[ET Trac] [Einstein Toolkit] #2030: Multi-block boundaries leave uninitialized boundary points

[Einstein Toolkit]

#2030: Multi-block boundaries leave uninitialized boundary points
------------------------------------+---------------------------------------
  Reporter:  eschnett               |       Owner:                     
      Type:  defect                 |      Status:  new                
  Priority:  major                  |   Milestone:                     
 Component:  EinsteinToolkit thorn  |     Version:  development version
Resolution:                         |    Keywords:                     
------------------------------------+---------------------------------------

Comment (by eschnett):

 Yes, this is the same problem.

 You can avoid it by using both {{{initial_boundary_condition="scalar"}}}
 and {{{rhs_boundary_condition="scalar"}}}, and by not using any
 symmetries. Any other multi-block case, i.e. either using bitant mode, or
 using extrapolation for the initial boundary condition or radiative for
 the rhs boundary condition, should trigger the issue.

 The second bug compounds the issue, because the initial boundary condition
 is not handled correctly even in the scalar case. I will propose a patch
 soon.

 As you say, the best current solution for this problem is to apply the
 boundary conditions twice, both before and after multi-block
 interpolation. The true solution, which is not supported by Cactus's
 boundary conditions, is to distinguish between faces, edges, and corners,
 and to apply some of these before and the others after multi-block
 interpolation. While this still requires boundary conditions twice in the
 schedule, they will be executed only once per boundary point.

-- 
Ticket URL: <https://trac.einsteintoolkit.org/ticket/2030#comment:2>
Einstein Toolkit <http://einsteintoolkit.org>
The Einstein Toolkit