Hi Ed -- Actually, your email below makes it sound like you'll have an easier time of it that I did, since it sounds like your existing render code doesn't make any assumptions about OpenGL state. If that's the case, then I'd think that simply wrapping your OSG cull/render traversal code with glPushAttrib/glPopAttrib should address the state problem. -Paul
> -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Ed > Sent: Thursday, November 09, 2006 10:37 AM > To: osg users > Subject: Re: [osg-users] How difficult is it to integrate OSG > code into anexisting OpenGL application? > > Let me further describe my situation. I think I may not have > as many issues as first thought. > > First, I mispoke. The existing OpenGL application is > actually a library. > > The ONLY reason I am investigating the possibility of putting > OSG code into this existing library, is that there is a > history, or traceability, of the library, based on several > years of government spending. If I can simply add another > interface function to add the ability to render scene type C, > then the current user base should, in theory, be able to drop > the new lib into their apps, and call my function and > voila.... Scene Type C. > > Granted I could just brute force straight opengl into the new > render code, but I think OSG will do the job better and faster. > > At the point where I take control of the opengl state with my > new OpenGL render function, I don't know that state control > will be returned to the original lib code. In other words, > the lib code establishes the X-Window, Render Context, Set > OpenGL State, etc. Then I render my OSG scene then quit. In > this case, can I simply wrap the OSG code, as Paul mentioned, > with gl(Push/Pop)Attrib? > > Does this additional information change any indications that > it is possible, but maybe more straight forward? > > Ed > > > Robert Osfield wrote: > > HI Ed, > > > > Others users have embedded the OSG into existing OpenGL > applications > > so it is possible. It it isn't all straight forward, mainly due to > > the issues of managing OpenGL state. > > > > I'm hoping Paul Martz will volunteer his experiences, as he > did just > > this at his previous job ;-) > > > > Robert. > > > > On 11/9/06, Ed <[EMAIL PROTECTED]> wrote: > >> How difficult is it to integrate OSG code into an existing OpenGL > >> application? > >> > >> Let's say I have an application which renders 2 different types of > >> scene. The app reads in object data, material information > etc. etc. > >> then enters a framing loop and calls the main Render function. > >> Something like this: > >> > >> InitializeXWindowGLContextEtc() > >> InitializeObjects() > >> InitializeRest() > >> while (still need to render) > >> RenderSceneA() > >> > >> Or maybe the app calls RenderSceneB() > >> > >> while (still need to render) > >> RenderSceneB() > >> > >> I want to add my own Render function which uses OSG > >> > >> while (still need to render) > >> RenderSceneC_OSG() > >> > >> Is this possible? Difficult? Are there examples? > >> > >> Ed > >> > >> > >> _______________________________________________ > >> osg-users mailing list > >> [email protected] > >> http://openscenegraph.net/mailman/listinfo/osg-users > >> http://www.openscenegraph.org/ > >> > > _______________________________________________ > > osg-users mailing list > > [email protected] > > http://openscenegraph.net/mailman/listinfo/osg-users > > http://www.openscenegraph.org/ > > > > _______________________________________________ > osg-users mailing list > [email protected] > http://openscenegraph.net/mailman/listinfo/osg-users > http://www.openscenegraph.org/ _______________________________________________ osg-users mailing list [email protected] http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
