residual calculation needs to be done when boundary degrees of freedom have not been removed (since their removal takes place in the Mult step in order to introduce the proper restoring force). Whereas jacobian calculation needs to always work from the boundary aware operators (these are sparse matrices)
The other thing to note here is that this seems contrary to a matrix free design. While true it is common practive to assemble linear terms since they are cheap. We still never assemble the non linear matrix form.
In order to maintain memory efficienty I have implimented a matrix free SchurComplement operator as well as an operator which uses a few iterations of GMRES to approxinate the inverse of any general operator.
hypre is required for parallel MFEM usage and hypre requires MPI. Both of these have been added as dependecies. MPI is added as a dependency in the same manner as boost (i.e. with an install script which checks for system installation and provides a way that the user can initiate a system installation). The MPI install script has been incorporated with the mk script. Hypre has been added as a subproject.
In general we may want to enforce that a config file is explicitly loaded before any access is requested. However, there are times when this is non ideal behavior. We introduce a compile time flag (CONFIG_HARSH, and CONFIG_WARN). If config hars is defined then a runtime error will be thrown if a config value is requested before the config file has been loaded. If Config warn is defined (and config harsh is not) then a warning will be printed, otherwise nothing will happen. If either warn or nothing is defined this means that the default values defined in the get methods will be used.
Note that the meson build system has had an option added -Dconfig_error_handling=["none", "warn", "harsh"] (default="none") which can be used to manage these compile time options. In general release builds should have this disabled while debug builts should have it set to harsh.
