hi all
I have a little probleme,
I have two node who represent two differents things,
and I would like to render one, and afer pushing a touch, changing on other
something like that
//init
osg::Group* root = new osg::Group();
osg::Group* root2 = new osg::Group();
viewer.setSceneData( root
i would be interested in
2007/2/8, Zach Deedler [EMAIL PROTECTED]:
Hello osg users/developers,
Are there any plans to do shadows using shaders? I know hardware
shadowing
is supported on newer video cards.
Zach
___
osg-users mailing list
HI Jason,
Did you disable depth test for this pyramid? If so don't this is why
you can see it through the terrain.
Robert.
On 2/7/07, Jason [EMAIL PROTECTED] wrote:
Attached file: help.JPG
Here is the terrain and my pyramid. The tip of the pyramid is actually near
the center of the terrain,
HI Bill,
On 2/7/07, Bill Prendergast [EMAIL PROTECTED] wrote:
Is user data safe for general use?
Yes, the usage you see of it in the OpenFlight load is what I'd
classify as general use.
If you don't want to overwrite existing UserData simply check for it
and then emit a warning that you are
Hi Zach,
Thanks for the feedback. The debugger problem is odd, osgViewer is
just C++ with a bit of Win32 code in the GraphicsWindowWin32.cpp
implementation. I wonder if they build options are have a an error
that is throwing off the VS7 compiler/debugger. Perhaps its just down
to bugs in VS7.
Hi Chuck,
On 2/7/07, Chuck Cole [EMAIL PROTECTED] wrote:
Once the viewer locks up, I have to kill the process. The display no
longer responds to key presses or mouse movements. Also, I disabled
hyperthreading and I got the same results (locks up on first press of 'm').
I'll rebuild the
Hi all.
If I have vertex shader with attribute attribute vec3 vTangent;
How can I set this attribute in OSG?
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
Hi Terry,
On 2/8/07, Terry Welsh [EMAIL PROTECTED] wrote:
I think I could get a good performance boost rendering shadow maps if
I didn't sort anything. Since shadow texels are either on or off with
no medium gray, I shouldn't matter what order in which geometry is
rendered. Is there a way I
Hi Eric,
You OpenGL code looks wrong, you shouldn't unbind the FBO then do
glReadPixels as this will mean that you are reading from the main
frame buffer not the FBO.
If you look at the RenderStage.cpp RenderStage::drawInner you'll see
the where is unbinds the FBO:
if (fbo_supported)
{
Hi elekis
You can use an osg::Switch node on top of your two scenes:
osg::Switch *switchRoot = new osg::Switch()
... your nodes...
switchRoot-addChild(root);
switchRoot-addChild(root2);
switchRoot-setSingleChild(0);
...
When you want change to the root2 scene only do:
Hi Gazi,
I'm a bit confused. I do debugging on the OSG all the time with gdb
under Linux and don't have to just through any hoops.
What I do is install links to the libs when I first build the
OpenThreads/Producer/OpenSceneGraph using:
make instlinks
Then any time after I just go to a local
Hi Andy,
OpenThreads doesn't at present have support to reentrant/recusive
mutexs so the behavior you've come across is expected, I need it in
the OSG a while back so implementated the osgDB::ReentrantMutex.
In general you shouldn't need ReenetrantMutex's, but in some cases
they are the only
Hi Zach,
On 2/8/07, Zach Deedler [EMAIL PROTECTED] wrote:
Hello osg users/developers,
Are there any plans to do shadows using shaders? I know hardware shadowing
is supported on newer video cards.
Many shadow effects are implemented with shaders, the osgdepthshadow
example is based on GLSL.
On 2/8/07, UnSeal [EMAIL PROTECTED] wrote:
Hi all.
If I have vertex shader with attribute attribute vec3 vTangent;
How can I set this attribute in OSG?
osg::Geometry has various set array methods, including vertex
attributes, but there isn't yet support for naming attributes. The
naming of
Hi,
With osgProducer::Viewer you'd be best toggling the using a Switch.
With osgViewer::Viewer it can handle a setSceneData/getSceneData()
without problems.
Robert.
On 2/8/07, elekis [EMAIL PROTECTED] wrote:
hi all
I have a little probleme,
I have two node who represent two differents
Hi Robert,
I am also doing some debugging on OSG. I am debugging osgdem source code
of which i am able to see in ddd debugger. But as soon as i want to go
the function of osgTerrain dataset through osgdem code i am not able to
do so. I have compiled OSG using make debug option and then i am just
thnaks, thats works, except in my code it's setSingleChildOn(0); (he didn't
recognize without On)
a++
On 2/8/07, Rafa Gaitan [EMAIL PROTECTED] wrote:
Hi elekis
You can use an osg::Switch node on top of your two scenes:
osg::Switch *switchRoot = new osg::Switch()
... your nodes...
On 2/8/07, Robert Osfield [EMAIL PROTECTED] wrote:
Hi,
With osgProducer::Viewer you'd be best toggling the using a Switch.
With osgViewer::Viewer it can handle a setSceneData/getSceneData()
without problems.
what is the difference between osgProducer and osgViewer?? whitch is the
best?
make instlinks
Simple creates symbolic links from the installed directories to the
orignal built libs.
On 2/8/07, RJ [EMAIL PROTECTED] wrote:
Hi Robert,
I am also doing some debugging on OSG. I am debugging osgdem source code
of which i am able to see in ddd debugger. But as soon as i want to
Hi Elikis,
On 2/8/07, elekis [EMAIL PROTECTED] wrote:
what is the difference between osgProducer and osgViewer?? whitch is the
best?
Do a search for discussions in th osg-users archives about osgViewer
over the past two months, there is huge amount of discussion on it.
osgViewer is new
Hello.
Recently we have tried to make a few very simplified houses in our
application.
In order to create the houses we used QUAD_STRIPS for the walls.
Normals are created for the walls and point out of the house.
The problem we see is that when creating the normals, only the first one is
Hi Werner,
BIND_PER_PRIMITIVE will require a normal per quad, even when its in a strip.
BIND_PER_PRIMITIVE_SET will require a norm pwer primitive set.
BIND_PER_PRIMITIVE will force the OSG down on to OpenGL slow paths as
non of the fast paths support this type of usage so I wouldn't
recommend
So splitting each quad in the quad strip down to single primitives/Quads
and creating a normal pr. vertex would be faster even though it creates
more data?
Werner.
___
Werner Lindgård, Programmerer
Telephone +47 67 81 70 00
Direct line +47 67
The most efficient way is to use per vertex normals, even though it
means more data.
On 2/8/07, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
So splitting each quad in the quad strip down to single primitives/Quads
and creating a normal pr. vertex would be faster even though it creates
more
Ok. Thanks. We will try this.
Werner.
___
Werner Lindgård, Programmerer
Telephone +47 67 81 70 00
Direct line +47 67 81 70 68
[EMAIL PROTECTED]
ViaNova Systems AS
Leif Tronstads Plass 4, P.O.Box 434
N-1302 Sandvika, Norway
Hi all,
having an eye on my cpu-usage-monitor I noticed a strange behavior while
rendering a relatively simple OSG-scene with osgUtil::Viewer: when zooming
out a little bit cpu-usage jumps from about 2% to 100%.
There's nothing special that becomes visible so I wonder, whats the reason
and -
Hi Frans,
It doesn't look like anything in your scene will push the cpu usage up
when zooming out, its a tiny scene so should tax the system very
little.
Is vysnc on? I believe this should be on by default for OSX, perhaps
Stephan Huber can clarify.
The other things you could look at is
Hi All,
To help speed the process of locking down the multi-threaded
stability, and hopefully provide some generally useful features, I
have add two items:
To osgviewer I have addd the command line options (extracted from
osgviewer --help)
Options:
--CullDrawThreadPerContext Select
elekis wrote:
hi all
I have a little probleme,
I have two node who represent two differents things,
and I would like to render one, and afer pushing a touch, changing on
other
something like that
//init
osg::Group* root = new osg::Group();
osg::Group* root2 = new osg::Group();
Robert,
That's correct. I have a P4 single core CPU. With hyperthreading on, it
behaves like a dual core or dual CPU machine, and with it off, it's like a
single core.
I've updated to the current CVS versions and I'm currently rebuilding the
debug and release versions. Apparently, there
hm, ok, thank you Robert.. but that leads me to another problem:
When I try to link ANY osg-Application against libosgViewer (and that's
where StatsHandler is located, isn't it?) the program hangs at start. I
can't even produce any output with printf() as the first command in main().
The
I forgot to mention, for me make instlinks didn't show the source in
gdb either. At least for OpenThreads. It was still looking for the
source in ../
-Gazi
On 08/02/07, Robert Osfield [EMAIL PROTECTED] wrote:
make instlinks
Simple creates symbolic links from the installed directories to the
Robert Osfield wrote:
On 2/8/07, UnSeal [EMAIL PROTECTED] wrote:
Hi all.
If I have vertex shader with attribute attribute vec3 vTangent;
How can I set this attribute in OSG?
osg::Geometry has various set array methods, including vertex
attributes, but there isn't yet support for naming
Unfortunately no, I did not disable depth test for this pyramid.
On 2/8/07, Robert Osfield [EMAIL PROTECTED] wrote:
HI Jason,
Did you disable depth test for this pyramid? If so don't this is why
you can see it through the terrain.
Robert.
On 2/7/07, Jason [EMAIL PROTECTED] wrote:
Attached
Hi Frans,
We've seen problems mixing osgViewer and QT. It seems the default
calling to XInitThreads in GraphicsWindowX11.cpp is the where cause QT
to hang. I don't know why QT is sensitive in this way yet though.
Making the XInitThreads call be done in a lazy way might a workaround
i..e. only
On 2/8/07, Jason [EMAIL PROTECTED] wrote:
Unfortunately no, I did not disable depth test for this pyramid.
Is depth write disabled? Are yo drawing the pyramid under a different
projection?
Robert.
___
osg-users mailing list
Hi Robert,
sorry for the delay, but I've been out of the office for a few days.
Robert Osfield wrote:
The second problem and outwardly more serious is that the low res
tiles that come up initially are all white, and just very simple
quads, and as the higher res tiles are paged in holes appear
Hi Stephan,
On 2/8/07, Stephan Kussmaul [EMAIL PROTECTED] wrote:
We tracked down the problem and it is because of the optimization, which
is done when loading the database into the OSG viewer.
We have a MatrixTransform as parent of the whole scene, which moves the
terrain to it's place in
Hello osg-pagers,
Anybody know if it is better to page a bunch of little files, or one large
file?
For example,
What would generally be better?
1) 1000 2MB paged files (contained in a 10x10km grid)
2) 50 40MB paged files (contained in a 10x10km grid)
My guess is 1 would be better, but not sure
Hi Zach,
On 2/8/07, Zach Deedler [EMAIL PROTECTED] wrote:
Anybody know if it is better to page a bunch of little files, or one large
file?
For example,
What would generally be better?
1) 1000 2MB paged files (contained in a 10x10km grid)
2) 50 40MB paged files (contained in a 10x10km grid)
My
Hi Staphan,
On 2/8/07, Robert Osfield [EMAIL PROTECTED] wrote:
This sounds like a bug. Making the MatrixTransform dynamic will be a
workaround, but a real fix will involved the
FlatenStaticTransformsVisitor being away of PagedLOD's and ProxyNode's
and not optimizing subgraphs with them in.
A
Zach,
In the project settings for Core osgViewer, change the C/C++ - General -
'Debug Information Format' setting to 'Program Database for Edit Continue
(/ZI)' (it is set to 'Program Database (/Zi)' by the CVS project file).
This is how VS8 is setup and I believe VS7 is the same, but the exact
Hi John,
You'll need to pass in the elements your 3x3 matrix one by one into
the MatrixTransform's Matrix.
Note, Matrix is typedef'd to Matrixd by default. See the
include/osg/Matrix for its setting.
Robert.
On 2/8/07, Whitfield, John M. (SED/TMI)
[EMAIL PROTECTED] wrote:
Classification:
This happens to me from time to time. Here's another thing to try, when you
startup the debugger, look in the output window look for a line like so:
'osgvislib_driver_console.exe': Loaded
'C:\bsi\3rdparty\osg\OpenSceneGraph\bin\win32\osgTextd.dll', No symbols
loaded.
Notice the part that says
Sherman,
I just noticed that what I had posted earlier only worked once and I could
no longer do the debugging. What you state is what finally fixed it for me.
The problem, as I understand it now, is when the osgviewer.exe (application)
is built, the program database file overwrites the
I need to perform some raw OpenGL operations. So I took the teapot example
and used that. So far so good. Problem is that the drawImplementation
function is const. Since it's const, I can't get a non-const reference to a
my vector container that is a member of this class. Like so:
LabelPoint
Hi Robert,
I just wonder how to debug such tings on the GPU ? Has someone played
with FBO's and VertexTexture this way ?
Which image-extensions (using Nvidia hardware) er best?
using osg from cvs.
after debugging inside of FrameBufferObject.cpp RenderStage.cpp
in FrameBufferObject.cpp:
By the way, if you relicense OpenThreads, I think you need to make
sure every single contributor that has changes in the current code
base agrees to the license change or their changes must be pulled from
the source base. (I remember PhysFS went through this and it was quite
an ordeal for
OK. Here's the solution that I was able to find that allows debugging the
osgViewer.dll and osgviewer.exe application.
For the osgviewer application, set the program database filename to
.\../../../bin/$(PlatformName)/osgviewerapplicationd.pdb (the current
setting is without the 'application'
The mutable keyword?
Alan.
On 8 Feb 2007, at 18:02, sherman wilcox wrote:
I need to perform some raw OpenGL operations. So I took the teapot
example and used that. So far so good. Problem is that the
drawImplementation function is const. Since it's const, I can't get
a non-const
Hi Sherman,
The drawImplementation is const as a hint to the developer that draw
can be multi-threaded so should be a const operation. If its not
const then you either have to multi-buffer, mutex or force single
threaded usage.
Use of mutable variable is the easiest way to get round a const.
Hi Markus,
I'm struggling to follow what bits work and what doesn't. Is it that
_image-getInternalTexureFormat() is not GL_RGBA32F_ARB?
Could you modify the osgprerender example to illustrate what works and
what doesn't. This might give me a chance to following whats going on
Robert.
On
Yes, well, I feel rather silly now. Here I was looking for a much more
exotic solution. Sorry for the post.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
Hi Eric,
On 2/8/07, E. Wing [EMAIL PROTECTED] wrote:
You need to check with all the copyright owners, not all the
contributors. If you don't add a specific note that you are a
copyright owner of a particular piece of code then you can't assert
rights over it. I didn't spot any copyright
I haven't had much experience dealing with these kind of issues, but
maybe this can help some. The license that we (SCEA) use for the COLLADA
DOM has a special clause about providing contributions that eliminates
this potential copyright problem. Something similar might be able to
work for OSG,
Our SVN server is up again at https://www.zerobar.net/csp/
If You have any further questions do not hesitate to ask.
Kind regards
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
Hi All,
One of the challenges with optimizing modern graphics apps is that
there is complex mix of parameters to
paging/viewers/database/threading experiment with and also huge range
of ways to measure performance/quality. Things like frame rates are
often a metric that graphics developers use
Hi Robert,
I have started looking at this threading issue yesterday evening and here
are some observations (warning: this is an on-going effort more
investigation is required):
* It looks like there are some threads being orphaned (but not terminated)
in the NVIDIA driver (calling
I think adding the full word application to the filename is pushing
the bounds of readability. When I worked around this issue locally, I
used _app instead.
And the problem isn't just for osgviewer/osgViewer. It exists for all
applications and examples that share names with libraries:
HI Andre,
On 2/8/07, André Garneau [EMAIL PROTECTED] wrote:
I have started looking at this threading issue yesterday evening and here
are some observations (warning: this is an on-going effort more
investigation is required):
* It looks like there are some threads being orphaned (but not
I think adding the full word application to the filename is pushing
the bounds of readability. When I worked around this issue locally, I
used _app instead.
And the problem isn't just for osgviewer/osgViewer. It exists for all
applications and examples that share names with libraries:
As far as readability, all that was changed was the program database
filename, not the application executable filename. This file is only
needed
for debugging anyway and is not directly used by the debugger (i.e., the
debugger doesn't ever need to open or do anything with the file; the
Hi Robert,
Sounds a bit odd, the osgViewer's threading should be decoupled entirely
from the driver, the only thing it'll know about things is that their are
makeCurrent() and
releaseContext() being called, unless everytime there is makeCurrent()
isn't kicking off a thread.
Are you able to
On Thursday 08 February 2007, E. Wing wrote:
Actually, I don't think that's true, at least in the United States.
I'm not a lawyer, but I recently sat in on a class dealing with
patents and copyrights for engineers (taught by a lawyer), and I think
the mere act of writing automatically assigns
Thanks Chuck,
That didn't work for me, though. Argh. I wanna help debug this threading
problem.
I think VS7 is not resolving namespace between osg::View and osgViewer::View
somewhere.
Zach
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Chuck Cole
Classification: UNCLASSIFIED
Caveats: NONE
Does anyone know when the fedora linux rpms that match up with what was
in CVS on 12/12/2006 or later will be available for download? Or are
they already?
Thanks,
-John
Classification: UNCLASSIFIED
Caveats: NONE
Hello,
On Thursday 08 February 2007, [EMAIL PROTECTED] wrote:
However, if any of Your contributions are included into the Software,
they will become part of the Software and will be
distributed under the terms and conditions of this License. Further, if
Your donated Contributions are
So did your lawyer say what happens in the case of modifications to an
existing copyrighted body of work? If somebody changes a character?
osg submission are often a couple of characters or a couple of lines
of fixed code.
So I think I'm weak on this, but I believe in general, each person
owns
Well I bet you thought I finally gone away but I have been trying to track down
some form of info that might be helpful on this but to follow up here is some
additional info.
Robert Osfield wrote:
I can't really comment much more, threading is heard to debug, next to
impossible when you don't
What you mean is the Berne convention,
http://www.law.cornell.edu/treaties/berne/overview.html
Yeah, that looks like it.
Actually, AFAIK, you cannot surrender your rights to the work. No matter
what
you sign away, you are still the author. However, you may sign away rights
for distribution
Hi Zach,
In case you do not subscribe to the osg-submissions list, just wanted to let
you know that a fix has been submitted for this issue. The issue was caused
by an unowned mutex being released in the code.
André
-Message d'origine-
De : [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] De
Hi Eric,
On 2/8/07, E. Wing [EMAIL PROTECTED] wrote:
Unfortunately, being practical about things isn't our right.
Copyright is to protect the creators of works, not the users (and keep
lawyers employed).
I think you miss my point. If you have a wide and collection of
copyright owners that
Hi John,
The binaries associated with the OSG are typically only updated on
full releases, not CVS. I dont' maintain the rpm's though so there is
chance then are more up to date than 1.2, but chances are they arn't.
Robert.
On 2/8/07, Whitfield, John M. (SED/TMI)
[EMAIL PROTECTED] wrote:
Hi Robert,
Curious.
What happens when makeCurrent runs from the same thread? What happens on
a relaseContext?
In both cases, nothing; it seems to be created only on first attachment to
it (i.e. only for the first wglMakeCurrent() for each thread). Doing
stopThreading() and startThreading()
74 matches
Mail list logo