Description of files in this directory
======================================

$Header$

This file contains (typically) one file for each model, containing
a Fortran-callable subroutine to compute the 4-metric $g_{ab}$ and
inverse 4-metric $g^{ab}$:
      subroutine Exact__model_name_coords(
     $     x, y, z, t,
     $     gdtt, gdtx, gdty, gdtz, 
     $     gdxx, gdyy, gdzz, gdxy, gdyz, gdzx,
     $     gutt, gutx, guty, gutz, 
     $     guxx, guyy, guzz, guxy, guyz, guzx,
     $     psi, Tmunu_flag)
      CCTK_REAL x, y, z, t
      CCTK_REAL gdtt, gdtx, gdty, gdtz, 
     $       gdxx, gdyy, gdzz, gdxy, gdyz, gdzx,
     $       gutt, gutx, guty, gutz, 
     $       guxx, guyy, guzz, guxy, guyz, guzx
      CCTK_REAL psi
      LOGICAL   Tmunu_flag
	...
      RETURN
      END

 psi  is normally unused, but can be set by the subroutine to explicitly
specify the Cactus static conformal factor.  At present only the
"Minkowski_gauge_wave" model does this

The subroutine should set  Tmunu_flag  to .true. if this model needs
to set the stress-energy tensor (i.e. if it's non-vacuum and/or it has
a cosmological constant, or .false. if it doesn't (i.e. if it's vacuum
and has no cosmological constant).



FIXME: Wouldn't it be cleaner to just compute $g_{ab}$,
       and use a 4x4 matrix inversion to get the other one?