The win32 pbuffer implementation returned an error unless both the
WGL_ARB_pbuffer and the WGL_ARB_render_texture functions were present.
This was too restrictive, as a pbuffer can usefully be created without
render-to-texture, e.g. for use with glReadPixels. The osg 1.2/Producer
pbuffers worked without RTT, and osgUtil::RenderStage has all the code to
handle both RTT and non-RTT pbuffers, doing a read and copy in the
latter case.
With these changes I have successfully tested the osgprerender example
on a graphics card which supports RTT, and one which doesn't. Plus
tested in my own application.
In order to aid diagnostics I have also added more function status
return checks, and associated error messages. I have included the win32
error text in all error messages output. And there were some errors
with multi-threaded handling of "bind to texture" and a temporary window
context which I have corrected.
These is one (pre-existing) problem with multi-threaded use of pbuffers
in osgViewer & osgprerender, which I have not been able to fix. A win32
device context (HDC) can only be destroyed from the thread that created
it. The pbuffers for pre-render cameras are created in
osgUtil::RenderStage::runCameraSetUp, from the draw thread. But
closeImplementation is normally invoked from the destructor in the main
application thread. With the additional error messages I have added,
osgprerender will now output a couple of warnings from
osgViewer::PixelBufferWin32::closeImplementation() at exit, after
running multi-threaded on windows. I think that is a good thing, to
highlight the problem. I looked into fixing it in osgViewer::Renderer &
osgUtil::RenderStage, but it was too involved for me. My own
application requirements are only single-threaded.
Unrelated fix - an uninitialised variable in
osg::GraphicsThread::FlushDeletedGLObjectsOperation().
"
"This is a fix for the issue reported by Anders a week ago (see \u201c[osg-users] BUG?: mouse coordinate changes after window move\u201d discussion thread on Sept. 20). The issue was that the initial implementation added a few months back was not converting the window coordinates to client-area coordinates resulting in a slight offset each time a decorated window was moved (caused by the window border). This was also causing windows to move out of their assigned screen."
and
"Attached is a fix for the taskbar repaint issue that occurs when a graphics window is toggled from full-screen mode to windowed mode (as identified by Gert van Maren a couple of weeks ago).
Also included is a fix derived from the \u201cEvents from the past\u201d discussion thread that took place on July 11."
Stephan: "attached you'll find some modifications to the GraphicsWindow-class and
their platform-dependant implementations.
The problem:
setWindowRectangle and setWindowDecoration do not update the
traits-object, so, if you call setWindowRectangle on a
not-realized-window it will open with another size when realized later.
getWindowRectangle reports possible wrong sizes if setWindowRectangle
called before.
My solution:
split the implementation in two parts:
GraphicsWindow::setWindowRectangle will update its traits-object and
call afterwards the virtual method setWindowRectangleImplementation
(which is implemented by the derived platformspecific classess). For
setWindowDecoration I am useing a similar mechanism.
I hope you'll find the submission useful, the Win32 and X11 changes are
not tested but should work."
Changes to this made by Robert are call of resized in setWindowRectangle
instead of setting of Traits, and use of a bool return type.
