Hi Paul, > If you could all take a few moments to build and run this with your 2.8.x > applications, I'd appreciate any feedback. Thanks!
I have just built 2.8.5-rc3 and it all went smoothly this time. I executed the runexamples.bat and everything worked fine except for : osgviewer glsl_simple.osg osgviewer glsl_julia.osg osgviewer glsl_mandlebrot.osg Which give black textures, so black shapes on the top row for glsl_simple.osg, and black quads for glsl_julia.osg and glsl_mandlebrot.osg. Curiously glsl_confetti.osg works fine. The osgshaders example also works as do other shader based examples. I have tried all these examples with OSG svn/trunk and they work fine. The files in question haven't change since they were checked in to OpenSceneGraph-Data so these won't be a variable in this regression. I'm afraid I don't recall doing this test for 2.8.4 so I couldn't guess at when the regression happened. To explore where the .osg plugin might be a variable I've just done an osgconv glsl_julia.osg test.osg for both 2.8.5-rc3 and svn/trunk and both version generated a test.osg file that different from glsl_julia.osg but both identical with each other... So reading and writing seems to be fine. The test.osg works fine under svn/trunk but still fails under 2.8.5-rc3 confirms that the changes between glsl_julia.osg and test.osg is effectively a cosmetic change. I have to head offline now - children to read stories too :-) Sorry to report some bad news again. Robert. Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
