fixing the build system

Splitting out ifstrstream into a header and an implementation.  This
means that much of the code that depends on it can be compiled only
once and the parallel I/O is concentrated into a few files.  To do
this, a number of files that relied on basic_ifstrstream to load the
mpi header had to be modified to manage their own headers.

More fortran removal, and trimming out unused code

compiles, builds and runs, but is very slow

Added MAW to the C++ side, removed a whole bunch more fortran. 

Doesn't build yet, but getting much closer...

Less busted than it was on last check-in, but still won't completely
build.

Changes to remove more of the low level stuff from the fortran side.

Somewhat working cmake build system now. Does not yet build a parallel target. Builds all non-mpi programs.

Converting Sticky over to C++

Added GB module to the C++ side, got rid of it on the fortran side.

fixed some destructor bugs dealing with the TypeContainers

builds

well, it compiles, but still segfaults

removing cruft (atom numbers, do_pot, do_stress) from many modules and
force managers

mostly whitespace issues

Fixed a typo in EAM, cleaned up EAM and sutton chen a bit

Fixed bug is Sutton-Chen force calculation in parallel.

removing MPI responsibilities from the lowest level force routines.  This is
in preparation for migrating these routines (LJ, electrostatic, eam, suttonchen,
gay-berne, sticky) to C++

Fixing a parallel bug in the self-self interaction.

Fixed odd bug in electrostatic for self-term forces with neutral molecules

General bug-fixes and other changes to make particle pots work with
the Helfand Energy correlation function

Fixes for Inversions for use with Amber

Inversion fixes and amber mostly working

Added support for scaled 1-2, 1-3 and 1-4 interactions.

fixed a strange compiler structure alignment problem with the
intel Mac compiler  (structs that cross between C and fortran
appear to need 8-byte boundary alignment on this compiler).

All your base are belong to us!

Changes required for Inversions and Base Atom types.  This will
break OOPSE badly for a few days or so...

Adding infrastructure for Amber force field.

Removed accidental print lines.

Checking in changes for Hefland moment calculations

Changing functional form

MnM interaction now works.

Changes to simparalell to remove MPI stuff.

Many fixes for Charmm-type torsions.

Changes to MetalnonMetal derivatives.

adding MetalNonMetal interactions

removed a stray is_MWM

Adding more missing files from metal-nonmetal interaction.

More changes to MnM.

Added metal-nonmetal forcefield.

Removed debug message from simParallel.

forgot the if (do_stress) condition on the previous commit

SF Lennard-Jones was added for everyones' enjoyment.  The behavior is tethered to the electrostaticSummationMethod keyword.

Massive update to do virials (both atomic and cutoff-group) correctly.
The rigid body constraint contributions had been missing and this was
masked by the use of cutoff groups...

Fixed a bug in NeighborListNeighbors.

Added interface to change number of neighbors in calculating neighbor list.

refined the reaction field dielectric passing

fixed an ordering issue in quadrupole-charge interactions

fixed d / sigma0 confusion

Added the multipole interaction routine.  It doesn't work yet, so it's not included in the build.

fixed forces and torques for GB

damping now works for charge-quadrupole and dipole-dipole

testing GB, removing CM drift in Langevin Dynamics, fixing a memory
leak, adding a visitor

Massive changes for GB code with multiple ellipsoid types (a la
Cleaver's paper).

Also, changes in hydrodynamics code to make HydroProp a somewhat
smarter class (rather than just a struct).

Adding single precision capabilities to the fortran side

electrostatic refinement

Adding spherical boundary conditions to LD integrator

Removed sqrt splines and shape stuff (doesn't work, so why was it in there?).  Changed some of the water samples to use shifted_force.  Probably should set the defaults to use the damped version now that we sped it up.

splined up sticky

Many performance improvements

Added interpolation module.

Fixed memory leak bug where SCList capacity was not reset to zero on destruction.

Removed MixingRule from Globals. Handled by forceField now.

Made some changes to electrostatics for shifted-potential and shifted-force with mono and multipoles

Sutton and Chen should be working now.

remove extra forceFieldOptions from SC_FF

More Sutton-Chen bug fixes.

Sutton-Chen bug fixes. Almost there...

Force options passed to fortran.

changed __FUNCTION__ to __func__ to match C99 standard, and then added
an autoconf test to check for __func__ usability.  Changed some default
compile flags for the Sun architecture

made some minor changes to allow compilation with the portland group
compilers

Wow thats alot o' files. Now passing forceFieldOptions to all of the parsers.

*** empty log message ***

Further fixes for compiler issues.

Added support for fortran force options.

Adding handler for CLAY force field

OptionSectionParser get compiled

included shifted potential self term

cleaned up the charge-charge interactions a bit...

Made preforce calc public in suttonchen.

Build Fixes for sutton-chen.

Sutton-Chen almost done. Just need to fix do_forces to use Sutton-Chen.

added in a 5th order polynomial switching function

removed a poorly commented section

More work on SC.

removed some test code

changing how we deal with summation and screening methods

Added suppport to atypes for MEAM and sutton-chen

still fixing up wolf...

reaction field looks to be working now - for charges and dipoles alike

added charge-dipole reaction field - don't know if it works...

Still had some globals toUpper problems - these changes should fix those...

fix an index mismathcing between c and fortran

merged reaction field with electrostatics.F90

bugtracking gb

Add parts for the GayBerne LJ

Breaky, Breaky: c-fortran interface now expects an array for pot.

More changes to MPI potential calculations.

Free some variables that weren't freed before.

Fixed MPI bug with Row and Column indexing of groupToGtype. We now have two seperate maps groupToGtypeRow and groupToGypeCol. GroupToGtypeCol is a pointer that is set to groupToGtypeRow in the single processor version.

just some random changes when testing

Added define for ifc 7 so derfc is external. Other compilers should treat erfc as intrinsic.

MPI fix for SimHasAtype in simulation module. We needed a seperate receive buffer.

Fixed "dum-dum" where we ignore the skin thickness and hardcode listSkin to be 1.0. We now get listskin from skin. This will get fixed to where we can manually set skin thickness.

Fixed a defaultCutoff bug (HEMES!)

reaction field seems to work now, still need to do some testing...

fixing some summation method issues

it builds now, but there are known issues - particularly with reaction field...

Working on adding WOLF

Fixed bug where gtypeMaxCutoff was not initialized after creation. When maxval(gtypeMaxCutoff) was called, the largest random garbage value was returned from the array.

some changes to activate the coulombicCorrection selector

fix to put back calculation of r in do_prepair

Added nGroupsInRow for mpi build.

bugfix on the grouptype finding algorithm

Some bug hunting

initialized atomTypeMaxCutoff(i) to zero

made some changes for implementing the wolf potential

added some probably nonfunctional get*cut routines

fixing some of the problems in the interactionHash and gtypeCutoff routines

breakage in progress

Breaky Breaky BREAKY breaky breaky

fixes in progress

Changed cutoffs... Segfaults nicely now...

Added subroutine to set cuttoff for Interaction map and function in simulation.F90 to determine if a particular atype is present in a simulation.

More breaking and destruction of force code. Does not build at this point...

Removed balance from the Darkside (files)

fix bug in NPAT and NPrT

Modifications to temper the dipolar strength in the first solvation shell for tap

added gb

OOPSE setup for TAP water.  It's not parametrized, but OOPSE will now let me run it...

Shapes is limping along with a array bounds overwrite (I think...). At least the parser loads the forcefield fine...

Working on the shapes parser...

Changed interface name for SGI Compiler

Added destroy method to clean up memory

no real changes here, maybe added a space by accident

Updates to deallocate object in fortran.

fixing of the quadrupoles.  look!  it's divide by 3 like stone says!

quadrupole interactions should be working with a remaining question of interaction magnitude

Make sure electrostatic_module provides data for reaction_field

fixed three typos

added some samples

using map to replace link list

fixing restraints

adding IndexFinder which is used to select the molecules; Seperate ElectrostaticAtomTypesSectionParser into
ChargeAtomTypesSectionParser and MultipoleAtomTypesSectionParser;remove print dipole option from Dump2XYZ;

removed #include lines for outdated charge and dipole interfaces

electrostatic unification project

adding soft potential to LJ Module

corrected filename collision for EAM on operating systems which are
not case sensitive (aka Mac OS X)

moved Package variable into subdirectory makefiles to speed up
the building process

separating modules and C/Fortran interface subroutines

merging new_design branch into OOPSE-2.0

fix a bug in WATER.cpp when initializing the new Charge AtomType

Fixed a mixing list bug that slowed down the force loop

current status of the debugging process

Breaky Breaky

Shapes looks like it's working

shapes rcut calculator added

forgot to advance lines in the shapeFile

bug fix

more char* -> string conversion

forcefield refactoring for shapes

Changes to Shapes force field reader

fixey fixey the breakey breakey

Added un-busticated fortran files and c/Fortran interfaces

Added newLJtype to lj module.

fixed a fortran name problem

Fortran/C++ interface de-obfuscation project continues

fixed include

fixing some broken fortran stuff

Added correct prototypes for C-fortran interface functions.

Added new files for fortran interface.

working on shapes for portability

move mpiSimulation_module to DarkSide

define DEV_ROOT as relative path

move fortran files around to solve dependency problem

change the #include in source files