------------------------------------------------------------------------
release 6.3

* PDBrestore interface and in-line pdbrestore.bash command are now available

* minimization module (run_minimize.f) adapted for PDBrestore  

* amber99sb lib expanded to include Li & <Mertz  metallic ions.

------------------------------------------------------------------------
release 6.2

* Dual topology in alchemical calculation can now be performed using the
shift potential in lieu of standard soft-core (still available).

* Intramolecular decoupling/recoupling has been implemented in
alchemical trasformation. Intra Coulomb and (optionally) LJ
interactions can now be turned off/on in the alchemical specie 

* Self-adjusing HREM is now availbale. The variable rem_adjust_ratio
is the third (optional) parameter in the PRINT(READ_REM) directive
(see DOI: 10.1021/acs.jctc.3c00867)

* tools scripts section created and documented. For help type 'help_scripts'
after installation

------------------------------------------------------------------------
release 6.1

* Introduced installation procedure in the etc directory
  (see ORACRC file). All scripts and tools are made available
  via installation.

* Improper torsions are now included in REM_SEGMENT  

------------------------------------------------------------------------
release 6.0.1

* no longer in beta state

* COM-COM distance harmonic potential for ligand-protein    

3* PrimaDORAC output for assigning missing parameters

------------------------------------------------------------------------
release 6.0

MAJOR CHANGES FROM 5.4.1:

* OpenMP/MPI multi-level implementation (still in beta state)

* Old force routines entirely rewritten and renamed. 

*  setup of the fftw interface. pme_code and pme_init translated to 
  fortran90 and renamed.   

*  Obsolete group scaling in constant P simulation eliminated. 

------------------------------------------------------------------------
release 5.4.1

** THIS IS THE LAST 5.X RELEASE BEFORE REINTEGRATION INTO TRUNK OF
BRANCH "OPEN_MP_PP", WHICH WILL MAKE RELEASE 6 **


MAJOR CHANGES FROM 5.4.0:

- only minor changes; as branch rel_5.4 is broken, a complete
  traceback is impossible


------------------------------------------------------------------------
release 5.4.0

MAJOR CHANGES FROM 5.3.3:

(`svn log -v ^/ORAC/branches/rel_5.4 -r1961:2492' for a full log)


* Ewald Alchemy, seems to work fine for ethanol. Comments are in
mts_furier_alchemy 

* Added scalingxy of the pressure in NPT ensemble for membrane simulation

* Added torsion of the kind cos(4*x+60) for membrane simulations;
  60 phase (n=2,4) added for membranes both for serial and rem simulation

* Added batteries in REM



OTHER CHANGES:

* tools: 
  	added makefiles to recent tool dirs; 
	enforced naming standards for dirs; 
	removed temp and obsolete files from repository

* tests:
	added a very basic Makefile to perform all tests

* tests/*:
	added 'mpiexec' command line format for MPICH2 1.4.1 (use
	"-machinefile" option and no 'mpdboot')

* New version of steer_path. Driver coordinates can be turned off.

* Jarzynski tests done with Intel 11.1 (they differ slightly from gfortran)

* src/mtsmd.f:
  Added velocity reset of infinite masses atoms for SMD with
  dynamical freezing.

* Added Simone's Qtcluster

* Added Simone's program MBAR  in tools/mbar_simone

* Added  RESP script in tools/scripts/symmetrize_ESP_charges

* src/fptors.f:
  Fixed rattling of coordinates when overflow occurs due to 
  EXACTLY planar arrangments. This happens when PDB confs are prepared 
  with programs such as GhemPiant or avogadro. 

* Added very fast bar program for REM (BENNET Z estimate from
  contiguous replica) in tools/SI_bar/src/strings.mod

* Added Bactereochlorophyll-pheophytin FF parameters (Ceccarelli et al.)

* Parameters set for organic molecules and amber99SB.prm FF with
string atom types.

* Oscillatory SGE simulation now working (always_accept
  true). Documentation and tests yet to come.

* Now print weights in xyz with REM always_accept

* rem.f90 now fixed:

   Work (always_accept) are now printed correctly on both 1.xyz and work file 
   Work (always_accept) are also printed in REM_DIAGNOSTIC file. 
   Frequency of REM_DIAGNOSTIC printing can now be controlled by end-user.
   
   N.B. in .DBG output last replica is never printed as low replica
   takes all the decision. However, work at target_replica is correct

   N.B. target_replica works with the default (1). User defined
   target_replica never tested (beta version).

* SGE:

  Cyclic SGE transferred to SGE. Only serial version tested; Beta
  version of parallel cyclic SGE (untested) 

  changed acc ratio printout;

  Tested on DIALA.pdb
  Added print_always_accept command for printout controls on target_replica

* tools/xyz2pdb:

  print out fragment coordinates with respect to the centroid of the system
  instead of absolute (MD) coordinates. 


* src/read_inout.f

  Do not apply PBC in xyz trajectories unless specified with new
  PLOT option

* src/mtsmd.f:

  Bugs on cocnst fixed (real instead of integer). noxccnst was
  INTEGER*2 instead of *4 yielding unpredictable behaviour.

* sge.f90:
	now compatible with "non-backspace" compilers:
	gfor < 4.6, or AIX xl fortran

* Added save_all_files in REM. Only replica 1 will be written out. 
When save_all_files is active REM *cannot* be restarted.

*    Added tools/voronoi:
     Voronoi polyhedra computation w/ solvent exposure 

*   Added tools/alltrans:
     all-trans protein generator

*  Added tools/build_pro

*  Added tools/gaussn1.0:
    Crooks based multi-comp. gaussian fit

------------------------------------------------------------------------

release 5.3.3

MAJOR CHANGES FROM 5.3.2:

(`svn log -v ^/ORAC/trunk -r1928:1960' for a full log)


* fixed paralell/serial MPI_FINALIZE problem for the `energy_then_die'
  option in mtsmd_avg_inc.f

* jarzynski tests: steer_path.in: BUG FIXED: decaalanine starting
  values for coords now set to values corresponding to starting
  configuration; ending value for end-to-end distance set to value
  consistent with ending value for angle (21.5 Ang, 90 deg)

* initial output with "STEER_PATH" input (Jarzynski) is now clearer


------------------------------------------------------------------------
release 5.3.2

MAJOR CHANGES FROM 5.3.1:

(`svn log -v ^/ORAC/trunk -r1853:1927' for a full log)

* system calls in parallel version:

  Removed unnecessary system calls when preparing parallel environment
  (PARxxxx directories, etc.). Set up code for platforms that strictly
  don't allow system calls (e.g. fermi.cineca.it)

  orac_mpi.f90: (now CPP precompiled, depends on -DNO_SYS_CALLS)
    in mpi_mkdirs:
	  CALL SYSTEM("cp ...") changed to a line-by-line read-write
	  CALL FORT_CSTRING changed with f95 standard TRIM()
	  CHDIR() and GETCWD() still call non-standard intrinsics (gfortran, Intel)	

       if NO_SYS_CALLS is defined: 
	  no "CALL SYSTEM(command)" used:
	  CALL SYSTEM("mkdir ...") changed to "CALL mkdir()" with
	  explicit INTERFACE to C. Needs std. module "ISO_C_BINDING"
	  (gfortran >=4.4, Intel>=10.1, xlf >= 14)
	  chdir() and getcwd() call C procedures through explicit INTERFACE

  Makefile:
	  new targets added, that use -DNO_SYS_CALLS
	  target BGQ_PARALLEL now uses -DNO_SYS_CALLS

* Explicitly SAVE and initialize to ZERO some local variables that so
  require

  Compilers have options for initializing to zero and SAVEing local
  variables between subroutine calls, but better not to rely on them:
  option "-finit-local-zero" with "-fno-automatic" is missing/buggy in
  gfortran < 4.5

   M /ORAC/trunk/src/comp_work.f
   M /ORAC/trunk/src/mts_test.f
   M /ORAC/trunk/src/mtsmd.f
   M /ORAC/trunk/src/steer_print.f

	diffbo,diffbe,diffto now initialized (with DATA) and SAVEd in
	comp_work.f and mts_test.f. This fixes the zero-and-save
	problem for both methods:

         "PLOT STEER_ANALYTIC" (mtsmd <-> comp_work  <-> steer_print) 

       and 
  
         "PLOT STEER" (mts_dump_inc <-> mts_test <-> steer_print)


* meta_tests: `make diffplot' will now display FES plots with gnuplot


------------------------------------------------------------------------
release 5.3.1

MAJOR CHANGES FROM 5.3.0:

(`svn log -v ^/ORAC/trunk -r1719:1852' for a full log)

* bugs/errors fixed:

 NEX was not initialized to 0 in assign_agbnp_types.f
 #idfef -> #ifdef in mtsmd_avg_inc.f
 tests/basic_tests/7c.in now runs in vacuo as declared


* SGE:

 Multi Window Tempering initialization changed (mtsmd.f)
 output format changed (sge.f90)
 DEO/SEO option added, tests changed accordingly
 manual chapter updated

* src/Makefile 
 various improvements
 BGQ* targets added

* tests/*
 all target output will be deleted on error

* compilation and test run reports added 
 file doc/etc/COMPILES and doc/etc/TESTS

------------------------------------------------------------------------
release 5.3.0

MAJOR CHANGES FROM 5.2.1:

- use AGBNP model for implicit solvent in minimizations
- AGBNP tests added to tests/basic_tests



------------------------------------------------------------------------


r1716:

** Start release 5.3 in branch  **



------------------------------------------------------------------------
r1672: 

release 5.2.1

------------------------------------------------------------------------
r1668:

**   merged r1565 through r1668 from trunk into branches/rel_5.2:  

- new MBAR program:

-- tools/mbar: removed

-- tools/MBAR: created, bugs fixed, comments and sample data added;

-- doc/etc/manual:
    MBAR mentioned explicitly 

- tools/hist2d added

- tools/analysis: added `puckering' coord (for sugar rings), with comments
  and sample data


- a few minor changes to Makefiles, etc





------------------------------------------------------------------------
r1557: 

release 5.2.0

MAJOR CHANGES FROM 5.1.1:

- SGE functionality added
- trajectory formats revised, DCD format added
- auxiliary tools added/improved
- rewritten some include files as fortran modules
- compilation targets added 

------------------------------------------------------------------------

5.2.r1540:
	'gfortran_DBG' added to list of targets

5.2.r1538:
   BUG FIXED: file name "META_DATA" garbled

5.2.r1528:
   BUG FIXED: array bounds exceeded in metadynamics

5.2.r1449:
   removed obsolete command `plot_xyz_only' (&INOUT)

------------------------------------------------------------------------

r1443: Start release 5.2 in branch
 
------------------------------------------------------------------------


5.1.1.r1442:
  tools reading PDB/DCD trajectories improved and updated to new formats

5.1.1.r1427:
  Changed format of .xyz and .pdb files (Ensemble no. now printed last)

5.1.1.r1297:
  The writings of pdb and xyz have been revised
  DCD: box side lengths printout added. 

5.1.1.r1061:
  src/Makefile:
   SP6 targets are now fourfold: plain, O4, vector, vector_O4
   now -Dvector is no more default for SP6xx; SP6_xx_vector targets added

5.1.1.r1057:
	added vector version of mts_forpp_rem
        (mts_forpp_rem_vector.f)

5.1.1.r1054:
   prints simulation box data in DCD file

5.1.1.r1015:
   config.H.0: static copy of config.H 
   config.h_dist: obsolete
   config.H.0 dimensions reduced, now requires ~ 42M

5.1.1.r981:
  /tools/analysis/src/output.f90:
   use two digits for CV index in filename

5.1.1.r976:
   print command name in SGE envirnment has been changed (PRINT_DATA
   -> PRINT_WHAM)

5.1.1.r956:
   file <dcd filename>.rem is created with the correct name
   tools/dcd2pdb  improved, added a man for dcd2pdb

5.1.1.r954:
    A bug in opening SGE.set for serial SGE-simulations has been fixed. 
    SGE initial writings  changed
    Diagnostic writings modified in sge.f90

5.1.1.r945:
   mtsmd_avg_inc.f:
     writing of the unscaled potential removed in xyz, pdb and unformatted files

5.1.1.r943:
  sge.f90: a bug in updating rem_pot has been fixed. 
  SGE-related printings have been moved in sge module. 
  SGE commands/variables changed

5.1.1.r876:
  tests of SGE, jarzynski, REM: improved/updated
  now default is gfortran

5.1.1.r855:
  orac_mpi.f90:
  fixed AIX (xlf90) bug: wdir had a trailing NULL

5.1.1.r852:
  mts_test.f:
    (1) a bug in REM environment is fixed. 
    (2) A bug in formatted writings of SGE environment is fixed (which
    appeared only when using gfortran compiler)
  SGE_tests updated

5.1.1.r840:
   print frequency of rem data now given in input
   dcd.f90 (DCD printing) added, along with tool dcd2pdb and examples

5.1.1.r825:
   tests/SGE_tests updated to r810

5.1.1.r810:
	blkdta.f
	  now kdump=10
	  kdump=0 gives error in AIX (sp0201 1 6 00C3E6444C00)

5.1.1.r809:
   sge.f90:   fixed compilation bug (gfortran)

5.1.1.r808:
 mts_test.f:
	splitted rem and simulated tempering test 
	neater headers for REM_DIAGNOSTIC

5.1.1.r792: 
   - All variables, subroutines and functions for simulated tempering
     now put into module sge
   - SGE printout modified

5.1.1.r726:
   PRINT_ENERGY in REM environment now prints rem energies with the
   same frequency as ASCII pdb plot

5.1.1.r716:
   tests/SGE_tests added

5.1.1.r680:
   new reference output files for REM_tests

5.1.1.r677:
   REWRITTEN SOME INCLUDE FILES AS FORTRAN MODULES
   all sources including old inc files modified to use now fortran modules
   old mtsmd_rem_inc moved into rem.f90 mod as rem_exchange subroutine
   old init_rem.f moved into rem.f90 mod

5.1.1.r633: 
	    doc/etc/chapter_replica-exchange-method.tex: corrected
	    	wrong formulas for solute tempering more corrections
	    	still needed for consistence with source program


5.1.1.r605:
     ~/ORAC/trunk/tests/REM_tests/1.in:
	removed spurious bending pot
	OUT_1, OUT_2 updated

5.1.1.r589:
	print REM energies every large time step on output energy-file


5.1.1.r564:
   tools/rmsd now reads PDB files starting at atom > 1 
   tools/analysis: small fixes
   tools/analysis:
    added new routines in cv.f90
    - centroid
    - fit_normal
    - line_fit
    - principal_axis
    - inertia_tensor
    added a new cv
    - line_fit
    tools/analysis: lapack routines in the sources
    tools/analysis:
      pdb.f90:
       fixed: will not hang if pdb ends with no END record
       fixed: will left-adjust atom_name

   
5.1.1.r573:
  src/verify_input.f:
	sets rem_print to an EXACT multiple of the
        MTS-timestep (like rem_step, etc.)

5.1.1.r537:
   crem.inc changed to module rem.f90

5.1.1.r525:
   In /ORAC/trunk/tests/REM_tests/:
   updated reference tests OUT_{1..6} (see rev 388 in mts_test.f)

5.1.1.r522:
   orac-post-out: fixed non-standard regexp


5.1.1.r515:
   changed input format for tools/analysis

5.1.1.r498:
   Added the awk  script 'orac-post-out' to tools


5.1.1.r497:
   changed output format for SGE runs


5.1.1.r480:
   added tools/analysis for post-run computation of several CVs

   SGE: added the feature to change on-the-fly the maximum number of
   estimates to be used in the calculation of the weighted average of
   free energy differences


5.1.1.r459
  &ST environment has been changed into &SGE
  Some output writings have been changed
  the default value of the parameter used to update free energies in SGE simulation has been changed
  SGE: modified tests and documentation
  fixed residue 'dum' in charmm_a10_fixed.tpg

5.1.1.r429:

  Change introduced in the context of simulated tempering. Specifically,
  the running average of the free energy differences is allowed to be
  performed with a fixed number of samples selected by input

5.1.1.r417:
   M /ORAC/trunk/tools/xyz2pdb/src/xyz2pdb.f90
   fixed bug: wrong PDB read format

5.1.1.r415:
  reference files, etc. updated in tests/

5.1.1.r409:

  Changed paths:
   M /ORAC/trunk/tests/REM_tests/Makefile
   M /ORAC/trunk/tests/jarzynski_tests/Makefile
   M /ORAC/trunk/tests/meta_tests/Makefile
   M /ORAC/trunk/tools/bennett/Makefile
   M /ORAC/trunk/tools/metafes/Makefile
   M /ORAC/trunk/tools/xyz2pdb/src/Makefile

  corrected non-POSIX operator ' >& ' in Makefiles

5.1.1.r403

  Now includes new feature of Solute Tempering / Serial
  Gen. Ens. method (merged branch TRY_Riccardo_SoluTemp back into
  trunk)

5.1.1.r386 
  version where steered torsion can be driven from -180 to 360;

5.1.1.r366:
  limit for torsp to 1000

5.1.1.r355:

  src/Makefile:
     better use of MPIFC combined with FC in parallel targets;  
     command-line "FC=" overrides
     better method for guessing FC with mpif90
     all parallel targets changed ; 
     comments on parallel targets added
     new target `gfortran', `PGI_PARALLEL'
     new service target `makedirs'

     tested on AIX 3.1.1 for 'default','IBM','PARALLEL','IBM_PARALLEL'


5.1.1.r345:

  Added tools for extracting .pdb replica trajectories from .xyz,
    and for computing RMSD

5.1.1.r270:

   M /ORAC/trunk/src/Makefile
   D /ORAC/trunk/src/recpr.f
   A /ORAC/trunk/src/recpr.f90 (from /ORAC/trunk/src/recpr.f:265)

   Better management of FORTRAN90. Now all FORTRAN90 files should be
   suffixed .f90

5.1.1.r260: 
   M /ORAC/trunk/src/orac.f
   MPI_FINALIZED moved at the very end


------------------------------------------------------------------------


5.1.1:

        tools/xyx2pdb: reduced big sample files; Makefile's added; doc added
	src/Makefile: vstart.c compilation now depends on Fortran compiler, not on OS
	src/Makefile: fixed IBM* targets, added IBM_PPC*
	tests/REM_tests/kill_orac_rem: fixed, comments added
        Added (print_xyz_only(&INOUT)) input feature, and associated xyz2pdb
	Documentation polished a bit


------------------------------------------------------------------------


5.1.0: Initial release