Hello Brian,
Thanks and sorry for the confusion and I hope this helps explain my thought
process.
Yes, I understand your thought process much better now, but I still
think that using an invalid option (i.e. --image with an .osg file) gets
you nowhere. It will just exit with an error at the earliest possible
time, so it does not test any significant code paths.
Instead, why not create an .osg file with a simple quad and a texture,
and then try to load that. That will reproduce what 'osgviewer --image'
does, but by loading an .osg file (without --image). You could create
such an .osg file by modifying osgviewer.cpp so that it calls
osgDB::writeNodeFile(...) right before calling viewer.run(), and then
running the resulting executable with '--image someImage.jpg'.
Another test could be to try a very simple OSG file (axes.osg from
OpenSceneGraph-Data, for example, has no texture and very simple geometry).
It does seem weird to me that you get problems with the osg plugin, but
not the image plugins. The osg plugin has no external dependencies, and
as long as it loads it should work fine. Other platforms use it without
problems.
I'm almost inclined to suspect (gasp!) a compiler bug / other
compiler-related problem. Perhaps cygwin's version of gcc/g++ has a bug
that OSG is running into? Are there multiple versions you could try out?
Also, I don't understand why I'm the only one to respond to you in this
thread. I have never compiled OSG in cygwin, nor do I use cygwin if I
have the choice. Surely someone else on this list has more experience
than me in this environment and could help you better?
Anyways, I hope I'm at least helping you go forward a bit...
J-S
--
______________________________________________________
Jean-Sebastien Guay [EMAIL PROTECTED]
http://www.cm-labs.com/
http://whitestar02.webhop.org/
_______________________________________________
osg-users mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
