[vtk-developers] vtkMPIController::Initialize() and wrapped languages

[Moreland, Kenneth]

Why not just call MPI_Initialized?  That would be a lot more reliable than
having the user manage a flag.

-Ken

   ****      Kenneth Moreland
    ***      Sandia National Laboratories
***********  
*** *** ***  email: kmorel at sandia.gov
**  ***  **  phone: (505) 844-8919
    ***      web:   http://www.cs.unm.edu/~kmorel




On 5/17/11 2:04 PM, "Shead, Timothy" <tshead at sandia.gov> wrote:

>Gang:
>
>As it's coded today, vtkMPIController::Initialize() (the no-argument
>version) assumes that MPI_Init() has been called elsewhere; this means
>for example that vtkMPIController.Initialize() can only be called from
>the pvtkpython executable - it won't work when called from a vanilla
>python interpreter.  This seems needlessly limiting, and greatly reduces
>the utility of VTK's multi-processing abstraction.  I propose adding an
>"int initializedExternally" parameter to vtkMPIController::Initialize(),
>and defaulting it to "true".  This would provide backwards-compatible
>behavior, allow VTK+MPI use with vanilla interpreters, eliminate the need
>for the pvtkpython executable, and would be consistent with the
>3-argument form of vtkMPIController::Initialize().
>
>Any thoughts?
>
>Cheers,
>Tim
>
>Timothy M. Shead
>Sandia National Laboratories
>1461, Scalable Analysis and Visualization
>
>
>_______________________________________________
>Powered by www.kitware.com
>
>Visit other Kitware open-source projects at
>http://www.kitware.com/opensource/opensource.html
>
>Follow this link to subscribe/unsubscribe:
>http://www.vtk.org/mailman/listinfo/vtk-developers
>
>