-
Bug
-
Resolution: Unresolved
-
P4
-
None
-
5.0
-
x86
-
windows_xp
I have a canvas and the canvas's paint method contains some complex renderings. I am getting all the GraphicsDevices from the GraphicsEnv and getting all the GraphicsConfigs from each of the GraphicsDev. Inside a loop which loops through all the GCs, I am creating a Window and adding the canvas to it. When showing the window, for some of the GCs, the window appears empty and the renderings are not rendered properly to the canvas. I am disposing the window at the end of the loop.
This is reproducible only on WinXP with Tiger (tested with b51 and FCS) and Mustang. Not reproducible on 1.4.2 and hence it is a regression on Tiger. Not reproducible on Solaris / Linux.
I have attached 2 sample tests along with few images (one using Canvas and other drawing directly to a window). Execute the test on WinXP (single monitor or a multi-monitor system). You would see a window with many renderings on it. The window will be disposed and recreated for each of the GCs. If any of the windows appear blank, the bug is reproduced. Sometimes, all the windows would appear properly. So try running the app 2-3 times to reproduce this. Not reproducible with Frames / Dialog.
This is reproducible even when rendering is carried out directly on the window and not on canvas.
###@###.### 2004-12-09 12:41:46 GMT
This is still reproducible on slower machines. I tried this with Mustang-b16 on WinXP and I am able to reproduce this consistantly.
Here is the configuration of the machine where i reproduced this issue:
CPU: Pentium - III
Intel 32bit
Single Processor
Processor Speed: 728 MHz
Operating System:
Windows XP (version 5.1 (Build 2600))
DirectX version: DirectX 8.1 or greater
DirectX display devices:
Device 0:
Resolution (width x height x depth): 1024 x 768 x 32
Video memory (Total, Free): 186691 KB, 186691 KB
Device driver ati2dvag.dll version 6.14.10.6396
GDI display devices:
Primary: Name = \\.\DISPLAY1, Description = RADEON 9600 SERIES
I reproduced this on a couple of other similar machines also, having onboard Intel graphics controller (D86845GL) and NVidia GeForce4 MX 440 respectively.
###@###.### 2004-12-20 13:14:41 GMT
This is reproducible only on WinXP with Tiger (tested with b51 and FCS) and Mustang. Not reproducible on 1.4.2 and hence it is a regression on Tiger. Not reproducible on Solaris / Linux.
I have attached 2 sample tests along with few images (one using Canvas and other drawing directly to a window). Execute the test on WinXP (single monitor or a multi-monitor system). You would see a window with many renderings on it. The window will be disposed and recreated for each of the GCs. If any of the windows appear blank, the bug is reproduced. Sometimes, all the windows would appear properly. So try running the app 2-3 times to reproduce this. Not reproducible with Frames / Dialog.
This is reproducible even when rendering is carried out directly on the window and not on canvas.
###@###.### 2004-12-09 12:41:46 GMT
This is still reproducible on slower machines. I tried this with Mustang-b16 on WinXP and I am able to reproduce this consistantly.
Here is the configuration of the machine where i reproduced this issue:
CPU: Pentium - III
Intel 32bit
Single Processor
Processor Speed: 728 MHz
Operating System:
Windows XP (version 5.1 (Build 2600))
DirectX version: DirectX 8.1 or greater
DirectX display devices:
Device 0:
Resolution (width x height x depth): 1024 x 768 x 32
Video memory (Total, Free): 186691 KB, 186691 KB
Device driver ati2dvag.dll version 6.14.10.6396
GDI display devices:
Primary: Name = \\.\DISPLAY1, Description = RADEON 9600 SERIES
I reproduced this on a couple of other similar machines also, having onboard Intel graphics controller (D86845GL) and NVidia GeForce4 MX 440 respectively.
###@###.### 2004-12-20 13:14:41 GMT