I think so, but to be sure, what is the way you use to run all the tests ?ickby wrote:are you guys able to run all FEM tests successfully with the occ7 branch?
on GUIvejmarie wrote:but to be sure, what is the way you use to run all the tests ?
The branch of vejmarie? To be honest I'ven't tried the unit test on this. Lazy boy I am ... The one I have here was compiled on May 5th, just after the first release of the vagrant file.ickby wrote:are you guys able to run all FEM tests successfully with the occ7 branch?
Yes it is exactly the getTuple problem with wrongly initialized meshes. I've seen that you fighted this already in FemMeshShapeNetgenObject::execute . I see that it works with your fix, but honestly I do not understand why I tried nearly 2 hours to change the Fem::FemMesh initialisation to make the function work without your addition, but now way. IMHO this is a good testing point, we need to make this function work with normal mesh initialization, than all other places with the same problem should be fixed too. Porting your fix everywhere else is not an option, it is a hack at best. But well, I do not understand SMESH very well, it is a badly documented beast. I have no idea how to fix the Fem::Mesh initialisation...vejmarie wrote:Ok, I think I have been able to reproduce the problem.
The PartMesher works for you ? (it is working for me, as well as Fem on MacOS)
You got a crash into VTK with getTuple function ?
If yes, there should be somewhere a Mesh = operation which is performed on a non properly initialized Mesh (probably the one who is getting the value), I had this issue within the Mesher code of FEM. It is crashing within my VM on this currently.
I have some time this afternoon to track this bug.