Re: [osg-users] Visual System Performance reply

2006-09-22 Thread Donald Tidrow

Speaking of the new drivers...

http://www.nzone.com/object/nzone_downloads_linux_display_x86_1.0-9625.html

Jan Ciger wrote:

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hello,

Wasileski, Bryan J. wrote:


I've messed with the configuration for a couple days now and I don't
believe I'm getting an SLI configuration. Anytime I set that option, the
video drops to a single DVI output (single screen)



This is normal with SLI.



and I see no
difference in performance. 



Well, this is also "normal" at the moment, at least with Linux - see the
article:
http://www.phoronix.com/scan.php?page=article&item=551&num=1

Quoting:
"A single NVIDIA GPU performs approximately the same under the Windows
and Linux drivers, however, once SLI is enabled the Linux drivers can
barely provide any advantages."



In fact, the only time I've been able to get
the second board to produce an output is if I configure for two
x-screens, then I only get 1 DVI output per board. In this configuration
I have to run separate processes, one for each GPU, so I'm not really
taking advantage of having the two GPU's.



Yes, normal - see the documentation for the nVidia's drivers - the
second board is "dead" when using SLI (not producing any output). This
works exactly the same in Windows. You *cannot* use SLI configuration
with multiple monitors, it is intended to accelerate rendering on one
output device (i.e. gamers are the target).



Do you know of any system
check to see if SLI is active or not? 
Right now, the only thing setting

the SLI configuration does for me that I can tell is kick off the dual
screen. Then I know that SLI has been set - what clue.



This is normal behavior and is correct. You can check whether SLI is
working by running 'nvidia-settings', going to 'OpenGL Settings' and
turning on 'Enable SLI Heads-Up-Display'. Then you should see two green
bars displaying the load balancing between the two graphic cards on the
screen whenever you run an OpenGL application. This feature may depend
on the SLI mode (AFR/SFR/SFRAA) selected, but if you see the bars, SLI
is working.

Regards,

Jan

- --

Jan Ciger
GPG public key: http://www.keyserver.net/

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Mandriva - http://enigmail.mozdev.org

iD8DBQFFFEjIn11XseNj94gRAsuzAJ4kMO02uDyjC3TfmQLVwr1KnCq6vQCg5ozC
MR50K4eAwdoCrWoqBRDhtfQ=
=46Gv
-END PGP SIGNATURE-
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/



--
"Democracy is two wolves and a lamb voting on what to have for lunch.
Liberty is a well-armed lamb contesting the vote."
- Benjamin Franklin

++
|  Don Tidrow|
|  Visualization Software Engineer   |
|  SAIC  |
|  ph: 703-253-1094  |
++
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Interactive documentation

2006-09-22 Thread Watson Ladd
I don't think so. If you need documentation a *cough* documentation
*cough* file should be enough. Just ship it with the application.
Robert Osfield wrote:
> A thought occured to me the other while I was waiting for a rebuild of
> the osgIntrospection wrappers, that perhaps we could have an option in
> osgIntrospection for having a documentation string for each namespace,
> class and member function/variable.  This documentation string would be
> one of the doxygen strings that are embedded in header files.   You'd
> want to make this optional, as it'd be a bit more bloat to wrappers, but
> it should be too hard to add since genwrapper is based around doxygen
> anyway.
> 
> So once you have a documentation string, and a suitable
> scripting/interpreted language like Lua or Phython you can have a script
> console window in your app that you could do something like:
> 
>help(node)
> 
> Or
> 
>help(node:getChild(i))
> 
> This was my thought.  Would it be useful?   That bit I can only
> speculate on.
> 
> Thoughts?
> 
> Robert.
> 
> 
> 
> 
> ___
> osg-users mailing list
> osg-users@openscenegraph.net
> http://openscenegraph.net/mailman/listinfo/osg-users
> http://www.openscenegraph.org/


-- 
They who would give up an essential liberty for temporary security,
 deserve neither liberty or security
--Benjamin Franklin



signature.asc
Description: OpenPGP digital signature
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Parallel-Split Shadow Map with osg [PSSM]

2006-09-22 Thread Terry Welsh

Hi Radu,
I tried to keep my shadow mapping code out of the scene setup business
as much as possible and have it concentrate on just rendering shadow
maps.  The only scene setup it does is add CameraNodes between a
user-specified root and scene.  The CameraNodes override scene shaders
with their own depth shaders to render shadow maps quickly.  The user
is still free to specify their scene as a subgraph of their whole
scene for the cases like you mention.  And the user is free to render
the final scene any way they want using the resulting shadow maps any
way they want.

There are so many ways to set up a scene that are application-specific
that I think it would be dangerous to let a shadow library do very
much of it.  It might make the library unusable for a lot of
applications.
- Terry



Message: 7
Date: Fri, 22 Sep 2006 17:38:57 -0400
From: Radu Mihai <[EMAIL PROTECTED]>
Subject: Re: [osg-users] Parallel-Split Shadow Map with osg [PSSM]
To: osg users 
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset="utf-8"

I haven't had time to move the files around the systems so it took me
some time to get back to posting the refactored osgdepthshadow:



Robert has mentioned that he's working on a shadow library, and I
think that, for me, now it's the time to have an interface to go
towards.
If there is something that can be used I would be ecstatic

I have to integrate this new depthShadows code in the company's app,
we already have this implemented, but it's based on pbuffer cameras
form OSG 0.99, and some of the code in there is plain ugly ...
The kind of issues I have to deal with now relate more to the
interface of this with the rest of the viewer:
- are there objects excluded from casting shadows ? like terrain,
that receives, but doesn't cast
this could be implemented straight-forward with a NodeMask on 
the
occluders Group
- objects that cast, but don't receive shadows ?
here changing shaders is needed
- is the shadow library going to be managing the various render-
stages of the app ?
or (link in the code above) it keeps track of the essential 
pieces,
and makes changes to subgraphs that are passed to it
The last point will be more important to PSSM, as the shadowLib has
to manage the splitting of the view fulstrum ( have to look @ Terry's
demo to see if some of these have been already dealt with)

--
Radu Mihai
[EMAIL PROTECTED]
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: osgPlugins not found on OS X

2006-09-22 Thread Radu Mihai


On 22-Sep-06, at 5:20 PM, E. Wing wrote:

So are you saying the prebuilt binaries work for you, but when you
compile yourself, it does not? Compiling it yourself shouldn't really
matter. Are you using the Xcode projects or Makefiles? This shouldn't
really matter either for plugins.



It seems to be an issue of the OpenFlight plugin not appearing in the  
Xcode project for the sources posted on the Download section, not so  
much a location problem for the binary.


I'm testing the cvs version and I'll post the results.
--
Radu Mihai
[EMAIL PROTECTED]


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


[osg-users] Re: osgPlugins not found on OS X

2006-09-22 Thread E. Wing

Okay, this is starting to worry me. This is all supposed to work out of the box.


I've noticed a similar problem. The /Library/... path is not set by
default.



I have solved this problem by putting the following near the top of
main():



#ifdef __APPLE__
   osgDB::FilePathListpluginPaths;
   pluginPaths.push_back( "/Library/Application Support/OpenSceneGraph/PlugIns" 
);
  osgDB::Registry::instance()->setLibraryFilePathList( pluginPaths );
#endif


No, this should all be set up already without you having to do
anything. There's a ton of code we added to osgDB/FileUtils.cpp that
does all this already. This code was written several years ago and I
don't think it has been touched since. The code is actually a bit
smarter because it queries the system for the Application Support
directory. I learned that the Application Support directory can be
changed by users.



I put the plugins in:
/Library/Frameworks/PlugIns or in ~/Library/Frameworks/PlugIns
The frameworks are in the same place as you have them.



I forgot why this one works. Officially we tell people to use:
/Library/Frameworks/Application Support/OpenSceneGraph/PlugIns (or ~/...)
We put in some redundancy early on which I think why
/Library/Frameworks/PlugIns works, but these paths are not the ones we
advertise.


Works fine for me, without setting LD_LIBRARY_PATH (I've been using
the app.bundles from the Xcode projects)



Btw . LD_LIBRARY_PATH is hardly used on OS X . you should use
DYLD_LIBRARY_PATH . then it would work just like you expect . but copying is
the cleaner solution I think J


You shouldn't have to set (DY)LD_LIBRARY_PATH or any other
environmental variables (sans one for the Intel/ATI bug). The whole
motivation for our FileUtils.cpp patch was so we could leave
environmental variables behind.


ps. on the intel mac there's a bug in the ati driver that creates
crashes when rendering font files loaded from the disk.



Yes, I documented this fairly extensively in the OSX_OSG_README.rtf file.



With the framework version I placed the Frameworks in "/Library/Frameworks"
and PlugIns in "/Library/Application Support/OpenSceneGraph/PlugIns".  This
failed to find the plugins the same way as when I compiled myself.


So are you saying the prebuilt binaries work for you, but when you
compile yourself, it does not? Compiling it yourself shouldn't really
matter. Are you using the Xcode projects or Makefiles? This shouldn't
really matter either for plugins. The difference is the latter builds
.dylibs instead of frameworks, but this doesn't affect plugin loading
which is controlled by osgDB. (Just don't mixup the two plugins,
because they link to different libraries (dylibs or frameworks)).

Are you using the standard Apple gcc toolset? The code base relies on
__APPLE__ being defined by the compiler. If this is not being defined
by your compiler, then building OSG is going to miss out on pretty
much all the Apple specific code paths.

-Eric
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


[osg-users] gdb cannot access memory on 64 bit linux

2006-09-22 Thread Mark Acosta



Hi guys,
 
    I just put together a brand 
spanking new linux box running 64 bit fedora core 5 and much to my suprise, I 
found that gdb was unable to access any memory while debugging an osg program. 
After much head banging, I tracked the problem down to the -gstabs+ compile 
option. I switched to -ggdb in osg, Producer and OpenThreads and in my own code 
and now gdb works fine. I was running 32 bit fedora core 5 before and there 
wasn't any problem with gdb there. I'm not sure if this is a 64 bit gdb or 
gcc/g++ bug or what. Anybody else ever run into this problem or even not have 
this problem? By the way, I'm using gcc 4.1.1 and gdb 6.3.
 
    Also, I noticed that when doing 
a "make instlinks" , osg will install in /usr/local/lib64, which I assume is 
correct, but Producer and OpenThreads will always install in /usr/local/lib even 
if I specify ARCH=64. I took a look at the make files but I don't see anything 
obviously wrong. 
 
Thanks,
 
Mark Acosta
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Visual System Performance reply

2006-09-22 Thread Jan Ciger
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hello,

Wasileski, Bryan J. wrote:
> I've messed with the configuration for a couple days now and I don't
> believe I'm getting an SLI configuration. Anytime I set that option, the
> video drops to a single DVI output (single screen)

This is normal with SLI.

> and I see no
> difference in performance. 

Well, this is also "normal" at the moment, at least with Linux - see the
article:
http://www.phoronix.com/scan.php?page=article&item=551&num=1

Quoting:
"A single NVIDIA GPU performs approximately the same under the Windows
and Linux drivers, however, once SLI is enabled the Linux drivers can
barely provide any advantages."

> In fact, the only time I've been able to get
> the second board to produce an output is if I configure for two
> x-screens, then I only get 1 DVI output per board. In this configuration
> I have to run separate processes, one for each GPU, so I'm not really
> taking advantage of having the two GPU's.

Yes, normal - see the documentation for the nVidia's drivers - the
second board is "dead" when using SLI (not producing any output). This
works exactly the same in Windows. You *cannot* use SLI configuration
with multiple monitors, it is intended to accelerate rendering on one
output device (i.e. gamers are the target).

> Do you know of any system
> check to see if SLI is active or not? 
> Right now, the only thing setting
> the SLI configuration does for me that I can tell is kick off the dual
> screen. Then I know that SLI has been set - what clue.

This is normal behavior and is correct. You can check whether SLI is
working by running 'nvidia-settings', going to 'OpenGL Settings' and
turning on 'Enable SLI Heads-Up-Display'. Then you should see two green
bars displaying the load balancing between the two graphic cards on the
screen whenever you run an OpenGL application. This feature may depend
on the SLI mode (AFR/SFR/SFRAA) selected, but if you see the bars, SLI
is working.

Regards,

Jan

- --

Jan Ciger
GPG public key: http://www.keyserver.net/

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Mandriva - http://enigmail.mozdev.org

iD8DBQFFFEjIn11XseNj94gRAsuzAJ4kMO02uDyjC3TfmQLVwr1KnCq6vQCg5ozC
MR50K4eAwdoCrWoqBRDhtfQ=
=46Gv
-END PGP SIGNATURE-
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


[osg-users] Visual System Performance reply

2006-09-22 Thread Wasileski, Bryan J.
 
Hi everyone.

Thanks for the responses. 

I am running the latest drivers for the NVIDIA cards. That was the first
priority once the machines arrived.
I did run the osgparametric suggested by Bob and I do get the 1M
triangle at 60Hz (fill).
Wire frame and lines drop to 30 - which to me is bewildering.

I've messed with the configuration for a couple days now and I don't
believe I'm getting an SLI configuration. Anytime I set that option, the
video drops to a single DVI output (single screen) and I see no
difference in performance. In fact, the only time I've been able to get
the second board to produce an output is if I configure for two
x-screens, then I only get 1 DVI output per board. In this configuration
I have to run separate processes, one for each GPU, so I'm not really
taking advantage of having the two GPU's.

Radu, I'm running the 64bit EMT workstation version with RedHat -
version 4. Those drivers are clearly labled on the NVIDIA site so I feel
pretty good that I have the correct drivers. Do you know of any system
check to see if SLI is active or not? Right now, the only thing setting
the SLI configuration does for me that I can tell is kick off the dual
screen. Then I know that SLI has been set - what clue.

Thanks again.

Bryan  


--

Message: 4
Date: Fri, 22 Sep 2006 13:34:16 -0400
From: Radu Mihai <[EMAIL PROTECTED]>
Subject: Fwd: [osg-users] Visual System Performance
To: osg users 
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset="us-ascii"


Also check the state of SLI in Linux 8x Nvidia drivers.

People are reporting that there a little performance gains from SLI  
in the current driver. That had been the case in Win-land too at the  
beginning, and the 'word' is that the 9x version will fix this issue  
for Linux.
--
Radu Mihai
[EMAIL PROTECTED]

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Paul Speed



Robert Osfield wrote:

On 9/22/06, [EMAIL PROTECTED] * 
<[EMAIL PROTECTED] > wrote:


Those haggis have been at the calendar in Callander again! This time
they've managed to squeeze three days into one. We'll have to start
eating
them to keep their numbers down. :)


:-)

The "correct" official dates for the highland gathering are:

   Wednesday 18th October
   Thursday 19th October
   Friday 20th October

Happy typo'ing!
Robert.



Phew... for some reason Henry Ford's quote "You can have any color you 
want as long as it's black." was running through my head.


-Paul
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] OSG -- near future

2006-09-22 Thread Sullivan, Joseph (CDR)
Sounds great. We'll look at pushing more material - class notes and such - that 
we use with the tutorials onto the web.  Hopefully others will be able to 
expand on posted material.


From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Robert Osfield
Sent: Thursday, September 21, 2006 6:26 AM
To: osg users
Subject: Re: [osg-users] OSG -- near future

On 9/21/06, Sullivan, Joseph (CDR) <[EMAIL PROTECTED]> wrote:
Speaking of introductory guides... Are any of the other educational institutes 
that are making use of the OSG able to collaborate on and publish introductory 
material?
Don and Robert,
Would making this introductory material available interfere with training that 
you provide? 

I have no problem with introductory material being provided and certainly 
support ways in which we can make this happen.

I personally wouldn't expect having better documentation will qwell peoples 
need for training too much.  Those who can afford training are the ones can 
afford to invest in speeding their employees learning curve, and to top up 
their skills, face to face training is still great for accelerating learning. 

There is the other side, the better introductory documentation we have the more 
likely the OpenSceneGraph will attract new groups to use the software, which in 
turn will mean more potential customers for training.  

Good documentation will also provide a more professional shine to the project, 
good code might impress developers, but pretty bound manuals and books may 
impress the decision makers more...

Robert.



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] Re: Highland gathering

2006-09-22 Thread Sullivan, Joseph (CDR)
Now we *seriously* re-think budgets!

-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Mike Weiblen
Sent: Friday, September 22, 2006 9:31 AM
To: osg users
Subject: Re: [osg-users] Re: Highland gathering

Um, this is a no-brainer, right?
-- mew
:-)


Robert Osfield wrote:
> Anybody a monty
> python fans?  The castle from "The Holy Grail" is just a few miles
down
> the road in Doune. 
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] OSG -- near future

2006-09-22 Thread Sullivan, Joseph (CDR)
Eric,
'Working with the art path' sounds of definite interest to us - I'd have
a hard time imagining it would be of interest to others. Just curious,
have you ever used Moodle for course development?  I have no experience
with it, but it seems like it might be good way to collaborate remotely
on course development. Is it me or does wiki seem to be the wrong tool
for this?  Anyone with any experience, other recommendations?
Thanks 
Joe

-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Eric
Maslowski
Sent: Thursday, September 21, 2006 6:46 AM
To: 'osg users'
Subject: RE: [osg-users] OSG -- near future

I would be happy to contribute to any guides that are written.
Currently, my
strongest points in OSG are: working with the art path, animation paths,
developing for multi-screen displays, and integration with VRJuggler. If
any
of this sounds appealing, just let me know and we can work out the
specifics.

E.

---
Eric Maslowski
Research Computer Specialist
University of Michigan 3D Lab

Autodesk 3D Studio Max Certified Trainer

email: [EMAIL PROTECTED]
office: 734-615-9699

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Sullivan, Joseph (CDR)
> Sent: Thursday, September 21, 2006 09:14
> To: osg users
> Subject: RE: [osg-users] OSG -- near future
> 
> Speaking of introductory guides... Are any of the other educational
> institutes that are making use of the OSG able to collaborate on and
> publish introductory material?
> Don and Robert,
> Would making this introductory material available interfere with
training
> that you provide?
> Joe S.
> 
> 
> 
> From: [EMAIL PROTECTED] on behalf of Robert Osfield
> Sent: Wed 9/20/2006 7:23 AM
> To: osg users
> Subject: Re: [osg-users] OSG -- near future
> 
> 
> So I've fleshed out a little why expanding the OSG users base is good
for
> the OSG project, its developers, community, the software quallty and
the
> wider industry, but I didn't touch on just how to go about it.
> 
> Well fixing things in the OSG that turn people off the project is one
> thing we need to address.  The big fat hairy one is obviously
> documentation.  Not the lack of information, but the lack of
introductory
> and advanced programming guides and in depth reference documentation.
The
> book thread address this so I'll not go into this any further here.
> 
> Next thing to fix is how easy it is to get the OSG compiled, installed
and
> running.  The sheer size of the project, and the numerous build
options
> make it very powerful, but it does raise the barrier to entry.  Could
> Cmake streamline it?  Scripts for building?
> 
> Next up is API size, its sheer breadth and depth of functionality is a
lot
> to tackle for the newbee.  OpenGL will eventaully become more
streamlined,
> so perhaps too can the OSG.  This is part of what I discuss in the
> Siggraph 2006 talk.
> 
> Next up is the relatively sophisticated use of C++, this is a hurdle
for
> some in terms of learning curve, partly because of the complexities of
> C++, and partly down to the fact that C++ no longer has developers
sole
> attention.
> 
> Then there are other languages Jave, Python, Lua, C# and all vying for
> attention of developers, we don't yet support these well.  If we want
to
> help these developers out then we need to embrace support for them
much
> more emphatically.  The C++ core here is a bit of hinderance when it
comes
> to intergration with other languages, with C being the comon
denominator.
> osgIntrospection can come to the resuce here, its a bit heavy weight,
but
> dynamically typed langauges its a good match.
> 
> Further to the integration story is that awkwardness of getting
different
> types of viewers setup in the multitude of windowing libraries that
exist
> today.  Making it easier to build 3d viewers into your existing or new
> apps is something that I feel strongly about.  This what the osgViewer
> library I've previously talked about is really about.
> 
> So to recamp things to fix and get done:
> 
>1) Documentation and book(s)
>2) Better viewer classes supporting multiple 3rd party windowing
API's
>3) Better 3rd party language support
>4) Streamlining of builds
>5) Streamlining of API
> 
> No doubt there are other items that need to be added to the list, and
> priorities to be, so fire away.  But remember that someone has to get
on
> and do it, so if you add an item be prepared to roll your sleeves up
and
> help get it done.
> 
> And on that note, please step forward those who'd like to pitch into
> various aspect of this work.
> 
> I'll create a few new pages on the wiki to help documentation the
tasks
> and who's available to make progress on the various items.
> 
> Robert.
> 


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] osgPlugins not found on OS X

2006-09-22 Thread Frank Bergmann








Btw … LD_LIBRARY_PATH is hardly used on OS X … you
should use DYLD_LIBRARY_PATH … then it would work just like you expect …
but copying is the cleaner solution I think J 

 

Frank

 





From: Radu Mihai [mailto:[EMAIL PROTECTED] 
Sent: Friday, September 22, 2006 10:35 AM
To: osg users
Subject: Fwd: [osg-users] osgPlugins not found on OS X





 





 



I put the plugins in:



/Library/Frameworks/PlugIns or in ~/Library/Frameworks/PlugIns





The frameworks are in the same
place as you have them.





 





Works fine for me, without setting
LD_LIBRARY_PATH (I've been using the app.bundles from the Xcode projects)





 





Tested on a MacBook Pro.





 





ps. on the intel mac there's a bug
in the ati driver that creates crashes when rendering font files
loaded from the disk.





 





--





Radu Mihai





[EMAIL PROTECTED]





 









 











On 22-Sep-06, at 12:28 PM, Steven R
Woolsey wrote:








I am working on
a program using osg and have it compiling under Windows, Linux, and OS X.
 It works great under windows.  I just started working on the OS X
version.  I get the program compiled with osg, but when I run the program,
it cannot find any of the plugins and exits.  I need to load an obj file,
a 3ds file, and a number of textures.  It always tells me it cannot find
the plugin.  I have tried the framework version of the binaries of osg
available on the site, as well as compiling the osg from scratch using gcc.


With the
framework version I placed the Frameworks in "/Library/Frameworks"
and PlugIns in "/Library/Application Support/OpenSceneGraph/PlugIns".
 This failed to find the plugins the same way as when I compiled myself.


I am also
unable to get the osg example programs to find any plugins.   

I am using osg
1.2, and this would be from a terminal window.  I have set my
LD_LIBRARY_PATH to "/usr/local/lib:/usr/local/lib/osgPlugins".
 My path is set to "/usr/local/share/OpenSceneGraph/bin".


I have tried
this on both a G5 and an Intel based Powerbook with the same results.  I
must be missing something.  I am not that familiar with OS X, so forgive
my ignorance if this is an easy fix. 

Any suggestions
would be greatly appreciated. 

Thanks, 

Steve Woolsey















___





osg-users mailing list





osg-users@openscenegraph.net





http://openscenegraph.net/mailman/listinfo/osg-users





http://www.openscenegraph.org/





 







 






___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] quick question about osgWrapper

2006-09-22 Thread Yanling Liu
Hi, just wondering what's osgWrapper for? Do I have to complie it when I complie OSG? Yanling
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Fwd: [osg-users] osgPlugins not found on OS X

2006-09-22 Thread Radu Mihai
 I put the plugins in:/Library/Frameworks/PlugIns or in ~/Library/Frameworks/PlugInsThe frameworks are in the same place as you have them.Works fine for me, without setting LD_LIBRARY_PATH (I've been using the app.bundles from the Xcode projects)Tested on a MacBook Pro.ps. on the intel mac there's a bug in the ati driver that creates crashes when rendering font files loaded from the disk.--Radu Mihai[EMAIL PROTECTED] On 22-Sep-06, at 12:28 PM, Steven R Woolsey wrote:I am working on a program using osg and have it compiling under Windows, Linux, and OS X.  It works great under windows.  I just started working on the OS X version.  I get the program compiled with osg, but when I run the program, it cannot find any of the plugins and exits.  I need to load an obj file, a 3ds file, and a number of textures.  It always tells me it cannot find the plugin.  I have tried the framework version of the binaries of osg available on the site, as well as compiling the osg from scratch using gcc.  With the framework version I placed the Frameworks in "/Library/Frameworks" and PlugIns in "/Library/Application Support/OpenSceneGraph/PlugIns".  This failed to find the plugins the same way as when I compiled myself.  I am also unable to get the osg example programs to find any plugins.    I am using osg 1.2, and this would be from a terminal window.  I have set my LD_LIBRARY_PATH to "/usr/local/lib:/usr/local/lib/osgPlugins".  My path is set to "/usr/local/share/OpenSceneGraph/bin".  I have tried this on both a G5 and an Intel based Powerbook with the same results.  I must be missing something.  I am not that familiar with OS X, so forgive my ignorance if this is an easy fix.  Any suggestions would be greatly appreciated.  Thanks,  Steve Woolsey ___osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/ ___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Fwd: [osg-users] Visual System Performance

2006-09-22 Thread Radu Mihai
 Also check the state of SLI in Linux 8x Nvidia drivers.People are reporting that there a little performance gains from SLI in the current driver. That had been the case in Win-land too at the beginning, and the 'word' is that the 9x version will fix this issue for Linux. --Radu Mihai[EMAIL PROTECTED] On 22-Sep-06, at 12:36 PM, Gordon Tomlinson wrote: Are you running in SLI mode ? if not they run independent and you will be only using one if on a single monitor Do you have Antialiasing, Anisotropic filtering on etc   That's 225m optimized non lilt non shaded non non textured etc.   You certainly should be able to do much better the 5hz   Also make sure you get the lasted drivers, drivers installed and supplied with card are normal quite OLD    Best Regards  Gordon__Gordon TomlinsonEmail : gordon.tomlinson @ overwatch.comYIM/AIM   : Gordon3dBritMSN IM    : Gordon3dBrit @ 3dscenegraph.com__Telephone (Cell): (+1) 214-477-8914Telephone (Work): (+1) 703-437-7651"Self defence is not a function of learning tricks but is a function of how quickly and intensely one can arouse one's instinct for survival" - Master Tambo Tetsura     -Original Message-From:   [EMAIL PROTECTED]   [mailto:[EMAIL PROTECTED]]On Behalf Of Wasileski,   Bryan J.Sent: Friday, September 22, 2006 12:29 PMTo:   osg-users@openscenegraph.netSubject: [osg-users] Visual System   Performance  Hi   all,     I recently   received a couple of new workstations which have the NVIDIA 4500 X2 video   cards installed (Dell machines). We intend to run all our new visuals on the machines   using OSG. I've been trying to gauge the graphics performance   and from   what I've seen so far I'm thoroughly disappointed. I'm assuming at this point   the boards are not properly configured. The issue I'm having is trying to find a   way to test the boards to determine if they are in fact improperly configured or just   don't measure up.     As an example, I   have a scene database that I load into OSG; at any given moment, the scene has   about 40K primitives. I'm only getting frame rates around 5-10Hz - which is   totally unacceptable. According to the specs, these GPU's are capable of 225M   triangles/sec. I feel certain that only 1 of the 2GPU's is being exercised - I   know of no way to verify it.     With all the   graphics/visual system wizards who monitor this sight I thought someone   out there might be able to provide some advice on getting these new boards   properly configured so I can actually test their   performance.     Additional info on   the new boxes is:  2 Intel Xenon   3.7GHz - Linux OS EMT64  16GB   memory     Also, kudos for   the flawless 64-bit OSG build. Worked without a single   glitch.        Bryan   ___osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/ ___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Chris Hanson

Robert Osfield wrote:
The haggis, lots of the little rascals around here.  It'll soon be the 
end of the fishing season soon so the little critters better watch out, 
they will be fair game once the "Highland Gathering" comes up.


  I swear, every time you say "Highland Gathering" all I can picture is Christopher 
Lambert waving a sword around. Stop it! ;)



Robert.


--
Chris 'Xenon' Hanson aka Eric Hammil | http://www.3DNature.com/ eric at logrus
 "I set the wheels in motion, turn up all the machines, activate the programs,
  and run behind the scenes. I set the clouds in motion, turn up light and 
sound,
  activate the window, and watch the world go 'round." -Prime Mover, Rush.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] consistent osg file format

2006-09-22 Thread Chris Hanson

Robert Osfield wrote:

On 9/22/06, *c sklu* <[EMAIL PROTECTED] > wrote:
I would expect the results to be the same on all systems as all the core 
OSG and plugins are just C++, and should be produce the pretty well the 
same results across, I can't think of any reason for any major differences.


  Could STL functionality on different platforms/compilers affect container storage and 
iteration order?



Robert.


--
Chris 'Xenon' Hanson aka Eric Hammil | http://www.3DNature.com/ eric at logrus
 "I set the wheels in motion, turn up all the machines, activate the programs,
  and run behind the scenes. I set the clouds in motion, turn up light and 
sound,
  activate the window, and watch the world go 'round." -Prime Mover, Rush.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] osgPlugins not found on OS X

2006-09-22 Thread Steven R Woolsey

Ahh..  that is a great solution.
 I like the idea of adding a path inside the application packaging
also.  That would probably be a good idea for my app with all the
platforms we am supporting.

Thanks for the information and quick
response.

Steve






Ross Anderson <[EMAIL PROTECTED]>

Sent by: [EMAIL PROTECTED]
09/22/2006 10:46 AM



Please respond to
osg users 





To
osg users 


cc



Subject
Re: [osg-users] osgPlugins not found
on OS X








I've noticed a similar problem. The /Library/... path
is not set by default.

I have solved this problem by putting the following near
the top of main():

#ifdef __APPLE__
    osgDB::FilePathList    pluginPaths;
    pluginPaths.push_back( "/Library/Application
Support/OpenSceneGraph/PlugIns" );
osgDB::Registry::instance()->setLibraryFilePathList(
pluginPaths );
#endif

I'm using a file list because I also push another path
to inside my application package (not shown here), so that I can have a
copy of the OSG Frameworks and Plug Ins bundled internally (allowing drag-and-drop
installs of the software).

---
Ross Anderson
[EMAIL PROTECTED]
Group 106 - Active Optical Systems
MIT Lincoln Laboratory
(781) 981-3344


On Sep 22, 2006, at 12:28 PM, Steven R Woolsey wrote:


I am working on a program using osg and have it compiling under Windows,
Linux, and OS X.  It works great under windows.  I just started
working on the OS X version.  I get the program compiled with osg,
but when I run the program, it cannot find any of the plugins and exits.
 I need to load an obj file, a 3ds file, and a number of textures.
 It always tells me it cannot find the plugin.  I have tried
the framework version of the binaries of osg available on the site, as
well as compiling the osg from scratch using gcc. 

With the framework version I placed the Frameworks in "/Library/Frameworks"
and PlugIns in "/Library/Application Support/OpenSceneGraph/PlugIns".
 This failed to find the plugins the same way as when I compiled myself.


I am also unable to get the osg example programs to find any plugins.  


I am using osg 1.2, and this would be from a terminal window.  I have
set my LD_LIBRARY_PATH to "/usr/local/lib:/usr/local/lib/osgPlugins".
 My path is set to "/usr/local/share/OpenSceneGraph/bin".


I have tried this on both a G5 and an Intel based Powerbook with the same
results.  I must be missing something.  I am not that familiar
with OS X, so forgive my ignorance if this is an easy fix.


Any suggestions would be greatly appreciated. 

Thanks, 

Steve Woolsey


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] osgPlugins not found on OS X

2006-09-22 Thread Ross Anderson
I've noticed a similar problem. The /Library/... path is not set by default.I have solved this problem by putting the following near the top of main():#ifdef __APPLE__    osgDB::FilePathList    pluginPaths;    pluginPaths.push_back( "/Library/Application Support/OpenSceneGraph/PlugIns" );	osgDB::Registry::instance()->setLibraryFilePathList( pluginPaths );#endif I'm using a file list because I also push another path to inside my application package (not shown here), so that I can have a copy of the OSG Frameworks and Plug Ins bundled internally (allowing drag-and-drop installs of the software). --- Ross Anderson [EMAIL PROTECTED] Group 106 - Active Optical Systems MIT Lincoln Laboratory (781) 981-3344  On Sep 22, 2006, at 12:28 PM, Steven R Woolsey wrote:I am working on a program using osg and have it compiling under Windows, Linux, and OS X.  It works great under windows.  I just started working on the OS X version.  I get the program compiled with osg, but when I run the program, it cannot find any of the plugins and exits.  I need to load an obj file, a 3ds file, and a number of textures.  It always tells me it cannot find the plugin.  I have tried the framework version of the binaries of osg available on the site, as well as compiling the osg from scratch using gcc.  With the framework version I placed the Frameworks in "/Library/Frameworks" and PlugIns in "/Library/Application Support/OpenSceneGraph/PlugIns".  This failed to find the plugins the same way as when I compiled myself.  I am also unable to get the osg example programs to find any plugins.    I am using osg 1.2, and this would be from a terminal window.  I have set my LD_LIBRARY_PATH to "/usr/local/lib:/usr/local/lib/osgPlugins".  My path is set to "/usr/local/share/OpenSceneGraph/bin".  I have tried this on both a G5 and an Intel based Powerbook with the same results.  I must be missing something.  I am not that familiar with OS X, so forgive my ignorance if this is an easy fix.  Any suggestions would be greatly appreciated.  Thanks,  Steve Woolsey ___osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/ ___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Visual System Performance

2006-09-22 Thread Robert Osfield
Hi Bryan,It sounds like a driver issue of some kind, you should get significantly better performance than that.  Try benchmarking with/without SLI.As far as the specs go, this is purely theortically, you should be able to kick out 50-100 million triangles/sec with a real model.  Try the OSG examples like osgparametric, I pushes through 1 million triangles as a single mesh.  I get 200Hz with vsync off on single 7800GT to two 7800GT, which put me in at a real throughput of 200 million triangles a second per card, 400 million a second for both.  This is running multipipe though, not SLI.
Robert.On 9/22/06, Wasileski, Bryan J. <[EMAIL PROTECTED]> wrote:





Hi 
all,
 
I recently received 
a couple of new workstations which have the NVIDIA 4500 X2 video cards installed 
(Dell machines). We intend to run all our new visuals on the machines 
using OSG. I've been trying to gauge the graphics performance 
and from 
what I've seen so far I'm thoroughly disappointed. I'm assuming at this point 
the boards are not properly configured. The issue I'm having is trying to find a 
way to test the boards to determine if they are in fact improperly configured or just 
don't measure up.
 
As an example, I 
have a scene database that I load into OSG; at any given moment, the scene has 
about 40K primitives. I'm only getting frame rates around 5-10Hz - which is 
totally unacceptable. According to the specs, these GPU's are capable of 225M 
triangles/sec. I feel certain that only 1 of the 2GPU's is being exercised -  I 
know of no way to verify it.
 
With all the 
graphics/visual system wizards who monitor this sight I thought someone out 
there might be able to provide some advice on getting these new boards properly 
configured so I can actually test their performance.
 
Additional info on 
the new boxes is:
2 Intel Xenon 3.7GHz 
- Linux OS EMT64
16GB 
memory
 
Also, kudos for the 
flawless 64-bit OSG build. Worked without a single glitch.
 
 
Bryan 


___osg-users mailing listosg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Re: Highland gathering

2006-09-22 Thread Gordon Tomlinson
If there is any one on the list that is NOT a Python fan they should be
ashamed and leave for ever 


Only Joking 


Best Regards



Gordon

__

Gordon Tomlinson
Email : gordon.tomlinson @ overwatch.com
YIM/AIM   : Gordon3dBrit
MSN IM:  Gordon3dBrit @
3dscenegraph.com

__

Telephone (Cell): (+1) 214-477-8914
Telephone (Work): (+1) 703-437-7651
"Self defence is not a function of learning tricks
but is a function of how quickly and intensely one
can arouse one's instinct for survival"
- Master Tambo Tetsura




-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of Mike Weiblen
Sent: Friday, September 22, 2006 12:31 PM
To: osg users
Subject: Re: [osg-users] Re: Highland gathering


Um, this is a no-brainer, right?
-- mew
:-)


Robert Osfield wrote:
> Anybody a monty
> python fans?  The castle from "The Holy Grail" is just a few miles down
> the road in Doune.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] Visual System Performance

2006-09-22 Thread Gordon Tomlinson



Are 
you running in SLI mode ? if not they run independent and you will be only using 
one if on a single monitor
Do you 
have Antialiasing, Anisotropic filtering on etc
 
That's 
225m optimized non lilt non shaded non non textured etc.
 
You 
certainly should be able to do much better the 5hz
 
Also 
make sure you get the lasted drivers, drivers installed and supplied with card 
are normal quite OLD
 


Best Regards 
 
Gordon
__Gordon 
TomlinsonEmail 
: 
gordon.tomlinson @ overwatch.comYIM/AIM   
: Gordon3dBritMSN 
IM    
: Gordon3dBrit @ 
3dscenegraph.com__Telephone 
(Cell): 
(+1) 214-477-8914Telephone (Work): 
(+1) 703-437-7651"Self defence is not a 
function of learning tricks but is a function of how 
quickly and intensely one can arouse one's instinct 
for survival" - Master Tambo Tetsura
 

  -Original Message-From: 
  [EMAIL PROTECTED] 
  [mailto:[EMAIL PROTECTED]On Behalf Of Wasileski, 
  Bryan J.Sent: Friday, September 22, 2006 12:29 PMTo: 
  osg-users@openscenegraph.netSubject: [osg-users] Visual System 
  Performance
  Hi 
  all,
   
  I recently 
  received a couple of new workstations which have the NVIDIA 4500 X2 video 
  cards installed (Dell machines). We intend to run all our new visuals on the machines 
  using OSG. I've been trying to gauge the graphics performance 
  and from 
  what I've seen so far I'm thoroughly disappointed. I'm assuming at this point 
  the boards are not properly configured. The issue I'm having is trying to find a 
  way to test the boards to determine if they are in fact improperly configured or just 
  don't measure up.
   
  As an example, I 
  have a scene database that I load into OSG; at any given moment, the scene has 
  about 40K primitives. I'm only getting frame rates around 5-10Hz - which is 
  totally unacceptable. According to the specs, these GPU's are capable of 225M 
  triangles/sec. I feel certain that only 1 of the 2GPU's is being exercised - I 
  know of no way to verify it.
   
  With all the 
  graphics/visual system wizards who monitor this sight I thought someone 
  out there might be able to provide some advice on getting these new boards 
  properly configured so I can actually test their 
  performance.
   
  Additional info on 
  the new boxes is:
  2 Intel Xenon 
  3.7GHz - Linux OS EMT64
  16GB 
  memory
   
  Also, kudos for 
  the flawless 64-bit OSG build. Worked without a single 
  glitch.
   
   
  Bryan 
  
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Mike Weiblen
Um, this is a no-brainer, right?
-- mew
:-)


Robert Osfield wrote:
> Anybody a monty
> python fans?  The castle from "The Holy Grail" is just a few miles down
> the road in Doune. 
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
On 9/22/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
Those haggis have been at the calendar in Callander again! This timethey've managed to squeeze three days into one. We'll have to start eating
them to keep their numbers down. :):-)The "correct" official dates for the highland gathering are:   Wednesday 18th October
   Thursday 19th October

   Friday 20th OctoberHappy typo'ing!Robert.


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Re: Highland gathering

2006-09-22 Thread Gordon Tomlinson



Dang 
those Haggis have been busy have they been playing with the tribbles again 
;)
 


Best Regards 
 
Gordon
__Gordon 
TomlinsonEmail 
: 
gordon.tomlinson @ overwatch.comYIM/AIM   
: Gordon3dBritMSN 
IM    
: Gordon3dBrit @ 
3dscenegraph.com__Telephone 
(Cell): 
(+1) 214-477-8914Telephone (Work): 
(+1) 703-437-7651"Self defence is not a 
function of learning tricks but is a function of how 
quickly and intensely one can arouse one's instinct 
for survival" - Master Tambo Tetsura
 

  -Original Message-From: 
  [EMAIL PROTECTED] 
  [mailto:[EMAIL PROTECTED]On Behalf Of Robert 
  OsfieldSent: Friday, September 22, 2006 12:24 PMTo: osg 
  usersSubject: Re: [osg-users] Re: Highland 
  gatheringHi All,Ok time for making preferences 
  known is up. The official dates for the highland gathering 
  is:   Wednesday 18th October   Thursday 18th 
  October   Friday 18th OctoberThe overall intent of the 
  three days is explore future paths of scene graphs in general, and the 
  OpenSceneGraph specifically.  I haven't allocated all the topics we'll 
  cover yet, we can do that as the list of who's coming and for what days 
  settles in. So feel free to go ahead a book flights!  Glasgow or 
  Edinburgh airports are equidistant from here so you can choose 
  either.Let me know what days you plan to be in Scotland, when/where 
  you might want picking up, and what days you'll need accommodation in 
  Callander. See you there (here?) 
:-)Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] osgPlugins not found on OS X

2006-09-22 Thread Steven R Woolsey

I am working on a program using osg
and have it compiling under Windows, Linux, and OS X.  It works great
under windows.  I just started working on the OS X version.  I
get the program compiled with osg, but when I run the program, it cannot
find any of the plugins and exits.  I need to load an obj file, a
3ds file, and a number of textures.  It always tells me it cannot
find the plugin.  I have tried the framework version of the binaries
of osg available on the site, as well as compiling the osg from scratch
using gcc.

With the framework version I placed
the Frameworks in "/Library/Frameworks" and PlugIns in "/Library/Application
Support/OpenSceneGraph/PlugIns".  This failed to find the plugins
the same way as when I compiled myself.

I am also unable to get the osg example
programs to find any plugins.  

I am using osg 1.2, and this would be
from a terminal window.  I have set my LD_LIBRARY_PATH to "/usr/local/lib:/usr/local/lib/osgPlugins".
 My path is set to "/usr/local/share/OpenSceneGraph/bin".

I have tried this on both a G5 and an
Intel based Powerbook with the same results.  I must be missing something.
 I am not that familiar with OS X, so forgive my ignorance if this
is an easy fix.

Any suggestions would be greatly appreciated.

Thanks,

Steve Woolsey
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] Visual System Performance

2006-09-22 Thread Wasileski, Bryan J.



Hi 
all,
 
I recently received 
a couple of new workstations which have the NVIDIA 4500 X2 video cards installed 
(Dell machines). We intend to run all our new visuals on the machines 
using OSG. I've been trying to gauge the graphics performance 
and from 
what I've seen so far I'm thoroughly disappointed. I'm assuming at this point 
the boards are not properly configured. The issue I'm having is trying to find a 
way to test the boards to determine if they are in fact improperly configured or just 
don't measure up.
 
As an example, I 
have a scene database that I load into OSG; at any given moment, the scene has 
about 40K primitives. I'm only getting frame rates around 5-10Hz - which is 
totally unacceptable. According to the specs, these GPU's are capable of 225M 
triangles/sec. I feel certain that only 1 of the 2GPU's is being exercised -  I 
know of no way to verify it.
 
With all the 
graphics/visual system wizards who monitor this sight I thought someone out 
there might be able to provide some advice on getting these new boards properly 
configured so I can actually test their performance.
 
Additional info on 
the new boxes is:
2 Intel Xenon 3.7GHz 
- Linux OS EMT64
16GB 
memory
 
Also, kudos for the 
flawless 64-bit OSG build. Worked without a single glitch.
 
 
Bryan 

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Getting a drawable's material

2006-09-22 Thread Robert Osfield
Hi Bill,The easist thing to do would be to osg::Node::getParentalNodePaths() to get the list all possible NodePath's to your node.  Then walk through each of these NodePath to compute the final state are the leaf node, honouring the override/protect settings.
Robert.On 9/22/06, Bill Prendergast <[EMAIL PROTECTED]> wrote:
> Although it may not be the most efficient, wouldn't it be easier to> start from the root of the scene and walk the path down to your node> and use the same StateSet application logic that OSG uses?
Perhaps so.  But consider this...Right at the node I would like to get a Material for, let's saythere is a stateset with OVERRIDE | ON for material.I'm just learning statesets, but it would seem to me that I
could stop right there because the node has its own uniquematerial.  I wouldn't need to traverse upwards.Similarly, if I ever hit any parent node where there is afixed stateset with material, I could stop.
The only problem with walking "down" to my node is knowingapriori the child path needed to get there.I still need to learn to always think about material as a "state"and not a property of a geometry.  Physically, I keep thinking
of a drawable as having faces, textures, and materials ;-)___osg-users mailing listosg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread John_Donovan
Those haggis have been at the calendar in Callander again! This time 
they've managed to squeeze three days into one. We'll have to start eating 
them to keep their numbers down. :)

John Donovan
Sony Computer Entertainment Europe
http://www.scee.com


[EMAIL PROTECTED] wrote on 22/09/2006 17:23:59:

> Hi All,
> 
> Ok time for making preferences known is up. The official dates for 
> the highland gathering is:
> 
>Wednesday 18th October
>Thursday 18th October
>Friday 18th October
> 
> The overall intent of the three days is explore future paths of 
> scene graphs in general, and the OpenSceneGraph specifically.  I 
> haven't allocated all the topics we'll cover yet, we can do that as 
> the list of who's coming and for what days settles in. 
> 
> So feel free to go ahead a book flights!  Glasgow or Edinburgh 
> airports are equidistant from here so you can choose either.
> 
> Let me know what days you plan to be in Scotland, when/where you 
> might want picking up, and what days you'll need accommodation in 
Callander. 
> 
> See you there (here?) :-)
> Robert.___
> osg-users mailing list
> osg-users@openscenegraph.net
> http://openscenegraph.net/mailman/listinfo/osg-users
> http://www.openscenegraph.org/


**
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
[EMAIL PROTECTED]

This footnote also confirms that this email message has been checked
for all known viruses.

**
Sony Computer Entertainment Europe


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
Hi All,Ok time for making preferences known is up. The official dates for the highland gathering is:   Wednesday 18th October
   Thursday 18th October

   Friday 18th October

The overall intent of the three days is explore future paths of scene graphs in general, and the OpenSceneGraph specifically.  I haven't allocated all the topics we'll cover yet, we can do that as the list of who's coming and for what days settles in.
So feel free to go ahead a book flights!  Glasgow or Edinburgh airports are equidistant from here so you can choose either.Let me know what days you plan to be in Scotland, when/where you might want picking up, and what days you'll need accommodation in Callander.
See you there (here?) :-)Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Getting a drawable's material

2006-09-22 Thread Bill Prendergast
> Although it may not be the most efficient, wouldn't it be easier to
> start from the root of the scene and walk the path down to your node
> and use the same StateSet application logic that OSG uses?

Perhaps so.  But consider this...

Right at the node I would like to get a Material for, let's say
there is a stateset with OVERRIDE | ON for material.

I'm just learning statesets, but it would seem to me that I
could stop right there because the node has its own unique
material.  I wouldn't need to traverse upwards.

Similarly, if I ever hit any parent node where there is a
fixed stateset with material, I could stop.

The only problem with walking "down" to my node is knowing
apriori the child path needed to get there.


I still need to learn to always think about material as a "state"
and not a property of a geometry.  Physically, I keep thinking
of a drawable as having faces, textures, and materials ;-)

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


[osg-users] Interactive documentation

2006-09-22 Thread Robert Osfield
A thought occured to me the other while I was waiting for a rebuild of the osgIntrospection wrappers, that perhaps we could have an option in osgIntrospection for having a documentation string for each namespace, class and member function/variable.  This documentation string would be one of the doxygen strings that are embedded in header files.   You'd want to make this optional, as it'd be a bit more bloat to wrappers, but it should be too hard to add since genwrapper is based around doxygen anyway.
So once you have a documentation string, and a suitable scripting/interpreted language like Lua or Phython you can have a script console window in your app that you could do something like:   help(node)
Or   help(node:getChild(i))This was my thought.  Would it be useful?   That bit I can only speculate on.Thoughts?Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Eric Maslowski
Hi Marc,
  Thanks for the example. I agree. Sharing resources can be really important
for some apps. I just need to change my way of thinking with respect to how
to build my sims. In the past, there would be "assets" which represented a
texture, geometry, animation, whatever. They were completely independent of
nodes/entities unless you assign it one. If you wanted to share, you would
just grab an existing asset from the repository. If not, you would copy the
asset and store the new one instead. Same basic idea. :)

Cheers

E.

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Marc Balzig
> Sent: Friday, September 22, 2006 11:46
> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms
> 
> Hi Eric,
> 
> it is very practical to have multiple instances of one object placed at
> various locations in your scene. For that you would have only one Node for
> the object but you would have one MatrixTransform for each instance of
> your
> object (each MatrixTransform would have the Node added as a child) - this
> saves you from loading one object to memory several times.
> 
> Marc
> 
> - Original Message -
> From: "Eric Maslowski" <[EMAIL PROTECTED]>
> To: "'osg users'" 
> Sent: Friday, September 22, 2006 4:48 PM
> Subject: RE: [osg-users] simple question regarding transforms
> 
> 
> > Hi Robert,
> >   Right, that's what I'm wondering. What additional functionality of
> > MatrixTransform is there that requires them to be a separate level in
> the
> > scenegraph and separated from the node it controls? If they are only
> used
> > for transforming nodes, and nodes almost always have a single
> > MatrixTransform directly above them, then why an interface like below
> isn't
> > used.
> >
> > node->setTrans(Vec3);
> > node->setOrientation(Quat);
> > node->setMatrix(Matrixf);
> > node->setXForm(MatrixTransform);
> >
> > Is it so that multiple nodes can share the same xform without needing to
> > parent each other explicitly? Currently, our environments have shown a 1
> > MatrixTransform to 1 Node relationship, and I'm just making sure I
> > understand the full intentions of the transforms so I can explain it to
> > others.
> >
> > Thanks
> >
> > E.
> >
> > > -Original Message-
> > > From: [EMAIL PROTECTED] [mailto:osg-users-
> > > [EMAIL PROTECTED] On Behalf Of Robert Osfield
> > > Sent: Friday, September 22, 2006 10:19
> > > To: osg users
> > > Subject: Re: [osg-users] simple question regarding transforms
> > >
> > > Hi Eric,
> > >
> > > Umm... its called OO programming, nodes can't be everything, you break
> up
> > > the classes to fit the functionality they deliver.
> > >
> > > Robert.
> > >
> > >
> > > On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> > >
> > > Thanks again, Robert.
> > >
> > > One more quick question regarding the use/purpose of
> > > MaxtrixTransforms. Is
> > > there a reason why the transforms for a node are separate from the
> > > Node*
> > > class? The reason I ask is that I'm finding myself using a
> > > MatrixTransform
> > > above many of my objects (static or otherwise), but not anywhere
> > > else. Are
> > > there other intentions or advanced uses for the MaxtrixTransform
> > > class which
> > > better demonstrate/explain the separation?
> > >
> > > Thanks again for your advice. If I have something worth sharing
> > > regarding
> > > the physics integration, I'll post it.
> > >
> > > Cheers
> > >
> > > E.
> > >
> > > ---
> > > Eric Maslowski
> > > Research Computer Specialist
> > > University of Michigan 3D Lab
> > >
> > > Autodesk 3D Studio Max Certified Trainer
> > >
> > > email: [EMAIL PROTECTED]
> > > office: 734-615-9699
> > >
> > > > -Original Message-
> > > > From: [EMAIL PROTECTED] [mailto:osg-users-
> > > > [EMAIL PROTECTED] On Behalf Of Robert Osfield
> > > > Sent: Thursday, September 21, 2006 16:57
> > > > To: osg users
> > > > Subject: Re: [osg-users] simple question regarding transforms
> > > >
> > > > On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> > > >
> > > >   Thanks Robert,
> > > > Unfortunately, I currently don't have the luxury of doing
> > > it at
> > > >   construction time, but this is something I will probably
> > > address in
> > > > the near
> > > >   future.
> > > >
> > > >   Is there an easy way to query whether or not a specific node
> > > already
> > > > has a
> > > >   MatrixTransform sitting above it? (casting the results of
> > > > getParent()
> > > >   perhaps?) I would like to avoid creating duplicate
> > > MatrixTransforms,
> > > >   especially for scenes that may be comprised of many rigid
> > > bodies.
> > > >
> > > >
> > > >
> > > > Sure, just use:
> > > >   for(all my parents)
> > > >   {
> > > >  if (!getParent(i)->asTransform())
> > > >  {
> > > >  need to insert a parent, but first put this in a list
> > > >  }
> > > >
> > > >   }
> > > >   // then change the parents over to av

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
On 9/22/06, Gordon Tomlinson <[EMAIL PROTECTED]> wrote:





Who 
stole the Thursday :)
The haggis, lots of the little rascals around here.  It'll soon be the end of the fishing season soon so the little critters better watch out, they will be fair game once the "Highland Gathering" comes up.
Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Re: Highland gathering

2006-09-22 Thread Gordon Tomlinson



Who 
stole the Thursday :)
 
 
 


Best Regards 
 
Gordon
__Gordon 
TomlinsonEmail 
: 
gordon.tomlinson @ overwatch.comYIM/AIM   
: Gordon3dBritMSN 
IM    
: Gordon3dBrit @ 
3dscenegraph.com__Telephone 
(Cell): 
(+1) 214-477-8914Telephone (Work): 
(+1) 703-437-7651"Self defence is not a 
function of learning tricks but is a function of how 
quickly and intensely one can arouse one's instinct 
for survival" - Master Tambo Tetsura
 

  -Original Message-From: 
  [EMAIL PROTECTED] 
  [mailto:[EMAIL PROTECTED]On Behalf Of Robert 
  OsfieldSent: Friday, September 22, 2006 10:44 AMTo: osg 
  usersSubject: Re: [osg-users] Re: Highland 
  gatheringThe current dates that seem to fit the most 
  people are:  Wednesday 18th - Friday 19th October.I'll 
  plumb for these dates, but will hold off for another for hour, so shout now if 
  you can't make the above dates. 
Cheer,Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] consistent osg file format

2006-09-22 Thread Robert Osfield
On 9/22/06, c sklu <[EMAIL PROTECTED]> wrote:
Hi Robert,The scenegraph is constructed programmatically (based on our own inputfiles) and I don't use any optimization since the structure needs tobe preserved so it can be modified at runtime. Does OSG have any
procedure for testing its file loaders?There is no formal test procedure, but for testing I use for testing is write it out, then read it back in, then write out again and compare the differences.
I would expect the results to be the same on all systems as all the core OSG and plugins are just C++, and should be produce the pretty well the same results across, I can't think of any reason for any major differences.
Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Getting a drawable's material

2006-09-22 Thread c sklu

Although it may not be the most efficient, wouldn't it be easier to
start from the root of the scene and walk the path down to your node
and use the same StateSet application logic that OSG uses?

On 9/22/06, Bill Prendergast <[EMAIL PROTECTED]> wrote:

> This sounds like a good idea, but a generic way to do this has the
> problem of multiple parents.  In order to maximize reuse of
> drawables/models, you might have the same model attached to multiple
> parents with possible different matrix transforms or statesets

Hmm... good point.  From top-down, the object will get rendered each
time it is traversed into by a parent, with THAT parent's prevalent stateset.

Going backwards would result in a "tree" of potential results, and could
be arbitrarily complex (multiple parents of multiple parents).

My needs fall on the "single parent" side, so I wasn't thinking along the
lines of multiple parents.


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Eric Maslowski
Hi Robert,
  You're right, it wasn't really a scenegraph as you described it.
Basically, each subsystem had the scene stored in a manner that was optimal
for its purposes. (e.g. collision/response stored spatially, entity
repository stored for quick name searches, renderer stored by visibility and
states/materials, etc.) It was an interesting system which provided some
nice features, but like you said it did make other things require a more
work.

E.

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Robert Osfield
> Sent: Friday, September 22, 2006 11:31
> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms
> 
> Hi Ericm
> 
> 
> On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> 
>   Thanks guys and thanks for the example, Don. With some of the past
> engines
>   I've worked on, an entity/node had access to everything that defines
> its
>   current state (xform, mat, etc.) and would just be derived from the
>   components that it needed.
> 
>   class Door: public Animatable, public Drawable, public Entity {};
> 
> 
> These weren't general purpose scene graphs then, perhaps not scene graphs
> at all.
> 
> A scene graph is princply about uses a directed graph to "compose" a
> scene, both logically and spatially.  It is an embodiement of the
> composite design pattern, where functionality is set up using "has a"
> relationships.
> 
> This is different from the approach above where entities are setup via "is
> a" relatioships.
> 
> In OO programming the "has a" is well know to be much more flexibility.
> However, with the flexibility that loose coupling brings the downside is
> that some operations require more work to obtain the effective state of
> subgraph.
> 
> Robert.
> 


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] consistent osg file format

2006-09-22 Thread c sklu

Hi Robert,

The scenegraph is constructed programmatically (based on our own input
files) and I don't use any optimization since the structure needs to
be preserved so it can be modified at runtime. Does OSG have any
procedure for testing its file loaders?

On 9/22/06, Robert Osfield <[EMAIL PROTECTED]> wrote:

Hi Chris,

There shouldn't be an order dependency of the writing out or reading of .osg
or .ive files.  My guess is that the Optimizer is what is probably jumbling
things around.  Try switching off the optimizer in your tests.

Robert.


On 9/22/06, c sklu <[EMAIL PROTECTED]> wrote:
>
> Hi,
>
> We're using the osgDB::WriteNodeFile functionality to write our
> scenegraph to a file which can then be used for testing purposes.
> Unfortunately and surprisingly, the layout of the file is not the same
> on different platforms. We are developing on: Windows, Linux (RH and
> SuSE), Irix, Sun, HP-UX, and AIX. I expected that the UniqueID field
> would be different which our testnig scripts can deal with, but we've
> found that the order of nodes can also be different across platforms.
> Is there any reason for this and is there any way of enforcing a
> certain order or is there any other better way to make regression
> tests for our osg-based application.
>
> Thanks,
> Chris
> ___
> osg-users mailing list
> osg-users@openscenegraph.net
> http://openscenegraph.net/mailman/listinfo/osg-users
> http://www.openscenegraph.org/
>


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] simple question regarding transforms

2006-09-22 Thread Marc Balzig
Hi Eric,

it is very practical to have multiple instances of one object placed at
various locations in your scene. For that you would have only one Node for
the object but you would have one MatrixTransform for each instance of your
object (each MatrixTransform would have the Node added as a child) - this
saves you from loading one object to memory several times.

Marc

- Original Message - 
From: "Eric Maslowski" <[EMAIL PROTECTED]>
To: "'osg users'" 
Sent: Friday, September 22, 2006 4:48 PM
Subject: RE: [osg-users] simple question regarding transforms


> Hi Robert,
>   Right, that's what I'm wondering. What additional functionality of
> MatrixTransform is there that requires them to be a separate level in the
> scenegraph and separated from the node it controls? If they are only used
> for transforming nodes, and nodes almost always have a single
> MatrixTransform directly above them, then why an interface like below
isn't
> used.
>
> node->setTrans(Vec3);
> node->setOrientation(Quat);
> node->setMatrix(Matrixf);
> node->setXForm(MatrixTransform);
>
> Is it so that multiple nodes can share the same xform without needing to
> parent each other explicitly? Currently, our environments have shown a 1
> MatrixTransform to 1 Node relationship, and I'm just making sure I
> understand the full intentions of the transforms so I can explain it to
> others.
>
> Thanks
>
> E.
>
> > -Original Message-
> > From: [EMAIL PROTECTED] [mailto:osg-users-
> > [EMAIL PROTECTED] On Behalf Of Robert Osfield
> > Sent: Friday, September 22, 2006 10:19
> > To: osg users
> > Subject: Re: [osg-users] simple question regarding transforms
> >
> > Hi Eric,
> >
> > Umm... its called OO programming, nodes can't be everything, you break
up
> > the classes to fit the functionality they deliver.
> >
> > Robert.
> >
> >
> > On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> >
> > Thanks again, Robert.
> >
> > One more quick question regarding the use/purpose of
> > MaxtrixTransforms. Is
> > there a reason why the transforms for a node are separate from the
> > Node*
> > class? The reason I ask is that I'm finding myself using a
> > MatrixTransform
> > above many of my objects (static or otherwise), but not anywhere
> > else. Are
> > there other intentions or advanced uses for the MaxtrixTransform
> > class which
> > better demonstrate/explain the separation?
> >
> > Thanks again for your advice. If I have something worth sharing
> > regarding
> > the physics integration, I'll post it.
> >
> > Cheers
> >
> > E.
> >
> > ---
> > Eric Maslowski
> > Research Computer Specialist
> > University of Michigan 3D Lab
> >
> > Autodesk 3D Studio Max Certified Trainer
> >
> > email: [EMAIL PROTECTED]
> > office: 734-615-9699
> >
> > > -Original Message-
> > > From: [EMAIL PROTECTED] [mailto:osg-users-
> > > [EMAIL PROTECTED] On Behalf Of Robert Osfield
> > > Sent: Thursday, September 21, 2006 16:57
> > > To: osg users
> > > Subject: Re: [osg-users] simple question regarding transforms
> > >
> > > On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> > >
> > >   Thanks Robert,
> > > Unfortunately, I currently don't have the luxury of doing
> > it at
> > >   construction time, but this is something I will probably
> > address in
> > > the near
> > >   future.
> > >
> > >   Is there an easy way to query whether or not a specific node
> > already
> > > has a
> > >   MatrixTransform sitting above it? (casting the results of
> > > getParent()
> > >   perhaps?) I would like to avoid creating duplicate
> > MatrixTransforms,
> > >   especially for scenes that may be comprised of many rigid
> > bodies.
> > >
> > >
> > >
> > > Sure, just use:
> > >   for(all my parents)
> > >   {
> > >  if (!getParent(i)->asTransform())
> > >  {
> > >  need to insert a parent, but first put this in a list
> > >  }
> > >
> > >   }
> > >   // then change the parents over to avoid invalidating parent
> > > positions/iterators.
> > >
> > > Robert.
> >
> >
> > ___
> > osg-users mailing list
> > osg-users@openscenegraph.net
> > http://openscenegraph.net/mailman/listinfo/osg-users
> > http://www.openscenegraph.org/
> >
> >
>
>
> ___
> osg-users mailing list
> osg-users@openscenegraph.net
> http://openscenegraph.net/mailman/listinfo/osg-users
> http://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] simple question regarding transforms

2006-09-22 Thread Robert Osfield
Hi EricmOn 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
Thanks guys and thanks for the example, Don. With some of the past enginesI've worked on, an entity/node had access to everything that defines itscurrent state (xform, mat, etc.) and would just be derived from the
components that it needed.class Door: public Animatable, public Drawable, public Entity {};These weren't general purpose scene graphs then, perhaps not scene graphs at all.A scene graph is princply about uses a directed graph to "compose" a scene, both logically and spatially.  It is an embodiement of the composite design pattern, where functionality is set up using "has a" relationships.
This is different from the approach above where entities are setup via "is a" relatioships.In OO programming the "has a" is well know to be much more flexibility.  However, with the flexibility that loose coupling brings the downside is that some operations require more work to obtain the effective state of subgraph.
Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Holtz, Corbin L
Eric,

In my application I maintain a list of "reference frames" (which are
just group nodes under a transform) to which I can easily add or remove
objects or subtrees from.  For example, I have reference frames for
J200, EarthFixed, Edwards Air Force Base, Runway 22, etc to deal with
placing spacecraft in the world.  I also have vehicle related frames for
structural frame, body frame, sensor bases, etc.  Each of these
reference frames have many, many nodes and other transforms under them.

Just some more info on applications of transforms :)

Corbin


-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Eric
Maslowski
Sent: Friday, September 22, 2006 9:48 AM
To: 'osg users'
Subject: RE: [osg-users] simple question regarding transforms

Hi Robert,
  Right, that's what I'm wondering. What additional functionality of
MatrixTransform is there that requires them to be a separate level in
the
scenegraph and separated from the node it controls? If they are only
used
for transforming nodes, and nodes almost always have a single
MatrixTransform directly above them, then why an interface like below
isn't
used.

node->setTrans(Vec3);
node->setOrientation(Quat);
node->setMatrix(Matrixf);
node->setXForm(MatrixTransform);

Is it so that multiple nodes can share the same xform without needing to
parent each other explicitly? Currently, our environments have shown a 1
MatrixTransform to 1 Node relationship, and I'm just making sure I
understand the full intentions of the transforms so I can explain it to
others. 

Thanks

E.

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Robert Osfield
> Sent: Friday, September 22, 2006 10:19
> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms
> 
> Hi Eric,
> 
> Umm... its called OO programming, nodes can't be everything, you break
up
> the classes to fit the functionality they deliver.
> 
> Robert.
> 
> 
> On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> 
>   Thanks again, Robert.
> 
>   One more quick question regarding the use/purpose of
> MaxtrixTransforms. Is
>   there a reason why the transforms for a node are separate from
the
> Node*
>   class? The reason I ask is that I'm finding myself using a
> MatrixTransform
>   above many of my objects (static or otherwise), but not anywhere
> else. Are
>   there other intentions or advanced uses for the MaxtrixTransform
> class which
>   better demonstrate/explain the separation?
> 
>   Thanks again for your advice. If I have something worth sharing
> regarding
>   the physics integration, I'll post it.
> 
>   Cheers
> 
>   E.
> 
>   ---
>   Eric Maslowski
>   Research Computer Specialist
>   University of Michigan 3D Lab
> 
>   Autodesk 3D Studio Max Certified Trainer
> 
>   email: [EMAIL PROTECTED]
>   office: 734-615-9699
> 
>   > -Original Message-
>   > From: [EMAIL PROTECTED] [mailto:osg-users-
>   > [EMAIL PROTECTED] On Behalf Of Robert Osfield
>   > Sent: Thursday, September 21, 2006 16:57
>   > To: osg users
>   > Subject: Re: [osg-users] simple question regarding transforms
>   >
>   > On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
>   >
>   >   Thanks Robert,
>   > Unfortunately, I currently don't have the luxury of
doing
> it at
>   >   construction time, but this is something I will probably
> address in
>   > the near
>   >   future.
>   >
>   >   Is there an easy way to query whether or not a specific
node
> already
>   > has a
>   >   MatrixTransform sitting above it? (casting the results
of
>   > getParent()
>   >   perhaps?) I would like to avoid creating duplicate
> MatrixTransforms,
>   >   especially for scenes that may be comprised of many
rigid
> bodies.
>   >
>   >
>   >
>   > Sure, just use:
>   >   for(all my parents)
>   >   {
>   >  if (!getParent(i)->asTransform())
>   >  {
>   >  need to insert a parent, but first put this in a list
>   >  }
>   >
>   >   }
>   >   // then change the parents over to avoid invalidating parent
>   > positions/iterators.
>   >
>   > Robert.
> 
> 
>   ___
>   osg-users mailing list
>   osg-users@openscenegraph.net
>   http://openscenegraph.net/mailman/listinfo/osg-users
>   http://www.openscenegraph.org/
> 
> 


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Eric Maslowski
Thanks guys and thanks for the example, Don. With some of the past engines
I've worked on, an entity/node had access to everything that defines its
current state (xform, mat, etc.) and would just be derived from the
components that it needed.

class Door: public Animatable, public Drawable, public Entity {};

Previously in Performer, and now OSG, my applications were typically very
simple in terms of interaction, so I never really thought about transforms
that much. Thanks for clearing it up.

E.
 
> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Donald Tidrow
> Sent: Friday, September 22, 2006 11:04
> To: osg users
> Subject: RE: [osg-users] simple question regarding transforms
> 
> On Fri, 2006-09-22 at 10:48 -0400, Eric Maslowski wrote:
> > Hi Robert,
> >   Right, that's what I'm wondering. What additional functionality of
> > MatrixTransform is there that requires them to be a separate level in
> the
> > scenegraph and separated from the node it controls? If they are only
> used
> > for transforming nodes, and nodes almost always have a single
> > MatrixTransform directly above them, then why an interface like below
> isn't
> > used.
> >
> > node->setTrans(Vec3);
> > node->setOrientation(Quat);
> > node->setMatrix(Matrixf);
> > node->setXForm(MatrixTransform);
> >
> > Is it so that multiple nodes can share the same xform without needing to
> > parent each other explicitly? Currently, our environments have shown a 1
> > MatrixTransform to 1 Node relationship, and I'm just making sure I
> > understand the full intentions of the transforms so I can explain it to
> > others.
> >
> > Thanks
> >
> > E.
> >
> There are many situations where multiple Groups will be placed under a
> single MatrixTransform, especially when rendering large terrain
> surfaces.  The transform will position the terrain block in the 'world'
> to reduce precision artifacts, and the terrain geometry will be broken
> up into smaller groups in a quadtree structure for efficient culling.
> 
> Don
> 
> --
> "A government that is big enough to give you all you want
> is big enough to take it all away." -- Barry Goldwater
> 
> +---+
> |  Don Tidrow   |
> |  Vis-Sim geek |
> +---+


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Donald Tidrow
On Fri, 2006-09-22 at 10:48 -0400, Eric Maslowski wrote:
> Hi Robert,
>   Right, that's what I'm wondering. What additional functionality of
> MatrixTransform is there that requires them to be a separate level in the
> scenegraph and separated from the node it controls? If they are only used
> for transforming nodes, and nodes almost always have a single
> MatrixTransform directly above them, then why an interface like below isn't
> used.
> 
> node->setTrans(Vec3);
> node->setOrientation(Quat);
> node->setMatrix(Matrixf);
> node->setXForm(MatrixTransform);
> 
> Is it so that multiple nodes can share the same xform without needing to
> parent each other explicitly? Currently, our environments have shown a 1
> MatrixTransform to 1 Node relationship, and I'm just making sure I
> understand the full intentions of the transforms so I can explain it to
> others. 
> 
> Thanks
> 
> E.
> 
There are many situations where multiple Groups will be placed under a
single MatrixTransform, especially when rendering large terrain
surfaces.  The transform will position the terrain block in the 'world'
to reduce precision artifacts, and the terrain geometry will be broken
up into smaller groups in a quadtree structure for efficient culling.

Don

-- 
"A government that is big enough to give you all you want
is big enough to take it all away." -- Barry Goldwater

+---+
|  Don Tidrow   |
|  Vis-Sim geek |
+---+



signature.asc
Description: This is a digitally signed message part
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] simple question regarding transforms

2006-09-22 Thread Robert Osfield
Hi Eric,For your particular usage model your finding that nodes often have transforms above, however, this is far from typical, typically one has a scene is a modest number of transforms in them relative to the total number of nodes in usage.  There a lots of different types of nodes available in the scene graph, you compose your scene graph with these to suit your purpose.
Robert.On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
Hi Robert,  Right, that's what I'm wondering. What additional functionality ofMatrixTransform is there that requires them to be a separate level in thescenegraph and separated from the node it controls? If they are only used
for transforming nodes, and nodes almost always have a singleMatrixTransform directly above them, then why an interface like below isn'tused.node->setTrans(Vec3);node->setOrientation(Quat);
node->setMatrix(Matrixf);node->setXForm(MatrixTransform);Is it so that multiple nodes can share the same xform without needing toparent each other explicitly? Currently, our environments have shown a 1
MatrixTransform to 1 Node relationship, and I'm just making sure Iunderstand the full intentions of the transforms so I can explain it toothers.ThanksE.> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-> [EMAIL PROTECTED]
] On Behalf Of Robert Osfield> Sent: Friday, September 22, 2006 10:19> To: osg users> Subject: Re: [osg-users] simple question regarding transforms>> Hi Eric,>> Umm... its called OO programming, nodes can't be everything, you break up
> the classes to fit the functionality they deliver.>> Robert.>>> On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:>>   Thanks again, Robert.
>>   One more quick question regarding the use/purpose of> MaxtrixTransforms. Is>   there a reason why the transforms for a node are separate from the> Node*>   class? The reason I ask is that I'm finding myself using a
> MatrixTransform>   above many of my objects (static or otherwise), but not anywhere> else. Are>   there other intentions or advanced uses for the MaxtrixTransform> class which
>   better demonstrate/explain the separation?>>   Thanks again for your advice. If I have something worth sharing> regarding>   the physics integration, I'll post it.>
>   Cheers>>   E.>>   --->   Eric Maslowski>   Research Computer Specialist>   University of Michigan 3D Lab>>   Autodesk 3D Studio Max Certified Trainer
>>   email: [EMAIL PROTECTED]>   office: 734-615-9699>>   > -Original Message->   > From: 
[EMAIL PROTECTED] [mailto:osg-users->   > [EMAIL PROTECTED]] On Behalf Of Robert Osfield>   > Sent: Thursday, September 21, 2006 16:57
>   > To: osg users>   > Subject: Re: [osg-users] simple question regarding transforms>   >>   > On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]
> wrote:>   >>   >   Thanks Robert,>   > Unfortunately, I currently don't have the luxury of doing> it at>   >   construction time, but this is something I will probably
> address in>   > the near>   >   future.>   >>   >   Is there an easy way to query whether or not a specific node> already>   > has a
>   >   MatrixTransform sitting above it? (casting the results of>   > getParent()>   >   perhaps?) I would like to avoid creating duplicate> MatrixTransforms,>   >   especially for scenes that may be comprised of many rigid
> bodies.>   >>   >>   >>   > Sure, just use:>   >   for(all my parents)>   >   {>   >  if (!getParent(i)->asTransform())
>   >  {>   >  need to insert a parent, but first put this in a list>   >  }>   >>   >   }>   >   // then change the parents over to avoid invalidating parent
>   > positions/iterators.>   >>   > Robert.>>>   ___>   osg-users mailing list>   
osg-users@openscenegraph.net>   http://openscenegraph.net/mailman/listinfo/osg-users>   
http://www.openscenegraph.org/>>___osg-users mailing listosg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-usershttp://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Eric Maslowski
Hi Robert,
  Right, that's what I'm wondering. What additional functionality of
MatrixTransform is there that requires them to be a separate level in the
scenegraph and separated from the node it controls? If they are only used
for transforming nodes, and nodes almost always have a single
MatrixTransform directly above them, then why an interface like below isn't
used.

node->setTrans(Vec3);
node->setOrientation(Quat);
node->setMatrix(Matrixf);
node->setXForm(MatrixTransform);

Is it so that multiple nodes can share the same xform without needing to
parent each other explicitly? Currently, our environments have shown a 1
MatrixTransform to 1 Node relationship, and I'm just making sure I
understand the full intentions of the transforms so I can explain it to
others. 

Thanks

E.

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Robert Osfield
> Sent: Friday, September 22, 2006 10:19
> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms
> 
> Hi Eric,
> 
> Umm... its called OO programming, nodes can't be everything, you break up
> the classes to fit the functionality they deliver.
> 
> Robert.
> 
> 
> On 9/22/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> 
>   Thanks again, Robert.
> 
>   One more quick question regarding the use/purpose of
> MaxtrixTransforms. Is
>   there a reason why the transforms for a node are separate from the
> Node*
>   class? The reason I ask is that I'm finding myself using a
> MatrixTransform
>   above many of my objects (static or otherwise), but not anywhere
> else. Are
>   there other intentions or advanced uses for the MaxtrixTransform
> class which
>   better demonstrate/explain the separation?
> 
>   Thanks again for your advice. If I have something worth sharing
> regarding
>   the physics integration, I'll post it.
> 
>   Cheers
> 
>   E.
> 
>   ---
>   Eric Maslowski
>   Research Computer Specialist
>   University of Michigan 3D Lab
> 
>   Autodesk 3D Studio Max Certified Trainer
> 
>   email: [EMAIL PROTECTED]
>   office: 734-615-9699
> 
>   > -Original Message-
>   > From: [EMAIL PROTECTED] [mailto:osg-users-
>   > [EMAIL PROTECTED] On Behalf Of Robert Osfield
>   > Sent: Thursday, September 21, 2006 16:57
>   > To: osg users
>   > Subject: Re: [osg-users] simple question regarding transforms
>   >
>   > On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
>   >
>   >   Thanks Robert,
>   > Unfortunately, I currently don't have the luxury of doing
> it at
>   >   construction time, but this is something I will probably
> address in
>   > the near
>   >   future.
>   >
>   >   Is there an easy way to query whether or not a specific node
> already
>   > has a
>   >   MatrixTransform sitting above it? (casting the results of
>   > getParent()
>   >   perhaps?) I would like to avoid creating duplicate
> MatrixTransforms,
>   >   especially for scenes that may be comprised of many rigid
> bodies.
>   >
>   >
>   >
>   > Sure, just use:
>   >   for(all my parents)
>   >   {
>   >  if (!getParent(i)->asTransform())
>   >  {
>   >  need to insert a parent, but first put this in a list
>   >  }
>   >
>   >   }
>   >   // then change the parents over to avoid invalidating parent
>   > positions/iterators.
>   >
>   > Robert.
> 
> 
>   ___
>   osg-users mailing list
>   osg-users@openscenegraph.net
>   http://openscenegraph.net/mailman/listinfo/osg-users
>   http://www.openscenegraph.org/
> 
> 


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] Getting a drawable's material

2006-09-22 Thread Bill Prendergast
> This sounds like a good idea, but a generic way to do this has the
> problem of multiple parents.  In order to maximize reuse of
> drawables/models, you might have the same model attached to multiple
> parents with possible different matrix transforms or statesets

Hmm... good point.  From top-down, the object will get rendered each
time it is traversed into by a parent, with THAT parent's prevalent stateset.

Going backwards would result in a "tree" of potential results, and could
be arbitrarily complex (multiple parents of multiple parents).

My needs fall on the "single parent" side, so I wasn't thinking along the
lines of multiple parents.


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
The current dates that seem to fit the most people are:  Wednesday 18th - Friday 19th October.I'll plumb for these dates, but will hold off for another for hour, so shout now if you can't make the above dates.
Cheer,Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
On 9/22/06, Roberts, Gareth <[EMAIL PROTECTED]> wrote:






I hope to be there, I can probably make 
either date.
I expect I'll come by car, which 
may help with transport.Great.  Where abouts are you based?Are there particular topics you interested in?  Anything you'd like to present yourself?Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Getting a drawable's material

2006-09-22 Thread Holtz, Corbin L
This sounds like a good idea, but a generic way to do this has the
problem of multiple parents.  In order to maximize reuse of
drawables/models, you might have the same model attached to multiple
parents with possible different matrix transforms or statesets (think a
forest of trees).  If your scenegraph is not so complicated and doesn't
have multiparenting, then you'll have a better chance of this working.

I think Robert will have to chime in on the actual specifics of
gathering StateSet info while traversing with a node visitor (maybe it's
done automatically).

Corbin


-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Bill
Prendergast
Sent: Friday, September 22, 2006 9:23 AM
To: osg-users@openscenegraph.net
Subject: [osg-users] Getting a drawable's material

Could someone please confirm if I'm understanding this right...

I want to know the material for an arbitrary node or drawable object.
The object may or may not contain a stateset of its own, but isn't
required to have one if it's inheriting from above.

Regardless, to find the material in play at that object requires that
I traverse up through the parents, taking the OVERRIDE and PROTECTED
flags into account until I find the last stateset change that takes
effect
before the object in question, correct?

(i.e., I need to emulate what OpenGL does, but traverse backwards
until I find the material that OGL would end up using from a root-down
traverse).

If so, is there a "get the current material" nodeVisitor already written
to do this?  I didn't see one from a cursory look.


[related follow-on Q: once I found the material, is there a way to set
it in the object's stateset in such a manner that OGL doesn't interpret
it as a state change?  Kind of a.. hmm... "comment" stateset, such
that a future "get material" query on the same node would avoid another
traversal? ]

Thanks!

Bill



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


[osg-users] Getting a drawable's material

2006-09-22 Thread Bill Prendergast
Could someone please confirm if I'm understanding this right...

I want to know the material for an arbitrary node or drawable object.
The object may or may not contain a stateset of its own, but isn't
required to have one if it's inheriting from above.

Regardless, to find the material in play at that object requires that
I traverse up through the parents, taking the OVERRIDE and PROTECTED
flags into account until I find the last stateset change that takes effect
before the object in question, correct?

(i.e., I need to emulate what OpenGL does, but traverse backwards
until I find the material that OGL would end up using from a root-down
traverse).

If so, is there a "get the current material" nodeVisitor already written
to do this?  I didn't see one from a cursory look.


[related follow-on Q: once I found the material, is there a way to set
it in the object's stateset in such a manner that OGL doesn't interpret
it as a state change?  Kind of a.. hmm... "comment" stateset, such
that a future "get material" query on the same node would avoid another
traversal? ]

Thanks!

Bill



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] simple question regarding transforms

2006-09-22 Thread Robert Osfield
Hi Eric,Umm... its called OO programming, nodes can't be everything, you break up the classes to fit the functionality they deliver.  Robert.On 9/22/06, 
Eric Maslowski <[EMAIL PROTECTED]> wrote:
Thanks again, Robert.One more quick question regarding the use/purpose of MaxtrixTransforms. Isthere a reason why the transforms for a node are separate from the Node*class? The reason I ask is that I'm finding myself using a MatrixTransform
above many of my objects (static or otherwise), but not anywhere else. Arethere other intentions or advanced uses for the MaxtrixTransform class whichbetter demonstrate/explain the separation?Thanks again for your advice. If I have something worth sharing regarding
the physics integration, I'll post it.CheersE.---Eric MaslowskiResearch Computer SpecialistUniversity of Michigan 3D LabAutodesk 3D Studio Max Certified Traineremail: 
[EMAIL PROTECTED]office: 734-615-9699> -Original Message-> From: [EMAIL PROTECTED]
 [mailto:osg-users-> [EMAIL PROTECTED]] On Behalf Of Robert Osfield> Sent: Thursday, September 21, 2006 16:57> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms>> On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:>>   Thanks Robert,
> Unfortunately, I currently don't have the luxury of doing it at>   construction time, but this is something I will probably address in> the near>   future.>>   Is there an easy way to query whether or not a specific node already
> has a>   MatrixTransform sitting above it? (casting the results of> getParent()>   perhaps?) I would like to avoid creating duplicate MatrixTransforms,>   especially for scenes that may be comprised of many rigid bodies.
 Sure, just use:>   for(all my parents)>   {>  if (!getParent(i)->asTransform())>  {>  need to insert a parent, but first put this in a list
>  }>>   }>   // then change the parents over to avoid invalidating parent> positions/iterators.>> Robert.___
osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] consistent osg file format

2006-09-22 Thread Robert Osfield
Hi Chris,There shouldn't be an order dependency of the writing out or reading of .osg or .ive files.  My guess is that the Optimizer is what is probably jumbling things around.  Try switching off the optimizer in your tests.
Robert.On 9/22/06, c sklu <[EMAIL PROTECTED]> wrote:
Hi,We're using the osgDB::WriteNodeFile functionality to write ourscenegraph to a file which can then be used for testing purposes.Unfortunately and surprisingly, the layout of the file is not the same
on different platforms. We are developing on: Windows, Linux (RH andSuSE), Irix, Sun, HP-UX, and AIX. I expected that the UniqueID fieldwould be different which our testnig scripts can deal with, but we'vefound that the order of nodes can also be different across platforms.
Is there any reason for this and is there any way of enforcing acertain order or is there any other better way to make regressiontests for our osg-based application.Thanks,Chris___
osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Anti aliasing

2006-09-22 Thread Stephan Huber

Hi,

Ivan Bolčina schrieb:

Hi. I have a question.

1. How to enable full screen anti aliasing


You can enable it with your video drivers or from code via:

osg::DisplaySettings * ds = osg::DisplaySettings::instance();
ds->setNumMultiSamples(numSamples);

before realizing your viewer.


2. Is it possible to enable anti aliasing only on some object?


On Mac OS X you can try to add a stateset to the given node:
node->getOrCreateStateSet()->setMode(GL_MULTISAMPLE_ARB, 
osg::StateAttribute::ON);


This does not work with every hardware, though, don't know about 
windows/linux.


HTH,
Stephan
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] simple question regarding transforms

2006-09-22 Thread Eric Maslowski
Thanks again, Robert.

One more quick question regarding the use/purpose of MaxtrixTransforms. Is
there a reason why the transforms for a node are separate from the Node*
class? The reason I ask is that I'm finding myself using a MatrixTransform
above many of my objects (static or otherwise), but not anywhere else. Are
there other intentions or advanced uses for the MaxtrixTransform class which
better demonstrate/explain the separation? 

Thanks again for your advice. If I have something worth sharing regarding
the physics integration, I'll post it.

Cheers

E.

---
Eric Maslowski
Research Computer Specialist
University of Michigan 3D Lab

Autodesk 3D Studio Max Certified Trainer

email: [EMAIL PROTECTED]
office: 734-615-9699

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:osg-users-
> [EMAIL PROTECTED] On Behalf Of Robert Osfield
> Sent: Thursday, September 21, 2006 16:57
> To: osg users
> Subject: Re: [osg-users] simple question regarding transforms
> 
> On 9/21/06, Eric Maslowski <[EMAIL PROTECTED]> wrote:
> 
>   Thanks Robert,
> Unfortunately, I currently don't have the luxury of doing it at
>   construction time, but this is something I will probably address in
> the near
>   future.
> 
>   Is there an easy way to query whether or not a specific node already
> has a
>   MatrixTransform sitting above it? (casting the results of
> getParent()
>   perhaps?) I would like to avoid creating duplicate MatrixTransforms,
>   especially for scenes that may be comprised of many rigid bodies.
> 
> 
> 
> Sure, just use:
>   for(all my parents)
>   {
>  if (!getParent(i)->asTransform())
>  {
>  need to insert a parent, but first put this in a list
>  }
> 
>   }
>   // then change the parents over to avoid invalidating parent
> positions/iterators.
> 
> Robert.


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


RE: [osg-users] Re: Highland gathering

2006-09-22 Thread Roberts, Gareth



Hi,
I hope to be there, I can probably make 
either date.
I expect I'll come by car, which 
may help with transport.
Gareth

  -Original Message-From: Robert Osfield 
  [mailto:[EMAIL PROTECTED]Sent: 22 September 2006 
  11:21To: osg usersSubject: Re: [osg-users] Re: Highland 
  gathering
  


  *** WARNING ***This mail has originated 
outside your organization,either from an external partner or the 
Global Internet. Keep this in mind if you answer this message. 
  Hi All,I've short listed a four guest 
  houses that are in Callander.  They are look nice from the outside, and 
  have websites to look at.  It is off season so I don't expect any problem 
  getting rooms.Middle of town (nearest to my home/office):  http://ardenhouse.org.uk/
  http://theknowecallander.co.uk/
  West side of town:
  
  http://lubnaighouse.co.uk/ http://brooklinn-scotland.co.uk/ 
  
  As the numbers arn't likely to be large I'm tempted to 
  hold the meetings at my house.  I have a pretty decent sized 
  conservatory, and will be able to set up net access for us in there.  
  There are other places around that I can hire rooms though so if numbers 
  increase then I'll book a place. I can book rooms if you wish.  I 
  think it would be ideal to get every one, or two places, and keep them near 
  each other, we wouldn't want you getting lost on the way home from the 
  pub...I would like to the set the dates today, so please ping me if 
  you have any preference. 
Robert.


This email and any attachments are confidential to the intended
recipient and may also be privileged. If you are not the intended
recipient please delete it from your system and notify the sender.
You should not copy it or use it for any purpose nor disclose or
distribute its contents to any other person.


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

RE: [osg-users] Anti aliasing

2006-09-22 Thread Holtz, Corbin L








Hi Ivan,

 

Full screen AA is usually enabled through
your OpenGL driver.  If you’re on Windows than I think you can access
through the control panels.  If you have an nVidia card under Linux or OSX you
can use environment variables (__GL_FSAA_MODE).  Look through the nVidia readme
that comes with the drivers.  If you’re using an ATI card than I’m
not sure, perhaps others can reply with ATI info.

 

I’m not sure about anti-aliasing
individual objects, but I think that is what the GL__SMOOTH (POINT, LINE,
POLYGON) are for, although I’ve not used them so I can’t say for
sure.  Again, hopefully someone else will add more info.

 



Corbin











From:
[EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On
Behalf Of Ivan Bolcina
Sent: Friday, September 22, 2006
2:31 AM
To: osg-users@openscenegraph.net
Subject: [osg-users] Anti aliasing



 

Hi. I have a question.

1. How to enable full screen anti aliasing

2. Is it possible to enable anti aliasing only on some object?

Thanks






___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] consistent osg file format

2006-09-22 Thread c sklu

Hi,

We're using the osgDB::WriteNodeFile functionality to write our
scenegraph to a file which can then be used for testing purposes.
Unfortunately and surprisingly, the layout of the file is not the same
on different platforms. We are developing on: Windows, Linux (RH and
SuSE), Irix, Sun, HP-UX, and AIX. I expected that the UniqueID field
would be different which our testnig scripts can deal with, but we've
found that the order of nodes can also be different across platforms.
Is there any reason for this and is there any way of enforcing a
certain order or is there any other better way to make regression
tests for our osg-based application.

Thanks,
Chris
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] MFC CWnd only renders after resize?

2006-09-22 Thread Alan Harris




John

I had a similar problem, which if I remember correctly was fixed by
adding a call to the function that displays the scenegraph in the
DoPaint() function of the CView.

Alan

Robert Osfield wrote:
Hi John,
  
I no nothing about MFC so can comment on specifics.  In the past I
tried integrating QT3 with osgProducer but it didn't work out well due
to isses with resizing amoung other things.  In the end just integating
with SceneView worked out well, but this does entail updating things
like Viewport's and projection matrices mannually when the window
resizes, there is little you can do to avoid this as SceneView is
completely anware of windowing API.
  
  
Robert.
  
  On 9/21/06, Argentieri, John-P63223 <[EMAIL PROTECTED]>
wrote:
  

Hello all.

I know how much the moderators love
MFC, which is a great deal more than I. We have designed our software
using sceneview and thus it is capable of running on other systems as
well, but for one purpose that we have in mind it must work on an
existing MFC project. We have to launch the rendering frame-by-frame in
a timer thread. For some reason, the frames do not show properly until
after the CWnd window is resized at least once. We have already
circumvented this issue by programmatically resizing the window, but
this doesn't seem like the correct thing to do and I think I am missing
something here that should have been done. Has anyone had any similar
issues that can help shed some light on this problem I am having? 
Your time and consideration are much
appreciated.

Sincerely,

John




___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

  
  
  
  

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

-- 
Regards
Alan Harris

ReSoft Ltd
7 Church Lane
Flitton
BEDFORD, MK45 5EL, UK
Tel: +44 (0) 1525 862616
Email: [EMAIL PROTECTED]
Web: www.resoft.co.uk


___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] CameraNode PIXEL_BUFFER/-_RTT fails for Release builds under WIndows

2006-09-22 Thread Stephan Maximilian Huber

Hi Robert,

Robert Osfield schrieb:

As for working in debug but not release, my best guess would be some
unitialized variable perhaps.  The message box perhaps in an indicator 
that
the setup of the Pbuffer failed in Producer so I'd recommend digging 
through
this code and placing stop points on the code to see what it does and 
where

the release vs debug difference might be comming in.

It might be that the glue code osgProducer that allows CameraNode to 
set up

the pbuffer is causing some difference. See
osgProducer::GraphicsContextImplementation for this glue code.


Robert, thanks for the pointers! I have a deadline next week, so no time 
to investigate it deeper, but it seems, that the problem is inside 
Producer. Part of my problem is that SetPixelFormat (in 
WGLExtension.cpp) fails. After doing some research it seems, that there 
are some missing flags when the temporary window gets created, I hope to 
do some more tests next week.


thanks,
Stephan



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/


Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
Hi All,I've short listed a four guest houses that are in Callander.  They are look nice from the outside, and have websites to look at.  It is off season so I don't expect any problem getting rooms.Middle of town (nearest to my home/office):  
http://ardenhouse.org.uk/
http://theknowecallander.co.uk/
West side of town:http://lubnaighouse.co.uk/

http://brooklinn-scotland.co.uk/

 As the numbers arn't likely to be large I'm tempted to hold the meetings at my house.  I have a pretty decent sized conservatory, and will be able to set up net access for us in there.  There are other places around that I can hire rooms though so if numbers increase then I'll book a place.
I can book rooms if you wish.  I think it would be ideal to get every one, or two places, and keep them near each other, we wouldn't want you getting lost on the way home from the pub...I would like to the set the dates today, so please ping me if you have any preference.
Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
Hi Paul,On 9/21/06, Paul Martz <[EMAIL PROTECTED]> wrote:





I am 
likely to attend if we can keep the non-airline expenses under 
$150US/day and not exceed a 7 day stay.The guest houses look to be around £30-35 ($60 - $70) a night.  That should leave some money over for food and beer, and tacky tourist nick nacks!
There is one place near full budget per night so I'd guess we'd have to skip that one, perhaps go there for lunch or evening meal... 
What I 
hope to get out of the event:

  Discussion of future OSG. I want to stay on top of any major 
  evolutionary efforts such as rewriting in C, shrinking OSG to create "OSG 
  Lite", etc.
  Prioritization discussions for any currently non-funded future work, 
  and any creative ideas for funding it.
  Increasing my overall 
  knowledge of OSG and how it is used. I'd love to see presentations by as many 
  attendees as possible. What drove you to OSG 
  and what are the key features that your software 
  requires?
  Networking with like-minded folk.
  Face-to-face with Robert for work on the documentation.
  Some 
  tasty beer.
  Brief 
  sightseeing -- castles have always fascinated me.I'm sure we can cover most of these including castles.   Anybody a monty python fans?  The castle from "The Holy Grail" is just a few miles down the road in Doune.  For a grander castle there is Stirling castle just 15 drive mins on for there.
Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
Hi Don,On 9/21/06, Don Burns <[EMAIL PROTECTED]> wrote:
OK.. I've seen Scotland in Spring and I've seen it in Summer.  Give us a bit of insight on how  to dress for November.

Colin has obvious given you the proper dress code.However, fur ye wooses oot there ye might wanna bring yurself a g'd water proof jacket, and some warm layers as the evenings are far drawing in and can be now bit a wee bit chilly in late October.
Day time temps are likely to between 10 and 15 celcius, guess that would 50-60 fahrenheit, chance of frost too on any cloudless nights.Robert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] Re: Highland gathering

2006-09-22 Thread Robert Osfield
On 9/22/06, Colin Dunlop <[EMAIL PROTECTED]> wrote:
Women tend to go for the longer length kilts as pictured, and usblokes just add a fashionable hat - but aside from that it's businessas usual ;-) I'm sure I caught site of Robert last week in his OSGTartan bonnet looking sleek as ever...
Thanks for the pictures and commentary, certainly caused amusment here across the hill.  och eye.
It is indeed a beautiful place to visitI cyled to school/nursery with my children this morning, the sun was shinning, the forests are just starting to turn to their autum colours, very very pretty indeed.  I took a detour to look where some of the local hotels/guest houses are, boy some pretty locations looking down over Callander to the surround hills and mountains.   Those who can make it are in for a treat :-)
It is Scotland though, and it has mainly rained since Tuesday. so the river teith is a metre up, flooding the low lieing local river side carpark, medow and childrens play park.  Its a pretty impressive torrent.  Still today it shines so all is forgiven.
So Colin, was it as wet holidaying on the west coast as it was here? and since I'm just over a kinda
small hill from Robert I can also help anybody out with transportto/from the airports - so jump on that plane and come visit Roberton his home turf if at all possible for a serious OSG gathering!
http://www.visitscottishheartlands.com/areas/trossachs/index.cfmhttp://www.visitscottishheartlands.com/areas/trossachs/callander.cfm
Yeehaww!!! 8-DRobert.
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] osgfadetext

2006-09-22 Thread Robert Osfield
Hi Anders,I've just doubled checked and osgfadetext is checked in ok.  Perhaps you missed checking out new directories.  I use the so updates using:     cvs -Pd updateTo make sure all new directories are added and redundent ones pruned.
Robert.On 9/22/06, Anders Backman <[EMAIL PROTECTED]> wrote:
Just did a cvs checkout of osg under windows:The OpenSceneGraph.dsw includes a project called osgfadetext, but theproject file (dsp) is missing, so is the code...
 Anders Backman   Email:[EMAIL PROTECTED] HPC2N/VRlab  Phone:+46 (0)90-786 9936 Umea university  Cellular: +46 (0)70-392 64 67
 S-901 87 UMEA SWEDEN Fax:  +46 90-786 6126   http://www.cs.umu.se/~andersb___
osg-users mailing listosg-users@openscenegraph.nethttp://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Re: [osg-users] rotate Object relative and absolute

2006-09-22 Thread Oliver Kutter




ok, forget it. It works,
it was my fault. 

Oliver Kutter schrieb:

  
  Hi,
  
I have a problem concerning the transformation of objects. For example,
I want to rotate an object for each frame. I don't want to rotate it
relative to its current position but absolute to its initial position.
But it does not work. 
  
And now my question (mt
is a MatrixTransform):
What's the difference between those two commands. 
  
  mt->setMatrix(osg::Matrix::rotate(osg::inDegrees(myAngle),
  osg::Vec3(0.0,
1.0,
0.0)) *
mt->getMatrix());
  mt->setMatrix(osg::Matrix::rotate(osg::inDegrees(myAngle),
  osg::Vec3(0.0,
1.0,
0.0)));
  
  I thought, the
first one
is to rotate an Object relative to its old position. And the second one
sets the rotation of a matrix and overwrites the existing rotation. But
the result is the same.
  
regards
Oliver
  
  
  

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/



___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] rotate Object relative and absolute

2006-09-22 Thread Oliver Kutter




Hi,

I have a problem concerning the transformation of objects. For example,
I want to rotate an object for each frame. I don't want to rotate it
relative to its current position but absolute to its initial position.
But it does not work. 

And now my question (mt
is a MatrixTransform):
What's the difference between those two commands. 

mt->setMatrix(osg::Matrix::rotate(osg::inDegrees(myAngle),
osg::Vec3(0.0, 1.0,
0.0)) *
mt->getMatrix());
mt->setMatrix(osg::Matrix::rotate(osg::inDegrees(myAngle),
osg::Vec3(0.0, 1.0,
0.0)));

I thought, the
first one
is to rotate an Object relative to its old position. And the second one
sets the rotation of a matrix and overwrites the existing rotation. But
the result is the same.

regards
Oliver




___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] Anti aliasing

2006-09-22 Thread Ivan Bolčina
Hi. I have a question.1. How to enable full screen anti aliasing2. Is it possible to enable anti aliasing only on some object?Thanks
___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

[osg-users] osgfadetext

2006-09-22 Thread Anders Backman
Just did a cvs checkout of osg under windows:

The OpenSceneGraph.dsw includes a project called osgfadetext, but the
project file (dsp) is missing, so is the code...


 Anders Backman   Email:[EMAIL PROTECTED]
 HPC2N/VRlab  Phone:+46 (0)90-786 9936
 Umea university  Cellular: +46 (0)70-392 64 67
 S-901 87 UMEA SWEDEN Fax:  +46 90-786 6126
   http://www.cs.umu.se/~andersb
 

___
osg-users mailing list
osg-users@openscenegraph.net
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/