J-S, I appreciate the input and insightful discussion. This is good information to have for both new and experienced developers.
Thank you! -Shayne -----Original Message----- From: osg-users-boun...@lists.openscenegraph.org [mailto:osg-users-boun...@lists.openscenegraph.org] On Behalf Of Jean-Sébastien Guay Sent: Friday, January 28, 2011 12:17 PM To: OpenSceneGraph Users Subject: Re: [osg-users] Linker error during execution Hi Shayne, > The setting that you amply described IS what we do. We do it quite often so > that things don't run so slow when we're debugging our app (not OSG). We may > need to run a while before we break into our code for debugging. Running > with OSG debug libs is too slow for this scenario so we use the release libs > instead with the settings set accordingly. Yeah, that explains it. So it really was because my wording of "don't mix debug and release" was not thoroughly accurate... Sorry about that. Still, I really recommend most users don't change the default settings, because then you're totally on your own. Any dependency you use, you'll have to recompile yourself from source, you won't be able to use any precompiled binaries. And for new users, I think it's easier just to say "link debug to debug, release to release and you're fine". See the OP of this thread for example, who was setting /MD in his debug build when OSG was built with /MDD. Plus, just how slow it is in debug is a matter of perspective... It is much slower but if you're not debugging too often then it might not matter. Honestly, you haven't seen anything until you've used boost::persistence in debug. THAT's slow. :-) And the other thing is, if you're not actively debugging, then you can run without attaching the debugger (use Ctrl-F5 instead of F5). Then it's much faster than running with the debugger attached (not quite as fast as release, but much faster). And then you can attach the debugger only when you get to a point where you want to debug the running code. That's how we survive using boost::persistence in debug - the persistence is used in loading/saving files, so we start without the debugger, load our scenario, and then attach the debugger. Loading is therefore much faster than with the debugger attached. But of course that doesn't work if we need to debug the actual loading of a scenario. Then we're screwed :-) Another approach could be to always run in release, and just have another configuration with symbols (and possibly with optimization turned off). Sure, why not. But I digress. To each his own. We use release and debug, with the default settings in debug, and just make sure not to mix the two. I recommend the same for new users, but more experienced users can do whatever they want and there are many different ways to get mostly the same result... It all comes down to knowing what you're doing, and if you don't want to know all the small details then you'd be better keeping the default settings I think. I'll make sure to be more precise next time I talk about not mixing A and B... :-) J-S -- ______________________________________________________ Jean-Sebastien Guay jean-sebastien.g...@cm-labs.com http://www.cm-labs.com/ http://whitestar02.webhop.org/ _______________________________________________ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
