[CIG-SHORT] pylith time stepping

[Brad Aagaard]

Christopher DiCaprio wrote:
> One more question:
> I notice that in Formulation.initialize there are calls to 
> integrator.initialize, constraint.initialize and output.open. All of 
> them pass numTimeSteps as an argument. As far as I can tell numTimeSteps 
> is not actually used by integrator.initialize and constraint.initialize 
> and the writer functions just need to know if there is more than one 
> time step. So it looks like I'm not breaking anything 
> by effectively making numTimeSteps incorrect. Is this right ?

Currently, there is only one DataWriter (DataWriterVTK). Other writers 
may need to know how many time steps are expected. If they do need to 
know the number of time steps at the beginning of the simulation, it 
isn't clear how this should be handled for variable time step 
simulations. In the meantime, things should be okay.

Brad