Sorry, the debug_buffer() is a new feature - from when I thought it was my patch. ;-)
Have you run this on debug mode? Reading the system from disk initialized the equation systems, and then you add another system? Conceptually this should work, but... That's what the asserts are there for! -Ben On Dec 12, 2012, at 3:25 PM, "Jens Lohne Eftang" <[email protected]> wrote: > On 12/12/2012 04:11 PM, Kirk, Benjamin (JSC-EG311) wrote: >> On Dec 12, 2012, at 3:07 PM, Roy Stogner <[email protected]> wrote: >> >>> On Wed, 12 Dec 2012, Jens Lohne Eftang wrote: >>> >>>> The system that is read from file has three variables. Now, when I >>>> iterate of the nodes of mesh, and do >>>> >>>> index_0 = node->dof_number(0, i, 0); >>>> index_1 = node->dof_number(0,0,0); >>>> >>>> I alwas get index_0==index_1 also when i = 1,2. >>> I assume you're running with the latest svn trunk or git master? >>> >>> If so, would you try with svn revision 6517? It's possible that Ben's >>> subsequent VariableGroup optimizations to the dof numbering regressed >>> something. >> Definite possibility. And I assume this is something that worked previously? >> >> If you are running master, could you add >> >> node->debug_buffer(); >> >> after the erroneous call you've identified and post the results? > I was using 6524, rolled back to 6517 but it didn't change anything. > node->debug_buffer() caused a compilation error ... > > Best, > Jens > >> Thanks, >> >> -Ben > ------------------------------------------------------------------------------ LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers http://p.sf.net/sfu/logmein_12329d2d _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
