Hi Js, I think the testing should be sufficient, and to warrent merging the change and to look for wider.
I'm not at my computer right (using my phone) now so will merge tomorrow On 1 Dec 2010 20:18, "Jean-Sébastien Guay" <[email protected]> wrote: Hi Robert, >> The answer though, is it does work, in the pbuffertest at least. Removing >> the makeCurrentImp... No, but at least it confirms that changing my submitted file by this would fix the observed problem that realizeImplementation() did a makeCurrent and then the GraphicsThread couldn't makeCurrent() successfully after that. So it's a partial test, not a complete one. :-) > running osgprerender or osgprerendercubemap with --pbufer should force > the use of wglShareList... I ran osgprerendercubemap --pbuffer, with a breakpoint at PixelBufferWin32.cpp line 664 (the call to wglShareLists()). The breakpoint was hit multiple times, and never caused any error (the OSG_NOTICE line just below). And the visual result (the reflection on the central sphere) looked correct. Is this sufficient testing to be confident the removal of makeCurrentImplementation() in realizeImplementation() is safe? Thanks, J-S -- ______________________________________________________ Jean-Sebastien Guay jea...
_______________________________________________ osg-submissions mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
