> From: Dirk Reiners [mailto:dirk.rein...@gmail.com] > Sent: Wednesday, March 24, 2010 00:59 > To: opensg-users@lists.sourceforge.net > Subject: Re: [Opensg-users] VRML and exit() woes > > Carsten Neumann wrote: > > > > I'm not really happy that it behaves so badly when objects outlive the > > FieldContainerFactory and some of the other internal infrastructure that > > gets taken down in osgExit and about the only positive thing about it is > > that it provides a fairly high incentive to not leak OpenSG objects when > > closing an application ;-/ > > I'm not very happy about that one, either. Can we just agree that if an FC is > destroyed after osgExit the FCFactory just prints the Warning and ignores the > request? At this point you're screwed anyway, and trying to clean up just > leads > to pain... > > Comments?
>From my perspective: I'm writing a DLL that (I hope) can be safely loaded, unloaded, and reloaded multiple times from one process, meaning multiple calls into osgInit() and osgExit(). If something isn't clean at exit, I need to know about it. Though I was somewhat surprised that the *RefPtr destructors were sensitive to whether osgExit() had already been called, the fix was easy. I was zeroing in on that solution, but hadn't quite gotten there yet -- distracted by the random nature of the failures down the pipe. Now that I know, I'm OK with it. -- Ted ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Opensg-users mailing list Opensg-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensg-users
