opensg-users-request,
>Send Opensg-users mailing list submissions to
> [email protected]
>
>To subscribe or unsubscribe via the World Wide Web, visit
> https://lists.sourceforge.net/lists/listinfo/opensg-users
>or, via email, send a message with subject or body 'help' to
> [EMAIL PROTECTED]
>
>You can reach the person managing the list at
> [EMAIL PROTECTED]
>
>When replying, please edit your Subject line so it is more specific
>than "Re: Contents of Opensg-users digest..."
>
>
>Today's Topics:
>
> 1. Problem with ProxyGroups (Manjunath Sripadarao)
> 2. Re: Problem with ProxyGroups (Dirk Reiners)
> 3. Re: nvidia driver slowness (Matthias Stiller)
> 4. Re: AW: [Opensg-users] Problems with VRML-Loader (Christoph Fuenfzig)
> 5. Re: New library (Dirk Reiners)
> 6. Win32 static library problem (Rune Laursen)
> 7. Re: Win32 static library problem (Andreas Zieringer)
> 8. Re: methods for assigning materials (Dirk Reiners)
>
>--__--__--
>
>Message: 1
>Date: Tue, 26 Apr 2005 21:09:19 +0530
>From: Manjunath Sripadarao <[EMAIL PROTECTED]>
>To: [email protected]
>Subject: [Opensg-users] Problem with ProxyGroups
>Reply-To: [email protected]
>
>Hello,
> I am trying to load the attached file in the tutorial program 10loading.cp=
>p.
>I am doing these steps, I run proxyBuilder on it
>proxyBuilder -t 1 cube.off -o cube
>I get cube.osb and cube_0.osb files.
>I run 10loading as
>./10loading cube.osb
>But I get this error
>LOG: Found no object named Scene.
>LOG: Found no object named TF_DETAIL (did you load the tie?).
>WARNING: MatrixPerspective: near -4.59626e-43 > far -4.59177e-40!
>
>WARNING: MatrixPerspective: near -4.59626e-43 ~=3D far -4.59177e-40!
>...
>
>What am I doing wrong ?
>I have both of these in the same directory and all that.
>
>Thanks,
>Manju
>
>
>--__--__--
>
>Message: 2
>Subject: Re: [Opensg-users] Problem with ProxyGroups
>From: Dirk Reiners <[EMAIL PROTECTED]>
>To: users <[email protected]>
>Date: Tue, 26 Apr 2005 11:12:05 -0500
>Reply-To: [email protected]
>
>
> Hi Manju,
>
>On Tue, 2005-04-26 at 21:09 +0530, Manjunath Sripadarao wrote:
>> Hello,
>> I am trying to load the attached file
>
>there was no attachment on that mail.
>
>> in the tutorial program 10loading.cpp.
>> I am doing these steps, I run proxyBuilder on it
>> proxyBuilder -t 1 cube.off -o cube
>> I get cube.osb and cube_0.osb files.
>> I run 10loading as
>> ./10loading cube.osb
>> But I get this error
>> LOG: Found no object named Scene.
>> LOG: Found no object named TF_DETAIL (did you load the tie?).
>
>Those are normal, they are specific to the tie.wrl model that's part of
>the dist.
>
>> WARNING: MatrixPerspective: near -4.59626e-43 > far -4.59177e-40!
>>
>> WARNING: MatrixPerspective: near -4.59626e-43 ~= far -4.59177e-40!
>> ...
>>
>> What am I doing wrong ?
>> I have both of these in the same directory and all that.
>
>This usually happens when nothing is loaded. The SSM tried to computer
>near/far for an empty scene and ends up getting zeros.
>
>Can you load the off file directly in 10loading, i.e. does the OFF
>loader works for this? If yes, could you send me all the involved files?
>
> Dirk
>
>--
>-- Dirk Reiners OpenSG Forum [EMAIL PROTECTED]
>-- The OpenSG Open Source Scenegraph: http://www.opensg.org
>-- Join the list at http://lists.sf.net/lists/listinfo/opensg-users
>
>
>
>--__--__--
>
>Message: 3
>From: Matthias Stiller <[EMAIL PROTECTED]>
>To: [email protected]
>Subject: Re: [Opensg-users] nvidia driver slowness
>Date: Tue, 26 Apr 2005 17:44:20 +0200
>Reply-To: [email protected]
>
>Hello Jan,
>
>I am talking about windows performance (Hardware is FX3000 resp.
>FX3400). I got feedback from a user who told me that other OpenGL-apps
>are experiencing similar slowdowns.
>
>Wonder if nvidia is already aware of this problem
>
>Matthias
>
>On Tuesday 26 April 2005 16:43, Jan Wurster wrote:
>> Hi Matthias,
>>
>> Matthias Stiller wrote:
>> > after upgrading from the last release driver of the 6x.xx series to
>> > the new release driver of the 71.xx series as well as beta driver
>> > 76.xx I have noticed serious performance drops up to factor 3.
>> >
>> > Can anyone else confirm this ?
>>
>> Are you using win32 or lnx? I've posted about my experience here a
>> couple of weeks ago. I don't see any comparable performance problems
>> with applications other than OpenSG under Linux, though ..
>> everythings fine & fast using win32, so I'm still at a loss as to
>> what might be wrong - I haven't had time to further investigate much
>> in the meantime, though.
>>
>> Best regards,
>> -.jan.-
>
>
>--__--__--
>
>Message: 4
>Date: Tue, 26 Apr 2005 18:48:08 +0200
>From: Christoph Fuenfzig <[EMAIL PROTECTED]>
>To: [email protected]
>Subject: Re: AW: [Opensg-users] Problems with VRML-Loader
>Reply-To: [email protected]
>
>
>Hi,
>
>as far as I know, there are still two versions of the VRMLLoader...
>And for one of them there exists a modification to handle
>PositionInterpolator,
>OrientationInterpolator, and Sound.
>What about integrating this into the CVS version VRMLLoader? Would this
>cause major problems in core development?
>Ok, PROTO are difficult and SCRIPT are (more) difficult to support at
>library level.
>
>Just my two cents,
>Christoph
>
>
>Johannes Behr wrote:
>
>>Hi,
>>
>>I just got a example from the MolScript page
>>(ras_gdp_gz.wrl) and while looking at the file
>>there are a lot stuff (e.g. Proto-declarations,
>>Routes, Text, ...) which the OpenSG-VRML loader
>>alone can not handle:
>>
>>(Proto)Ball: 79
>>(Proto)Cyl: 5
>>(Proto)Label: 8
>>(Proto)Stick: 76
>>FontStyle: 9
>>Text: 9
>>Billboard: 9
>>Collision: 1
>>Transform: 181
>>Coordinate3D: 28
>>Cylinder: 6
>>Extrusion: 77
>>Group: 1
>>IndexedFaceSet: 23
>>IndexedLineSet: 10
>>Appearance: 205
>>Material: 205
>>NavigationInfo: 1
>>Shape: 205
>>Sphere: 80
>>Viewpoint: 4
>>WorldInfo: 1
>>ProtoDeclaration: 4
>>----------------------------------
>>Nodes: 1227
>>Routes: 2546
>>
>>Sorry, but you have to use a system which supports
>>quite a lot of the full VRML/X3D spec.
>>
>>Regards,
>> Johannes
>>
>>On Tue, 2005-04-26 at 14:36 +0200, Khai Binh Duong wrote:
>>
>>
>>>Problem found!
>>>All of these VRML files are generated with MolScript.
>>>I deleted some tags and now the model is correct displayed.
>>>The problem is not solved, but I see that the VRMLLoader is
>>>not able to handle scripts in the VRML files.
>>>
>>>
>>>
>
>
>
>--__--__--
>
>Message: 5
>Subject: Re: [Opensg-users] New library
>From: Dirk Reiners <[EMAIL PROTECTED]>
>To: users <[email protected]>
>Date: Tue, 26 Apr 2005 12:04:50 -0500
>Reply-To: [email protected]
>
>
> Hi Bjoern,
>
>On Tue, 2005-04-26 at 09:30 +0200, Bj=C3=B6rn Harmen Gerth wrote:
>> Hello,
>>=20
>> I have a few questions to setting up a new Library in
>> OpenSG/Source/Experimental. I know of
>> http://cvs.berlios.de/cgi-bin/viewcvs.cgi/*checkout*/opensgplus/HowToExte=
>nd/,
>> but my I think my questions are a bit more specific.
>
>you can say that again. ;)
>
>> 1) I have a template class, say Foo.
>...
>>=20
>> And now the problem: The makefile tries to compile
>> OSGFoo.cpp. How can I exclude that from compilation?
>> Can I write something into common.libs.in or common.mk
>> so that this file will be excluded?
>
>Just don't call it OSG... The build system looks for all files starting
>with OSG and compiles those, and only those.
>
>> 2) There is a test application in the library. It's a
>> QT app, it consists of the files testLib.h/.cpp,
>> testLibWidget.h/.cpp, testLib.ui, testLib.ui.h, the
>> latter two from the QT designer. Some files must be
>> automatically generated from the ui file, and the .h's
>> must be moc'ed. Can I still put the test app into the
>> library dir and get it compiled?=20
>
>You can use your own build system, which is probably not a good idea for
>something that at some point might go into the core. The OpenSG build
>system will moc files that end with _qt.h (e.g.
>OSGQGLManagedDesignerWidget_qt.h). We don't have any .ui files yet,
>therefore there are no rules to update those automatically.
>
>> Do I have to do
>> something similar to
>> OpenSG/Source/WindowSystem/QT/SConscript? How does
>> this script get executed?
>
>These are for the new build system we're planning. It uses an
>alternative to make called SCons (http://www.scons.org/). It's
>functional, but not fully done yet.=20
>
>> 3) I recently updated my OpenSG tree from a CVS
>> version in December to a CVS version a month ago. Now
>> I have a lot of libOSG*D.so files, but when I run make
>> dbg the system looks for libOSG*.so files. My quick
>> fix was to generate symlinks, but is there a proper
>> way to solve this? I did run configure, but is there a
>> new parameter or something?
>
>Hm, that's weird. That might be a bug in the build system. Gerrit, can
>you take a look at that?
>
>Thanks
>
> Dirk
>
>
>--=20
>-- Dirk Reiners OpenSG Forum [EMAIL PROTECTED]
>-- The OpenSG Open Source Scenegraph: http://www.opensg.org
>-- Join the list at http://lists.sf.net/lists/listinfo/opensg-users
>
>
>
>--__--__--
>
>Message: 6
>Date: Tue, 26 Apr 2005 20:31:43 +0200
>From: Rune Laursen <[EMAIL PROTECTED]>
>To: [email protected]
>Subject: [Opensg-users] Win32 static library problem
>Reply-To: [email protected]
>
>Hello,
>
>I'm in the process of encapsulating an OpenSG cluster client and head
>tracking functionality in a application class, so hopefully it will be
>easy for students to get programs running in our VR arenas and head
>mounted displays. I'm trying to get this class compiled into static
>library on Win32 which is causing some problems.
>I have used an existing Tutorial project file to get all the correct
>settings for building normal OpenSG project and then just changed the
>Configuration Type to static library which build a .lib without errors.
>This lib is linked with the main application (also based on a Tutorial
>proj file) without any erros, but when runninng it, it just crashes
>before getting to main().
>Anybody got some advice on how to setup this static lib project for
>CL.NET 2003?
>
>/Rune
>
>
>--__--__--
>
>Message: 7
>Date: Tue, 26 Apr 2005 21:11:50 +0200
>From: Andreas Zieringer <[EMAIL PROTECTED]>
>To: [email protected]
>Subject: Re: [Opensg-users] Win32 static library problem
>Reply-To: [email protected]
>
>Hi Rune,
>
>the linker on windows just links the referenced symbols in your exe.
>There are many classes in OpenSG without any reference, they are created
>via static intilization. I managed to compile a static version of the
>OpenSG activex plugin. You can look in OpenSG/Examples/AXPlugin/main.cpp
>how to create some missing references.
>
>Andreas
>
>> Hello,
>>
>> I'm in the process of encapsulating an OpenSG cluster client and head
>> tracking functionality in a application class, so hopefully it will be
>> easy for students to get programs running in our VR arenas and head
>> mounted displays. I'm trying to get this class compiled into static
>> library on Win32 which is causing some problems.
>> I have used an existing Tutorial project file to get all the correct
>> settings for building normal OpenSG project and then just changed the
>> Configuration Type to static library which build a .lib without errors.
>> This lib is linked with the main application (also based on a Tutorial
>> proj file) without any erros, but when runninng it, it just crashes
>> before getting to main().
>> Anybody got some advice on how to setup this static lib project for
>> CL.NET 2003?
>>
>> /Rune
>>
>>
>> -------------------------------------------------------
>> SF.Net email is sponsored by: Tell us your software development plans!
>> Take this survey and enter to win a one-year sub to SourceForge.net
>> Plus IDC's 2005 look-ahead and a copy of this survey
>> Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix
>> _______________________________________________
>> Opensg-users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/opensg-users
>>
>>
>
>
>
>--__--__--
>
>Message: 8
>Subject: Re: [Opensg-users] methods for assigning materials
>From: Dirk Reiners <[EMAIL PROTECTED]>
>To: users <[email protected]>
>Date: Tue, 26 Apr 2005 15:30:02 -0500
>Reply-To: [email protected]
>
>
> Hi Terry,
>
>On Thu, 2005-04-21 at 14:36 -0500, Terry Welsh wrote:
>> >From the tutorial and example programs, it appears there are two ways
>> to assign materials in OpenSG: either assign them to each individual
>> Geometry node or make a Node with a MaterialGroup core. Assigning
>> materials to every Geometry node seems a bit cumbersome to me. I'm
>> told the MaterialGroup core is dangerous because it overrides
>> materials beneath it on the graph. Please correct me if any of this
>> is wrong.
>
>Nope, that's correct.
>
>> When I speak of materials, I'm not just talking about traditional
>> OpenGL material properties but also textures and shaders. I believe
>> this is the same way OpenSG thinks about materials.
>
>Yes and no. Currently almost everything is Chunk-based, which is very
>close to OpenGL. But I don't think that's the best way to think about
>it, I would like to see the Material as more of a black box (to abstract
>away things like multi-pass, multi-stage, system-dependent stuff etc.).
>
>> Anyway, I'm accustomed to the model used in other scenegraphs where
>> you can assign individual material properties (not just complete
>> materials) to any node,
>
>Hm, which one might that be... ;)
>
>> and those properties affect all nodes beneath
>> them. So the material properties closest to the geometry leaf nodes
>> are the ones used, unless some override method is in use. I always
>> found this method of assigning materials very easy to use, so I was
>> wondering if there is a similar method in OpenSG that I just don't
>> know about or if there is a plan to make one.
>
>There isn't one right now. I've been thinking about adding one, but I
>haven't come up with a satisfying approach yet.
>
>The obvious way would be to just have a set of active chunks during
>traversal and add/subtract from the set when encountering either a Node
>with Materials or a StateGroup (or whatever you want to call it). There
>could be some performance issues with keeping track of all those groups,
>but I guess with enough optimization that wouldn't be a big issue.
>Sorting becomes more expensive as you can't use a single key any more,
>but you could probably fix that by sorting only on a few critical Chunks
>(GLScry is showing some interesting results for different state
>changes), or using pipeline buffers for state change minimization.
>
>What most people want this for is changing the rendering characteristics
>of a part of the graph (doh!). But in general, especially if shaders are
>involved, these changes cannot be mapped 1:1 to OpenSG state.
>Transparency is the most common one, but with multiple passes wireframe
>or color changes become interesting, too. As long as you're happy to
>manage all the details yourself (which I know you probably are), you can
>manage, but as a general solution for people who just want to have it I
>don't see this as a good path.
>
>My current idea is to have manipulators that can be put at certain nodes
>that are collected during traversal and are available to the
>shaders/materials to deal with. Implementation has not started on this,
>though.
>
> Dirk
>
>
>--
>-- Dirk Reiners OpenSG Forum [EMAIL PROTECTED]
>-- The OpenSG Open Source Scenegraph: http://www.opensg.org
>-- Join the list at http://lists.sf.net/lists/listinfo/opensg-users
>
>
>
>
>--__--__--
>
>_______________________________________________
>Opensg-users mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/opensg-users
>
>
>End of Opensg-users Digest
>
= = = = = = = = = = = = = = = = = = = =
致
礼！
Jiejie Zhu
[EMAIL PROTECTED]
2005-04-27
�������������������������������������������^甸��X���'���u��M�e禾ê�(~���^介h�gФ�Z�嫌jG��+,夯奚┹z{^��0�v��矧y�旃�hJ��q�h����O��
/燠M9��$��^i支u�(�* ��碴�{ ���!z翻⑺Z�哚囤���0螟�r� s�(��
�鲡��.�鞑�'�Na� ������������������������������������^���呵��f��)��+-:�Р畋觎�X�断膦贶y�+�镧z�m�����b槽����q玷�����+-躇b藏�~�瑗殪���
