normals when no corresponding indices
are specified. It uses vertex indices now, compliant with the VRML spec.
- - Added colour per vertex support.
- - Added group node support.
- - Changed the code to use osg::ref_ptr instead of naked pointers to
avoid memory leaks.
- - Fixed breakage for loading files specified by relative path."
"I have adapted to osgShadow the soft shadow map technique described in "Efficient Soft-Edged Shadows Using Pixel Shader Branching" by Yury Uralsky, Chapter 17 of GPU Gems 2 (Matt Pharr ed. Addison-Wesley).
Here is my code in attachment: basically, it works in the same way as osgShadow/ShadowMap (core code is copied from it) but implements a specific GLSL shader for the soft rendering of penumbra.
I have tested it under Linux with a NVidia graphic card, but there should be no dependency on platform nor on the graphics driver (as far as they support GLSL 2). Screenshots attached show the current results (frame rate bound to v-sync, but the shader takes actually not much time)."
to the view to be done during syncronous updateTraversal().
This feature can be used for doing things like merging subgraphs that have been loaded
in a background thread.
Attached a modification that read the HeigthField position and X,Yintervals.
I also removed the limitation to 1024*1024 to 4096*4096, because when you are preprocessing your data with OSG, it can be useful to read large images/heigthfields. Is there a reason (other than hardware limitations for textures) for this limit ?"
various operating system differences between socklen_t and int have
broken the FreeBSD build. Change was to add __FreeBSD__ to the list of
defines that are checked."
I just add ReinterpretCastConverter in the Reflector to convert void* in T* and T* in void*
files joint :
OpenSceneGraph/include/osgIntrospection/Reflector // modified file
OpenSceneGraph/src/osgIntrospection/Reflector.cpp // modified file
"
of the source file:
The "delete [] path" appears before the "osg::notify", causing the data pointed to by
"filename" to be deleted before access causing an access violation.
...
I have put a comment on
line 521 where I have moved the "delete []path" below.
"
multiple GraphicsWindows, this singleton is accessable via GUIEventEvent::getAccumulatedEventState().
Added use of this new singleton in GraphicsWindow* implementations.
Added WindowSizeHandler to osgkeyboard to help with debugging of event state
between windows.
Created a new GraphicsThread subclass from OperationThread which allows the
GraphicsContext specific calls to be moved out of the base OperationThread class.
Updated the rest of the OSG to respect these changes.
completed the new registration of the plugin-readerwriters
("REGISTER_OSGPLUGIN") according to your osgstaticviewer-example (see
attachment, based on today's svn)."
Added and cleaned up DeleteHandler calls in osgViewer to help avoid crashes on exit.
Changed DatabasePager across to dynamically checcking osg::getCompileContext(..)
Updated wrappers.
The current code takes the mouse cursor position and adds it to the
window (left,top) position, and sends the mouse cursor there. But this
doesn't take into account the window decoration.
The new code converts the given (x,y) coordinates from the client area
coordinate system to the screen instead using ClientToScreen. I think
this is the natural windows way to do it.
Tested on XP with osgviewer
"
Note from Robet Osfield, made a few changes to layout to make it more consistent
with the rest of the OSG and used #if 0 instead if (0) blocks.
"
Found in the join file the fix for the bug found by Rafa.
Problem :
osgIntrospection::Value grp(new osg::Group);
osgIntrospection::ValueList vlcall;
vlcall.push_back(osgIntrospection::Value("toto"));
const osgIntrospection::MethodInfo *m =
grp->getType.getCompatibleMethod("setName", vlcall, true);
if (m)
{
m->invoke(grp, vlcall); // ** SEGFAULT here
}
Algorithm explanation :
The "invoke" method try to convert "grp", which reflect an
"osg::Group*", in a
"osgIntrospection::Value", which reflect a "osg::Node*".
This because
the "setName(const char *)" method found by
"grp->getType.getCompatibleMethod"
is an "osg::Object" type method.
When osgIntrospection do this conversion it try :
- to found a "osgIntrospection::Converter" to convert
from "osg::Group*" to "osg::Node*"
- to found a chain of "osgIntrospection::Converter" to convert
from "osg::Group*" to "one or many type" to "osg::Node*"
- to converte an Enum to int or unsigned int
- to convert the value in its "value string representation",
then converte this string in the destination value
Else it throw a "TypeConversionException".
Bug :
1)
When osgIntrospection try to found a chain of
"osgIntrospection::Converter"
It could do any downcast or (Type to SuperType) or upcast
(SuperType to Type).
This mean the the chain could be :
osg::Group to osg::Transform to osg::Camera to
osg::CullSettings to osg::CullStack to
osg::CollectOccludersVisitor to
osg::NodeVisitor to osg::Referenced to osg::Object
During the convertion with this chain, A METTRE failed and
the pointer in
"grp" is set NULL. But the "grp" is always a valid
"osgIntrospection::Value"
and so, osgIntrospection accept the conversion. Then it try
to use this pointer
to call the "setName" function. And Bing SEGFAULT.
2)
In "bool Reflection::accum_conv_path( ... )"
the convection path isn't accumulate in the recursive loop.
this cause multi request of a conversion path, and a
slowdown in the
conversion algorithm.
3)
Use of the last conversion way in a conversion from
pointer to pointer
this mean you can do this :
"osg::Node*" to " value string representation" to "osg::Material*"
What a bad thing !!!
Solution :
1)
Introduce the concept of dynamic_cast and static_cast.
now, to do a conversion, osgIntrospection does this :
- to found a "osgIntrospection::Converter" to convert
from "osg::Group*" to "osg::Node*"
- to found a chain of "osgIntrospection::Converter" to convert
from "osg::Group*" to "one or many type" to "osg::Node*"
only with static_cast, downcast (Type to SuperType)
- to found, if the source and the destination are two pointer,
a chain of "osgIntrospection::Converter" to convert
from "osg::Group*" to "one or many type" to "osg::Node*"
only with dynamic_cast, upcast (SuperType to Type)
- to convert an Enum to int or to unsigned int
- to convert the value in its "value string representation",
then convert this string in the destination value
Else it throw a "TypeConversionException".
Add the "enum CastType" to distinguish the static_cast or
dynamic_cast converter.
Add file OpenSceneGraph/include/osgIntrospection/CastType
2)
add a line to accumulate converter in converter Path.
3)
add a line to check if source and destination are pointer.
"
COLLADA modules STLDatabase, LIBXMLPlugin and stdErrPlugin are
statically included in the main COLLADA library on Linux and shouldn't
be linked separately - those libraries do not exist in the default Linux
build and the compilation will fail.
Second issue - the current version of the COLLADA plugin (both current
HEAD in Subversion and the one in stable 2.0) do not work right with the
stable COLLADA DOM 1.4.1. I am getting the following error:
"
is not the usual OpenGL BOTTOM_LEFT orientation, but with the origin TOP_LEFT. This
allows geometry setup code to flip the t tex coord to render the movie the correct way up.
bugs in the Carbon-implementation of GraphicsWindow, so here's another
fix for setWindowDecorationImplementation, now updates the
titlebar-height, and fixes some display-issues when switching the
decoration on/off.
"
Stephan: "attached you'll find some modifications to the GraphicsWindow-class and
their platform-dependant implementations.
The problem:
setWindowRectangle and setWindowDecoration do not update the
traits-object, so, if you call setWindowRectangle on a
not-realized-window it will open with another size when realized later.
getWindowRectangle reports possible wrong sizes if setWindowRectangle
called before.
My solution:
split the implementation in two parts:
GraphicsWindow::setWindowRectangle will update its traits-object and
call afterwards the virtual method setWindowRectangleImplementation
(which is implemented by the derived platformspecific classess). For
setWindowDecoration I am useing a similar mechanism.
I hope you'll find the submission useful, the Win32 and X11 changes are
not tested but should work."
Changes to this made by Robert are call of resized in setWindowRectangle
instead of setting of Traits, and use of a bool return type.
R3G3B2, R5G6B5, A1R5G5B5, X1R5G5B5, A4R4G4B4, X4R4G4B4, R8G8B8 (now
without swaping of red and blue), A8R8G8B8 (also w/o swapping),
X8R8G8B8, A8B8G8R8, X8B8G8R8, A2R10G10B10, A2B10G10R10, L4A4 (not work
on my machine), L16A16, L16, A16B16G16R16, A16B16G16R16F,
Q16W16V16U16, R32F, R16F and A32B32G32R32F.
And these ones are correctly detected, but prints "unsupported" using
osg::notify(osg::WARN) and are not loaded:
A8R3G3B2, G16R16, G16R16F, G32R32F and CxV8U8.
Also added checking of not supported DDPF_BUMPDUDV (V8U8, V16U16,
Q8W8U8L8, A2W10U10V10 etc.) and DDPF_BUMPLUMINANCE (L6V5U5, X8L8V8U8,
etc.) pixel formats.
Mipmap handling is slightly modified and now support all additional formats.
"
changes I made:
+ put a warning in the console if a nonexistant screen is requested
+ add getters for the aglcontext and pixelformat -- I need access to
them in my own code.
"
old loader, but appear very, very wrong with the new one. I traced the
problem to the handling of the palette override flags in the external
reference records. The current behavior for handling the palette
override flags for external references has different offsets for
different OpenFlight version (2 bytes for 14.2-15.1 and 4 bytes for 15.2
and later). However, I believe this behavior is incorrect.
I know that the original 14.2 OpenFlight spec (dated April 1995)
specifies 2 bytes between the filename and the override flags, and the
15.4 and later specs specify 4 bytes. However, I also found a 14.2.4
OpenFlight spec (dated January 1996) that changes the specification to 4
bytes. Also, the databases in question were created using an old IRIX
version of MultiGen II, which wrote OpenFlight 14.2 files natively.
These files also have 4 bytes between the filename and flags.
Furthermore, these databases have always worked properly under earlier
versions of OSG, under Performer, and in every MultiGen product we've used.
This leads me to believe that the original 14.2 spec was incorrect (the
14.2.4 spec corrected this error), and there should be 4 bytes between
the filename and flags for all OpenFlight files version 14.2 and later.
The attached fix modifies the OpenFlight loader to behave in this way."
Currently, if the texture attribute file doesn't explicitly specify an
internal format, the loader will force it to use GL_RGB, which keeps
translucent textures (eg. GL_RGBA textures) from showing up properly.
This patch changes the default behavior to simply use the image's format
instead of forcing a particular format."
libraries listed under TARGET_EXTERNAL_LIBRARIES.
The removed libraries are not needed when linking the plugin, they are
loaded during runtime by Performer.
The modified file is attached."
