James Dickson schrieb:
Fetching external item into 'OpenSceneGraph/include/OpenThreads'
svn: PROPFIND request failed on
'/svn/osg/OpenThreads/trunk/include/OpenThreads'
svn: PROPFIND of '/svn/osg/OpenThreads/trunk/include/OpenThreads': 301
Moved Permanently ( http://www.openscenegraph.com)
Paul Martz napsal(a):
[...] The naïve approach -- just let OSG load it -- appears to cause a
guaranteed frame rate hit.
I imagine you could break the texture up into smaller chunks and let OSG
page each in incrementally as it comes into range.
Yes, breaking the texture into the smaller ones
Hi Renato,
Are you using 2.0? If so update to one of the 2.1.x series or SVN
version as this fixes a problem with frame stamps that CompositeViewer
had in 2.0.
Robert.
On 9/11/07, Renato Silveira [EMAIL PROTECTED] wrote:
I'm trying to use osgCal with osg::CompositeViewer. I did exactly what
On 9/10/07, Paul Martz [EMAIL PROTECTED] wrote:
Hi Robert -- Assuming you've massaged your database so that the highest res
textures only get loaded/displayed at close ranges, how would you use OSG to
avoid the (seemingly inevitable) cost of loading the large texture from disk
and sending it
On 9/11/07, Johan Johnsson [EMAIL PROTECTED] wrote:
Hello !
I want to mirrorize a specific slave counted from the mastercameras view.
At the moment i use a offset matrix
addSlave(camera.get(),osg::Matrixd(),offsetMatrix); that rotate it 180
degrees in Z-axis, well thats good now i look
On Tue, 11 Sep 2007 10:40:31 +0200, Robert Osfield
[EMAIL PROTECTED] wrote:
On 9/11/07, Johan Johnsson [EMAIL PROTECTED] wrote:
Hello !
I want to mirrorize a specific slave counted from the mastercameras
view.
At the moment i use a offset matrix
On 9/11/07, Johan Johnsson [EMAIL PROTECTED] wrote:
You just need to scale the y axis by -1, then flip the front face to
prevent the wrong faces from being culled.
addSlave(camera.get(),osg::Matrixd::scale(1,-1,1),offsetMatrix);
it's get upside down :S
Opps... try the x axis...
In producer there was a method for specifying if one should get (or not)
auto repeat for keyboard events when a user holds a key down...
RenderSurface::setAutoRepeatMode(bool flag);
Is there something similar available is osgViewer?
I found a post from June discussing this behaviour in X11, but
HI Anders,
There isn't such a feature right now. The changes back in June were
down to making X11 behave more like Windows and OSX w.r.t keyboard
repeats.
I presume switching off keyboard repeat will require a platform
specific call to the windowing system to switch it off. I don't know
about
On Tue, 11 Sep 2007 11:10:09 +0200, Robert Osfield
[EMAIL PROTECTED] wrote:
On 9/11/07, Johan Johnsson [EMAIL PROTECTED] wrote:
You just need to scale the y axis by -1, then flip the front face to
prevent the wrong faces from being culled.
Is it possible to have a viewer with a RTT camera render with one
GraphicsContext, then to use that texture with another viewer and a different
GraphicsContext? It works ok if all the cameras use the same GC.
What I'm aiming for is to render to texture in a completely offscreen buffer,
then use
hi, all
The platform: windows xp, vc 71
I want download file with .net plugin,and write to the disk before
osgDB::readfile.
So I add some code :
if( reader != 0L )
{
I'm wondering how to go about layering manipulators so that by default the
terrain manipulator is used except for a small corner of the screen where I
want a custom manipulator (like Google Earth's compass manipulator). Is there
already a mechanism for doing something like this?
Thanks,
-
If you wish to reuse texture objects generated by one context then you
need to share the various contexts with each other. You can set this
via the GraphicsContext::Traits::sharedContext variable.
On 9/11/07, John Donovan [EMAIL PROTECTED] wrote:
Is it possible to have a viewer with a RTT
Lilinx schrieb:
hi, all
The platform: windows xp, vc 71
I want download file with .net plugin,and write to the disk before
osgDB::readfile.
but I find that the result is wrong. it is seen that some unwanted '0D' was
writen into the file.
why?
do you mean ASCII hex 0D? Then
There is no explict support for this, you'll need to filter the mouse
events yourself and pass them on to the manipulator you want to be
active.
On 9/11/07, Nick Prudent [EMAIL PROTECTED] wrote:
I'm wondering how to go about layering manipulators so that by default the
terrain manipulator is
I think that the best way would be to create a manipulator,
encapsulating the terrain manipulator and a compass manipulator ( to be
created also ) and forwarding the handle() callback to the right
manipulator regarding the mouse position.
Nick Prudent a écrit :
I'm wondering how to go about
Robert -- Thanks for the reply. I'm wondering how many people here are grabbing
the Blue Marble datasets and trying to do a Google Earth clone just for fun ;)
Remy -- I like this solution. I'll try it. Thanks.
- Nick -
Date: Tue, 11 Sep 2007 13:48:31
I know that similar question has been asked before, but I was wondering, if
anyone has a solution to this. Or at lease steps I need to implement.
Thanks
___
osg-users mailing list
osg-users@lists.openscenegraph.org
Robert -- Thanks for the reply. I'm wondering how many people here are grabbing
the Blue Marble datasets and trying to do a Google Earth clone just for fun ;)
Remy -- I like this solution. I'll try it. Thanks.
- Nick -
Date: Tue, 11 Sep 2007 13:48:31
Hi all,
I would like to put in my GLSL shader a uniform array variable,
I know how to put a simple float or int value with osg::Uniform, but for an
array I don't know
Is anyone can help me?
thanks
Mael.
___
osg-users mailing list
Hi mael
never try but probably with osg::Uniform::setArray(FloatArray * array);
David Callu
2007/9/11, Mael Blanchard [EMAIL PROTECTED]:
Hi all,
I would like to put in my GLSL shader a uniform array variable,
I know how to put a simple float or int value with osg::Uniform, but for
an
Hi Ivan
You need to found or write a Read/Write plugin for your encrypted
compressed file format.
David Callu
2007/9/11, Ivan Bolčina [EMAIL PROTECTED]:
I know that similar question has been asked before, but I was wondering,
if anyone has a solution to this. Or at lease steps I need to
I'm using 2.0.
I will update now.
Thank you!
2007/9/11, Robert Osfield [EMAIL PROTECTED]:
Hi Renato,
Are you using 2.0? If so update to one of the 2.1.x series or SVN
version as this fixes a problem with frame stamps that CompositeViewer
had in 2.0.
Robert.
On 9/11/07, Renato
Robert Osfield wrote:
If you wish to reuse texture objects generated by one context then you
need to share the various contexts with each other. You can set this
via the GraphicsContext::Traits::sharedContext variable.
Ah! OK, that makes sense. But I am having problems with it...
I have the
How can i process this error? Thanks! I must compile success,because i'll
use OSG to cope with my graduate experiment! Help me!
IRIS 27% make
[ 0%] Building CXX object
src/OpenThreads/sproc/CMakeFiles/OpenThreads.dir/SprocThread.o
Hi John,
I can only guess that you first need to realize the window before you
can share it.
For an example of sharing contexts see src/osg/GraphicsContext.cpp and
look for the createCompileContext() code segment.
Robert.
On 9/11/07, John Donovan [EMAIL PROTECTED] wrote:
Robert Osfield
Robert Osfield wrote:
I have merged quite a few changes to the OSG over the last few days,
these could break the build, so I'd like some feedback on how the OSG
is build and running on various platforms before I go ahead the weekly
dev release (2.1.10).
I'm rebuilding now and will test a bit
Hi Robert,
I'm really curious in the error I'm getting trying to compile
svn version:
OSG-SVN/OpenSceneGraph-g++-4.1/include/osg/Uniform:169: error:
'GL_SAMPLER_1D_ARRAY_EXT' was not declared in this scope
OSG-SVN/OpenSceneGraph-g++-4.1/include/osg/Uniform:170: error:
'GL_SAMPLER_2D_ARRAY_EXT'
Robert Osfield wrote:
BTW, is there a reason why you aren't just using a single context and
a FBO RTT Camera?
Not as such. But what I'm playing with is some GPGPU stuff where the results
won't necessarily be rendered anywhere. I'm really only using the textured quad
to check the results,
Hi Rafa,
yes you are right.
I attached a patch which should correct this.
Robert, could you please merge it.
Rafa you could just use this file, until it is merged
in.
Cheers, Art
--- Rafa Gaitan [EMAIL PROTECTED] schrieb:
Hi Robert,
I'm really curious in the error I'm getting trying
to
John Donovan wrote:
...results won't necessarily be rendered anywhere.
On screen, that is.
-J
__
This email has been scanned by the MessageLabs Email Security System.
For more information please visit
I get latest version and build under VS2005.
But in these line:
TextureRectangle.cpp Line 561, TextureCubeMap.cpp Line 387,
Texture3D.cppLine 459,
Texture2DArray.cpp Line 467, Texture2D.cpp Line 411,
Texture1D.cpp Line 417
These lines VS2005 reports:
error C3861: log2 can't find symbol
Is there
I'm seeing the same thing:
error C3861: 'log2': identifier not found
-Paul
_
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of sicong he
Sent: Tuesday, September 11, 2007 8:21 AM
To: Public OpenSceneGraph Users discussion list.
Subject: Re: [osg-users] Please test SVN
Hi Wolfgang,
This is fixed in osgDotNet SVN -- I have successfully built it against
2.1.9. Beware though that there are some unresolved runtime stability
issues with regard to object destruction and the garbage collector with
the OSG 2.1.x versions. If you're not doing anything terribly complex
I'm getting same compiler error on FreeBSD:
Texture1D.cpp:417: error: `log2' was not declared in this scope
I did:
# svn update
# ./configure
# make
-john
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of sicong
he
Sent: Tuesday, September 11,
Should be defined in the cmath or math.h
Hmm, the header is not included properly. gcc seems to
compile it correctly...
Art
--- sicong he [EMAIL PROTECTED] schrieb:
I get latest version and build under VS2005.
But in these line:
TextureRectangle.cpp Line 561, TextureCubeMap.cpp
Line 387,
It seems that VS2005 c++ library do not have log2() function defined. So I
change log2() function to log()/log(2.0), I don't know if it is slower than
log2() but it really works.
2007/9/11, sicong he [EMAIL PROTECTED]:
I fixed this by change all log2(x) to:
log((double)maximum(...))/log(2.0)
yes, it seems that math.h does not define the log2
function on some systems.
Sorry, for this.
My proposition is to patch the osg's Math file to
support for log2 method. I attached the patched file,
which contains log2 function.
Best regards, Art
--- sicong he [EMAIL PROTECTED] schrieb:
It
Hi cedric
http://www.openscenegraph.org/projects/osg/wiki/Community/WikiLogIn
David Callu
___
osg-users mailing list
osg-users@lists.openscenegraph.org
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
Hi All
Fedora 7, gcc 4.2.1, all build fine !!
2007/9/11, Art Tevs [EMAIL PROTECTED]:
yes, it seems that math.h does not define the log2
function on some systems.
Sorry, for this.
My proposition is to patch the osg's Math file to
support for log2 method. I attached the patched file,
Hi Rafa,
I had the same problem 1 week ago.
I use Linux Fedora 7.
When compiling, the default glext.h used is the glext.h provide by the
mesa package.
I have a Nvidia graphic card and the nvidia package for fedora provide an
up to date glext.h.
But this header is in the
It should be the bad VS2005 problem
2007/9/11, sicong he [EMAIL PROTECTED]:
My VS2005 report me:
Compiler Error C2668. 'log' : ambiguous call to overloaded function.
I see in VS2005, function log have only works with double or float:
double __cdecl log(__in double _X);
So you should
Thank you David
David Callu wrote:
Hi cedric
http://www.openscenegraph.org/projects/osg/wiki/Community/WikiLogIn
David Callu
___
osg-users mailing list
Under VS8, everytthing fine apart from linking of osgversion, need to
explicitely link OpenThreads
so add
SET(Target_ADDED_LIBRARIES OpenThreads) after SET(TARGET_SRC
osgversion.cpp )
hope it helps
___
osg-users mailing list
Hi Art and David,
With the new GL2Extension works fine thanks! ;) I think
robert will check in the changes.
David, the package of ubuntu doesn't provide a glext or
I didn't find it on the system! :), but I think patching
GL2Extensions is a better and more generic solution.
Thanks anyway David.
Hello,
I'd like to flip (vertically, horizontally or both) each rendered frame... I
thought to use a MatrixTransfrom and to scale the y-coordinates by -1.0 but
as an effect the normals of each normal also got flipped... how can I
restore the normals? Or maybe is there another way to flip each
Thanks Art, now merged and submitted to SVN.
On 9/11/07, Art Tevs [EMAIL PROTECTED] wrote:
yes, it seems that math.h does not define the log2
function on some systems.
Sorry, for this.
My proposition is to patch the osg's Math file to
support for log2 method. I attached the patched file,
Thanks Luigi, addition now checked in. In future could you send the
whole file as copying and pasting is not very reliable. Cheers,
Robert.
On 9/11/07, Luigi Calori [EMAIL PROTECTED] wrote:
Under VS8, everytthing fine apart from linking of osgversion, need to
explicitely link OpenThreads
so
intel core 2 duo
gentoo 2007.0 x86
gcc (GCC) 4.1.2 (Gentoo 4.1.2 p1.0.1)
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
compile correctly, i
On 9/11/07, Paul Melis [EMAIL PROTECTED] wrote:
I'm rebuilding now and will test a bit later. Did you see my message
yesterday about osgviewerFLTK lockup at exit? Don't want it to get lost...
I did but it doesn't mean I've had time to fix it...
___
Hi Robert,
After seen and changed the patch provide by Art Tevs on the
GL_EXT_gpu_shader4 extension.
I have remarked that the extension guard
#ifndef GL_EXT_gpu_shader4
don't do
#define GL_EXT_gpu_shader4
I go through the glext.h and I see that it is define with the
Hi David,
We don't need to import all the extension prototypes and defines
provided in glext.h, only the ones that the code uses. We also
aren't trying to #define everything as per glext.h to work as a
replacement, so its not required to do a #define GL_extension_name
within each block of
Define values are always the same on all platform so why.
Moreover this seem clean and prevent redefine warning like I would
fix in my other post.
I am agree with this solution.
Cheers
David
___
osg-users mailing list
osg-users@lists.openscenegraph.org
Hi James,
An svn update should fix your problem... I had reported the same
problem a few hours before.. and now is fixed in the svn
Rafa.
On 9/11/07, James Dickson [EMAIL PROTECTED] wrote:
I have been trying to compile OSG under ubuntu 7.04 and I am having some
difficulties.
Having run
Hmm, seems to be a good idea, since it would really
define only those things, which are not defined.
Now we need somebody to merge this for every define in
the osg ;-) I think it would be faster to write a
script to search for #define GL_ and to replace it
with the proposed macro.
Art
---
#define GL_DEIFINE(name, value) \
#ifndef name \
#define name value\
#endif
Then in the header have:
GL_DEFINE(GL_SAMPLER_1D_ARRAY_EXT, 0x8DC0)
Thoughts?
I agree with this solution... the only problem is to pay attention on all
GL_* definitions inside osg:
Hi, Sicong.
The proposition you made before are not really a good
solution, since you force the compiler to use the
double version of the log function, which might be
slow.
For the integer version of the log2, I have another
proposition. I have attached a newly patched file. The
integer log2
On 9/11/07, Rafa Gaitan [EMAIL PROTECTED] wrote:
#define GL_DEIFINE(name, value) \
#ifndef name \
#define name value\
#endif
Then in the header have:
GL_DEFINE(GL_SAMPLER_1D_ARRAY_EXT, 0x8DC0)
Thoughts?
I agree with this solution... the only problem is to pay
Great! So gpg provides decrypted streams, which can be then used directly be
osg?
2007/9/11, Mihai Radu [EMAIL PROTECTED]:
Hi,
To get an encrypting Read/Write plugin, you can always layer an
encryption over a regular Read/Write plugin, for example use gpg
libraries to pipe in the data from
Hi,
just updated my copy..
My redhat WS-5 64-bit linux box seems ok.
My redhat WS-3(taroon update 9) 32-bit linux box can't even compile
OpenThreads..
This is my first time on this 32-bit box, and couldn't find anything
on the list about this..
Guess I should upgrade to ws5?
ml
Hello Art,
Yes, I was watching the discussion about log2 on osg-submissions. My
concern is more about fixing building issues than code improvement, so
whichever way you decide is best works for me.
Just wait couple of days or make your proposition of how to fix the
problem (see
Oh yeah.. I remember now.
I have to add -DCOMPILE_USING_TWO_PARAM_sched_setaffinity
to the compile flags..
thanks, sorry for the bother.
ml
On Sep 11, 2007, at 2:06 PM, Robert Osfield wrote:
Hi Mike,
I think the error you have is down to RedHat shipping
sched_setaffinity with the a
INCLUDE_DIRECTORIES( ${INVENTOR_INCLUDE_DIR} ) seems to fix include
problem, anyway INVENTOR_LIBRARY_DEBUG, although defined, is not used
for debug linking - as a result, debug build is broken while release
seems ok.
I think, something is broken in the general OSG scripts related to
*_DEBUG
64 matches
Mail list logo
