Works correctly here
Linux Debian
NVidia GF4 Ti 4200 drivers 53.36
no DRI
Jdk 1.5.0 beta2 + J3d 1.3.2 b4
jdk 1.4.2_04 + J3d 1.3.1
Alessandro
--- Russell East <[EMAIL PROTECTED]> escreveu:
---------------------------------
Quite likely it is the ATI driver, but what can I do? I am unableto
downgrade to FC1. I suppose I have a workaround - remove DRIsupport from
Xorg.conf. Everything in Java3D works in that case.
Attached is a sample app (DirectRenderIssue.java) that demos a problemwhich
*may* be related to what I'm seeing with the main applicationthat we're
developing.
I wonder if anyone else can reproduce the symptoms I see?
The application constructs a relatively simple TriangleFanArray anduses a
TextureLoader to partially display an image (smile.png) drapedover the fan.
I've tried running this app with jdk 1.4.2_05 + J3D1.3.1, and 1.5.0 beta2 + J3D
1.3.2, with identical results. WithoutDRI, the app works correctly, with DRI,
the app works correctly withoutthe texture, and incorrectly with the texture.
After compiling the app, you can run it either as:
java DirectRenderIssue t
which displays the texture (you can see what this looks like in theattached
smiley3d.jpg) or
java DirectRenderIssue w
which displays the wire-frame (see wireframe3d.jpg)
Actually, the app works fine with or without DRI *to begin with*.
If DRI is loaded, the problem occurs if you take focus away from theapplication
window, then give the window focus once more. In thiscase, I see what is shown
in the attached wrong3d.jpg). I have "focusfollows mouse" set for Windows
behaviour, but similar things occur for"click to focus".
When the app starts, it looks correct. I move the mouse into anotherwindow -
which removes focus from the app, and the app still looks ok. But when I move
the mouse back into the app window, it immediatelylooks wrong, as shown.
As stated, I don't know if this is the same issue that I'm seeing asreported
earlier, but it certainly seems like a bad one.
-- Russell
Paul Byrne wrote:
I suspect this problem is being caused by the ATI drivers. We have hada number
of problems with ATI's OpenGL and would love to work with themto resolve the
issues, anyone from ATI out there.....
Java 3D is a large system and makes extensive and fairly optimized useof
OpenGL so it's probably not the fact that you are trying to usetextures but the
way in which those textures are being used on thespecific geometry type.
I have had a lot of problems installing the ATI drivers on my Suse 9.1box, I
ended up making kernel modifications, recompiling the kernel,recompiling the
drivers and I still don't get reliable results.
My recommendation, go back to Fedora 1 until ATI updates the drivers
Rgds
Paul
Russell East wrote:
I triedusing "MahJongg Solitaire 3D"
http://www.reto-schoelly.de/mahjongg3d/downloads.html
It is a c++ application, and just a quick look at the code suggests it
is using textures. It is working fine for me, when DRI is enabled -
not a bad game actually. :-)
Attached are some screen shots of what is happening in my Java3D-based
application:
Wireframe.jpg - this shows the wire-frame I am using. it works with
or without DRI
WithoutDRI.jpg - shows the app with textures, working fine when DRI
is disabled
WithDRI.jpg - shows the mess when DRI is enabled.
I'm a bit inexperienced with 3D stuff in general. Does the last
screenshot mean anything?
-- Russell
Alessandro Borges wrote:
Russel,
glxgears doesnot uses textures.
Try another 3D application with textures. I guess this bug is notJava3D
related.
Try Quake, Doom or Xith WebStart demos available at
http://www.xith.org/demos/com.xith3d.test.php
Some upgrades are real downgrades...
Alessandro
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA3D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
