wine users - Mar 2007 - Java-1.4.2 not installing - video issues.

I took another pass at trying to install this, and failed again with
video issues.
I have Fedora 5 and a nvidia MX-440 video card (To the uninitiated
that's the sort they give away in breakfast cereal boxes - quite old 4x
agp w/64MB)

This time, I ran as windows 2000, ran under winedbg to slow things
slightly. So I uninstalled fedora's highly cooked current video drivers,
installed a debug 2.6.20 kernel and debug flavour of highly cooked
nvidia-legacy drivers (Version 1.07<splash>). Glxinfo got offended at
being asked, and glxgears didn't. Then I found that I had "nofb"
as a
kernel boot option, and removing that sorted things. glxinfo says nice
things, and loads of extensions. glx & glu are version 1.3, opengl is
1.5.3, and glxgears gives 2263 - faster than with the newer drivers.
But it still bombs. Any suggestions on a suitable WINEDEBUG variable
setting to get to the bottom of this video error?

WINEDEBUG=<your bright idea>?

Smile Duane - I got your final two errors in addition to the snipped
errors everyone gets
as you can see. Can anyone suggest sensible traces to


[SNIP OF COMMON ERRORS]
fixme:msi:ACTION_HandleStandardAction unhandled standard action
L"RemoveDuplicateFiles"
fixme:msi:ACTION_HandleStandardAction unhandled standard action
L"RemoveFolders"
fixme:font:WineEngAddFontResourceEx Ignoring flags 10
fixme:win:EnumDisplayDevicesW ((null),0,0x33eaa0,0x00000000), stub!
fixme:win:EnumDisplayDevicesW ((null),1,0x33eaa0,0x00000000), stub!
fixme:d3d:IWineD3DDeviceImpl_GetAvailableTextureMem (0x1d3840) : stub,
simulating 64MB for now, returning 64MB left
fixme:ddraw:IDirectDrawImpl_SetCooperativeLevel
(0x529dc0)->((nil),00001008)
err:d3d_surface:IWineD3DSurfaceImpl_BltOverride Wrong surface type for
BLT override(not on swapchain) !
fixme:d3d_surface:IWineGDISurfaceImpl_Blt Can't handle DDBLT_WAIT flag
right now.
fixme:d3d_surface:IWineD3DSurfaceImpl_UnlockRect unsupported unlocking
to Rendering surface surf@0x521810 usage(WINED3DUSAGE_RENDERTARGET)
err:d3d:state_multisampleaa Multisample antialiasing not supported by gl
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
err:d3d:tex_resultarg WINED3DTSS_RESULTARG not supported yet
fixme:d3d_surface:IWineD3DSurfaceImpl_LoadTexture
>>>>>>>>>>>>>>>>> 501
from glCopyTexImage2D @ surface.c / 1800
fixme:d3d_surface:IWineD3DSurfaceImpl_LoadTexture
>>>>>>>>>>>>>>>>> 501
from glCopyTexImage2D @ surface.c / 1800

This last line line repeated a total of 32 times then

fixme:ddraw:IDirectDrawImpl_SetCooperativeLevel
(0x529dc0)->((nil),00000008)
fixme:d3d:IWineD3DStateBlockImpl_Release Releasing primary stateblock
err:heap:HEAP_ValidateInUseArena Heap 0x110000: in-use arena 0x160ef8
next block has PREV_FREE flag
fixme:msi:ACTION_HandleStandardAction unhandled standard action
L"ScheduleReboot"
Invalid address (0x7b8702ee start_process+0xee) for breakpoint 0,
disabling it
Process of pid=000a has terminated

On Mon, 2007-03-19 at 21:47 +0000, Declan Moriarty
wrote:
> I took another pass at trying to install this, and failed again with
> video issues.
> I have Fedora 5 and a nvidia MX-440 video card (To the uninitiated
> that's the sort they give away in breakfast cereal boxes - quite old 4x
> agp w/64MB)
> 
Hi. This finally went in, although the test page on java.com fails. It
seems they don't know about firefox :-/.

Anyhow, the secret was to note that all the errors were coming through
channel msi, and set WINEDEBUG to msi, and it skated through on the seat
of it's pants.

For the uninitiated, a list of what those channels do might be helpful.
There's 187 of them :-o.