
These changes are always immediately updated in the "molden5.x.tar.Z"
source file, however the platform dependent executables in the
directory "/pub/molgraph/molden/bin" on the ftp site are not always
in tune with the latest changes, check the date on the executable.

------------------------------------------------------------------------
Molden5.9.6:

Thu Apr  4

- The first structure of a .mol2 file was not represented
  correctly in the multi-mol display, specifically the name field.
  This was corrected.

- A number of problems with docking were addressed:

  - With the docking example for 4dfr the hydrogens on residue 34:B
    (asparagine) were not correctly assigned. This was corrected.

  - when reading the ligand, while NOT using the folder button,
    the ligand was actually not processed. Corrected this.

  - after the docking finished and using the still open file reading
    window and after updating the directory, the directory seemed to be empty.
    This was corrected.


Wed mrt 27

- removed a bug introduced on 25 march

Mon Mar 25

- The "pronation State" entry from the residue menu popup for Histidine
  residues allowed the user to choice between HID, HIE and HIP states.
  Using this option from the Z-matrix editor for proteins, showed faulty
  connections. Fixed it.

Fri Mar 08

- Added checking for fingerprints in OpenBabel

Thu Feb 28

- Reading .sdf files and adding them to let us say to a protein with
  the 'add' button from the read file window, showed up the first
  structure form the .sdf file twice. This was corrected.

Mon Feb 25

- Removed a bug that would stop the ambfor Molecular Dynamics executable
  (ambmd) from working correctly

Wed Feb 20

molden5.9.4

- added the ability to add hydrogens to .sdf multi-mol files,
  by incorporating this functionality availabe from obabel.

- added the writing of "vector_on": 0, to the molden written
  .json files. This disables the directionality constraints.
  The .json files are used by the 'pharmer' package to specify
  a pharmacophore.
  With this change pharmer now works as expected:

  - if you increase the radii on the features you get more hits.

  Thanks to Dr. David Koes for looking in to this.

- also added keyword to the pharmer dbsearch facility that makes
  pharmer return one conformer per hit.

Mon Feb  4

molden5.9.3

- expanded the multi-mol reading capability of molden
  to also display 2D molecules for multi mol2 files. 
  Previously, 2D molecules displays were only supported for .sdf files. 

- now these multi-mol .sdf/.mol2 files can be editted by hitting the 
  "Delete" key. These editted list can be saved as .mol2/.sdf files.
  This implicates sdf->mol2 and mol2->sdf conversions, the first also
  includes generating charges for .mol2 files.

- Now the descriptor fields in an .sdf file can also be saved.

Thu Dec 20

molden5.9.2

- Corrected reading of .plt files created by open3dqsar.
  The .plt files were checked with the maestro program,
  that naturally supports these files.

- Improved capturing errors of missing programs when using open3dqsar.

- Updated assigning charges to the mol files from an .sdf file.
  In turned out that the oxygen in an 6- or 5- membered ring were not
  assigned a mol2 atom type. This in turn led to charges not being assigned.

Tue Dec 11

molden5.9.1

- Updated the NwChem interface via the ecce.out file.
  The ecce.out of the Hydrogen/Carbon atoms are now working.
  The orbitals look OK.
  The orbitals of a z-matrix optimisation of a z-matrix with dummys 
  now look ok.
  Reading the acce.out now also the beta orbitals are read and a beta
  density matrix is created.
  When editing a molin.nw, a new molin.nw file is written to /tmp
  and copied to a newly created directory with the name supplied a jobname.

- the popup box with help on the z-matrix editor variable fields now
  only pops up once. They were getting in the way of an unimpeded functioning
  of the z-matrix.

- Adding a button to undo the last applied changes to the z-matrix.
  Also added the possibility to cancel selection of atoms to be connected
  the new line, by hitting the 'escape' key.

- Added a new button to the file selector window. Besided replace and add there
  is now a additional button "show", which shows the text content of the
  file.

- Added new functionality to the file selector: 

  a) Now the buttons "Home", "End" "Page Up" and "Page Down" are now 
     functional. 

  b) Also you can scroll through the window with scroll button of the mouse.

  c) You can now hit for instance the "a" key and the window will show the
     first file beginning with the letter "a".

  d) The "return" button now also works as a selector of files.

- Finally got the full binary distribution for molden5.9 on ubuntu
  to run all external programs (pharmer,obabel,open3dqsar: 
  molden5.9.full.ubuntu.64.tar.gz).

Wed Nov 21

- Discovered an error in submitting gaussian jobs.
  Molden would crash in such a case.
  This error has been there at least since version molden5.7 !
  In addition the correct gaussian environment variable was not
  passed onto the rungau subroutine.
  Fixed it.

- also added processing for G16 subnission.
  
molden5.9

Tue Nov 1

- Added a molden interface to open3DQSAR

- Use environment variable OPEN3DQSAR_PATH to point to the directory
  where the open3dqsar executable is located.
  Proper operation of open3dqsar relies on the presence of openbabel,
  so this has to be installed as well.

- You will find the new option "Opem3DQSAR" under the "Surfaces icon"
  (the 7th icon from above middle column).
  When clicked, this will pop up the QSAR interface window.
  Here the minimum inputs are:

  1) a .sdf file with the pre-aligned structures
  (ligands or small molecules), you will find ref.sdf in the test directory,

  and 

  2) a text file with activities of the structures in the .sdf file,
  you will find activity.txt in the test directory.

  Now you can click "Run" and a the display window will show:

  1) the aligned structures from the .sdf file
  2) and four surfaces: 
     green and yellow are associated with the van der Waals
     light blue and red are associated with the electrostatic field

  You can hit the "Show Log" button to display the output of open3dqsar.
  the "Show File" window will pop up. You can select to see output of
  open3dqsar commands with the "Go To ->" button.

  The "Edit" button pops up the "open3dqsar edit commands" window, that
  shows the list of commands being sent to open3dqsar. When you select
  an entry from this list, the entry will be copied to the edit box.
  You can use the cursor keys to edit the command. When save is clicked
  the edited command is saved to the list of open3dqsar commands to be
  executed when the "Run Commands" button is clicked.

- Added the -Y commandline flag, telling molden to use a different
  form of fractional to cartesian conversion. One that is in line with
  vmd.


molden5.8.2

Tue Aug  14
- Docking was introduced in molden5.8, but this code did not always
  yield the same results on the docking tests in the test/dock directory. 
  This in contrast to results on my development machine. 
  molden5.8.2 fixes some of these issues.

molden5.8.1

Fri Aug  3

- Made some adjustments to the C code (xwin.c) to make it compile
  better on MacOSX

Thu Jul 26

- This version of molden, the electron density files for proteins are
  obtained from a different site. The old one: the Electron Density
  Server in Sweden is being phased out. Now we have to obtain electron
  density files from www.ebi.ac.uk/pdbe. However the old site used
  amongst others the BRICK/OMAP format of density files. The new site uses
  the CCP4 format of density files. These are stream files, generated
  by C programs. Fortran 2003 can read stream files, but some older
  versions of fortran can not (f77,g77). These version of compilers
  will generate errors.

- Heikki Tuononen reported a bug in molden5.8 with writing xyz files from
  the Z-matrix editor.

- Also on recommendation of Heikki Tuononen we incorporated tables of 
  bond radii based Pyykkö's and Atsumi's values.
  http://www.chem.helsinki.fi/~pyykko/Radii09.pdf

Molden5.8


Wed May 23 2018

- Three main functionalities were added in the molden5.8 release:

  - molecular docking
  - the updating of the nwchem functionality
  - the writing of lammps input files from an ambfor .xyz file
    by the ambfor executable by usage of the -l keyword

  Molecular Docking

  The docking functionality is a hybrid scheme of the FlexX docking
  and a force field (amber+gaff) optimisation.
  The same scheme as meployed by FlexX was used to generate the placement
  of the base fragment. So first distance tables are generated to store
  all the distances between the interaction points in the protein.
  The protein has to be supplied in a pdb format with hydrogens present.
  Molden can be used to generate such a pdb file. For now, the orientation
  of for example OH groups (Serine,Threonine,tyrosine) have to be predefined.
  The user can define a sphere around a to be picked point, which encloses
  all the interaction points which are relevant for the docking.
  These interaction point can be either H(ydrogen Bond)_Acc(eptor),
  H(ydrogen Bond)_Don(onor) or aromatic/ch(2,3). The user selected
  amino acid residues are then assigned interaction points.
  For each combination of interaction groups, distances are collected and 
  stored in a distance table.
  Next a ligand  must be supplied in ambfor .xyz format (molden own force 
  field format), also charges must have been assigned for proper force
  field scoring.
  Next interaction points are assigned to the ligand. From these interaction
  triples are generated. This list of triples is searched against the distance
  tables to find matching protein interaction triples, 
  these are called base fragment
  For the ligand two to three interaction points are generate for each 
  OH group.
  Initially the ligand is divident up into fragments. Fragments are typically
  connected by single SP3-SP3 or SP2-SP3 bonds. These fragments are stored 
  internally by molden. This fragmentation is used to generate different 
  conformations of the ligand by rotating around the dihedral angles of the
  single bonds separating the fragments. The bonds that are not strict
  SP2-SP3 bonds are kept in order to carry out a complete scan of the energy
  with respect to these dihedrals. For regular SP3-SP3 bonds the standard
  angles of the staggered conformations of ethane are employed.
  When the debug button is pressed in the "Docking Start" window a file 
  called "conformers.mol2" is written.
   The orientation of the base fragment is very important for the following
  steps in the docking procedure.  That is why the base fragment placement is 
  optimised to obtain a near perfect placement. If we had a base fragment
  consisting of all Hydrogen bond acceptors and donors, the van der waals
  part of the force field scoring is reduced, so that the orientation is
  dominated by electrostatic part of the force field scoring.
  If we were dealing with a base fragment with one pair of Hydrogen bond 
  acceptors or donors and one hydrophobic interaction point, we rotate
  around the axis formed by pair to find the energetically optimal angle for 
  the remaining hydrophobic interaction point with respect to the pair.
  Because the hydrophobic interaction between hydrophobic interaction points
  are much smaller than between a acceptor and donor pair (round about
  seven times weaker) and are much less well defined in space, the optimal
  placement is rarely the one found in the original base placement, but
  will be impacted more by the rest of the ligands placement.
    The main scheme assumes a base fragment that has 
  interaction points belonging to the same fragment. This is by far the 
  fastest docking scheme. Here we map the base fragment onto the earlier
  created conformers (Debug: mconformers.mol2), followed by an optimisation
  of the conformers along the dihedrals of the non-pure single bonds
  that molden stored earlier.
    However when a ligand triplet with two or more Hydrogen bond acceptor/donor
  interaction points can not be formed from the same fragment, a different
  docking scheme is employed. Now we use the whole of the conformations of the
  ligand to sample ligand points, but to make the algoritm more efficient we 
  only keep the unique ones (unique by location of the interaction points 
  origin) and store the conformations it belongs to with these interaction 
  points. After optimisation of the base fragments, the ligand poses are 
  generated from the combined confomer information of the three interaction
  points. If all three interaction points are from the same conformer, 
  this conformer is mapped onto the base fragment, followed by an optimisation 
  of the conformers along the dihedrals of the non-pure single bonds
  that molden stored earlier.
  The force field scoring is stored into the multi-mol result dock.mol2
  file. 
  The source distribution of molden5.8 has a 'dock' subdirectory of the test
  directory, with three different docking: 

   - ligand dtst.xyz (part of MTX) into the 4dfr protein.
     (fastest: all three interaction points on the same fragment)

   - ligand est.xyz into the 1a52 protein.
     Slower because the default interaction point separation of 10 angstrom
     has to be increased.

   - ligand rsb.xyz (part of RAL(oxifen)) into the 1err protein.
     (slowest: NOT all three interaction points on the same fragment)
  
  This is the launching of a prototype that will undergo extensive testing 
  in the coming months.

  CAVEATS/Missing functionality:

  - the orientation of the for example OH groups (Serine,Threonine,tyrosine) 
    have to be predefined in the PDB file.

  - Structural and Non-structural waters are ignored in the docking.

  - Docking into an active site with metals is not yest supported.

  - No ring conformations are being searched



  Updated nwchem functionality

  Molden NWCHEM functionality has been updated. Previous versions of
  molden could not display orbitals or density from NWCHEM because the 
  required information was missing from the NWCHEM output. However
  NWCHEM is bundled with the ecce program, this program can display
  the orbitals and other properties by incorporating ecce_print
  command into the nwchem input file. Molden can now also read these
  ecce.out files. This opens up a lot of functionality compared to 
  that of only the nwchem output file information being available.

  Some caveats: 

                The scf convergence information written to the ecce.out
                is not correct. At least in the somewhat old combination
                of ecce and nwchem we used (nwchem 6.3)
  
                Molden can not correctly calculate esp charges when dummys
                are being used

  Beside the reading of ecce files, molden can now also submit nwchem jobs
  via the z-matrix editor (use the NWCHEM button in the format section
  followed by "Submit Job"). The command nwchem must be defined for this to 
  work correctly.

  WRITING OF LAMMPS INPUT FILES

  Large-scale Atomic/Molecular Massively Paralell Simulator (LAMMPS)
  input files can be generated from an Ambfor .xyz file, by Molden
  own force field helper programs ambfor by typing:

  ambfor -s -l test.xyz

  After this a file called 'gaff.lammps' will have been written.

  MISCELLANEOUS

  - The reading of molden format files (such as vib.molden) has been repaired

  - When reading multi-mol file such as a .mol2 file contaning multiple
    structures, followed by reading an other multi-mol file when still having
    the multi-mol window open, would lead to adding the entries to the entries
    already present from the previuos read. This has been fixed.
  
  
Wed May 17 2017

- This is mainly an update to deal with the in my view 'crazily' increased
  strictness of the gcc 5.4 compiler. This compiler spews out a ridiculous
  number of warning messages. To cope with this, proper function proto typing
  was applied to xwin.c. (Language purists always give me the creeps !)


Molden5.7


- Gaussian 16 was not recognised, which led to improper processing
  of Guassian16 output/log files. Updated rdgaus.f to deal with this.


Thu Jan 26 2017

- Adjusted the dropdown list to print "<- to top" at the end of long
  lists, previously  it would say "more ..".

- gmolden: adding caps to proteins from the 'protein view' of the 
           z-matrix editor would effectively work, however the screen
           was not updated. Fixed

- The ambfor .xyz files offer the possibility of keeping some atoms fixed.
  However, molden/gmolden did not transfer this info to the ambfor program.
  Also, the ambfor program did not deal with these atoms correctly. Fixed

Wed Sep 21

- I have updated the auxiliary files in the plush directory, needed by
  molden to add hydrogens to PDB ligands (HETATM 's).
  I plan to do this each half year.

- during this update problems appeared with specific PDB entries, of which
  a number have been resolved:

    Ligand 3TV in pdb entry 4wr7:

      HIS 119 complexes with Zn, but had its H pointing towards positive Zn


    Ligand 3VU in pdb entry 4wz4:
    Ligand 3Y6 in pdb entry 4x6t:

      COOH group deprotonated, charge should be -1, molden said 0
      This caused by a boron with 4 connections in the ligand.
      We created a isBH4() function, if it is BH4, nht is not decreased !


    Ligand 3XT in pdb entry 4x6h:

      Ligands 3XT and I37 have overlapping coordinates and therefor
      are interconnected ! So the number of connections do NOT correspond to
      that of the plush 3XT file !!! Causing molden not to add the hydrogens
      and print:

      Decreasing formal charge by -38 units !!!!!!!

      We now check for overlapping ligands and alert the user.


    Ligand 3XZ in pdb entry 4tys:

      Molden has a routine for adding a Hydrogen to a tertiary nitrogen
      on a ligand. However in this case it puts a Hydrogen on a tertiary 
      nitrogen in a flat 5 membered ring, so the N should have a sp2


    Ligand 3YO in pdb entry 4rx5:
 
     The plush file for 3YO had one Hydrogen specified in the pdb file:

     HETATM 2322  H27 3YO A 706     -34.242  29.736  -8.804  1.00 15.26 

     This is very unusual.
     Just let molden add the other.

    Ligand 1WD in pdb entry 4yw8:

     1WD coordiates with an Mn, Lys 244 also coordinates with the Mn, 
     there one H of the lys should be removed.  We already have this 
     functionality for HIS residues, now we added it for lys.
     We had to increase the van der Waals radius of Mn a tiny bit to 
     get this right.

Fri Aug 12

- Added the possibility of editable text fields to be longer than the visible
  string. Now for example file names can be 1024 charachters long. 
  The visible (and editable) part of the string is a movable window on top
  of the whole string. When you get to the edges of the string (using the 
  left and right arrow keys) the windows position will move up or down one
  position with respect to the whole string.

- Concerns only protein PDB files:
  
  Molden already had the ability to pop up a window with commands, when
  clicking with the second mouse button on a residue on screen.
  Now I have added a "select range" command to this window. This will let
  the user select a range of amino acids, starting from the amino acid
  just clicked on, moving up or down in residue number, following the tip of
  the red arrow on screen. The selection is ended by hitting the escape key.

- Particulary the first option required a lot of changes to the code and
  there are bound to be some problems to be ironed out !
  Hence the version of molden.

Molden5.6

Sun jul 24

- The Mac OSX version had the problem that it did not load the OMAP
  electron density file for proteins. Fixed now. (not in the .dmg file)

Wed jul 20

- I found out that the displaying of the electron density ontop of a 
  protein (pdb file), only really worked for cells with three angles
  of 90 degrees. I now have adapted the program to also deal with maps
  where one angle is not 90 (this at least covers the most abundant
  spacegroup in the PDB databank). I still have to work on triclinic
  spacegroups.

- Fixed problem with clipping protein surfaces, now the clipping is 
  independent of the amount of zoom. Also the clipping did not cover 
  the whole protein surface. Fixed it.

- Fixed the annoying problem that a window which you swiped your mouse over
  to reach the window of interest pops up help on a button, making the
  window you are currently working on unresponsive.
  (Does NOT yet work for the windows version of gmolden).

Fri Jun 24

- removed a bug introduced with the reading file in memory functionality.
  When optimising a structure in the background, after reading a pdb file
  the reading routines were still in memfil mode, this caused problems.
  Fixed it.

- when reading files that contain multiple structures when opening the
  z-matrix editor the z-matrix of the first structure was always displayed.
  Now the z-matrix of the selected structure is displayed.

Thu Jun 23

- Provided binaray distributions containing besides 
  molden,gmolden,ambfor,ambdm,surf the pharmer and obabel executables.

  pharmer has to be ported to windows, which was no easy feat.

  You can find the binary distributions under the following URL's:

  Linux ubuntu 32 bits: 

  ftp://ftp.cmbi.umcn.nl/pub/molgraph/molden/bin/Linux/molden5.6.full.ubuntu.32.tar.gz

  Linux ubuntu 64 bits: 

  ftp://ftp.cmbi.umcn.nl/pub/molgraph/molden/bin/Linux/molden5.6.full.ubuntu.64.tar.gz

  Windows (32 bits):

  ftp://ftp.cmbi.umcn.nl/pub/molgraph/molden/bin/Windows/molden_native_windows_full.rar

  MacOSX (64 bits)*:

  ftp://ftp.cmbi.umcn.nl/pub/molgraph/molden/bin/MacOSX/molden5.6_MacOSX_full.tar.gz

  
  *) you may need to install 'wget' to be able to download pdb files from the
     file browser. Install home brew for Mac and install wget:
     brew install wget
  *) This version was installed on El Capitan with the latest gfortran
     version it contains all necessary libraries and may therefor also run
     on older MacOSX iterations.

- Fixes a bug in calculating the electron density from the EDS protein 
  crystallograpy server files.

- Fixes a bug that would result in molden not displaying all H positions correct
  when generating rotamers from a PDB file generated by the MOE program.

- Uses fork instead of vfork on MacOSX: now a molecule will optimise in the 
  background also on Mac OSX.

Wed May 18

- fixed a bug introduced in molden5.4. 
  With a zmatrix created for an pdb file, the dihedral for rotation
  of the OH/SH was missing from the protein view of the z-matrix
  Also in the full atom mode of the z-matrix the corresponding dihedral
  was not behaving properly.

Wed May 11

- Added an interface to the fingerprint searching capabilities of openbabel.
  Openbabel version 2.3.0 or higher is required (URL: openbabel.org).
  When reading in a .sdf file from the file selector 
  (NOT from the command line!), molden will check if the obabel executable
  is available (in /usr/local/bin). If it is there it will try to generate
  a 2D (SVG) representation for each molecule in the .sdf file and display
  them in the multi-mol listbox. 
  This is accomplished via a background process. When it is finished, it will
  write out the 2D representations with the original file name of the .sdf
  file, but now with extension .svgs
  The next time you read in this .sdf file molden will first look for the
  .svgs file and read it in.
  
  When the obabel executable is found molden will add two extra buttons
  to the "Multi-mol file" window:

  . "create fingerprint index" will use obabel to create a fast searching index
    of this .sdf file (with the same file name but with the .fs extension).

  . Clicking "Search" will add a few query boxes to the "Multi-mol file" 
    window. The query file must be one of the formats supported by openbabel.
    (.mol file will work). The query is searched against the fast searching 
    index of the .sdf file. Entries from the original .sdf file with a tanimoto
    coefficient greater than the value supplied by the user (0.6 default), will
    be stored int the results .sdf file (default res.sdf).
  
- Added an interface to the pharmacophore searching capabilities of pharmer.
  Pharmer is an open source program developed by David Koes.
  Pharmer powers ZINCPharmer, an online pharmacophore search engine for a 
  multi-conformer library of the ZINC database. 

  Because pharmer is a nightmare to install, precompiled versions 
  (32 and 64 bits) will be supplied together with the binary distribution 
  of molden5.6 for ubuntu (openbabel will also be included).

  When the pharmer executable is found (default /usr/local/bin) the option
  "Pharmacophore Editor" will become available from the "Surface" icon
  (7th from the top of the central icon column).

  This will run pharmer to create the annotation of your molecule.
  That is it will create a list of features for you to choose from
  that you think is essential for interaction with the target protein.

  From the pharmacophore Editor window you will be able to add features by
  clicking the "Create" button, after you have first clicked the feature in
  molecule display. A selected feature stores the associated coordinates,
  you can then select other features available in the annotation for these 
  coordinates. (sometimes multiple features are available with the same 
  coordinates eg: aromatic and hydrophobic). You can also adjust the radius,
  which defines a tolerance sphere around the feature.

  When you are done with selecting features you can save the selection
  by clicking the "Save" button. You can also restore previous selections
  by clicking the "Open" button. Selections are stored in the json format.

  When you click "Search" will open the "Pharmacophore Search Database"
  window. This allows you to select a .sdf database file (Open button)
  You can then create a Pharmer Database (index) by clicking
  "Create DB Index". This will create a directory with the same directory 
  name as the original filename of the .sdf file, but with the .db extension.

  Finally, clicking the "Search" button will search the Pharmer database
  with the stored .json query and return the results in the .sdf file specified
  in the "Results:" query box (default: phrm.sdf)


Molden5.5

Wed Apr  6

- The routine that parses strings from files read in memory expected
  strings no longer than 137 characters. Sometimes .sdf files contain
  much longer lines. This would make molden crash. Fixed it.

- PDB files written by MOE would contain an label unknown to molden ('HC")
  Molden can now handle it.

Thu Mar  8

- Fixed the option ADD file for .mol2 file added to a PDB file.

- Added the ADD file option to ADD .sdf file to a PDB file
  Both these options are handy for looking at docking results, stored in
  a .mol2 file or a .sdf file.

- when having multiple structures, selecting one in the upper listbox
  of the "structures" windows and ONE in the LOWER listbox; 
  the LOWER listbox could give problems with the scrolling. Solved by
  making the listbox 4 entries deep in stead of 3.

Thu Mar  3

- In the PDB HETATM records there are sugar like compounds that can make 
  connections to other sugar HETATM records, often the connecting atom would
  be an C-OH which after connecting would miss one hydrogen.
  Fixed this by adding a NOTH field is the HETATM files in the plush
  directory. The plush files contain information on the number of Hydrogens
  attached to the non hydrogen atoms.

- Some sugars connect to an ASN residue, often there would still be two 
  hydrogens on the C=ONH2 part plus the connection to the HETATM sugar.
  Fixed this.

Fri Feb 26

- A number of bug fixes:

  In the "Residue command" windows, typing eg: 1-40:A would also
  erroneously activate waters. Fixed

  PDB entry 471d: choosing optimise waters would lead to a neverending loop.
  Fixed.

  PDB entry 3jaa: would have wrong secundary structure connections on DNA

  Reading a PDB file from the file manager, now reads the file into memory.
  We hoped this would spead up the parsing of PDB files, however the speed up
  was dissappointing.

Fri Feb  5

- bug fix:

  When reading in two pdb structures, do backbone on the second,
  delete active structure. When reading in the next pdb sturcture 
  and doing backbone, the helix, beta and coil were not visible.

- A lot of minor bug fixes pertaining to PDB HETATM structures parsing:

  pdb entry 4phg: HETATM 2UH makes connection to CYS, to connecting C atom
                  had no Hydrogens attached.

  pdb entry 3c0g,4w9l,4w9k,4w9o,4w9p and probably more:
                  There two HETATM molecules with the same name and 
                  same number. They were treated as one big molecule.
                  Now we also read the chain identifier, and now they are
                  treated as different residues.
  
  pdb entry 3qn6, 4ylk and probably more:
                  HETATM molecule 3QN makes covalent bond with ASN259, 
                  forming a petide bond: the hydrogen connected to the N 
                  was pointing in the wrong direction. The fix makes the H ly
                  in a plane with N,C,C=O

 HETATM PRX the hydrogens were not correctly added.

Mon Jan 25

- Removed a few small bugs that would cause problems with reading
  fdat files, and gamess-uk output.

Fri Jan 22

- Adapted the molden code so that it can now read files in to memory.
  More correctly: in can do the old fortran reading and reading of files
  in memory, side by side.
  Currently, this is only done for files with the .sdf extension.
  This was done, to address the fact that reading large .sdf files was
  extremely slow and a max. of 100 entries would be stored.
  The reading of .sdf files is now completely dynamic, you can read any
  .sdf files no matter how many entries it contains. That is to say, untill
  your memory runs out :-)
  This only works from within the file manager, NOT from the command line.

  NOTE: This is all very new and errors are to be expected in the beginning.

- Support for reading in charges from the pubchem.sdf files.


End Molden5.4

Tue Jan  5

- Fixed a bug in the routine reading mopac .aux files. This could cause
  some atoms to have exponents with a zero value.

Fri Dec 11

- added a born shell script register_extension.sh, to be run from the makefile
  or by hand, which links gmolden to a number of file extensions and add
  gmolden to dash. Currently the makefile only activates it for the ubuntu
  linux distro. Probably also works on debian, havent tested that though.

- added the option clean and install to the makefile:
  
  make install

Tue Dec 1

- for protein structures the first HETATM molecule had NO charges applied
  when selecting Hydrogen icon -> "Optimise Hydrogens" 

- added new keywords to run molden with a command file:
  
  keywrd: WRXYZ=1-12

  0       xyz file
  1       mol file
  2       msf file
  3       tinker xyz file
  4          ''    ''
  5       pdb file
  6       chemx format
  7       msi format
  8       molden format file
  9       mol2 file
  11      ambfor xyz file
  12      mopac xyz input file

  This command allows to read in any by molden supported formats and
  write out the specified format in file 'mol.xyz'

  Sample molden command file:

  title line
  file=any_input_file wrxyz=11
  
  This will write a ambfor xyz file. As input you could specify a pdb file.
  The ambfor xyz file, will have hydrogens added, incomplete residues 
  corrected and charges and hydrogens added to ligands and water molecules.

  keyword: WRMOPXYZ
    
    Write an mopxyz input file.

  keyword: CHARGE=-1

    Specifies a total charge for the molecule to be written in conjuction 
    with the WRXYZ keyword.

  keyword: WRTAUTO=1-12

   0       xyz file
   1       mol file
   2       msf file
   3       tinker xyz file
   4          ''    ''
   5       pdb file
   6       chemx format
   7       msi format
   8       molden format file
   9       mol2 file
   11      ambfor xyz file
   12      mopac xyz input file

   Generates enol tautomers from keto forms and writes them as tau01.xyz,
   tauto02.xyz etc.

   QMPSA calculates the polar surface area

Wed Sep 23
- Robert Gyepes reported a bug that would ignore the fill setting
  in the space filling (3D) orbitals rendering in the Xwindows mode.
  Fixed this.

Tue Sep 15
- fixed a bug when adding hydrogens to HETATM residues from the PDB,
  using the plush directory. The first atom in the HETATM residue in the
  plush directory would have its hydrogens omitted.

Sat Sep 12
- fixed a bug that would not let you add characters to a string after
  you did a backspace on the start of a string

Fri Sep  4
- repaired a minor bug that would not label Hetatm molecules correctly
  when no water was present in the pdb file.

Wed Sep  2

- Now when a water box is added, water and added ions can be displayed
  independendly from the "HetAtm" button.
- Names of HETATM molecules from pdb files are passed on to ambfor .xyz and
  .arc files and are displayed at such after ready these .xyz files into
  molden

Thu Aug 20

- Restored proper visualisation of water in the water box added.
- When creating a crystal from scratch., by first creating a molecule
  with the z-matrix editor and the chosing "pack cell" from the "cell"
  icon, chosing "edit cell parameters" (choose P-1). The second molecule
  related by the inversion symmetry was not shown. Now it is.

Tue Jul 21 

- when reading in a pdb file, followed by optimise hydrogen positions,
  then starting a force field optimisation with water box, connections
  of water would be wrong. Fixed!

Fri Jul  3

- removed a bug that corrupted shadows
- the thousends of water on an amb(for/md) xyz file, are now represented
  as just one HOH in the HETATM drop down list.

Thu Jul  2

- removed an other bug from pdb file writing routine for non-pdb structures.
  This bug would cause a non-pdb structure to write a pdb file with only
  a HEADER line.

- When having added a water box and subsequent optimisation with ambfor,
  the resulting _opt.xyz would crash ambfor when using it to do further
  optimisations. 

- When viewing a ambfor xyz file with a water box added, selecting "backbone"
  the box would disappear and oddly only reappear when switching on the first 
  HETATM residue. Fixed.
  Also the color of the box would change to the selected color for the first 
  HETATM residue.


Tue Jun 30

- When writing PDB files with labels from the original PDB file
  this would lead to wrong element labels after z-matrix editing of the
  structure. Now the PDB file writing routine no longer uses these labels
  after the z-matrix editor was used.

- The newly introduced labels on a residue basis would run out of its array 
  bounds, especially when many waters were present, causing a segment 
  violation. Fixed.

Molden5.4

- beginning of June 2015: EBI changed their MSD web interface for
  there PDBE interface. As a result, the CML files of HETATM molcules
  in the PDB were no longer available.
  I therefor decided to create a local repository for Molden to get
  information on hydrogens to add on HETATM molecules.
  Now a 'plush' directory containing this information is added.

- Added the possibility to clip surfaces. This is especialy interesting
  for molecular surface of proteins.
  After creating a molecular surface, you click the "palette icon' ->
  surface -> clip

Molden5.3

Sat May 30

- After Copying a hetatm molecule to a separate structure: 
  and then saving as pdb file, produced a file containing only
  the HEADER line: fixed

- The 'plotplane' keywords in the opengl version of molden (gmolden)
  did not update the screen, giving the wrong impression that nothing had 
  happened. Now the screen does update after plotplane commands such as
  PHASE.

- In pdb entry 1fj3, Zn has a spurious connection to a H2o, of which
  only half bond (the zn part) is displayed, the O side is not visible
  on "contact" of Zn. This is fixed, now also this water is displayed.

Fri May  22
 
- Sometime back I introduced adding an extra hydrogen on tertiary
  nitrogens, when adding hydrogens to ligands. Now this will only be done 
  if the three connected atoms to nitrogen do not lie in a plane.

- Added the Label On/Off entry to the popup menu for residues and hetatm
  molecules. This lets labels be switched on/off on a residue basis
  rather than on the whole molecule.

Thu May  7 


- Adjusted rotamer generation for the different form of Histidine
  (HIE,HID and HIP).

May 1

- Added a alternative way to generate rotamers. 

  Previously rotamers could only be changed after generating a z-matrix 
  for a protein. This was because rotamers change was accomplised changing 
  certain dihedral angles in the protein z-matrix.
  The alternative method, aligns the stored rotamers of a particular type
  of residue to the actual residue in the protein. To be more precise, it 
  aligns the backbone atoms (Calpha, N, COO and Cbeta).
  The rotamer editing features are now also available without a z-matrix
  having been built.
  As a side effect the option to edit residues have been moved down in
  the small residue popup window.

- Aligning and Combining two protein strcutures, now conserves displayed 
  residues in the backbone mode.

- The option restore incomplete residues now also makes use of the new
  rotamer generation scheme.

------------------------------------------------------------------------
Molden5.2.4

Wed Apr 15

- Changed a minor detail in the source code for surf, that stops it from
  core dumping wtih the latest versions of Ubuntu.

Fri Mar 20

- added option to make complete the incomplete protein residues.
  It is available from the 'H' icon -> Optimise H postions.

- Previous, adding atoms to one of the twenty known protein residues
  in the z-matrix editor these atoms would be flagged "NewRes".
  However this was coded in such a way that while changing a amino acid
  for another one this was also done, which is not correct.
  Secondly the string for water in the hetatm arrays was overwritten.
    This has been solved. In addition instead of "NewRes" now a string like
  "gly.mod" is used to denoted the added atoms.

Fri Mar 13

- With combine/Align structures, now a new structure is created, 
  leaving the two original structures alone.

- Increased the speed of the optimsing water subroutine substantially

Wed Mar  5

- When loading PDB files -> zmat editor -> out if zmat editor -> add H's to
  HETATM molecule -> zmat editor: the residue variable where not updated and
  actually displayed the wrong variable. Fixed !!

- when animating residue variables in the z-matrix, in backbone mode:
  only the animating residue was displayed, the rest disappeared (only molden
  ,gmolden was OK). Fixed !!

- made changes to the makefile where it calls the makefile in the ambfor
  directory: no longer are the FFLAGS and CFLAGS exported.
  When they are exported, they can cause ambfor to thow a lot of NaN
  (Not a Number: the computer words can not hold the unexceptionally 
   large or small values of the varaible).
  Until this is completely solved, use a gradient norm not smaller than 0.03

Wed Mar  4

- when writing a pdb file from Molden, the orginal labels of the HETATM records
  would not be written, causing the add Hydrogens to HETATM not to work with
  the newly written file.

- In the case of a pdb file read into Molden and subsequently opening the
  z-matrix editor a z-matrix is made for the protein. When after leaving the
  z-matrix editor hydrogens are added to HETATM molecules (via clicking with
  the second mouse button on the molecule on screen, and selecting 
  "add hydrogens"), a new z-matrix has to be created including these newly
  added hydrogens. This fix creates a new z-matrix in such a case.
  

Fri Feb 27 

- When doing a charge calculation on a hetatm (PDB) molecule, only
  -3,-2,-1,0,1,2,3 were offered to pick from. Now the oprion "other" is added.

- In the "molden file select" window an extra button has been added "Search PDB"
  Clicking it will bring up a window that lets you search the PDB for keywords.
  For example "quorum sensing", a list of matching pdb codes will be displayed.
  When a pdb code is clicked a few lines from the description are displayed.
  When "Load PDB code" is clicked the pdb structure will be downloaded and 
  displayed.

- a few help strings have been added.

- In the z-matrix when you click a field with the second mouse button.
  A popup window lets select "animate", which would subsequently animate the
  result of variable being increased/decreased.
  The new way will open a window in which the variable can be increase or
  decreased by one frame as well as animated. 
  During that the window is displayed all other windows will become
  unresponsive, except the main drawing window, in which you will still
  be able to rotate/translate your molecule.

- When opening the z-matrix editor on a PDB file, adding hydrogens to
  HETATM molecules would nolonger work after the mutation of a residue.
  This has been fixed.

Wed Jan 28

- Molden text editing input handling routines have been updated.
  They now support dynamic editing: each entry field upon activation
  now shows a cursor to indicate its position. On that position you can now
  type characters or delete characters. Whereas previously you could only
  add or delete at the end of the string.

- make molden now works again (no more unsatisfied external reference ogsel)..

- fixed a small bug with the gif snapshot routine. The top line of the
  gif image was black or had garbage.

Molden5.2.3

Mon Jan 19

- The optimise hydrogen functionality is now split into to separate parts
  which can be switched on and off in a small window.
- GLN/ASN flips will now only be done on complete residues
- Molden now checks before adding hydrogens to residues if the hydrogen
  positions is not taken by a bonded atom (covalently bound hetatm).
- NO Hydrogens are added any longer to PO4 groups from the hetatm section 
  of a PDB file.
- NO Hydrogens are added any longer to COO (carboxyl) groups from the 
  hetatm section of a PDB file.

Molden5.2.2

Fri Nov 28

- When aligning two molecules the center of rotation was akward; fixed it.

- Molden can now also read gaussian input files with cartesian geometry
  specification

- Molden can now read in z-matrices with only two lines. Use commandline
  flag -d to get old behaviour.

Molden5.2.1

Mon Nov  3 2014

- From the Molden interface you can now create .avi movies.
  This functionality depend on a third party package avconv.
  (on linux: sudo apt-get install libav-tools).
  Screen snap shots are made with the .bmp format.

- Fixed bug it the saving bmp function. This bug would only show it self
  if the screen width was bigger than the screen height.

- Mopac2012 aux files are now better supported. 
  In the submit mopac job window, the keyword AUX is added by default.
  If you use an older version of mopac, you might want to remove this
  keyword again. (see xwin.c look for 'AUX').
  Added method PM7.

Molden5.2

Wed Oct 22 2014

- Molden can now calculate the Electron Localization Function (ELF).
  
- Molden can now write .WFN style of wavefunction files.
  It does not work with the use of 5D or 7F functions.

- When using Molden with gamess-us input files that contain cartesian 
  coordinates, Molden now recognises the use of coordinates specified in
  angstroms. So now the connectivity is correctly displayed in such cases.

- with commandlineflag -J [0,1,2] molden can now write other screen dump 
  formats than gif [0] such as rgb [1] and bmp [2]. Gif is the default.
  The high definition formats rgb and bmp are specially useful for
  creating movies, such as with the ffmpeg package:

  ffmpeg -i mol%03d.bmp -target dvd -s 800x800 movie.avi

- rdbas previously standard read 3dgridfile.
  Now it can read any file by spcifying it after the rdbas
  command:

  rdbas=anyfile

- Created two new commands to specify the upper and lower limits of the
  electrostatic potential treated as polar. This is in aid of the 
  calculation of polar surface area. POLUP=value POLLOW=value
  
  Example of a keyword file mapping an electrostatic potential grid (mapfil) 
  onto on electron density grid (rdbas):

  title
  FILE=boxazepam_631g.out RDBAS=d0.3dgrid
  MAPFIL=m0.3dgrid POLLOW=-0.075 POLUP=0.100 SPACE=0.01

- Molden now has rudimentary support for Mopac2012 output.
  Full support is provided through the auxilliary file (.aux) mopac
  creates through the use of the keyword AUX.

Molden5.1

Wed Jan 22 2014

- Updated the ambfor and ambmd codes to run optimisations and MD
  on proteins faster. This hold for the optimisations with cutoff
  They now use the Damped Shifted Force protocol by Fennell
  Gezelter. Also creation of a water box has been sped up.

Molden5.0.7

Fri Jan 10 2014

- Paul Fleurat-Lessard did an update of the molden routines reading
  VASP POSCAR and XDATCAR files. This was necesary to support a newer
  version of these files.

- added the keyword 'wrmolf' to ne used in the molden command file.
  'wrmolf' triggers the writing of a molden format file 'mol.molf'

- update the opengl secundary structure renderings (gmolden) of protein
  being edit in the z-matrix (and in animated variables). This is important
  for the psi and psi angles. Now the changes in psi and psi angles are
  immediately applied to the structures on screen.

- molden compiled with gfortran would sometimes crash on an unexpected
  end of line. Fixed it.

- molden format files with multiple sets of coordinates would not be displayed
  correctly.

- When reading ambfor .xyz files, molden would not set the internal 
  (molden) variable ipdbt() for C-Terminal OXT residues. This would lead
  to a faulty .xyz file written by molden, for instance on subsequent use
  of the original file to start further ambfor/ambmd calculations.

- When doing from molden a MD run in the background (with ambmd), the first
  read in .xyz file (by tnkfst in runjob.f) would lose the reordering information
  when more memory was allocated. This resulted strange water connectivity.
  Fixed it.

Molden5.0.6

Tue Apr  9

- The functionality of adding hydrogens to hetatm residues was broken due to
  change in the web service pdbechem. With this version it is again functional.

- gmolden: shader bug when using occlusion and Blur caused the screeen
           to not be updated correctly on some systems.

- gmolden: Added a proper "balls & sticks" mode, next to the already existing
           sticks mode (in gmolden incorrectly named as balls & sticks).

Molden5.0.5

Thu Dec 20

- Added to possibility to localize orbitals (boys scheme).
  Next to the 'orbitals' button there is now a 'local' button.

Molden5.0.4

Mon Nov 26

- Added a ramachandran plot for proteins. It can be found as last option
  under the 'Surface' icon.
- Fixed a bug that made the Average CH3 button disappear in the NMR spectrum.

Fri Nov  9

- Restructured file selection so that it can now be generally used.
  Used the file selector in the "Read Grid" button in the density mode
  and also on the "Vr" button where you can map an other grid onto the
  first grid.
  Added a filter to the file selector, nothing fancy just a substring match.

- Cleaned to code of molsint.f. Now when calculating distributed multipoles
  Molden will report when mxsite is exceded. This usually only happens when
  NO overlap contributions are shifted to atomic sites.

- Somebody reported problems on his self constructed gaussian style files.
  Fixed this.

- Added a charges scaling option to ambfor (for example : ambfor -S 0.7)
- Ambfor will now parse a file called 'param' with alternative parameters
  (If it exists of course).

 
Thu Sep 20 

- Molden now by default no longer searches for mpi executables at start up.
  On machines with terabytes of disks this can take very long.

- Updated reading of Orca files, so that geometry+gradient single point runs
  should now also be supported.

Thu Sep 13

- Molden can now also read Gamess-US inputfiles with cartesian coordinates

- Adjusted the writing of ambfor .xyz files so that the file can still be
  read by molden if there is an undefined atom type.

Fri Jul 13

- Molden used to read in the first point with orbitals for Gaussian09
  This was because in the beginning of its release, the orbitals
  of the stationary point were not printed. 
  Currently Gaussian09 does print the orbitals of the stationary point.
  Now Molden will display the orbitals of the stationary point.

Fri Jun 15

- Molden now check then openGL vendor string to circumvent a bug
  in the graphic drivers of intel video cards.
  This caused a segment violation on _mesa_generate_mipmap

Thu Jun  7

- the .jdx file written with the use of the keyword PLECTRUM now
  also contains information for use in combination with jspecview
  and jmol.

Thu Apr 26

- included wrting of a .jdx file with the keyword PLECTRUM
- keyword SPACE=value now is honoured when reading cube/grid files

Thu Mar 15

- Made the generated web pages for spectra (keyword PLECTRUM) compatible
  with Internet explorer.

Tue Feb 28

- Improved generation of the html page created with the use of the
  PLECTRUM keyword. (see below Tue Feb 14).
  The command used to convert the postscript file now is:

  gs -sDEVICE=jpeggray -g1000x500 -dBATCH -dNOPAUSE -sOutputFile=spec.jpg spec.ps

  The index.html file now also contains javascript code to view the
  active normal mode as a red line drawn on the spectrum.
  It also contains a table of available frequencies together with intensities,
  they can be clicked.
  
Tue Feb 14

- The molden keyword PLECTRUM calculates a spectrum for files that have
  frequency and intensity information, in addition it converts the
  postscript file that contains the spectrum to a jpg file. 
  For the conversion to work the ghostscript program must be installed
  and defined. The folowing command is used to generate spec.jpg:
  gs -sDEVICE=jpeg -r500x250 -dBATCH -dNOPAUSE -sOutputFile=spec.jpg spec.ps
  It also creates the files index.html, mol.xyz and freq???.xyz. 
  The last files contain animations of the normal modes. These can be used 
  for interactive spectra in combination with the jmol viewer.

- Rudimentary support for NWchem output files.
  
Mon Feb 13

- The xyz format already supported dummys with the 'xx' element label.
  Now Banquo ('Bq') sites are also supported. Both are displayed with the 'xx'
  label.

Fri Feb 02

- Added molden keywords SPECTRUM (and PLECTRUM) to calculate a spec.ps
  from a molden commandfile. Only for files that have spectra information.
- Molden now visualizes a file with just one atom better.

Fri Jan 20

- Added keywords EXTPOSCHG and EXTNEGCHG which specify a single pointcharge
  to be added for the calculating of the QM electrostatic potential.
  This can be used to model a counter charge when calculating the Elec.
  Pot. of an anion or cation.
  EXTPOSCHG=(r.rr,r.rr,r.rr) where r.rr is a real number representing
  the cartesian coordinates in Bohr.

- Added commandline flag -U to switch of using shaders. On some graphics
  cards the opengl version (gmolden) can hang on setting up the shaders.

Mon Jan  9

- Some optimisation to shader code for the benefit of older graphics
  cards under Ubuntu Natty

Sat Dec 31

- The lines representation of a surface in Opengl shader language
  gave problems. Fixed it.

Thu Dec  1

- Molden uses the pdbechem service to find out what hydrogens to add
  on ligands. Unfortunately, the URL for this service has changed.
  I have updated Molden to point to the new URL.

Mon Nov 28

- Fixed blocky noise when using occlusion.

Mon Oct 24

- amfor xyz files with cell information were not correctly stored in
  the internal common blocks of molden.
  As a result writing a just optimised structure and comparing it with the
  structure on screen would give different results.
  Fixed it !
  
Sun Oct 23

- Reading molden format files that contain only data for MO coefficients
  that are non-zero. When this file is not the first file read in
  it could occur that coefficinet that should be zero were actually non-zero.
  Fixed this.

Fri Oct 21

- Fixed small bug that would leave a atom selected in solid mode hovering
  around when switched to line or stick mode.
- gmolden using shaders would sometimes not draw a surface as transparent

Mon Oct 17

- Added a bigger font for the labels of atoms and residues for the opengl
  version gmolden.

Wed Oct 12

- added the commandline option '-k int' (int=0-15), to specify the
  color used to plot lables. Only available for the openGL version (gmolden).

Thu Sep 29

- The by molden created *.ogl files can now also be read from the commandline:
  ./gmolden molden.ogl
  Previously, these files could only be read from the filemanager.

Wed Sep 28

- fixed a bug that would leave the outvector of the plotplane essentially
  as a nul vector. This would only occur if you let molden figure out
  a default plane AND if the first three atoms would be on a line. 

Fri Sep 23

- Built in a button to make a gif screen shot in the density mode.
- Pressing 'R' for shadows now does not make the background black when
  no shader capability is found.
- when writing a ambfor .xyz file with crystal information, now also
  expands the cell before writing.

Thu Sep 15

- GAFF atom type assignement: switched from default aromatic Nitrogen 
  type nb to nc. (nb is only for pure aromatic compounds such as pyridine)
- also truncated negative values (see Wed Sep 14).

Wed Sep 14

- fixed problem with writing cube files on 64-bits platforms.
  Function values that are less than 1.0d-99, are 'truncated' to
  1.0d-99

Mon Sep  5

- Fixed a bug in the C code that calls getpoi (Thanks to Ben Roberts,
  University of Florida for fixing this). This bug could affect the 
  processing of Gaussian files.

- When there is only one (hetatm) molecule in a PDB file, molden would
  not pass on the total charge to the routines that calculate partial charges, 
  even when it offered the possibility to set it. Fixed.

Thu Sep  1

- when the main window was resized (or -geom commandline option was used)
  in the opengl version (gmolden), the rubber selection band for selecting
  atoms in the z-matrix editor was wrongly displayed. This is now fixed.

- when assingning charges to hetatm molecules of pdb files, molden would
  try to retrive the total the total charge from the pdbchem. If no valid
  three letter residue identifier was defined it would return charge option
   -15, as a result the default total charge in the "assign total charge" 
  window would be "quintet" in stead of "0". Such a situation would accur
  when a ligand was added by combination of two structures in the structures
  window.

Thu Aug 18

- When using xming as xwindows/opengl emulator, in the opengl mode when
  using the "select by pointer" mode in the z-matrix editor, the rubberband
  would not be drawn. This is now fixed.

Tue Aug 16

- When writing a Gamess-US style z-matrix and when variables were specified
  as constants, Molden now writes an ifreez record and nzvar.

Wed Aug 10

- Added visualisation of calculated UV-spectra from TD-DFT calculations with
  Gaussian.  Please feel free to suggest improvements.

Mon Aug  8

- Fixed a bug which could make molden crash when having multiple structures.
- The output of the internally run 'locate' command on linux is now redirected
  to /dev/null. So unsuspecting users are not confused.
- Added small tweaks to the gaff forcefield types assignment (nitrogen in 
  cyanide is now typed).
- Added a lot of ifdef statements to make sure older linux distributions
  do not include the newer code for OpenGL shaders.

Mon Aug  1

- Fix for reading gaussian 09 output when coordinates were frozen.

Wed Jul 13

- reading of esp charges on gaussian output has now been put before reading of
  basis-set. This way when no basis-set is written the esp charges
  will still be read.

Mon Jul 4

- The 3D wireframe drawing (X-windows) of the density/orbital mesh was
  no longer drawn with equal width and height when using the commandline
  -geom 1600x800 qualifier (example). Fixed it.
  The use of shadowmaps currently only works if you dont resize.

Thu Jun 30

- Introducing Shadow mapping, for proteins, the secondary structure cartoons
  cast shadows on them selves. See the palette icon, option shadows.
  Or just hit 'r' in the drawing screen.

Tue Jun  7 

- Introduced a bug in reading multi xyz formats, some month ago. Fixed it.

Tue May 31

- There was a bug reading Mopac2009 output

Mon May 16

- Molden now prevents the jobname being the same as the file being currently
  read. This can cause problems.

Wed May 12,13

- Removed a bugs from the mpi MD code.

Thu May  5

- gmolden will draw proteins faster in the space filling mode.

Wed May  4

- gmolden would be killed due to too much memory consumption, during an MD
  run, fixed.
- xyz files containing more than 1000 structures, are now all accounted
  for in the geom conv. window.

Thu Apr 28

- There is a bug detected in the fast water code of ambmd. I have
  applied a temporary quick fix.

Tue Apr 12

- Added postprocessing options: Blur and Occlusion to the colorpalette
  icon. These options work as toggles. It will slow this down a bit,
  certainly on less powerfull systems (graphics card dependant).
  This is only available if you run molden locally.

- Changed the default for gmolden (mixed openGL/Xwindows executable)
  to be fullgl. You can still get the old mixed behaviour by using the 
  commandline option -F.

- Molden does now also better support the mol2 file dialect used by
  the popular docking program FlexX.
 
- When calculating the contacts between a ligand and residues in opengl
  mode, the distance labels where not correctly displayed at far away 
  distances. Fixed it.

Wed Apr  6

- input z-matrices having variable names longer than 8 characters were
  not read in. I have now increased the maximum variable lenght to 10.

Wed Mar 16

- Molden now has per pixel lighting built in. This looks a lot better,
  especially the specular lighting. The per pixel lighting is accomplished
  with shaders. Windows emulators like Xming and Xwin32 do not support shaders,
  so you will not see a difference there.

Mon Mar 14

- Molden compiled with gfortran had problems with reading the protein
  density file, fixed it.
- Molden compiled with gfortran sometimes crashed when reading in a 
  second structure, fixed it.

Wed Mar  2

- Added a cutoff scheme for water-water interaction, see:

  Is the Ewald summation still necessary? Pairwise alternatives to 
  the accepted standard for long-range electrostatics
  Cristopher J. Fennell and J. Daniel Gezelter, 
  Journal of Chemical Physics. 124, 234104 (2006)

- Adapted the makefile to be a little more intelligent.
  It now tries to figure out on what os it is. This will cover most
  flavors of linux, MacOSX and cygwin.

Fri Feb 25

- Added support for the Gaussian 09, new IRC mode HPC.
  By request of Prof. Ugo Cosentino

Thu Feb 17

- added a routine for water-water interections. It uses a lookup table.
  It is unfortunately only a little bit faster (factor 0.66)

Fri Feb 11

- removed a bug that would not copy all of the force field information
  after allocating memory for a water filled box.

Thu Feb  3

- Adding ions is now done depending on the electrostatic potential.
  You can still invoke random placement with the commandline argument
  -z on ambfor/ambmd (in combination with -b).

Wed Feb  2

- replaced use of fork by vfork in an attempt to stop forking errors
  on windows7 (64bits version still has problems).
- When adding a water box in ambfor or ambmd automatically ions are
  added to neutralise the box. The placement of the ions is accomplished
  by replacing waters randomly

Fri Jan 28

- there is now also a multiprocessor version of the crystal optimiser.
- removed a small bug in the opengl version when reading a jaguar cube
  file.

Tue Jan 25

- The molden/gmolden interface will now automatically detect whether the 
  openmpi executables have been built. If so an option to use multiple
  processors will be availabe in the window that fires ambfor/ambmd jobs.

- Removed a bug in the multiprocessor version of the MD program, ambmd.
  Each subprocess would have its own initial velocities, which in turn
  would lead to different coordinates in each subprocess. Fixed it.

Thu Jan 13

- Finally got the interactive optimisation working flawless on
  WindowsXP.
- Fixed continued rotation during interactive optimise.
- optimised update time during optimisations

Wed Jan  5

 - Added atom type 'cz' to the GAFF force field, this is needed for guadinium
   and arginine. Also added improper torsion to keep the C connected to the
   three N's in one plane.

Tue Dec  7

- Crystal optimiser was only working for non-covalently bound crystals.
  Current version has been tested on organic crystals and seems to work.

Molden5

- Added a crystal optimizer, based on the gaff force field.
  Periodic conditions are enforced by translation symmetry copies.
  NO ewald summations ! This is a beta test.

- Gamess-US irc data, bug fix.  first atom was not displayed

- While adding support for the calculated NMR spectra, a bug was introduced
  that made it impossible to set the maximum intensity in a spectrum.
  fixed.

- When doing a combine structure on two proteins, the ligands are now also
  colored by structure.

- When using combine + align on two none protein structures, the user
  is asked to specify 3 pairs of atoms. Each pair consisting of one
  atom from the first and one atom of the second molecule. The code
  has been changed in such a way that only atoms can be selected
  of the first/second structure.

- Adapted molden to be more vista/windows7 friendly.

- When aligning two structures no longer the pmf score window pops-up
  after pressing escape two or more times.

- the interface to qchem and orca was only partly functional. Fixed it.

- The crystal information of a mol2 file was nolonger parsed, fixed it.

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

Molden4.9

Wed Sep 22

- Updated molden to read gaussian09 IRC caclulations outputs.

Wed Aug 11

- Updated Updated reading of Gamess-us output, kindly supplied by 
  Mariusz Klobukowski. The number of electrons when using pseudopotentials
  is also calculated correctly.

Mon Jun 21

- Updated reading of Gamess-us output, kindly supplied by Mariusz Klobukowski

  1) to read results from hessian calculation in the latest version of
     gamess-us (lines with mode symmetry information were added)

  2) to read basis sets in semiempirical calculations (we changed gamess-us
     to be able to see the semiempirical MOs, with the STO basis sets expanded 
     in terms of STO-nG expansions).

Thu Jun 17

- The reading of a gridkont file (Program GRID) has been moved from the
  density mode to the molecular mode. Under the surface icon, there now is 
  an option "gridkont". The benefit of this move is visualisation of the
  GRID molecular interaction together with the full opengl protein 
  visualisation and the use of clipping planes.
- added a box around the grid for gridkont and protein electron density.

Wed Jun  2

- Added the option to for the electron density to follow the current residue.
  This build on a feature of molden to show the label of the residue
  the pointer is currently over. In addition now also a block of
  electron density follows the residue the pointer is currently over.
  You can fix the postion of the block, by hitting the spacebar. 
  Hitting it again will make the block follow the pointer movement again.
  This all in the context of the latest hanges to molden of Fri May 28.
  See below.

Fri May 28

- Under the surface icon, there was the option electron density map.
  Clicking this would try to read a file contaning protein electron
  in the so called 'omap' or DSN6 format. These files are available
  from the Electron Density Server at Uppsala University:

  http://eds.bmc.uu.se/eds/

  This functionality has been extended. When reading a pdb file the pdbid
  is automatically parsed and stored. When clicking the "elec. Dens. Map'
  option the pdbid is used to automatically retrieve the omap file from
  the electron density server. (the program wget is used for this).

  After the file is read, a window will pop up, where you can specify the
  electron density contour level. The rendered electron density volume
  can also be clipped in three directions, for clarity.

- on request from Paolo Tosco, molden4.9 now allows upto four keyword lines
  in a molden keyword file.

Wed Apr 21

- When reading a protein ambfor .arc file, the secundury structure would be 
  all coils, fixed this.

Mon Apr 19

- When interactively omptimising a structure with a water box, molden
  sometimes would not show the waters. Also aborting the optimisation
  by pressing escape, would sometimes disrupt the X protocol.
  Fixed.
- Updated the windows version.

Wed Apr 14

- When a ambfor xyz file was loaded after a pdb file, its secundary
  would be all coils. Fixed.

Tue Apr 13

- When reading a ambfor xyz file, nolonger the hydrogen bonds are
  calculated by default. In the presence of a big water box this becomes
  very slow.

- Going into the zmatrix editor the waters were stored. After quitting
  the zmatrix editor, the waters were restored. However in the presence
  of a large water box, the restoration was very slow. Fixed.

- Using the default optimiser(lbfgs) now also the waters that water out
  of the box are come in to the other side of the box.

Mon Apr 12

- When interactively optimising a small molecule , in the presence of 
  water box, the waters and box would not be visible in the line drawing
  mode (solid was ok). Fixed.
  The box itself is now also visible.

- In the residue command window, when multiple chains were available,
  you would have to specify the chain also in order to visualise residues.
  This is now changed. You can however still use the chain identifier:
  *:A for example will switch on all residues belonging to the A chain.

- When reading an ambfor xyz file containing a protein, and subsequently
  writing a PDB file, every first residue of a chain had residue number 1.
  This caused probelms when there were chains consisting of 1 residue.
  Such a pdb file would not parsew correctly. Now all residues have a unique
  number.

- reading an ambfor xyz file with a box specified, would lead to an empty
  screen when opening the z-matrix editor. This is fixed.

- When optimising the Hydrogen positions, waters that are close to a metal
  center now also have hydrogens added.

- Water boxes containing more than 5000 waters would have a corrupt [RESIDUE]
  section in the ambfor xyz file. This would lead to artifact in visiulisation 
  with molden.

Wed Apr  7

- ambfor: made ff99sb the default protein force field. The unmodified
          amber 99 force field will be used when specifying the -x
          commandline flag
          
- ambfor: for rna/dna modelling the ff99bsc0 modification of the amber
          force field  will be used when specifying the -y 
          commandline flag

- fixed a bug that would visually screw up the secundary structure, when
  in the xwindow mode.

- in the standard molden xwindow version the residue popup windows,
  would have options, when clicked would activate the wrong action.
  This bug was introduced the week before.

Thu Apr  1

- absolute scores for histidines were still to big, relative scores
  were already OK.

- fixed a bug is the secundary visualisation, which under very special
  circumstances would not show part of the coil.

- The small amino acid menues you can bring up by clicking with the second
  mouse button now also contain an option to change the protonation states
  for histidine, and a flip option for asn and gln

- In ambfor the total charge on non amino acid residues is checked and
  corrected if it does not add up to a discrete integer.


Thu Mar 25

- fixed a major bug in calculating the scores for histidine (D,E,+),
  under the optimise hydrogen positions.

- added prediction of asn/gln flips, under the optimise hydrogen positions

Fri Mar 19

- Fixed a bug that could make the hetatm residues disappear in the z-matrix 
  editor, when previously "Opt H pos." was used.

- when aligning two protein structures, both structures have a different color.
  Also in the backbone mode the secondary structure visualisation will have a
  different color per structure.

- when combining two small molecules from the structure window, you will now
  be asked to move the two structures into visible range.

- when using command from the residue command window in combination with
  the SURF keyword, molden will now display opengl surfaces in opengl mode.

- cysteines missing an HG atom will now have partial charges adding up to -1.
  This also implemented in ambfor.

- some pdb structures containing hydrogens with labels not recognised by 
  molden are now recognized. There are probably more, let me know 
  if you find more.

- fixed a bug that would make some surfaces not being displayed in the 
  surfaces window. This could also cause: surface OFF, not to work.

Mon Mar  8

- added ambfor support for modified nucleotides such as:

  1MA,5MC,OMC,2MG,M2G,7MG,OMG,YG,H2U,5MU,PSU

  NOT supported are:

  1MG,I,+U

- water oxygen atoms connected to the protein are disconnected before
  writing ambfor xyz files.


Wed Feb 24

- fixed a minor bug that would give a superposed atoms in z-matrix,
  after doing optimise hydrogen positions on a dna molecule.

Tue Feb 23

- for rna to z-matrix molden now takes an OXT atom in to account.

Mon Feb 22

- Added parameters for nucleotides to the amber force field implementation
  in molden and ambfor.

- corrected charges for water, which did not completely add up to zero.

Mon Feb  1

- G09 has different zmatrix output than G03. Molden can now deal with this.

Tue Jan 26

- Molden4.9 introduces to optimisation method lbfgs as default optimisation
  algorithem for Ambfor. Together with some other code optimisations, this
  increases the speed of ambfor with a factor 4.
- changed param.f, such that less continuation lines are used.
  The many continuation lines caused problems with compilers such as ifort.

Wed Dec 23 
- more bug fixes concerning interaction between Molden with Tinker and Ambfor.
  
Tue Dec 22

- MAJOR Bug fixes concerning interaction between Molden with Tinker and Ambfor.
  For ambfor the bug was introduced on 27 November, when the writing of ambfor
  xyz files was changed. This bug is now fixed.

  Ambfor will now write a file 'ambfor.wr' to indicate it is writing 
  intermediate optimisation files, when done writing it will remove this
  lock file. This improves the interaction with Molden, which in some cases
  would begin to read the file when its writing was not yet complete, resulting
  in incomplete structure on screen.

  The interface with Tinker was compromised earlier when ambfor started to
  write intermediate files in binary form. Tinker and Ambfor now have their
  own reading routines.
  The jobname.run file, a csh script now has a -f added to the #!/bin/csh
  line. This is because .cshrc files could contain commands preventing the
  succesful executing of this file.
  An number of buttons not applicable to Tinker jobs are now nolonger visible
  in the Tinker optimise window.
  
Sat Dec 19

- fixed a small bug that interfered with wget, which molden uses to
  fetch pdb and hydrogen positions to add to ligands.

- fixed a bug that would make ambfor crash when using the -N option
  in combination with a liagnd being present

Thu Dec 17

- The switching function for the electrostatic used with cutoff with ambfor
  caused distortions of the final protein geometry. I switched to a 
  shifting function for the electrostatics to solve this problem.

- The use of switching/cutoffs (commandline flag -N) in ambfor/ambmd
  is now twice as fast.

Mon Dec 14

- Open shell stuctures are now also read from a Mopac2007 .aux file.
- Molden now tries to find the environment variable g09root
- Fixed a small bug that could make Molden crash after reading in two
  large structures.
- Fix that treats negative J-couplings correctly (NMR from gaussian)

Thu Dec 10

- In addition to 1H NMR, molden now also calculates the 13C NMR spectrum.
- The magnetic shielding and J-coupling corresponding with rotationally
  equivalent Hydrogens can be averaged by clicking the corresponding
  Button.

Wed Dec  9

- Molden can now display caculated H NMR spectra from gaussian
  You need to have caculated magnetic shieldings and J-couplings

Mon Nov 30

- Fixed bug in reading nmr data from gaussian output.

Fri Nov 27

- Ambfor can now be run using cutoff's and switching functions: use the -N
  commandline flag to use cutoffs and switching. This can be faster than
  all atom optimisation ( a lot).
  The window form which ambfor optimisations can be fired from Molden
  now also has a button to use cutoff's and switching.

- When writing ambfor files the order is now such that ambfor can
  work with contigous molecules. The ambfor file now contains an extra
  sections '[RESIDUES]' which contains the begin of amino acids and
  hetatm ligands (including water).

- The ambfor and ambmd now catch the user's Control-C and now flush the buffer
  before exiting.

- Ambfor now handles better finding equivalent parameters for covalently linked
  ligands to aminos acids.

- Ambfor can now be run using cutoff's and switching functions: use the -N
  commandline flag to use cutoffs and switching. This can be faster than
  all atom optimisation ( a lot).


Fri Nov 20

- When using Gaussian with the keyword NMR, the nuclear shieldings
  are read and can now be displayed as a label. If also J-couplings are
  available and extra "J" button will appear in the "calculate" section.
  When clicking it, select two atoms and the their J-coupling is displayed
  in the calculate window.

Sun Nov 15

- Removed a bug that affected the prediction of HID/HIE histidine
  residues.
 
Tue Nov 10

- Optimisation of Hydrogen positions under "H" icon, now also includes
  getting the formal charges of HETATM residues, adds hydrogens if not added
  already and calculates the partial charges of the hetatm residue.
  It then optimises the OH positions of THR, TYR, SER resiudes.
  Then it optimises the the choic between HIE and HID histidine residues.
  Finally it adds hydrogens to the water molecules , that are in close contact
  with the protein.

- the atom attributes window now also offers an editable field for the atom
  number. Which makes it easier to identify an atom on screen, when only
  its atomnumber is known.
  
Wed Nov  4 

- The generation of charges on a pdb file is now more verbose. It also
  does nolonger consider the amino acids, since they have charges from
  the ambfor forcefield. For each HETATM residue, the charges are 
  calculated separately.

- Molden now also writes HETATM residue names, when writing a PDB file
  via the "write" button.
 
- When charges are calculated for a HETATM residue, the VDW surface 
  colored by electrostatic potential is now supported.

Tue Nov  3

- The Hydrogens added to HETATM residues, are now put directly after
  the the residue. This way it does not interfere with other fucntionality
  such as interactive docking and interactive optimalisation.

Sun Nov  1

- Removed bug parsing PDB entries like 1eat, which have more than
  two alternative residues
- Made routines to add hydrogens to HETATM residues crash proove
  when no residue is recognized.

Thu Oct 29

- added routines to add hydrogens to HETATM residues from the PDB.

  The default uses the web server from PDBeChem. 
  It uses the external program 'wget' for this.
  If you dont have it, you can retrieve it from:
     http://ftp.gnu.org/gnu/wget/

  A second fallback method adds hydrogens by looking at the angles of the
  atom it coonects to. This method is far from ideal or unambiguous

- added the possibility to retrieve a PDB file from the internet. It uses the
  external program 'wget' to do this (see above).
  You will find a  "Get PDB"  button and text entry field in the file selector.

- When element information is available from columns 77,78 in the HETATM
  record of the PDB file, this is now used to determine the element type.
  The old method, which uses the atom label to determine the element type,
  is still used for old style pdb files.


Wed Oct 21

- Sometime back the MS-DOS fileformat check was introduced, for files
  supplied at the commandline (for files via the file select window
  this was already in place).
  However, when a non-existent file was specified, an empty file with
  that name was mistakenly created. This is now fixed.

- In case that there are intensities together with frequencies, both are
  now listed together in the frequency list box.

Tue Oct 20

- when reading a pdb structure from the commandline, followed by
  2nd mouse button click on and hetatm residue -> "Mol. -> Structure"
  molden would crash. Fixed it.

Mon Oct 19

- Added optimisation of water to the H optimise option.
  Waters acting as donors/acceptors to a residue, have hydrogens added.
  Other waters are deleted.

- An error in reading AMBFOR .xyz files no longer leads to an exit from
  the program.

- when reading an AMBFOR .xyz file, the error printing has now improved.

- removed a bug in ambfor (which can reveal it self after new memory
  allocation.

Mon Oct 12

- Changed the default for a histidine residue to be Neutral (HID).
  The option under the H icon, "Opt. Ser/Thr/Tyr/His H", will now
  also select which histidine is the best HID or HIE (both neutral).
  positive HIS has to be created via the z-matrix by adding a H to
  the neutral histidine.

- In Opengl mode structures would not be updated after zmatrix editing
  of a protein, fixed.

Fri Oct  9

- removed bug that would switch hetatms to visible when in backbone mode, 
  after setting a distance monitor.

- when switching from stick to solid in gmolden (opengl) already set
  distance monitors would no longer be visible. Fixed it.

- Fixed a bug that sometimes would ambfor crash when allocating more
  memory

- added the option the optimise the hydroxyl hydrogens positions of the
  threonine, serine and tyrosine residues. Under the H icon.

Mon Oct  5

- removed bug which under certain conditions would not correctly show 
  the opengl secondary structure of the second structure read in (gmolden).
  
Fri Oct  2

- Some more fullgl opengl version of gmolden updates.
- removed opengl helices/strands/coils during an optimisation, since
  they were not updated.

Thu Oct  1

- The fullgl opengl version of gmolden (as opposed to the mixed xwindows/opengl
  version) has a few bugs fixed. In the lines mode, during an protein movie
  the residues would not be updated.  Now they are.

- The distance monitor now also works during a protein movie.

Mon Sep 28

- when creating a water filled box with ambfor/ambmd, some of the waters
  at the edges could be very close to waters in neighbouring cells.
  When using this box in MD this would result is waters with very high
  velocities.  This problem is fixed now.

Thu Sep 24

- Gaussian09 MP2 energies are now correctly read.

Fri Sep 18

- Fixed a problem introduced in molden4.8 with reading AMBFOR xyz files.
  specifically recognising protein structure.
- fixed some of the things that dont work when running the full opengl 
  version of gmolden (-F commandline) flag. There still remain
  some issues.
- The zmatrix is NO LONGER updated during a movie
- Scroll wheel now also work in interactive optimisation mode 

Thu Sep 17 

- more updates to swooth the interaction between molden/gmolden and
  ambfor/ambmd
- You can now add a water box to both force field and md simulation.
- When ambfor xyz/arc files are read in, the element types are mapped
  from the integer atom type, and nolonger from the atom label

Fri Sep 11

Molden4.8

- Molden4.8 introduces ambmd:
  Added a simple isotropic MD program to the ambfor program.
  The Molden interface can now fire these MD jobs via the optimise (AMBFOR)
  window. You can now select 'ambmd' in addition to ambfor.

- the ambfor and ambmd programs now have a -b commandline flag, which
  will add a waterbox to your structure

- ambmd produces a .arc file as the trajectory file. The reading of
  this file in molden has been speeded up.

Wed Sep  2

- When adding the possibility of contraints to optimisation with
  ambfor, it was overlooked that in the event that no restraints
  were set, all atoms in the ambfor xyz file would be treated as fixed.
  This has now been corrected.

- Water was not handled well in ambfor, fixed it.

Thu Aug 27

- Added a simple isotropic MD program to the ambfor program.
  The Molden interface can now fire these MD jobs via the optimise (AMBFOR)
  window. You can now select 'ambmd' in addition to ambfor.

Thu Aug 20

- When reading G09 output, molden now correctly deals with the fact
  that G09 only prints orbitals for the input geometry and nolonger
  prints orbitals of the stationary point.

- Molden now also recognizes MSDOS files supplied via the commandline
  and automatically converts them to unix formatted files.

Wed Aug 12

- reading back in ambfor generated xyz files, the amber atom types
  have automatically charges assigned

Tue Aug 11

- further compatability fix for reading g09.

Thu Jul 23

- fixed serious bug in reading gaussian output, while creating support
  for g09. Fixed !

- reading back in ambfor generated xyz files, the supported amber ions
  are now correctly classified by element.

Tue Jul 21

- first support for gaussian 09
  single point, simple optimise, simple force/freq jobs supported
  More updates are probable to follow

Sun Jul 19

- Removed a number of bugs concerning reading the latest Gamess-UK
  output.

Fri Jul 17

- Reading gamess-uk output with more than 99 points in the 
  geometry optimisation, all point above 99 would not be updated, fixed it.

- The energies from the latest Gamess-UK output are picked by molden
  and shown in the geom. conv. window.

- The fullgl version of molden would not update the background color when
  changed with the paleete icon: fixed
  

Wed Jul 15

- fixed recently introduced bug, reading 3dgridfile

Mon Jul 13

- Partial optimisation can now run in BFGS optimisation mode in ambfor
  upto 330 flexible atoms.

- Molden reads back in ambfor .xyz files better.

- Updating labels no longer lags in slow connection opengl mode.

- Resizing main screen in opengl is now one update.

Tue Jul 7

- Added the possibility of partial optimisation. If you want to optimise
  a docked pose, select only the ligand and the surrounding residues
  flexible. And optimise with the the amber/gaff force field.

- Molden now supports the scroll wheel to zoom in and out.

- When using starnet X-Win32 X-windows server, gmolden switches
  to use full OpenGL (not mixed xwindows/opengl mode).
  The newer versions of X-Win32 ver 8, have problems with
  zooming in via the Ctrl key, in this case use the right Control key
  or the scroll wheel.

- Molden can not determine which version of X-Win32 server is active.
  You can use the molden commandline flag -Q, to return to the mixed
  OpenGL/Xwindows mode. This will work better with older versions of
  X-Win32 (6.1 and older).

- When aligning two small molecules and using the align and combine
  option, there was an extra copy of one of the molecules. Fixed it.

Wed Jun  3

- Removed bug from the handling of the multi mol2 and sdf files

Fri May 28

- Added the option to compile a complete GL version of gmolden.
  Uptill now gmolden did some graphics in xwindows and most graphics
  in OpenGL. In the MacOSX version of gmolden, the xwindows graphics
  were not visible. I therefore created the option to let gmolden
  do all graphics with OpenGL calls. You request this new functionality
  by adding the -F command line flag. 

Mon May 4

- fixed annoyance that during animations, the animation would stop because
  of the help window popup on the button started the animation.

Tue Apr 28

- fixed bug that could screw up the displaying of secundary structure
  in the opengl version of molden, in the case of multiple proteins
  being loaded.

Mon Apr 27

- Multi molecule files, such a .sdf files and .mol2 files will now
  open an extra window in which each separate molecule is clickable.
- In addition you can now combine .mol2 files conataining poses for
  a ligand with the protein being docked into:
  First load protein. Second, read multi mol2 file with "Add" selected
  in the read file window.

Wed Apr 16

- added the calculation of topological polar surface area (PSA).
  Molden control window -> surface icon -> totpol. PSA

- added the calculation of the Quantum Mechanical PSA,
  fisrt calculate a 3dgridfile with the density, second calculate
  3dgridfile with QM electrostatic potential (slow),
  Start up molden, go to the density mode, read in density gridfile
  map esp on density, Vr icon. -> mapped

- fixed bug, after animating normal modes going into the zmatrix
  editor, clicking one of the variables would move the structure
  allow a normal mode.

- added the reading of multi molecular sdf files.
  click the 'M' button to select other structures in the sdf file

- fixed a bug that would mess up windows on centos, when keeping
  the pointer over a button and make the help text appear.


Thu Mar  5

- fixed bug causes the atoms not to be connected in the density mode
  in combination with gamess-us and qchem outputs

- fixed a bug that would shown an empty structure for orca outout.

- surface off would sometimes also switch off some parts of the coil
  in opengl mode; fixed it

Wed Feb 25

- MAYOR: BUG FIX FOR AMBFOR the standalone amber/gaff force field
  optimizer which is called by molden.
  The bug caused some torsion force constants to be to big and 
  could lead to distorted optimised geometries.
  eg anything with a -so4- group.

- fixed a bug that would let molden crash when deleting all active structures
  from the structures window, when the atom attributes window was displayed.

- some programs out there write pdb files with lowercase element and residue
  labels. Molden can now handle these.

- fixed a bug that caused read in ambfor .xyz files, to be treated as
  if they had no charges

Thu Jan 15

- added 64 bits executables for linux on our anonymous ftp directory
  ftp://ftp.cmbi.umcn.nl/pub/molgraph/molden/bin/Linux/64bits

- fixed a minor bug in the 64 bits version when rendering helices

The Dec 18

- Zoom is now kept when editing a zmatrix
- Updated the makefile somewhat, let me know if you have problems with it

Tue Nov 25

- When a protein was edited via the z-matrix all waters were gone
  except those with a connection to the protein. Now the waters are
  restored when exiting the z-matrix editor.

- removed a a bug in an earlier undocumented implementation of the
  above that would make molden crash during the animation of variables
  in the z-matrix editor.

Thu Nov 20

- Added the posibility to color by chains (opengl only).
- fixed a bug that would render the molecule as opengl sticks after
  a HETATM molecule of a protein was removed via the 2nd mouse click
  popup window.

Wed Nov 12

- fixed a bug that could make molden freeze when a help window on a
  button was activated
- fixed a bug with rendering forces and the in/de-crease scale buttons.
- fixed a small bug with reading non protein mol2 files, which would cause
  problems when combining the molecule with a protein structure.
- Fixed bug which caused crystal information (biosym car,arc), to be
  wrongly symmetry expanded if it was not the first structure be read in
- fixed a bug that would let molden crash when clicking H-BONDS in the
  stick mode
- fixed a bug that would let molden hang on reading a VASP CONTCAR file
- fixed a bug that would let molden crash when as a first file a freq
  jobs was read in, and consequently the spectrum was displayed.
  (introduced very recently).

Mon Nov 10

- Fixed bug in memory allocation associated with the writing of .ogl
  files in the surfaces window.
  When you do a combine between a small molecule and a protein, the memory
  would be corrupted. fixed it.

Fri Nov  7

- Fixed bug writing "charges derived" electrostatic potential mapped
  surfaces in opengl surfaces window

- Fixed minor bug in writing mopac surfaces in density mode in
  opengl surfaces window

Thu Nov  6

- Fixed bug which caused filetypes with crystal information (biosym car,arc)
  and multiple structures to be treated as containing just one structure.

- when only surfaces but no molecule was present in a .ogl, 
  the scene would not be properly displayed. In perspective mode
  you would be inside the surface with no possibility to zoom in.
  In the non-perspective mode nothing would be displayed.
  Fixed it.

- fixed bug in new .ogl writing routine, that sometimes wrote too many
  connections.

Tue Nov  4

- when choosing contacts to be displayed followed by a distance
  measurement, the part of the distance window where the distance
  should be displyed stayed black, fixed it.

- in pdb files with dna/rna the label OP2,OP1 where not supported,
  now they are.

Fri Oct 31

- The newly introduced surfaces window also now allows to write all
  surfaces to a .ogl file. You can for instance now create several
  orbitals and write then to a .ogl file

- .ogl files ca now be read back in to gmolden.

Fri Oct 24

- A new window lets you manage the opengl surfaces created by molden.
  In the molden control window -> color palette icon -> surfaces
  You can delete surfaces, change color. Separate helices can now
  be colored differently. The surfaces window is also available in
  the denisty mode. It automatically pops up once the opengl icon
  is clicked. Surfaces can now be activated and deactivated.

- The .ogl files when selected in the read file window used longer
  trigger the launch of the external moldenogl helper program.
  This behaviour still goes for the Xwindows only version of molden.
  For the mixed opengl/Xwindows version of molden (gmolden), the .ogl
  files are dispayed within gmolden itself

- Removed a bug that would overwrite the last read file with postscript
  code, when the character 'p' was pressed in the display window of
  molden.

Wed Sep 17

- The align button in the structures window now also works for small
  molecules.

- When viewing gromacs trajectory the backbone view is maintained if
  chosen. Also the labels of the cell are no longer displayed in this case.

- You can now create a separate structure from the pdb hetatm molecules
  by clicking with the 2nd mouse button on the drawing screen.
  This makes it easy to save these molecules in a separte pdb file.

- Added cyclopentane and pyrrole as fragments in the zmatrix editor.

Fri Sep 12

- Added the option to map the electrostatic potential to the solvent
  accessible surface

Tue Sep  9

- Added the possibility to align proteins

Fri Aug 29

- Removed another bug that made the program crash when freeing allocated memory

Thu Aug 28 (My birthday :-)

- Removed a bug that made the program crash when freeing allocated memory

- mol2 files containing multiple structures are now also supported.
  An extra button (M) in the points section of the molden control window
  will appear. Clicking this button will allow you to select an other 
  structure in the mol2 file.

- Two structures can now be combined. Select a structure, supply a 
  structure number in the field next to the "Combine" button, and click
  'Combine'. The structure corresponding to the number will be appended
  to the active structure.
  
Molden 4.7 first version Aug 22 2008

- Molden4.7 can keep multiple structures in memory. Each time a new file is
  read in this is a new structure. You can switch betweeen structures 
  in the newly added structure window. The currently selected structure
  is displayed. You can also delete structures in this window. The original
  one structure at a time behaviour is still available when specifying
  the -O command line flag.

Molden 4.6

Tue Aug 12

- Fixed a bug when going up one directory (..) in the filemanager.
  This one managed to escape my attention for a long time.

- Minor changes to scrolling

Wed Jul 6

- Added dynamic memory allocation voor mopac2007 .aux files

Wed Jul 2

- Removed a bug that lets molden crash when you add the first line
  of a new structure in the z-matrix (windows,macosx)

Tue Jun 24

- The mopac2007 .aux format is now supported.
  The mopac .aux format can hold geometry optimisation data, charges, dipole
  and orbitals information.
  The MP6 and RM1 methods have been added in the submit job interface.

- Corrected a bug in displaying the dipole moment.

Thu May 29

- Activating the picture icon will allways cause a gif image to be
  written, also in OpenGL mode. The latter requires a 24 bit to 8 bit 
  conversion, which has now been incorporated into molden.
  For those who would like the rgb picture back, let me know.

Tue Apr 29

- Removed bug in gmolden that when adding the first atom in a new 
  z-matrix would make the atom invisible (solid/stick mode only).

- Added the ability to cap residues 

- Added support for neutral asp, glu and lys (ambfor force field)

- Improved zmatrix editing of proteins: adding H to histidine will
  now be recognized as the proper HIS form.

Wed Apr 16

- Small annoyance fixes.
  Switched atom selecting back to the fisrt mouse button again in the 
  atom properties windows.
  When doing a optimisation, selected atoms now become unselected.
  
  
Thu Apr 3

- Added small help boxes describing the use of button. If you keep the pointer
  over a button and dont move, after 1.5 seconds, this helpful window 
  will appear. NOT yet implemented for all windows !!
  Gradually help for more windows will be added.

- Added possibility to close/abort query box

- When the user would try to kill the main drawing window by clicking
  on the x in the border supplied by the window manager, molden would popup
  a window saying you should kill molden by clicking on the skull button.
  On some Xwindow servers, this popup window would be off screen. Fixed it.

Tue Mar 18

- In the window to specify the atoms and methods for Oniom, the possibility
  of writing amber charges as part of the element specification for Gaussian
  has been added.

Fri Mar 14

- While reading tinker/ambfor xyz files, a proline could be misclassified
  as a lysine. The most visible direct effect was strange behaviour of
  the spline to indicate the coil secundary structure.
  Fixed.

Thu Mar 13

- Labels of Cell axes are now also visible in the opengl mode of gmolden.
- Labels of XYZ axes are now also visible in the opengl mode of gmolden.

Mon Mar 10

- Fixed bug in 'edit cell parameters'

Fri Feb 22

- Added axes, only in gmolden with solid mode. It can be called by pressing
  '+' or '='. No labels of axes yet: red=x, green=y, blue=z

- Added the possibility of calculating molecular surfaces, with help of the
  program 'surf': 

   ``Linearly Scalable Computation of Smooth Molecular Surfaces",
   A. Varshney, F. P. Brooks, Jr., and W. V. Wright, IEEE Computer Graphics
   and Applications, September 1994 [invited submission].

  You will find the option under the 'surface' icon -> Mol. surf.
  This option is mainly intended for use with proteins.

Fri Jan 25

- corrected bug in reading molden format files when non integer occupation
  numbers would be present.

Mon Jan 21

- Added reading/writing of CIF files

Thu Jan 10

- Added writing of Gaussian cube files (reading was already present).
  This can also be used to store data on a 3D grid which are slow to 
  calculate.

Mon Dec 17

- fixed a bug which could make the automatic z-matrix conversion crash
  for very small molecules with few connected atoms

Wed Dec 12

- Fixed working with nptsx,nptsy,nptsz, edy,edy,edz in combination
  with opengl 

- Going to full screen in OpenGL mode would leave the molecule not
  centered on the ceter of the screen, fixed it.

- Refined creating connections between backbone connections.

- Refined handeling of unclassified pdb symbols in the ATOM records
  of a PDB file.

- Fog is now automaically disabled in the density mode.

Mon Dec 10

- Fixed atom selection by clicking on an atom with the second mouse button 
  while the "Atom Properties window" is open. This functionality was
  destroyed when introducing the popup window for HETATM molecules on
  Aug 8

Fri Dec  7

- Atomcolor is now switch on for all but protein structures after
  reading a new file
- Better default setting of shade for proteins

Sat Dec  1

- Fixed Molden in such away that it is more compatible with the free
  Xming Xserver for windows. 

- Fixed a bug that would let the structure rotate after an buttonpress
  in the main screen and clicking on of the amino acid options.

Fri Nov 23

- fixed bug with displaying G-functions

Wed Nov 14

- Added optimisations for displaying the opengl stick & balls for all atoms
- added support for gaussian counterpoise optimisations.
- added a commandlineflag -S to let molden start with shade off.

Fri Nov  9

- Added support for gaussian mp2 optimisations
  Previously the HF energy would be listed in the geometry convergence
  window, now the MP2 energy will be displayed.
- Added support for gaussian IRC outputs, previously all points were displayed
  in the geometry convergence windows, now only the optimised geometries
  along each point of the reaction coordinate.

Wed Nov  7

- added a window that lets you specify the total charge when using
  the EEM charges.
- added two new parameter sets to calculate EEM charges with, PESP and
  ESFF force field parameters
- Changed the default for automatically calculating charges for forcefield
  optimisation with ambfor to PESP. This is because the previous default
  only had parameters for HCNOF. In addition a string will be printed
  when no parameters for an element are available.
  You can always assign charges interactively and switch off automatic
  charge assigning in the AMBFOR optimise window.
- Improved GAFF forcefield typing a bit.
- Removed a bug that would occasionally make molden crash when doing a
  interactive GAFF/Ambfor optimisation and charge labels were also printed
  (Recursive IO bug).
- when the number of atoms exceeded 10000 the writing of tinker xyz files
  would not print connectivity correctly, fixed it.


Tue Oct 30

- added option to interactively change the parameters that determine
  when molden draws a hydrogen bond. You will find a new entry 'H-bond
  parameters' under the H icon, the 8th icon from the top from the 
  middle column in the molden control window.

Tue Oct 23

- fixed a bug where the reading of Gaussian files would go wrong if
  the following line was present:

   RelInt: Using uncontracted basis, .... etc.

Wed Oct 17

- added a 'prev' button in addition to the existing 'first', 'next' and
  'movie' buttons. 
- added to molden makefile, the feature to build molden, gmolden and
  ambfor/ambfor when you type 'make'
- added option to convert gaussian log files to zmatrices:

  gmolden molin

  contents of file molin:
  title
  file=gaussian.log wrzmat=2

  The zmatrix will be written in the file called 'zmat.out'

- added possibility to scan a part of a proteins rotamer space 
  (max. 6 residues). The best 10 rotamer combinations will be remembered.
  The initial scoring is done via the dfire pmf score. The output can
  be saved and (re)loaded via the file 'resrot'. The rotamer combinations
  can be rescored with the amber force field.
  To activate the window click on a residue with the second mouse button
  and select the last option in the menu (add flexible residue). 
  Subsequently do the same for all the residues you want to be flexible.
  However you first have to create a zmatrix for the protein fisrt by clicking
  the zmatrix button, otherwise the new option will not be presented.

Thu Sep 13

- Fixed bug when applying edited cellparameters
- Added option to calculate charges for hetatm molecules from the PDB
  (in the popup menu, activated with second mouse button click)

Fri Sep  7

Fix bug ambfor memory allocation

Mon Sep  3

Fixed bug that caused the zmatrix-ordering of atoms to be applied
without also changing the charges ordering accordingly, when
starting an ambfor (amber/gaff) optimisation.

Thu Aug 30

Fixed bug when writing zmatrices bigger than 100 atoms.

Wed Aug 29

- improved GAFF typing

Thu Aug 23

- fixed a bug that would not display a residue when replaced (in opengl)

- fixed a bug that could occur when memory is allocated for proteins
  larger than 20000 atoms.

Fri Aug 17

- fixed bug that displayed "(null)" string when the cursor was above
  an atom when displaying a non pdb file

Tue Aug 14

- fixed a bug that caused molden not to make a correct z-matrix for
  a pdb file containing nucleotides

Wed Aug  8

- added support with ORCA through reading of the ORCA output, currently
  supported are:

  single point, geometry optimisation and force/frequency jobs

  No support for reading electron density and orbitals

- Added support for hetatm moelcules from the pdb file, which label
  will now light up when the pointer is over the hetatm molecule.
  Clicking the middle mouse button will now bring up a popup menu
  with options to center on the molecule, create a surface of the
  molecule and to find neighbouring aminoacids.

Mon Jun 11

- Fixed bug in displaying residue labels for pdb files containing DNA

- Added the option to have a background color with a gradient.
  When clicking the 'Palette' icon the first option now is 'Background Mode'
  You can choose between "solid color" (the way it was) and ' color gradient'
  To implement this we changed color 13 of the 15 basic molden colors to dark 
  blue, which is used for the default gradient color.
  Of course this affects the .moldenrc file so you had best remove it and 
  reset your defaults. The background mode is also saved in the .moldenrc
  file.

- Added the reading of electron density maps in the O format (also known as 
  DSN6 format) these are available form the electron density server in Upsala:

  http://eds.bmc.uu.se/eds/

  These maps are only displayed in the OpenGL mode (so gmolden).

Wed May 30

- Fixed a bug in writing .mol filee, of elements with a two letter abbreviation
  only the last was written.

- added reading of DSN6 electron density maps, these can be read in from
  the surface icon -> last option and can be displayed on top of a pdb 
  structure

- added information how to compile under gfortran/gcc4.0 and higher to
  the makefile and molden.f file (uncomment external gfortran_iarc)

Fri May  4

- Fixed bug of displaying the dipole moment, the dipole moment
  was only correctly displayed for single point QM calculations, not
  QM optimisations. Fixed it.

Tue Mar 20

- Removed a bug affecting the processing of biosym .car files.
  Bug was introduced when adding dynamic memory allocation for protein
  coordinates

- Removed a bug affecting reading of xyz files being added to an existing
  structure.
  Bug was introduced when adding dynamic memory allocation for protein
  coordinates

Mon Mar 12

- Removed a bug affecting the visualisation of vibrations of gamess-us
  and Qchem introduced when adding dynamic memory allocation for protein
  coordinates

Fri Mar  9

- Removed a bug in the processing of gaussian outputs, the bug is
  'fixed' but be aware that it is not exhaustively tested as of yet

- made a minor bug fix to the reading of QChem output

Fri Mar  2

- Performance update: OpenGL density mode now makes less unnecessary 
  screen updates. Only important for working with remote displays.

Tue Feb 20

- Created dynamic memory allocation for proteins (all formats that support
  protein information, pdb, mol2, tinker xyz, .gro)

- Display of dipole moment, if QM electron density is available, the QM dipole
  moment is first calculated, if charges are available (but no QM electron 
  density), the dipole moment is calculated from the charges.

- Added possibility to rotate ligand within an active site:

  The align icon (middle column, 9th from the top) Interactive docking -> 
  Pick molecule from screen -> pick a molecule from the HETATM records of 
  the Pdb file.

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

Molden 4.5

Fri Dec 2

- Optimised protein display in the opengl version:

  - residues now have their own gllist, when you display a whole
    protien as solid you will see the protein being build up residue
    per residue. Solid protein display will also display somewhat faster.
  - when clicking the backbone button you will immediately see the
    secondary structure elements
  - animation of variables in the zmatrix editor in solid mode will
    be a lot faster, also replacing rotamers.
  - you can now choose between the richardson and dunbrack rotamer
    libraries

Tue Oct 17

- Did some improvements to the QChem interface

Thu Sep  7

- Did some optimisation of orbitals and regular density space filling
  plots.

Thu Aug 31

- changed the fortran code so that use of the -O2 compiler flag on linux
  nolonger causes incorrect display of the surface in xwindows lines mode
  (the surface looked incomplete)

Thu Aug 17

- The opengl version of molden (gmolden) would show dark blue lines
  in the density-3dgrid mode with certain xwindow emulators. Fixed it
  so that the lines are white again i

Wed Jul 12

- added a button "to Xmgr" which when clicked will right a file with
  pairs of frequency,intensity on each lines. These files can be imported
  into xmgr or xmgrace

Mon Jul 10

- when displaying vasp POSCAR or CONTCAR files, and there is a XDATCAR
  file present molden will ask to use the XDATCAR, the multiple structures 
  can be called with the 'fisrt','next' and 'movie' buttons.

- when doing interactive docking sometimes extra hydrogens were added
  to the second molecule with erroneous connections to the first molecule,
  fixed it.

- when doing interactive docking you are now asked if you want to align
  the centers of both molecules or not (yes used to be the default)

- the make option 'make noxwin' now should work again. This can be used
  to create a version of molden without xwindows and opengl, you can
  only control this version with a keyword file.

Wed Jun 28

- The internal opengl representation of mopac .gpt files had errors
  in representing the connections of the molecule.

- Tinker xyz files which use the amber or charm forcefields are now
  correctly visualized when atoms like F, P Cl and Br are used.

Tue Jun  6

- B. van Eijck added the keyword CHADD, which will read extra charge
  centers from a 'molden.def' file.

Tue May 30

- Removed some small bugs, reported by Robert Gyepes
- added a special makefile for cygwin makefile.cygwin

Fri May 19

- Added 'best rotamer' option, this calulates for a particular residue,
  the best scoring rotamer according to the dfire protein-protein score.

- Removed a bug which on some platforms would make molden crash when
  writing out a z-matrix

- Some rudumentary support for gfortran (you still have to edit molden.f
  and comment out the line 'external iargc' )

Wed Mar  8

  Spherical G-functions (9G) in combination with electrostatic potential
  and mulliken charges are now also fully supported

Mon Mar  6

- Support for Gaussian cartesian G-functions (previously i was unaware of
  G-functions in the Gaussian series of packages).

- Support for the Gaussian03 revision D, the output format of the basisset
  has slightly changed. Adpated molden to deal with it.


Tue Feb  7

- First support for QChem outputs: For molden to be able to read orbital
  and basis-set information, supply the following qchem keywords in the $rem
  section:

  PRINT_ORBITALS          TRUE
  PRINT_GENERAL_BASIS     TRUE

  You can supply output of composite qchem jobs, molden detects it and will
  allow you to select the different job outputs by clicking on the 'M' button
  which appears in the "Select Point:" section in the molden control window.

  NOTE: support is still rudimentary, i dont have qchem, only a few outputs
        sent to me by a user. If you think this functionality is important
        than let me know when it doesnt work

Thu Jan 19

- Support for dynamic memory allocation for the part that shows geometry
  convergence, implemented for use with gromacs and cpmd.

- In popop menus you always had to choose a selection (sometimes cancel)
  Now you can click besides the popup window to get out of the popup and
  cancel the selection. Therefor the option "Cancel" there where used to 
  be offered has been removed.

Tue Dec 20

- Support for the latest Gamess-US version (tested 27 JUN 2005 (R3) version).
  The basis-set print out of Gamess-US version had changed.

Fri Dec 16

- When writing PDB files, in case of no chain information read in 
  molden will now generate and write chain labels

- With gromacs files the generation of connectivity is now much quicker
  also solvent molecules will be rendered in less detail in opengl molden
  which will speed things up a little

Fri Dec 9

- Molden can now read the binary gromacs files .trr and .edr. Previously
  molden would first convert them to their ascii counterparts with the gmxdump
  utility. This way the very large ascii files are nolonger needed.

Tue Nov 29

- Added molden keyword 'MOGL' which will let you create a molden.ogl
  file from a molden keyword file.

Mon Nov 28

- When replacing one amino acid residue by another, followed by writing
  a pdb file, the replaced residue number would be wrong: fixed

- When writing .mol2 files, sometimes a double bond to many was written
  (conjugated systems), since molden does not do anything with bond types
  this was not serious, but other programs (sybyl) do work with bond types.

- When Cbeta or backbone O (C=O) was missing on a pdb residue, connections
  to a non existing atom at cart.coord (0,0,0) would be drawn in the opengl
  version of molden (gmolden)
 
Mon Nov  7

- With gmolden the aminoacid currently under the pointer is shown by
  temporarily showing the label, when clicking with the second mouse button
  a popup window with a few options pertaining to amino acids would appear,
  every now and then this window would not be visible untill you moved the
  pointer over it (only on certain xservers), fixed it.

Wed Nov  2

- Fixed writing of Gaussian inputfiles for oniom calculations (from the submit
  Gaussian Job windows). This bug was introduced when charges for proteins
  were added (amber forcefield) in molden4.1 (Why does no one ever tell
  me it does not work ???)

- Changed the working of the residue commands window:

  128-130:A   would mean residue 128 through 130 of chain with Label "A"

  NOW all numbers refer to residue numbers used in the PDB file:
  128-130:A now actually means the same as 128-130.
  You still can use the chain label for things like coloring:

  *:A col

  Will let you pick a color with which all residues belonging to chain "A"
  are colored with this color

  128

  residue with PDB residue number 128 will be shown.



Fri Oct 28

- fixed a bug that would override user assigned sybyl atom types, when
  writing mol2 files

- added the possibility to change the potential levels at which certain colors
  where used for the colorcoded opengl esp potential 

- The PC-UNIX version of Gamess-US misses the first column with respect
  to the regular Unix version, made some adaptation to let molden work
  with this output to.

- Fixed geom conv for a series of xyz structures where so of them had 'heat'
  0.0

 
Fri Sep 30

- reorder Z-matrix, clicking a few atoms and then hitting 
  escape, this would create a zmatrix in the order of clicked atoms
  and the rest would be figured out by molden. This did not always
  work. I fixed it in such a way that it can handle nearly all cases
  but there are still some conditions not supported, but this is not
  so relevant since you can always do it again.

- When doing interactive docking, you can now change the center around
  which the ligand rotates (in the same way as the overall rotation center).
  And you can now reorder the z-matrix of the docked ligand.

Wed Sep 23

- Added the possiblity to do a surface per residue in the surface window
  (opengl version only), takes longer but also looks a lot better.

- When aligning or docking and then writing as a PDB file some of the atoms
  were missing, corrected it.

- When docking a molecule, the to be docked molecule is placed at the center
  of the molecule docked to.

Wed Sep 21

- Fixed running tinker in the background for the opengl version of molden
- Made residue commands concerning surface also work for the opengl version
- Added Fit Rotamer option, which will replace the amino acid conformation
  with the closest rotamer in the rotamer library

Mon Sep  5

- Added support for the tinker amoebapro force field
- reading in tinker xyz amber files would be retyped to charmm types
  fixed this

Mon Aug 15

- Fixed xwin.c for molden without OpenGL

- Fixed writing of GIF files for using Xservers with 16 bit planes

Fri Jul 15

- Added interactive opengl support for the density mode (only available
  with 'make gmolden').

- Fixed bug which can lead to not displaying gaussian outputs correctly
  in some cases.

- Fixed a number of small annoyances (dont worry, there are enough left).


Wed Jun 15

- fixed bug that let f-functions run in to the code meant for the g-functions
  resulting in reporting number of orbitals is wrong

- fixed reading MOL files, this bug was introduced with support for reading
  gromacs files

Thu Jun  2

Addition of creating opengl surfaces in the molecular mode, when clicked
add to the backbone (protein structure only), and clicking with the second 
mouse button on a backbone atom a drop menu will appear, which let you
do things like (un)display the residue, center it, and add to the surface
(of course only available when molden is compiled with opengl support (gmolden))

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

		Molden4.4

Mon May  9

The main reasons for this new minor version of molden are:

- Introduction of the incorporation of interactive Opengl in the main
  interface of molden (molden already had opengl helper program moldenogl)
  You can build the gmolden version by typing; 

      make gmolden

  This version is still under development

- Reading of NLMO's of the NBO analysis in gaussian

- Partial support for G-functions. As far as orbitals and electron density
  G-functions are supported. Also mulliken charges, except when using 9G
  functions. The electrostatic potential is not yet supported.

  Addition of a new keyword to the molden format: [9G]

  which specifies the use of spherical G-functions instead of gaussians.

  The order in which molden expects MO coefficients to pe printed is:
  For spherical G-functions:

  G 0,G+1,G-1,G+2,G-2,G+3,G-3,G+4,G-4

  For cartesian G-functions:

  xxxx yyyy zzzz xxxy xxxz yyyx yyyz zzzx zzzy 
  xxyy xxzz yyzz xxyz yyxz zzxy

  The request for support for G-function came from the molcas development
  people, but also Gamess-US will benefit from this support.

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

		Molden4.3

Fri Apr 29

- introduced number of small bug fixes:
  the environment g03root was not recognised and g98 was written in the
  .run files created by molden

- premature end of file in gaussian can let molden crash, removed one
  instance on which it could crash

- Introduction of the incorporation of interactive Opengl in the main
  interface of molden (molden already had opengl helper program moldenogl)
  You can build the gmolden version by typing; 

      make gmolden

  This version is still under development

- Reading of NLMO's of the NBO analysis in gaussian

- addition of dynamic aminoacid labels and clicking on a backbone
  atom will bring up a small dropdown menu with visualisation options

Wed Feb 23

- Fixed bug that effectively disabled the aligning of two molecules

- Fixed a bug in the z-matrix editor that would not correctly show all
  atoms in the when the 'stickcolor' mode was disabled.

Fri Jan 28 2005

- Lots of small bugfixes, as there are:
  
  Fixed editing z-matrix of ligand when doing interactive docking,
  Not edit operations where bug free

  Fixed updating of properties when editing the z-matrix

- Added support for the Gromacs (gmx) file format .gro:

  Molden will read the file name.gro and will then look for the file
  name.top to read connectivity, charges and atom types. When not found
  Molden will create connectivity.

  Next molden will look for a trajectory file name.ttr. If found it will
  use gmxdump to convert it to an ascii trajectory file name.trra.
  Molden will not do the conversion if a ascii trajectory file name.trra
  already exists.

  Next molden will look for an energy file name.edr. If found it will
  use gmxdump to convert it to an ascii energy file  name.trra.
  Molden will not do the conversion if a ascii trajectory file name.trra
  already exists.

  If a trajectory file is found, then molden can animate the MD run by
  clicking on the movie button.

- Hitting the escape key when running an animation will exit the animation.
  Hitting the space key will stop the animation and when hit again will resume
  the animation.

Fri Jan 14 2005

- Some more work on the writing of mol2 file for proteins.

- Incomplete amino acids are now reported.

- changed xwin.c so it does better on 64 bits AMD/Opteron systems.

Tue Dec 21 2004

Introduction of Molden4.3:

- The reading of Gaussian output files has been restructured, it is
  possible that this may temporarily be a step back in molden reliablity.
  PLEASE inform us of any problems related to reading gaussian outputs.

- Did some work on writing better mol2 files for proteins, there are
  still problems, but at least molden can now read it back and create
  a z-matrix for the protein without crashing.
  Sybyl still has problems with these files ! It needs more work.

- Added forces information to the molden format, example:
 
  [FORCES]
  point    1
  3
    -0.000038    -0.000160     0.000430
     0.000834     0.000150     0.000047
    -0.000263     0.000008    -0.000017
  point    2
  3
    -0.000505     0.000339     0.000987
     0.001268    -0.000634    -0.000242
     0.000123    -0.000017    -0.000640
  ....
  point   10
  3
    -0.001299     0.000970     0.001904
     0.002139    -0.001328    -0.001049
     0.000512     0.000004    -0.001636
  
 The number of points must match the number of points used in the [GEOMETRIES]
 section.
 Second form includes coordinates, this is only needed when the forces are
 not in the same orientation as the geometries supplied in the [GEOMETRIES]
 section. For instance, the Gamess-US forces are not in the z-matrix orientation
 , so when [GEOMETRIES] ZMAT is used the forces would not be in the correct
 orientation.

  [FORCES] COORD
  point    1
  3
    -0.000038    -0.000160     0.000430   0.0  0.0  0.0
     0.000834     0.000150     0.000047   0.0  0.0  1.0
    -0.000263     0.000008    -0.000017   0.0  0.0 -1.0
  ....

  NOTE: this extension is on trial, it may not be the final solution.
        Developers who write their own molden format should not be
        overanxious to implement it.

 - fixed problems with reading circulair peptides.

 - increased the max number of geometries and scf iterations molden can handle
   to 2000 and 1000 respectively.

------------------------------------------------------------------------
		Molden4.2

Tue Oct 05 2004

- Corrected a few fortran syntax errors, thnx to Gyepes R. 
- on some platforms molden would not write a z-matrix, after stating
  that it just did. fixed

Thu Aug 05 2004

Molden 4.2 introduces quite some radical changes in the way proteins
can be manipulated with the z-matrix editor. Where in previous versions
any change to the z-matrix was followed by a complete update of the 
cartesian structure, with molden4.1 only those parts of the cartesian 
structure are updated that have actually changed. 
The advantages are that protein manipulation is much faster and now
larger proteins can be manipulated, these manipulations include,
deletion, insertion and replacement of amino acids. Also the animation
of variables such as the chi sidechain rotamer angles can now be done
interactively. A new feature is the possiblity to change side-chain rotamers.
The procedure is now:

1) load in a PDB file
2) click the zmatrix button and create a zmatrix for this protein
3) click the backbone button
4) NEW click on the backbone with the SECOND mouse button and a popup
   window with side-chain commands will appear: 
B

   switch on display of this amino-acid, replace amino acid, replace rotamer

5) Switch to full zmatrix to edit individual atoms
6) click in the zmatrix on the Chi1-4 edit boxes with the second mouse
   button and select "animate", click the edit box again to end the animation.

7) click in the Molden Control window the button "Hetatm" and switch on
   Helix, Beta sheet, and coil and optionally any hetatm records.
8) click on the Vr icon (7th from the top) and select OpenGL
   (you need to have the molden opengl helper program installed for this,
   moldenogl) to get a high quality opengl display of your protein.

The z-matrix is only used for manipulation of the protein, a zmtrix of a
protein should never be saved, instead the updated result should be stored in
PDB file.
The disadvantage of this new approach is a much more complicated algorithm,
in which undoubtable some bugs have found there way home.


------------------------------------------------------------------------
		Molden4.1

Wed Jul 21
		- The zmatrix conversion for proteins containing nucleotides
		  has been improved, it will now create a zmatrix.
		  Still No manipulation of nucleotides however.
 

		- The windows version of molden now start the opengl
		  helper program mogl, when you click a *.ogl file
		  in the read file window.

Thu Jul 15
		- Now when you apply the tinker amber force, automatically
                  the charges associated with the atoms belonging to 
                  amino acid residues are set.

		- A user reported that within gaussian03 you can have the
		  condition that you use pure D functions (5D), but
		  use cartesian F functions (10F). Molden did not handle
		  this correctly, it would assume the use of all pure 
		  functions (5D and 7F) or all cartesian (6D and 10F).
		  Molden can now handle this mixed use of pure and cartesian
		  D and F functions.
		  This also has ramifications for the molden file format:
		
		  Previously the default was 6D en 10F, the use of the keyword
		  [5D] would mean 5D en 7F. The new format uses:

                  - default is    6D and 10F
		  - [5D7F]  means 5D and  7F
		  - [5D10F] means 5D and 10F
		  - [7F]    means 6D and  7F

 		  for backwards compatibility, [5D] still means 5D and  7F.

		- Fixed a bug that would molden not read the Pseudo
		  Potential information of a G03 run (older versions
		  were ok).

		- Added the [PSEUDO] section to the molden format:

 		  [PSEUDO]
		  Fe 1 16
		  ....

 		  This is elementname followed by the number of the atom
		  and the effective core charge.

Wed Jun 16
		- fixed bug that would let the windows version of molden
		  crash when you go to the density mode without having
		  read in a file (for the benefit of reading gaussian 
		  cube files)

		- Molden now checks whether the contraction coefficient
		  it reads in via gaussian output files and molden format
		  files contain the normalisation of the contracted
		  gaussian.

		- increased the maximum number of atoms in the density
		  mode from 300 to 2000, this to the benefit of oniom 
		  calculations. (this is independend of the max number of
		  basis functions in molden).

Wed Jun  9
		- Added interactive docking:

		  - start molden with a PDB/mol2 file containing the pocket
		  - Click the icon with the two phenyl rings overlayed:
		    (In the Molden Control window, the nineth icon from the
		    top). Choose 'Interactive docking'		  
		  - Supply the name of the PDB file containing the ligand.
		  - Follow on screen instructions for rotation/translation
		  - the Docking Score window will appear, when you rotate/
		    translate the ligand, the PMF score will be updated
		    in this window.
		  - the window also allows you to store and write combinations
		    of rotation/translation of the ligand (pose).
		    The filename pose.xx will be used for written poses.
		  - The AtomInf button lets you display info on the PMF score
		    contributions of a picked ligand atom: click on one of
		    the ligand atoms. The PMF score window will pop-up with
		    a list of the top negative (=favorable) and top positive
		    atom-pair contribution for the slected ligand atom.
		    When clicking on a list entry the corresponding protein
		    atom of the ligand atom - protein atom pair lights of
		    on the screen (clicking again will switch it off).
		  - activating the 'ligand color mode', will color ligand
		    atoms according to the amount the contribute to the total
		    PMF score for the ligand.

Thu Mar 25
		- In the z-matrix editor, you can now use the second mouse
		  button to select in molecule drawing screen an atom.
		  Now a change fragment popup box will appear to facilitate
		  structure manipulation.
 
		- added the ability to read the high level layer orbitals/
		  basis-set of an oniom calculation. Apparently the orbitals
		  are only printed at the first point, NOT at the stationary
		  point of an optimise run.

		- improved the interface to run amber within gaussian.
		  Parameters are automatically assigned (for amino acid atoms!)

		- When using ONIOM with High and Low layers, the charge & 
		  multiplicity check went wrong, fixed it.


 		PLEASE LET ME KNOW if you use the oniom interface in molden
		and have suggestions for inprovement.

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

		Molden4.0

Wed Mar 17
		- When there is an error reading the vectors (orbitals)
		  in gaussian, molden would automatically skip the reading
		  of ESP charges from gaussian. Fixed this.
		- The print format for ESP charges has changed going from
		  g98 to g03. The new format is now supported.
`
Mon Mar 15
		Fixed a bug that lets molden think that any file that
		begins with an empty line is a mopac .gpt file.

Mon Mar  1
		- Molden didnt correctly parse Gaussian oniom output:
		  corrected it.

		- Added better support for the MacOS system, by adding
		  an entry to the makefile for the moldenogl program
		  and adding #ifdef statements to the corresponding
		  C code, ogmol.c and silly.c

Tue Feb 17
		- Fixed the reading of gaussian outputfiles when using
		  the amber forcefield.

Fri Feb 13
		- Fixed small bug that wouldnt let you display beta
		  orbitals with the molden commandfile.
Thu Feb  5
		- Changed common block /freq/ in such a way that
		  compilers no longer complain

Mon Jan 19
		- Fixed a problem with the new Multi molden_format,
		  In case of multiple molecules with vibrational
		  information, this information was not updated.

		- The name of the file read via the filemananger
		  will now be displayed as the name of the main
		  graphics window. Same holds for named [MOLECULE]
		  sections of a multi molden_format file.

		- Changed the code of wrzmat.f, this gave problems
		  on some platforms (AIX).

Wed Jan 14
		- Fixed a bug that presented it self with g98,g03
		  xyz optimisation of a molecule with more than 50
		  atoms. When clicking in the geometry optisation window
		  on a point, the shown geometry would be out of sync
		  with the next/movie button geometries.

Mon Jan 12
		- Fixed a bug that made molden crash every once in
		  a while on LINUX !!!

		- Reduced the flickering of windows when
		  an other window is moved over it for LINUX.

Thu Jan 08
		- support for gaussian with pseudopotentials:
		  Now the proper mulliken charges and density are calculated

		- In rare cases gaussian reduces the number of MO's,
		  making it smaller than the number of basis functions.
		  Molden can now handle this.

Wed Jan 07	- fixed bug reading Molden Format files:
		  when reading two Molden Format files after each other
		  and the number of occupied orbitals in the first was
		  bigger, the second file would have the wrong density
		  matrix due to some old occupation numbers being present.
		  
		- Molden allocates memory for the orbitals dynamically
		  but not so for the basis-set. This could lead to problems
		  with very large calculations. I increased the storage
		  for the basis-set with a factor of 4.

		- Added new feature to the Molden format:

		  Molden format files can now be concatenated in to
		  one file, provided they are separated by the
		  [MOLECULE] keyword line. In addition each molecule
		  can have a [TITLE] line, example:

		  [MOLDEN FORMAT]
		  [MOLECULE]
		  [TITLE]
                  title molecule 1
		  .... rest as usual
		  [MOLECULE]
		  [TITLE]
                  title molecule 2
		  .... rest as usual

		  If Molden reads such a file, a extra button will appear in
		  the "select point" section of the molden control window.
		  When pressed, the titles of the molecule sections will
		  be listed and when selected the appropriate molecule
		  section will be read.
		  Each [MOLECULE] section can have all the previously supported
		  sections/keywords.

2004

Tue Nov 25
		- When reordering a Z-matrix nolonger all atoms have to be
		  clicked. When hitting the escape key molden will try
		  to figure out the ordering of the remaining atoms.

		- Gaussian03 no longer prints just electronic energy
		  during SCF, but total energy, when using the #P keyword.
		  Molden wrongly subtracted the nuclear repulsion energy.
		  Fixed.

		- Added a button near the "first, next and movie" buttons.
		  When clicked this button will open a window in which the
		  delay time (in milliseconds) between one structure and
		  the next. This delay time also has effect on the display
		  of normal vibrations.

Mon Nov 17
		- Molden now fully supports 5D and 7F orbitals.
		  Previously, only orbitals and density were supported,
		  NOT mulliken charges, electrostatic potential and
		  multipole expansion.

Fri Oct 17

		- Via the button "read", there is now the option "add"
		  available as opposed to "replace" which stand for the
		  original behaviour. "ADD" adds a structure to the 
		  existing structure on the screen.
		  Currently this option is ONLY supported for the following
		  file types: pdb, mol2, and xyz
		  This option is mainly useful for comparing different
		  conformations (protein) AND offers the ability to
		  read in a protein and overlay different docked ligands.

		- A mayor overhaul of the PDB editing capabilty via the
		  Z-matrix editor. This will probably need more work, but
		  is quite an improvement over the existing situation.

		- The Z-matrix editor will now try to keep the original
		  orientation and translation when translating zmat to
		  cartesians.

		- Molden picks up the normal modes symmetry tag for
		  Gaussian and Mopac output.

		- prelimenary support for Tinker4.1. Tinker4.1 has
		  different names for the amber and charmm parameters files.
		  Also a few (non residue) atom types have changed, this is
		  NOT YET supported.

		- use of the gaussian keyword hpmodes caused problems 
		  for molden: fixed


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

		Molden3.9


Tue Aug 26
		- calculated spectrum now also supported for mopac

Mon Aug 25
		- corrected bug that didmt read old gaussian freq
		  outputs

Wed Aug 20
		- Corrected a bug in the new calculated spectrum part.
		  This bug only manifested itself when the number of 
		  frequencies exceeded 200.

Mon Jun 30
		- Corrected Gaussian03 frequency reading

		- Now with PDB file writing, the original residue number
		  is maintained

		- You use an expression like *:A to turn on all residues
		  of the A chain

Tue Jun 17

		- Added a feature to display an interactive spectrum

		- Added to the Molden format an item:
		  [INT]
		  intensity1 [raman_intensity1]
		  intensity2 [raman_intensity2]
		  ...
		  intensityn [raman_intensityn]

		- Added the support for gamess-us Natural orbitals

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

		Molden3.8

Thu May 22
		- Gaussian03 basis-set output has slightly changed
		  with respect to Gaussian98, this would cause error
		  in molden with molecules with more than 9 atoms.
		  Fixed it.

Wed May 21
		- When forces or displacement were to large, Gaussian
		  would print *******
		  Molden couldnt deal with that, now it can.
Tue May 20
		- Support for mopac2000 ASCII gpt files
Tue Apr 29
		- 3Ddrc gamess-us IRC support
Wed Apr 16

		- First support for Gaussian03

Tue Apr 01

		- PDB residue numbering is now consistent with those
		  in the actucal PDB file
		- MOPAC grid search output is now somewhat supported
		- Drawing errors for coil secundary structure are
		  reduced
		- In the Z-matrix for PDB entries, HETATM's will
		  be attached to the backbone instead of the closest atom. 

Mon Feb 17
		- Added the ability to write MDL mol files.
		- Added the ability to calculate EEM charges,
		  a sort of gasteiger charges, by Patrick Bultinck
		- The reading of PDB files now handles additional
		  Hydrogen labels

Wed Dec 04
		The oniom stuff had a bug for the linux version, fixed.

Thu Nov 28
		Added an update to the ONIOM window.

Tue Nov 05
		- Complete update of the Gamess-UK dedicated code by
		  Huub v. Dam:

		  - Reading of SCF-convergence for CASSCF and DFT
		  - Reading of CASSCF orbital occupancies
		  - Support for HESSIAN jobs added
		  - and more ...
Thu Oct 17
		- Added support for the CPMD program
		  Distributed with the permission of TEODORO LAINO
		  NEST Laboratories - INFM - Scuola Normale Superiore
		  Some work still needs to be done.

Fri Oct 11
		- Added makefile entry for Mac OSX, thanks to Eric Brown.

Wed Sep 25

		- Molden3.8 has support for tinker3.9, older versions of
		  tinker are not supported (Molden3.7 support tinker3.6)
		  (Thanks to Nicolas Ferre)

		- Nicolas Ferre added support for ONIOM calculations with
		  gaussian98. 

		- In submit gaussian window you can specify
		  you want to write cartesian coordinates instead of z-matrix.

		- You can now run amber from gaussian, in the following way
		  Read in a pdb file, click the z-matrix button and create
		  a z-matrix, assign tinker amber atom types (the FF button)
		  Go to submit job option gaussian, choose method "amber"
		  "Write XYZ" form the "Gaussian Job" window.
		  In principle this can be combined with the oniom method
		  but this is not tested yet.

		- When introducing Mopac2000 support a bug was introduced
		  which interfered with the reading of tinker xyz files, fixed.


		Molden3.7

Wed Sep 25
		- When introducing Mopac2000 support a bug was introduced
		  which interfered with the reading of tinker xyz files, fixed.

Mon Jul 08
		- Correction added to the postscript plot.
		- added F and Cl as hydrogen-bond donors

Mon Apr 08	
		- Gamess-US version 16 FEB 2002, has a slightly different
		  basis-set section. Molden now supports this.

Fri Apr  5	- Added support for reading Mopac2000, out and arc files.

Wed Mar 20
		- added some small changes to the automatic xyz->z-mat
		  conversion 

Wed Mar 13
		- when mol2 files would miss residue information
		  molden would display the structure, fixed

		- when reading xyz files containing multiple structures
		  containing energy values in the title line, energy values
		  of zero, would not be correctly read on the second time
		  a multiple xyz structure was read. Fixed this.

Tue Mar 05
		- The introduction of a extra line "FORCE CONST.:" when
		running a force job with gamess-us, caused molden
		to not find the normal coordinates. Fixed this.

		- When running a Gaussian job using F functions, molden
		incorrectly calculated the ESP charges. Fixed this.

		- added an option to read the ESP charges calculated by Gaussian
		Under the surface icon there is a new option "read ESP charges"


YEAR 2002


 
Thu Nov 15
		Adapted the script multi_vrml_orbitals.csh to deal with
		the changed VRML orbitals.

Tue Oct 02
		The Gamess-UK orbital occupancy section has changed,
		molden now deals with it.

Thu Aug 30

		- Assigning sybyl force field atom types sometimes failed
		  to assign aromatic types, fixed it
		- Now support for different spacegroup setting in mol2
		  file (the CRYSIN line, setting field (75-142 alternative
		  settings are NOT supported)).
		- Molden now reads mol files, only 3D tested. The mol file
		  is recognized when the string V2000 or V3000 is present.

Wed Jul 25
		The Gamess-UK optxyz output has changed, molden gave errors
		with these new outputs, fixed it.

Tue Jul 17
		- Added a conversion program to convert CADPAC output to
		  molden format (cad2mol). This program was written by
		  Mariusz Klobukowski (Mariusz.Klobukowski@ualberta.ca).
		  This program is now distributed together with the molden
		  source, to compile it type: 

		    make cad2mol

		  To run it:

		  cad2mol 'cadpac_output'.out

		  A file 'cadpac_output'.mdn will be created, molden will
		  be able to read it.


Mon Jul 09
		- Molden did not allow to subsequent subtractions of 2D grids
		  (gridfile) from the current (density) grid.
		  Rho_AB - Rho_A - Rho_B was not possible. Corrected for it.
		- Molden nolonger recalculates the density before reading /
		  writing a 2D grid, this make it possible to write 
		  subtracted grids.

Fri Jun 29 
		- Added the possibility to render orbitals transparent
		  in VRML2.0 by clicking on them
		  Commandline flag -k adds background + a light.
		  This is to overcome a very dark scene on some terminals.

Thu Jun 21	- added small correction to the reading of
		  GAMESS-US outputs (when SCFTYP= and RUNTYP=HESSIAN
		  were not on the same line in the input file, molden
		  would not display the vibrations).
		  Thanks to Georgy Salnikov, NMR Group,
		  Novosibirsk Institute of Organic Chemistry, Russia

Thu May 31	- Added support for writing mopac2000 input
		  containing crystal information (Tv vectors)

Tue May 15
		- Added support for mopac2000 output containing crystal
		  information (Tv vectors)

Mon May 14
		- Removed a bug in reading pdb files containing just
		  HETATM records.
Thu May 10
		- REMOVED MAYOR BUG in reading new basis-set section
                  of GAMESS_US !!!!!

Thu May  9 	- Molden can now launch a G98W session on windows 

Thu Apr 19
		- Addded new makefile section for HP-UX
Mon Apr  2
		- The Z-matrix editor option "substitute atom by fragment"->
		  "Read" sometimes passed the wrong filename, fixed it.

Tue Mar 13
		- Molden sometimes crashed on the OPTXYZ output of
		  Gamess-UK. Fixed thanks to Huub van Dam.

Wed Mar 07
		- Molden now deals with changed writing o MCSCF orbitals
		  of Gamess-US.

		- Gaussian98 by default only prints the Z-matrix and resultant
		  coordinates ONLY for 50 atoms or less. However molden
		  didnt respond correctly when this default was overridden
		  with the keyword: IOP(2/11=1). Fixed.

Tue Dec 19
		-Added Option to specify opt=z-matrix in the "submit job"
		 for Gaussian, available from the Z-matrix editor.
		 This forces Gaussian to use the Z-matrix when optimising.
		 The default is to use a nonredundant set of 
		 internal coordinates

		- Molden should now nolonger crash on empty files

Mon Nov 27
                Molden format in combination with both alpha and beta
                spin orbitals gave problems when more than 128 orbitals
                were present solved it.

Tue Oct 31

		Added the ability to visualize protein secondary
		structure elements such as helices,strands and coil 
		with the OpenGL helper program. The procedure is
		as follows:

		- you read in your protein

		- you click the "BackBone" button
		  (If you do not, all of your protein atoms will
		   visible in the OpenGL, making it extremely slow,
		   except for small polypeptides)

		- you select ligands via the "HetAtm" button
		  No need to select "Helix","Beta","RNA/DNA" or "coil",
		  they will automatically be included in the OpenGL.

		- select separate residues to be displayed by clicking
		  "Residue" and then clicking on the backbone on the
		  screen.

		- Decide wether you want the residues and ligands to
		  be displayed as tubes (default) or as spacefilling
		  The latter can be switched on by clicking "Solid"
		  -> "Space Filling"

		- Decide wether you want the residues and ligands to
		  colored by their elements or by residue/ligand color
		  This is controlled by the "StickColor" button.			

		- Now click the "Vr" icon and select "OpenGL"
		  Molden will create a file called "molden.ogl"
		  and start up the OpenGL helper program "moldenogl"
		  Remember to rename this file if you want to keep
		  it, otherwise molden will overwrite it the next time.

		- At anytime you can visualize this file again, without
		  having to start molden by typing:

		  moldenogl 

		  or:

		  moldenogl molden.ogl

		- use the moldenogl cammand line flag "-r" is you
		  want to use high quality spheres and cylinders.
		  This will increase the graphics quality at the expense
		  of speed of rotation.

		- Use the left mouse button to rotate your molecule.
		  In combination with the "Ctrl" and "Shift" button
		  respectively, a zoom and translation are performed.

		- The third mouse button will bring up menu with
		  lots of option, of which the "Screen Capture"
		  option allows you to save your picture as screen
		  dump postscript or "Vector Postscript". The latter
		  take longer but will yield a better picture on a 
		  native postscript printer.

		- molden has a new commandline option "-z", this
		  will let molden use more polygons and will
		  result in a better opengl picture (at the expense of
		  speed, but ideal for a good postscript picture
		  in combination with the moldenogl -r flag). 

		- you can edit the "molden.ogl" file,
		  specifically the following section is important:

		  [COL STRANDTOP] 0.0 0.0 1.0
		  [COL STRANDBOTTOM] 1.0 0.0 1.0
		  [COL HELIXOUT] 0.0 1.0 0.0
		  [COL HELIXIN] 0.6 0.6 0.6
		  [COL RNA] 0.5 1.0 0.5
		  [COL COIL] 1.0 1.0 1.0

		  This defines the color of the beta strand (first two),
		  the helix outer and inner surface (following two),
		  and the RNA and Coil parts respectively.

		  They are used further down the file:

		  [RIBBON] HELIXOUT

		  Alternatively you may supply a color per ribbon,
		  by supplying rgb triples (red,green,blue):

		  [RIBBON] COLOR 1.0 0.0 0.0


		In molden you can now also select "coil".
		In addition you can click on the helix,strand or
		coil to activate a residue (in combination with the
		"Residue" button.

		Also added the writing of POVRAY files of the ribbon
		cartoons as desribed above. Follow the same procedure
		to generate then, except that you have to choose
		"Povray" under the "Vr" icon.

		Molden3.6:

Fri Sep 15
		- Gamess-us (After Jan10 2000) outputs gave problems
		  on 64-bit Alpha-linux systems, corrected it.
		  G77 compiler flag -mieee has to be added.

Tue Sep 12
		- Added screen capture possibilities to the OpenGL 
		  helper program moldenogl. Three options:

		  - IRIS RGB format (key F10)
		  - greyscale encapsulated postscript (key F11)
		  - PPM bitmap file (key F12)

		  added a menu option "screen capture" holding the
		  above three options
		  
		  Thanks to Pedro Vazquez for supplying the code for
		  postscript and PPM writing.

Mon Aug 21

		- Problem with reading shelx files generated by
		  the CSD's conquest code:

		  SYMM .5+x, .5+y, z

		  Molden expected all capitals, fixed to handle 
		  lower case. (All commands must still be in uppercase
		  however).

		- Trying to edit the cell parameters of a structure
		  read in as shelx file, made the program core dump, fixed.

Thu Aug 17
		- Added keywords to enable mapping of properties
		  onto a isosurface with Gaussian Cube files, example:

		  title
		  GAUCUB=meto_dens.cube MAPFIL=meto_esp.cube VRML2.0 
		  SPACE=0.001 MAPVAL=(-0.3,-0.1,0.0,0.1,0.3)

		  GAUCUB for the Gaussian cube containing the density
		  MAPFIL for the Gaussian cube containing the property
		         (like the electrostatic potential)
		  MAPVAL use to change the default property values
		         associated with the 5 basic colors used
			 for interpolation
			 (default values: (-0.1,-0.05,0.0,0.05,0.1))

		  
		  Remember currently mapping only works with VRML2.0
		  and OpenGL output.

		- Added a program called 'short_wrl' written by
		  Andreas Klamt of COSMOlogic, to remove redundant
		  vertices from moldens VRML2.0 files
		  To build it type: make short_wrl
		  To run it:  short_wrl molden.wrl
		  Now molden.+.wrl contains the new VRML2.0 file

Fri Aug 11
		- The GAMESS-US basis-set output section has changed
		  molden has been updated to deal with this

		- Molden now also writes a '[LINES]' section to
		  the molden.ogl file. This is the input file for the
		  molden OpenGL helper application moldenogl.
		  Older versions of moldenogl may have problems with
		  this new section.
		  The [LINES] section has been constructed to visualize
		  the cell and alpha helices and beta strands.

		- Added a few options to the moldenogl, when clicking
		  with the third mouse button in the main drawing window
		  you find a menu with as new option 'Animation'
		  this options under it will allow you to stop/start the 
		  animation, advance by one point (also key 'n' or 'N')
		  and lets you slow down animations.

Fri Jun 30

		- GAMESS-US has a changed writing when using pure 5D
		  molden now deals with in (Thanks to Fred Arnold)

		- the moldenogl helper program can now also be called
		  from the 'molecular mode', it will do normal vibration
		  animations and animations of molecular reactions.
		  (Via the VR icon->Opengl).
		  the molden.ogl format now accepts more molecules.
		  Added the option -l,-p,-b number to moldenolg

		  -l = low resolution molecule (faster)
		  -p = perspective mode on
		  -b number = choose background color

		  Added also to the popup menu, including a fullscreen
		  option, and a light2. (three lights 0,1,2)

Mon Jun 26

		- The routine reading XYZ files had a problem reading
		  actinium 'ac', fixed it
	
		- Added a number of things to the opengl helper program
		  moldenogl, accesible via the menu (third mouse button click):

		  + transparency toggle (or 't')
		  + switching on/off of lights 1 through 3 (or '1','2','3')
		  + fullscreen option (or 'f')
		  + perspective mode (or 'p')
		

Thu Jun 15

		- Added the possibility to create an isodensity surface
		  colorcoded with the electrostatic potential.
		  Needed are two gaussian cube files or two molden
		  3dgridfile's (you can create these by clicking
		  'write grid' after a 'space' calculation. 
		  Molden will always create a file called '3dgridfile')
		  Currently only "Opengl" and VRML2.0 output is supported
		  You will need the updated version of 'moldenogl'
		  to view the OpenGL output.
		  For an explanation how to to this see URL:

		  http://www.cmbi.kun.nl/~schaft/molden/mapped.html

		  For an example VRML2.0 file:

		  http://www.cmbi.kun.nl/~schaft/molden/wrl/gamuk.wrl

		  Added to 'moldenogl' opengl helper is a menuitem
		  (Third mouse button click) 'transparency'
		  In the molden file select window, a click on a file
		  ending with the name '.ogl' will bring up moldenogl
		  with this file.

Fri Jun 09
		- When a job required more than the default 256
		  orbitals, molden allocated more memory, but as a result
		  the HOMO and LUMO buttons of the xwindows interface
		  nolonger worked correctly. Fixed it.

Mon Jun  5
		- Jeffrey Frey and James Hess have suggested a small
		  adjustment of the [MOLDENOGL] format.
		  You can now adjust the transparency of a surface by adding
		  a number:

		  [SURF] TRANS 0.75

Wed May 31

		- When reading molden format files with more than 256
		  orbitals, some arrays were overwritten, fixed it.

		- In z-matrix editor: the check box to specify
		  Gamess-US style z-matrix writing now print a
		  small explaining message when operated

Fri May 12
		- Molden generated Povray orbitals had swapped
		  colors with respect to VRML and OpenGL orbitals
		  This is corrected.
		  (blue = positive, red= negative)


Thu Apr 27

		- BIOSYM/MSI car files for non-periodic work again

		- In z-matrix editor: instead of "clear status all variables"
		  you now get the option to set the status to variable
		  or constant

		- In z-matrix editor: Added a check box the specify
		  Gamess-US style z-matrix writing

		- In z-matrix editor: when typing "c" in the main drawing
		  window molden will look for a file "mapfile" which
		  contains coordinates to orient cartesian coordinates
		  after z-matrix -> cartesian conversion
		  Example "mapfile"

		  1 50.006  81.343  67.412
		  2 50.599  80.892  66.194
		  3 49.514  80.759  65.162

		  Contains coordinates (angstrom) of atoms numbered 1,2,3

Thu Mar 30

		- Molden now nolonger always creates the file 'plot'
		  The xwindow interfaces doesnt, here you generate
		  a postscript plotfile by clicking the corresponding button.
		  Only when supply you molden with a keyword file it will 
		  create the plot file

Thu Mar 23
		- Added primitive support for Jaguar cube files
		  No molecular geometry is displayed however.
		  Start molden without a filename on the commandline
		  Go to the density mode, click "Read Cube", chose "Jaguar"
		  supply a filename, then click the "space" button and supply
		  a contour value.

Tue Mar 14

		- Fred Arnold added support for reading Gamess-us 
		  orbital localisation schemes; Pipek-Mezey and 
		  Edmiston-Ruedenberg

Thu Mar 09

		- Changed povray writing didnt work for orbitals,
		  corrected by Fred Arnold

Wed Mar 08

		- Under edit cell parameters, you can now also specify
		  a spacegroup name, instead of just a spacegroup number.

Fri Mar 03

		- support for Gamess-us version with long contraction
		  coefficients. 

Tue Feb 29
		- Fred Arnold updated the writing of povray files.

Thu Feb 24

		- Molden now nolonger dies when you kill one of its windows

		- Added a restore color options, under the color palette
		  icon

Wed Feb 23
		- molden used to center a molecule by working
		  on the atomic coordinates directly.
		  Now this is accomplished via a viewing transformation
		  the coordinates are not affected.

		- In principle now the conversion of a z-matrix to
		  cartesian takes into account the original orientation
		  of the molecule in cartesian space

Fri Feb 11
		- Added writing of CML molecules (primitive)

		- With the introduction the new style of rotation
		  there was a clash between the use of control 
		  for zooming in and for temporarily escaping selecting
	  	  of atoms, under the "Reorder Z-matrix" button of the
		  Z-matrix editor

Wed Feb 02

		- Added dynamic memory allocation for reading mopac .gpt
		  files.

		- corrected reading of *.ogl files by moldenogl, 
		  created with mopac .gpt files.

Mon Jan 31
		- added commandlineflag to adjust the maximum number
		  of Gif files molden will write (default 100)
		  molden -j 2000
		  Molden will write 2000 files at the max.

Fri Jan 28
		- While adding OpenGL support some errors were 
		  introduced for the platforms VMS, CRAY and AIX

Mon Jan 24

		- Full support to view orbitals/density etc. in OpenGL
		  This is accomplished via a helper program called
		  'moldenogl', this needs to be build separately.
		  (make moldenogl), you need to place this program
		  in a directory which is in the default path, say
		  /usr/local/bin. You can evoke it from the Xwindows
		  interface under the 'Vr' icon, option OpenGL.
		  Molden will write a file 'molden.ogl', which will
		  be read by the helper program. If you want to save
		  this file you will have to rename it before you
		  choose OpenGL again. You can run 'moldenogl' standalone.

		  moldenogl file.ogl
		  
		  This has been tested for Silicon Graphics and linux.
		  The latter needs the MESA graphical library
		  have a look at URL: http://www.mesa3d.org/
		  You may need to activate the 'LIBSG' line for your
		  platform. 

		  HOW TO OPERATE MOLDENOGL:

		  Use the mouse to rotate the scene, drag while holding
		  down the left mouse button.
		  Further one letter commands:

		  Left Mouse button           - rotate
 		  Left Mouse button + Shift   - translate
		  Left Mouse button + Control - zoom in / out
		  Right Mouse button          - bring up options

		  m                           - toggle molecule
		  x,y,z,X,Y,Z                 - move position of light


		  PS. On linux the mesa libraries are installed in the
		      the directory /usr/local/lib. This may give the
		      problem that moldenogl cant find these libraries,
		      you can remedie this by either copying the libraries
		      to /usr/lib, or setting the environment variable;
		      LD_LIBRARY_PATH, in say /etc/csh.cshrc
		      (setenv LD_LIBRARY_PATH "/usr/lib;/usr/local/lib")


		 You generate .ogl file from a molden keyword file by use
		 of the keyword OPENGL.

		- added comandline option -t, which will let molden read
		  chem3d style ascii .gpt files (mopac93)

		- added a threshhold for plotting displacement vectors
		  in the postscript output of a normal mode.
		  You can set it via the -y commandline flag:

		  molden -y 2.0 

		  (default the threshhold is 0.0, every arrow is plotted)
 
		  Thanks to:

		   Valentin P. Ananikov
		   NMR Group
		   ND Zelinsky Institute of Organic Chemistry
		   Moscow, Russia
		   e-mail: val@cacr.ioc.ac.ru

		   (URL: http://nmr.ioc.ac.ru/Staff/AnanikovVP/molden-t/
		    PS: the path uses the -t commandline flag, this is now -y)


Thu Jan 13

		- The color editor didnt work well on linux, fixed it.

Wed Jan 12	
		- Added first primitive support for OpenGL,

		  THE BELOW IS OUTDATED HAVE A LOOK AT THE ENTRY FOR Jan 24

		  read the makefile on how to activate it.
		  It only works for Silicon Graphics machines and
		  linux machines. The latter needs the MESA graphical
		  library. Currently only a the equivalent of the 3dx mode
		  is available.
		  You can only activate the OpenGL via a molden keyword file,
		  and the use of the keyword 'silly', file 'molin'

		  Title
		  file=test/adf.molf silly homo

		  ./molden molin

		  Use the mouse to rotate the scene, drag while holding
		  down the left mouse button.
		  Further one letter commands:

		  m toggle molecule
		  . zoom in
		  , zoom out
		  x,y,z,X,Y,Z move position of light

		- Added the another type of molecular rotation model.
		  It is available via a new icon in the lower right
		  corner of the 'Molden Control Window'.
		  In this model you rotate the molecule by dragging
		  the mouse while holding down the left mouse button.
		  Doing the same while keeping the 'Shift' button pressed
		  lets you translate the molecule. 
		  Doing the same while keeping the 'Control' button pressed
		  lets you zoom in/out.

Wed Jan 05

		- Added maxmimum line width selection under:
		  the palette icon, the option: 'shade depth'.

Wed Dec 22

		- Added depth of shade editor, under the palette icon
		  there is a new option: 'shade depth'.
		  This option has no effect on the solid drawing mode
		  with perspective off, only the stick mode and 
		  solid+perspective modes are affected.

Tue Dec 21

		- Added the ability to edit colors, under the palette icon
		  there is a new option: 'edit colors'. This option will
		  edit the basic 15 colors available in molden. It will
		  write the colors toe the file '.moldenrc' in the users
		  home directory. The use of 'molden -r colorfile' takes
		  precendence over the reading of .moldenrc.
		  To reset to the defaults just remove the ~/.moldenrc file

Thu Dec  9

		- the WRZMAT keyword now writes the last z-matrix of
		  say a geometry optimisation, instead of the first

		- added the commandline flag '-=', which causes a 
		  gamess-us dialect of gaussian zmatrix to be written,
		  it has the = sign between variable names and values
		  For use with COORD=ZMAT

		- Gamess-UK6.2 energies of the geometry convergence
		  were not read properly, fixed it
 
Wed Dec 08	- Updated the makefile entry for CRAY unicos, for
		  running the f90 compiler

Thu Dec 02
		- writing of non-periodic MSI files used atomic units
		  rather than angstroms, corrected it.

		- When writing postscript in the molecular mode, molden
		  will check for dummy atoms and will ask you if you
		  want to include them in the plot. In the current situation
		  dummies where not plotted.
		
		- In the molecular mode you can change the connectivity
		  of the atoms, but not so in the density mode. This proved
		  a problem for people generating VRML orbitals of metal
		  complexes, i have solved this ONLY for the VRML writing:

			Before it will write the VRML molecule, it will check 
			for the existence of a file called 'molden_connect' 
			(be sure you have the permission to read it, 
			ls -l molden_connect), this file should hold the 
			connectivity of the molecule, per line two number; 
			atom1 atom2 If atom2 is negative, molden will read an 
			extra line with 5 numbers it will average the 
			coordinates of these 5 atoms and create a bond
			between atom1 (of the first line) and the average over 
			the 5 atoms (of line two), it will use the color 
			associated with the first atom of line two. If you want
			to average over less then 5 atoms just supply
			zeros for the remaining atoms 
			(so avg over 3: i1 i2 i3 0 0)

		  DONT FORGET TO REMOVE OR RENAME THE 'molden_connect' FILE
		  AFTERWARDS.

		- When you select atoms in the Z_matrix to 'cut out' a part
		  of the atoms to create a new z-matrix, properties like
		  atoms types, forcefield types, color, residue and charge
		  are now inherited.

Thu Nov 18
		- Added keyword POVRAY, to generate the povray file molden.pov
		- Added support for Ampac style cube files

Thu Nov 11	- Fetched some minor problems:

		  - some msi files were not read correctly, fixed
		  - gamess-US, sometimes doesnt end coordinates
		    with an empty line, but with:
			'...... END OF ONE-ELECTRON INTEGRALS ......'
		    dealt with
		  - the geometry convergence window sometimes didnt
		    update correctly, fixed it + made them smaller (for PC)
		  - when moving atoms in a cell you couldt use the
		    distance button anymore, fixed it

Tue Oct 12	- Solved problem when replacing last amino acid
		
Thu Oct  7	- Reading msi files generated stray atoms when H-bonds
		  where present in the msi file, resolved.

Thu Sep 30

		- First support for GamessUK6.2, now coping with changed
		  orbital writeout

		- Added Button to the Z-matrix window to make it smaller.

		- GLX and ASX PDB residues are mapped to GLN and ASN
		  GLX and ASX caused problems in protein z-matrices

Mon Sep 20

		- fixed bug: when creating a peptide via the z-matrix
		  closing the z-matrix window and then opening it again
		  caused molden to crash.

		- Added support for MOL2 protein files, small molecules
		  were already supported.

		- Added support for reading/writing MSI Cerius files,
		  The reading is NOT robust, it will read simple, periodic
		  structures 

Thu Sep  9

		Molden now uses dynamic memory allocation for the
		size of the grid, this will allow you to read in 
		any size gaussian or vasp cube files.
		It also allows you to redefine the size of the grid
		in points from the 'plotplane' window, via the commnds
		nptsx= nptsy= nptsz=
		Previously the maximum was nptsx=nptsy=nptsz=61

		The molden version created with 'make noxwin' is NOT
		dynamic.
Mon Sep  6

		Molden now uses dynamic memory allocation in some
		parts of the program:

		- For a z-matrix of a protein
		  (Not for reading a z-matrix of a protein)

		- For basis-sets larger than 256 orbitals

		Not done are for instance the maximum size of a grid
		(which is 61*61*61), so cube files with larger dimensions
		are still a problem

		The molden version created with 'make noxwin' is NOT
		dynamic.

Mon Aug 23

		Linked Z-mat variables were not updated onscreen.
		Non bonded atoms in the stick mode were not correctly
		shown in the postscript output

Thu Aug 19

		A word of WARNING:

		Molden3.6 has gone through quite a few
		drastic changes, it can therefore be expected to have a few
		bugs introduced, so for the time being use this version
		at your own risk, and PLEASE report any problems you
		encounter (schaft@caos.kun.nl)

		- Improved filemanager, it now also works correctly in 
		  directories other that the current.
		  A directory select window has been added.

		- More labels options: added "pdbsymbol","residue"

		- Molden now treats any color depth greater or equal 8
		  as having 256 color cells

		- Molden "Author" icon has been replaced by the "FF" icon
		  which when clicked pops up the ForceField type assign
		  window. You can still invoke the author window by typing
		  "a" in the drawing window.

		- Typing "u" in the drawing window, brings molden secondary
		  windows ontop of the drawing window. (This may not work
		  with the gnome window manager).

		- In the stick drawing mode, atoms which are not bonded
		  used to be invisible, they are now represented as crosses.

		- When saving a structure as PDB file it will ask you
		  if you want to write the Hydrogens or not.

		- Under the "H" icon an option has been added to
		  recalculate the H-bonds.

		- The biggest addon is the pdb to z-matrix conversion,
		  in essence this was already there in the 3.5 version,
		  but was very buggy (there will probably be some more :-))
		  In the 3.5 version the whole z-matrix of the protein was
		  shown, which is way to much and cumbersome to handle.
		  In 3.6 the default is to create a z-matrix representation
		  which list the aminoacids with its phi,psi and chi angles.
		  You can click on the Aminoacid button and choose from a 
		  number of options:

		  Switch On
		  Switch Off
		  Center
		  Contacts
		  Delete
		  Insert
		  Replace

		  The one letter codes are also displayed. You can search
		  with a string of one letter codes. There is a button to
		  go to the "full Z-matrix"

		  Unlike the old behaviour, ONLY variables which were changed 
		  on screen are updated in the internally stored z-matrix

		  IMPORTANT:

			MOLDEN WILL ONLY PROMPT YOU TO CREATE A Z-MATRIX
			FOR A PDB FILE WHEN IT HAS ENOUGH ROOM TO STORE
			THE Z-MATRIX, DEFAULT IT WILL ACCOMODATE 1000 ATOMS,
			NOT ENOUGH FOR MOST PROTEINS. SO YOU HAVE TO CREATE
			A BIGGER VERSION OF MOLDEN:

			- edit the file "change_parameter"
			- replace the body:

				1,\$s/maxorb=256/maxorb=1024/
				1,\$s/numatm=150/numatm=300/

			  by:

				1,\$s/maxat=1000/maxat=20000/

			- save this file and execute it:

			  ./change_parameter


			- change_parameter only works on the fortran files
			  you have to edit the C code manually:

			- edit xwin.c, replace:

				#define MAXAT 1000

			  by:

				#define MAXAT 20000
			
			- save the file, and type 'make'




		  	   
		Molden3.5:

Mon Jul 19

		- Fixed bug which let you not save charges you edited,
		  when you start of with a file which has no charges

Thu Jun 17

		- removed small bug in rdgaus.f which made some compilers
		  stop.

		- added the capability to write and read files with custom
		  spherical atom densities to be used in difference density
		  plots (bonds keyword).
		  
		  molden inputfile

		  inputfile:

		  some title
                  file=gaussian_atomic.log genat

		  a file 'basiinf.mbi' will be written, these can be 
		  concatenated and read back in to molden, fi:

		  molden -x atomicdens

		  Thanks to Bjoern Pedersen who implemented it.
		  (Bjoern Pedersen, Technische Universitaet Muenchen
		   Institut fuer organische und Biochemie, 
		   Bjoern.Pedersen@ch.tum.de)


Thu Jun  3	

		- Added resizebility of the geometry convergence window

		- Removed SERIOUS BUG introduced Fri May 21


Wed Jun  2	Added commandline opiton -m, which turns off bell sounds

Fri May 21	Added scale factor for normal mode displacement vectors
		in the frequencey window.

Fri May  7
		GAUSSIAN:

		- COMPOUND jobs, (optimisation, followed by a FREQ job)
		  should now be oke.

		- Gaussian98 doesnt print 'input orientation', for a
		  molecule with more than 50 atoms, in this case molden
		  couldnt show intermediate structures of an optimisation
		  run, FIXED. (The problem could be fixed by supplying
		  IOP(2/11=1) but molden now always takes the 'standard
		  orientation' and disregards forces (for more than 50 atoms))

Tue Apr 27
		- Support for reading Gaussian cube files via the
		  molden keyword file (new keyword GAUCUB):

		  tlite
                  gaucub=test/test_homo.cube space=0.1 vrml

Thu Apr 22
		- Molden can write VASP POSCAR files, under the 'cell' icon,
		  'write' -> 'write VASP', Molden can also read these files
		  (Use always 'direct or selective' keywords fullout)
		  (Use the first line not as title but put the element
		   names in there).
		  Read support is rudimentary; POSCAR files have all atoms
		  lying in the cell, which may give a scattered outlook
		  of the molecule in the cell.
		  Molden can also read the VASP density grid files, although
		  they tend to be too big for the default molden version.
		  Presently only grids which axes are under 90 degrees 
		  will be accepted.

		- Changed under the 'Cell' icon: the seperate write options;
		  'write DMAREL', 'write SHELX', 'write CRYSTAL95', into
		  one options 'write', which lets you select out these
		  options + 'write VASP'
Thu Apr 15
		- Molden can now generate Gif's for 24-bits displays

Fri Mar 26
		- Added Possibility to delete and move atoms in a
		  crystal, under the CELL icon, the last two options.
		  It was already possible to translate the atoms in a cell
		  but this was interactively, the new translate option
		  allows discrete translations (say 0.5 0 0) of selected
		  atoms in the assymetric unit.
		  
Fri Mar  5

		- Added commandline flag "-geom XXXxYYY-xxx-yyy"
		  where XXX and YYY are the size of the window in pixels,
		  and xxx and yyy are the position of the window in pixels.

Thu Mar  4

		- Molden can now read Gaussian Cube files:

		  restrictions:

		  - these must be formatted cube files
		  - supported are: density, potential and orbital cubes
			(orbital cubes only with ONE orbital!)
		  - the grid MUST be rectangular 
		  - grid can only be read in the 'density mode'

		- Molden now also supports grids which dont have an
		  equal number of points in all three directions,
		  and also the radius in the three directions can be
		  different, under the 'PlotPlane' window there are now
		  6 new keywords available:

		  npts1,npts2,npts3 and edx,edy,edz




		Molden3.4:

Fri Feb 12

		- Added support for G98 optxyz output

		- Added support for Pluton SPF format

		- Added some aminoacid sequence editing stuff

Tue Jan 19

		- Tinker optimisation of structures in molden is done
		  in the Z-matrix ordering of atoms, even if the read in
		  structure is a Tinker xyz file. The last case was not
		  handled properly, it is fixed. You can optimise in the 
		  original ordering by suppressing the generation of a
		  Z-matrix by using the commandlineflag -a

Tue Jan 12

		- Fixed bug in writing chemx .cssr files
		- Fixed bug in creating the super cell
		- Fixed bug in interpreting molden-format when [5D]
		  was used.
		- Added calculation of electrostatic energy, a new
		  button 'E q' was added and a field for specifying
		  the residue number (Eq also calculated interresidue
		  electrostatics), both in the 'Atom Attributes Window'
		  this is popped up when you click with the middle
		  mouse button on a atom.
		  Shift Mouse click in the residue field lets you set
		  this residue number to all the atoms you will subsequently
		  select.
		- Added -R npts commandline flag this allows you to
		  downsize the size of the grid in points. 
		  You can compile with a bigger number of maximum points
		  (Use the script 'change_parameter':
			1,\$s/mx3d=61/mx3d=200/
		   to change the grid size, memory is a consideration
		   since an array of the size mx3d*mx3d*mx3d will be used.)

		  Thanks to:

		  Bjoern Pedersen
		  Technische Universitaet Muenchen


Tue Dec  1
		- Molden treated Gaussian semi-emirical optimisations
		  as single points, fixed it

		- With Tinker xyz (charmm,amber), the caps are nolonger
		  (faultily) assigned as residues
		
Mon Nov 30	
		- Fixed bug that makes molden crash on WinNT/95 and linux
		  when a mopac GRAPH file is read and the orbital select
		  window is activated

Mon Nov  9
		- For the submit gaussian job, molden now also looks
		  for the g98root environment variable.
Tue Oct 20
		- support for cssr.arc files.

Tue Oct 13	
		- Removed bug introduced Sep 23, which disabled the
		  mapping of files onto the z-matrix
		- Further integration of wijnand Mooy's xtinker.

Mon Oct  5

		- Removed a bug which would in some cases now
		  show a structure in the "molecular mode" for
		  Guassian98 files.

		- Implemented tinker "amber" force field
Tue Sep 29

		- Removed bug that occured when viewing MO's from MOPAC/VAMP
		  in the "Molecular Mode" created in the "Density Mode";
		  the connectivity of the atoms would be messed up.

Mon Sep 28
		- Implemented tinker preferences window.

Wed Sep 23

		- Implemented, real time structure optimising with
		  tinker mm3 (small molecule) or tinker charmm (proteins).
		  the molecular display is updated with the intermediate
		  structures. At the end a tinker archive file is created.

		- Support for tinker archive files.

		- Vrml from the elctrostatic potential derived from charges
		  in the density mode, now shows both positive and negative
		  lobes.

		- Adding of fragments can now be done at the start of a
		  new z-matrix

Fri Sep 18	- In submit job (Gaussian, Gamess) the check for charge
		  and multiplicity gave an error when there was an dummy (X)
		  present.

Mon Sep  7	- Added VRML equivalent of the 3D-X plotting mode,
		  first select this mode from the Molden Control Window,
		  then click the "Vr" icon.

		- Added creation of peptides.
		  Click "ZMAT Editor", select "Substitute atom by Fragment",
		  select "Sequence" and the "Build Sequence Window" will 
		  pop up. Select what kind of conformation you want; alfa
		  helix, or beta (sheet) alternatively a turn. This selects
		  the PHI and PSI angles used. You can still change them 
		  in the corresponding PSI and PHI fields. Now specify an
		  amino acid sequence by clicking on the residue buttons,
		  a status bar shows the last 16 residues selected.
		  You can correct an error with the "Undo" button.
		  When ready click "Build". 
		  Alternatively you can read in a protein file, which should
		  contain per line:

		  '3 letter amino acid code' Phi Psi [Omega] [Chi1] [Chi2] 
			[Chi3] [Chi4]
 
		  Everything between [] is optional.
		  The files used by tinker to create peptide can be used.
		  Molden will remember the last amino acid added, so that
		  new partial sequences can be added without selecting an
		  atom to which it should be attached.
		  When ready you can write it out as a PDB file, in the
		  "Molden Control" window click "write" select "PDB".
		  You can also write it as a tinker charm file, close the
		  ZMAT-editor, click with the middle mouse button on an atom
		  this will bring up the "Atom Attributes Window", click on
		  The Force Field button and select "Tinker Charmm", now click
		  on the write button int the "molden Control" window and
		  select "Tinker".


Tue Aug 25	
		- Fixed bug that let molden crash on certain pdb, tinker
		  charmm files.


Tue Aug 11
		- Added support for Gaussian98
		- Molden now adds hydrogens to the amino acids and
		  nucleotides of a PDB entry.
		- Added checking to user selection of atom types,
		  for Tinker MM3 aand Tinker Charmm.


		Molden3.3:

Thu Jul 23	- Added actinide elements, to atoms property editor
		- Fixed small bug which wrongly switch label button
		  when FFtype was selected.

Wed Jul 22	- Fixed bug in code to deal with ADF orbitals
		  (Fabio Mariotti)
		- Gamess-uk/us uses different sequence of F orbitals
		  than gaussian, molden always did the gaussian way,
		  corrected.

Mon Jul 20
		- Fixed small bug which prints lots of arrows on a molcular
		  postscript plot when normal coordinates are available.
		- the ADF routine assumed exponents and normalisation
		  constants read in from a molden format file as written
		  by the ADFrom program to be in atomic units, this was
		  not the case, the routine has been adapted.

Tue Jul 14
		- By default molden always tries to make a z-matrix if 
		  there is none, occasionally it will have to add a 
		  dummy atom to accomplish this. This could cause
		  problems reading some of the frequencies of a frequency
		  job. Noe Fixed.

Mon Jul 13	
		- The window that lets you edit forcefield types,
		  is now brought up when you select the label "ForceF.Type"
		  When you click an atom with the middle mouse button
		  as before this window is also popped up, but in addition
		  the labelling is set to "ForceF.Type"
		- When using "Optimise" from the z-matrix editor it
		  checks wether atoms typing has already been done.
		- Added some tinker forcefield parameters for acids, and esters
		  to the .key file written when selecting "Optimise with Tinker"

Wed Jul  8

		- Added almost all spacegroups.
		- for molecules with no crystal information, it will
		  be generated when you click on the CELL icon
		- Added reading of biosym arc files, use the movie
		  and next button in combination with the "Geom. Conv."
		  button.
		- Under the CELL icon, option "edit cell parameters";
		  you can now change the spacegroup of the cell
		- FDAT files containing multiple entries can now
		  be viewed with the "Next" button.
		- Added "fit multipoles" under the surface icon
		  (molden has to be compiled with "make molden.mpfit"
		   for this option)
		- Added "Optimise xtinker" option under the CELL icon
		  this writes input files for and executes an adapted version
		  of xtalmin, an crystal structure optimiser from the tinker
		  suite of programs. Adaptation by Wijnand Mooij, not yet
		  available in the tinker distribution.

Mon Jun 15	- Added plotting of distance monitors to postscript
		- Added more types for tinker mm3, automatic assigning
		  still restricted to the old types.

Thu Jun 11	- Added writing of Chemx(CSSR) files.
		
		IMPORTANT !!!!!!!

		- Fixed BUG in reading GAUSSIAN94 frequency outputs.
		  It was introduced on 11 april 1998.

Wed Jun  3	- Added Scrollable scrollbars
		- Added atom dynamic typing, click with the middle
		  mouse button on a atom.

Tue May 19	- Fixed a bug in molden.f which caused some compilers
		  not to compile it. (Introduced at Thu May 14)

		- Writing of animated vibrations in VRML2.0:

		  molden -w3 test/mopac.freq

		  Click "Norm. Mode" button, select a frequency, say 11 
		  a file "freq11.wrl" will have been written now.
		  
Thu May 14	- Added routine to fit multipoles to the electrostatic
		  potential, stuff written by Wijnand Mooij 
		  (w.t.m.mooij@chem.ruu.nl), use keyword 'MPFIT'
		  together with keywords described in file 'mpolefit.f'
		  type 'make molden.mpfit' in the source directory to
		  add this feature to molden.

		- The new Gamess-US prints frequencies a little different
		  which caused molden not to display the right frequencies
		  this is solved

Wed May  6

		- Milan Hodoscek (xvibs) supplied code to plot
		  normal coordinates as arrows in the postscript
		  plot. EG: molden test/gaussian.freq
		  click on Norm. mode button, choose the frequency, 
		  then click on Norm. mode button again, now click the
		  postscript button. You can use the -s commandline flag
		  to scale the amplitudes (eg: molden -s4.0)

Mon Apr 27

		- Improved the writing of VRML2.0 animations.
		  These are invoked by:
  
		  molden -w3 geometriesfile

		  Then clicking the "Movie" button. At completion
		  the VRML2.0 file 'movie.wrl' is written.
		  It include a file via the WEB:

		  http://www.caos.kun.nl/~schaft/molden/wrl/animator.wrl

		  You can edit the movie.wrl via and subtitute the 
		  EXTERNPROTO declaration by the contents of animator.wrl.

		  Other adjustable parameters are the size of the animation
		  window: Window [2,1] (Within the DEF MOLDEN_ANIMATOR
		  definition). The cycleInterval in DEF MOLDEN_TIMER
		  this is the time the animation should take. 

Wed Apr 15

		- Added possibility to 'read' and 'write' basis set
		  3D grids, a file 3dgridfile is created.

		- Added choice of solid molecule draw:

		  "molden radii" (The old molden solid)

	          "vdwaals radii"

		  This on user requests.


Tue Apr 14

		- Added reading of BOYS localised orbitals for Gamess-US

Fri Apr  3

		- The non Xwindows version of molden now works again

		- Crystal visualisation support for chemx .cssr format
		  and biosym .car files

		- Through the Z-matrix editor you can build z-matrices
		  which have translation or screw-axis symmetry.
		  Read in the repeating unit, then click on 
	
		  "Substitute atom by Fragment" 

		  pick "Periodic", pick "Translation" or "screw-axis",
		  specify the number of repeats.

Mon Mar 23

		- Support for the new(?) gamess_us PC version is 
		  has been added.

		- Added animation of z-matrix variables.

		- Added writing of POVRAY files

Fri Feb 13

		- enabled atonnumbering in postscript printing

		- Electrostatic potential mapped to an isodensity
		  surface is now possible. Also charges derived this way.

		- Writing and reading of surfaces.

		- Orbitals can be displayed as surfaces as well

Tue Jan 27

		- Enabled writing of 'ATOM' records in the PDB
		  format.

                - Enabled writing of Tinker Charmm files

		- Postscript files written by molden, now allow
		  the turning of of the zero contour

                - Molden Postscript supports non solids drawings and
		  shading on these sticks

Fri Jan  9

		- Added posibility to create a super cell

		- Added read/write support for the crystallographic
		  SHELX format (LATT 3, not yet supported)

Thu Jan  8
		- Constants read in from the constants section of a
		  gaussian/gamess-uk input file are flagged as such
		  in the z-matrix editor, constants which have other
		  variables linked to it, are written to the constants
		  section of an gaussian/gamess z-matrix

Wed Jan  7 

		- Fixed small bug in the gaussian output reading routine.
		  When reading frequency output and encountering a
		  'exceeding maxnum of orbitals' condition, the wrong
		  coordinates were read.

Fri Dec  5

		- You can now specify the order of atoms in which a
		  Z-matrix is to be created

		- When working with files which contain crystal information
		  (FDAT,MOL2,MSF) it is possible to map a z-matrix on file
		  onto the cartesian coordinates. Your z-matrix file has
		  to contain extra mapping information, so after the
		  Z-matrix itself you have to supply the keyword 'map'
		  followed a line per z-matrix atom which indicates
		  to which atom it corresponds in the cartesian file.
		  The name of this special Z-matrix file has to be 'mapfile'.
		  You then start molden with the commanline flag -Z:

		  molden -Z test.fdat

Mon Nov 24	

		- removed bug introduced a week ago, that messed up the
		  connolly surfaces

		- Added distance monitors

		- Added editing of cell parameters.

		- Added the possibilty to align two structures:

		  Read in stucture 1.
		  Click the align icon in the center of the molden control
		  window. Then supply the filename of the file containing
		  structure2. Now use the standard ways of rotation,translation
		  and zoom in/out (z-translation)  to align stucture2 to
		  structure1. You can also use the origin icon to change
		  the center of rotation in structure2.
		  You can switch between rotating both structures or
		  structure2 by using the Esc key.
		  Use the Tab key to align by selecting three couples of
		  atoms (a couple =  atom structure1 (red), atom structure2 
		  (white)
		  
		- Reading Support for the tinker .xyz  format (mm3/charmm)

		- It is now possible to work with crystals and change
		  the position and orientation of the packed molecuul
		  via the "Cell Rot On/Off" option under the cell icon.
		  You can switch between rotating both structures or
		  structure2 by using the Esc key.
		  At the same time you can change the packed molecule itself
		  via the z-matrix editor.

		- Molden can now write input files for the Crystal95
		  package. Basis sets of previous read in gamess/gaussian
		  jobs are transferred to the Crystal95 input.
		  Works of course only on fdat files and mol2/msf files
		  containing crystal structures.

		- molden -f pdbfile
		  would create connectivity from scratch, however the
		  HETATM records where not dealt with correctly.

		- You can change 14 of the colors of the color palette
		  "molden -r colorfile", where colorfile could be:

 		  atom color(1-15) VandeWaalsRadius, (- = skip)
		  background color(1-15)
		  palette red #CF54FD ...   (14 colors)

		  You can continue the last line on the next line when using
		  the "\" character. You can use color names or Hex red-green
		  -blue triplets

Wed Nov  5
		- Added more spacegroups for use with mol2/msf files.
		  (100 spacegroups supported)

		- Improved crystal building code

		- Support for Gamess-US IRC files

		- Electrostatic potential due to point charges can
		  be calculated/displayed on a connolly surface
		  See Surface icon.

		- On mopac output the ESP charges are read in when available.

		- Writing of simple pdb files, via the 'HETATM' record.

		- Fixed a bug that caused variable names such as 'e1', 'd1'
		  in a z-matrix input to be not handled correctly.
		  Bug was introduced with the introduction of the molden format.

Mon Oct 27	- Introduction of the molden_format

		- fix of molden format on linux

		Molden3.2:

Fri Oct 24	- Introduction of the molden_format, have a look at
		  the "write" button. 

		- Normal modes with a sequence number greater than 99
		  where not handled correctly

		- The background color can be set via the molden -r
		  flag, see molden -h; molden -r

Mon Oct 13	- Gaussian Beta Orbitals were not always visible
		  specially when read of a checkpoint file:

Fri Oct 10	- Support for ADF, an intermediate program which
		  reads ADF's tape21 can produce files readable
		  by molden. This program was developed by Fabio Mariotti
		  from Prof. Alessandro Bencini group in florence.

		  This intermediate program will eventually become part
		  of the molden distribution, but since it is still
		  under development, you should contact Prof. Bencini:

			Prof. Alessandro Bencini
			Universita di Firenze
			Dipartimento di Chimica, Via Maragliano 75/77
			50144 Firenze, Italia

			Tel: 39-(0)55-354841
			e-mail: sandro@chim1.unifi.it
					bencini@dada.it

Wed Oct  8	- Fixed bug in Gamess-UK submit job scheme, the
		  proper 'vectors' option was not always chosen.

		- The Postscript color of the unit cell is now black
		  per default, it can be changed by in the postscript
		  file by setting /cellcol

Wed Sep 24 	- Added spin density for UHF calculations
		  Use keyword SPINDENS together with IPSI=0

		- Improved dmarel interface

		- Removed multiple expose events on the zmatrix editor

Mon Sep  8
		- more Gamess-UK version 6.0 support

Fri Sep  5

		- removed bug introduced day earlier, resulting
		  in an floating exception on at least Linux
Wed Sep  3

		- Gamess-UK version 6.0, added first support.

		- Adapted postscript driver to handle editing of
		  atom display radii.

Thu Aug 28	- Enhanced atom color editor to atom property editor
		  Now color, vdw and display radii can be altered.

Thu Aug 21
		- Added Submitting of Gamess-UK Jobs.
		  GAMESS_ROOT has to be set and the 'rungamess' script
		  must be operational.

Mon Aug 18
		- Added atom color editor.

Fri Aug 15	

		- Removed bug which when molden was started without
		  a file name would cause the molden control window
		  to never pop up, certainly on AIX.

		- Gamess-US: when using print options to print
		  'AO MULLIKEN POPULATIONS IN EACH MOLECULAR ORBITAL'
		  molden didnt read the occupancies correctly, 
		  resulting in the density to be calculated to low,
		  and mulliken charges to be incorrect, orbitals ok.
		  Fixed it.

		- Added Submitting of Mopac Jobs, still very primitive.
		  MOPAC_ROOT has to be set

Mon Aug 11	- Removed small bug whith respect to the PLANE/ROT
		  versus LINE/CENTER keywords.
		  Using the PLANE keyword the orientation of the calculated
		  rectangle of points is such that, the first atom specified
		  with the PLANE keyword together with the center of the 
		  plot lie parallel with Y-plot axis, when LINE/CENTER
		  or even ROT is used this scheme is abandoned.

Wed Jul 23	- Added Submitting of Gaussian Jobs, still very primitive.

Thu Jul 17	- Fixed bug that was dormant but became serious
		  on a few platforms after introducing reading of mol2
		  file on July 13. Files would not be correctly recognized.

Wed Jul 16	- Added writing of tinker .xyz files.

		- Added under "XYZ->curr.zmat" in the Z-matrix editor
		  the possibility to optimise a structure with, sybyl
		  quanta or tinker.
		  The corresponding environment variable has to be set
		  first (TA_ROOT, QNT_ROOT or TNK_ROOT)
		  TNK_ROOT is not used by tinker itself, but molden needs
		  it. (eg: setenv TNK_ROOT /compchem/tinker) 
		  The mm3 tinker parameter set will be used.
		  Running quanta is shakey.

		- Removed a bug that applied to Z-matrices automatically
		  generated for XYZ files, changing variables in the Z-matrix
		  would screw up the molecule.

Fri Jul  4	- Added writing of Quanta/Charmm MSF files.

		- Added atom row containing AG etc. in z-matrix editor.

Tue Jul  1	- Added reading of Quanta/Charmm MSF files.

		- You can now optimise you molecule with sybyl
		  from the z-matrix editor. The environment variable
		  TA_ROOT has to be set first. Choose the "XYZ->curr. zmat"
		  button.

Wed Jun 25	- Removed two bugs by which seemingly only the
		  OpenVMS version was affected.

Mon Jun 23	- The tolerence for partioning multipole expansion
		  in symmetric molecules was to low, leading very
		  occasionally to wrong multipoles, for higly symmetric
		  molecules. Fixed it.

Fri Jun 13	- Incorporated reading and writing mol2 files.
		  For protein just coordinates are read.
		  Writing of Mol2 is very primitive, with respect to
		  atom types, and type of bonds is always single.
		  But you can convert fdat to mol2 files this way,
		  including charges, just first calculate charges on
		  a job with the same atom order as the fdat file, 
		  then read in the fdat file and write mol2 (you will
		  find it under the z-matrix editor, cartesian coordinates.
		  Only the top-ten spacegroups have been incorporated,
		  so 90 % of crystal structures will be read.

Wed Jun  4	- Fixed bug introduced Fri May 30, affecting the proper
		  writing of zmatrices.

		- Fixed also some bug related with "delete line" in the
		  z-matrix editor, having to do with maintaining the right
		  status of an internal variable(=constant,=variable,=linked)

		- In mopac DRC calculations the potential energy is now
		  picked up.

Fri May 30	- Rotation around the z-axis is now possible in both
		  directions, clicking the mouse in the middle of the
		  screen will give positive rotation, doing the same
		  while holding down the Control or Shift key will
		  give negative rotation

		- Molden now supports writing of VRML2.0.
		  Apart from the keyword VRML you can now use VRML2.0
		  The -W commandline flag does the same.
		  The various VRML buttons now let you choose  between
		  VRML1.0 and VRML2.0. 
		  The -w commandline flag now knows an option 3.
		  When the movie button is pressed a VRML2.0 animation
		  is written (movie.wrl). This can be viewed with a VRML2.0
		  compliant browser such as Silicon Graphics Cosmoplayer
		  plug-in (free available, also fro PC's). Clicking on
		  the molecule will start the animation.

		- typing "g" in the display screen will let you read in
		  and XYZ file which will be mapped onto the current
		  z-matrix, provided the number and the order of the elements
		  is the same in both.

Tue May 27	- Added interactive choice between mono and color postscript

Fri May 23	- Added Pasting of strings capability to query boxes
		  (For now button 3 will do this)

		- The interactive calculations of:
	
			True ESP derived charges (Very Slow)
			DMA esp derived charges
			Mulliken Charges

		  Via the Surface button.

Thu May 22	- Removed a small bug in calculating the multipole
		  derived electrostatic potential

Fri May 16	- Charges are printed in postscript when the label
		  "atom+charge" is active

Wed May 14	- VRML of the elctrostatic potential "space" plots
		  are now handled correctly.

Tue May  6	- The Z-matrix scrollbar can now be clicked it, it
		  will trigger pageup or pagedown movement.

Fri May  2	- electrostatic potential on a connolly surface:

		  When clicking on the color level rectangle in the
		  upper right corner with the second (or middle) mouse
		  button you can change the colors used.

		- Added keyword DMACH to calculate charges fit to
		  Distributed Multipole Derived electrostatic potential.

Tue Apr 29	- the true or multipole derived electrostatic potential 
		  on a connolly surface via keyword file now includes a
		  postscript plot (-C to generate color postscript):

		  molden -C molin > molden.out

		  For the true electrostatic potential:

		  molin:
		  ----------------------------------------------------------
		  title
		  file=test/gamess_us.h2o_631g_opt.out postscript vdws elpot
		  ----------------------------------------------------

		  The new keywords XANG,YANG and ZANG controls the 
		  orientation viewed, usage XANG=45 (45 degrees)

		- added the EST parameter set for writing input for the DMAREL
		  program. (molden -e)

		- added a button "H" to undisplay hydrogens, this is handy
		  in combination with the CSD fdat files when displaying
		  multiple cells. Getting rid of the hydrogen gives a
		  clearer view of the crystal packing

		- Dummy atoms will no longer be displayed in postscript
		  molecule plots.

Thu Mar 13	- Molden could already calculate the true or multipole
		  derived electrostatic potential on a connolly surface,
		  also VRML could be generated for this but only
		  interactively, this can now be done with keywords:

		  molden molin > molden.out

		  For the true electrostatic potential:

		  molin:
		  ----------------------------------------------------------
		  title
		  file=test/gamess_us.h2o_631g_opt.out vrml vdws elpot
		  ----------------------------------------------------

		  For the multipole derived electrostatic potential:

		  molin:
		  ----------------------------------------------------------
		  title
		  file=test/gamess_us.h2o_631g_opt.out vrml vdws
		  ----------------------------------------------------------

		  A file molden.wrl will be written.


Wed Mar 12	- Added ability to display a cell with an Xmol xyz
		  cartesian file:

                  After regular atoms add cell points, valid
		  labels:

		  OO,OA,OB,OC,AB,BC,AC,ZZ

		  These cell points must be included in the number
		  of atoms.
		
		- The geometry convergence window together with
		  the movie button have been updated:
		
		  You can now stop a movie by clicking "first", "next"
		  or by selecting a point in the geometry window.
		  The tick marks are adapted so as not to overlap
		  when there are a large number of points.

		- added a file epint.f.g77, gnu fortran has problems
		  compiling the original epint.f. So before typing
		  "make", copy epint.f.g77 to epint.f.

		- Linux2.* has problems with Molden (And Gaussian94)
		  this probably due to an operating system bug affecting
		  memory management.

		- Removed bug in calculating the q31s octupole moment
		  (Only when using D-functions)

		- Added commandline flag to keep the order of atoms
		  when creating a z-matrix from cartesian coordinates

		  molden -A

		- Molden now writes an file "esp.xyz" when doing an
		  ESP charges calculation, it contains the atoms +charge
		  in xyz-coordinates

		- added support for Gamess-UK rhf localised orbitals.

		- Added beta-test support for DMAREL, a package that
		  can calculate lattice energies from a multipole
		  expansion.

		  New Commandline flags:

		    -E        DMAREL input: use coordinates from multipoles
		    -i opt    fdat files: 
		              opt=1 standardise H-C, H-N
		              opt=2 1 + standardise phenyl rings

		  Other Relevant flags:

		    -A        Keep order of atoms when creating a Z-matrix
		    -D opt    DMA mode:
		              0 = atomic sites only (default)
		              1 = atomic+halfway-bond sites

		  When creating a z-matrix for an optimisation run to be used 
		  with DMAREL you must keep the order of the atoms as in the 
		  CSD FDAT file and turn of symmetry to keep it this way:

		  molden -A -i 2 azijnzuur.fdat -> azijnzuur.zmat

		  To use the optimised coordinates instead of the FDAT 
		  coordinates in writing DMAREL input, use

		  molden -E azijnzuur_opt.out azijnzuur.fdat

		  if you want to use bondcenters as well,

		  molden -E -D 1 azijnzuur_opt.out azijnzuur.fdat

		  If you want to use fdat coordinates with bondcenters,

	          molden -D 1 -i 2 azijnzuur.out azijnzuur.fdat
			
		  In molden click on the surface button, choose 
		  "Multipole Derived", then read in the FDAT file,
		  click on the "Cell" button and choose "Write DMAREL"
		  a file "dmain" will be created.


Wed Feb 12	- In the density mode you could zoom in on the plot
		  with the second mouse button, sometimes old plot
		  were mixed with new plots, fixed it.

Tue Feb 11	- GAMESS-US: when using the HSSEND keyword to do a
		  normal vibration analysis at the end of a geometry
		  optimisation, the default is to display geometry
		  optimisation information, but molden -H will
		  trigger visualisation of normal vibration analysis.
		- GAMESS-US: when imaginairy frequencies were present
		  molden would skip the rest of the normal modes in
		  that line of the output.

Mon Feb 10	- Fixed bug in the 'space' mode of orbital plotting
		  A space plot consists of the density/orbital-amplitude
		  plotting and the plotting of the molecule in the
		  stick mode. If the density/orbital-amplitude was to low
		  in a slice then also the part of the molecule in this
		  slice was not plotted. (This did not affect the VRML
		  space filling plots)

Wed Dec 18	- Incorporated writing of inputfiles for aresp:
		  Using the keyword 'aresp' will let molden write
		  the files 'resp.in' and 'esp.in'
		  Example keyword file:
		
		  line1: test esp
		  line2: file = water.out espch debug aresp

		- The keyword occu dint work correctly, fixed
Wed Dec 11
		- Molden can now do gaussian natural orbitals as generated
		  by the keywords:
		  pop=(naturalorbitals) gfinput iop(6/7=3)
		- Molden with Gaussian94 and outputs with only cartesian
		  coordinates would say 'reference to an undefined center'
		  but would still work correct, solved

Mon Dec  9	- Molden didnt recognise certain Gaussian94 UHF jobs.
		  Resulting in only one set of orbitals being displayed

		- Dihedrals with a negative value where not properly
		  written as constants in write zmatrix

		- In zmatrix editor in the line drawing plot mode
		  the molecule become invisible after adding line two
		  this because th view direction was along the axis of the
		  molecule, fixed it.

		- some mopac output reads would hang

		- added a little more control over the postscript file



Fri Oct 18	Removed following bugs:

		- GAMESS-US input files with zmatrix and point group C1
		  failed

		- PDB file containing only non-protein and non-rna werent
		  handled well

		- FDAT files with NRAD=0 failed

Thu Oct 17	Removed following bug, when building a structure
		with the zmatrix editor the display would show the mirror
		image of the real structure. When writing out this structure
		and then reading it back in to molden the structure would
		be displayed in the right way.

Wed Oct  9	Added the possibility to do a Distributed Multipole Analysis
		and the calculation of Mulliken charges.
		For now, Mulliken charges are calculated only for single point
		Gamess/Gaussian runs.
		The Distributed Multipole Analysis can be activated by use
		of the keyword "MOLPOT". Interactively it can be activated
		either from the "molden control window" by pushing the
		"surface button" (The 6th icon in the middle of the window)
		or pushing the "Dens. Mode" button and consequently choosing
		the plot function "Elec. Pot" in both cases the user is then
		presented with the choice between the true electrostatic
		potential and the electrostatic potential derived from a
		Multipole expansion. The latter is an order of a magnitude
		faster than the first.

		Three types of DMA are available, they can be selected via
		the "-D opt" commandline flag, where opt can have the values:


			0 = atomic sites only (default)
			1 = atomic+halfway-bond sites
			2 = no shift of overlap dens. of conn. atoms
		
		NOTE: In previous versions the keyword "MOLPOT" was used
		      with a DMA analysis read from a Gamess-UK outputfile.
		      Use keyword "RDMULT" to get the old functionality.






		Molden3.1:

Thu Aug 29	Calculating the Laplacian via a keyword file, core dumped
		It has been resolved.

Thu Aug 22	- When visualizing csd fdat files in the "multiple cells"
		  problems arose when there were Hydrogen bonds present.
		  This has been resolved.

Wed Aug 21	- Support for the laplacian of the electron density,
		  (the trace of the hessian).
		  Two commandline flags were introduced to aid the
		  space type of plots with the laplacian.

		   -L   display both neg. and pos. contour in space
			plot of the laplacian. Default only the value
			specified with "space=value" will be rendered.
		   -1	use only the lower half of the cubic grid
			used for the space type plot
		   -2   use only the upper half of the cubic grid
			used for the space type plot

		  The latter two were introduced to increase the visibility
		  of the inner (enclosed by other surfaces) surfaces.

Thu Jul 25	- Corrected a bug which made you have to click buttons
		  twice

Fri Jul 12	- Corrected problem with postscript

		- Electrostatic potential on connolly surface added.
		  Activated by Clicking on the appropriate icon in the
		  middle of the molden control window.
		  Clicking on points of the surface brings of a popup
		  window with its electrostatic potential value.
		  In the upper right corner of the screen there is a
		  color code rectangle, clicking in it lets you change
		  the increment value between different colors.
		  The 'molden -G 0.5' commandline flag controls the
		  spacing between the lines that make up the connolly
		  surface.

		- Added the capability to write VRML
		  (molden -V filename, specifies density VRML file,
		   default: molden.wrl)
		
		- You can now create a Z-matrix with symmetry, use
		  mouse button two in the zmatrix editor, to:
			- link variables
			- to make constants
			- to mark or unmark

		Molden3.0:

Thu May 23	- Corrected problems on some platforms with the fdat format
		  and the reading of XYZ format without the -X flag

Wed May 22	- Rotation in the density window is now attached to button
		  one, rotation will proceed until the button is released,
		  cropping is now attached to button two

Wed May 15	- Molden now also recognizes XYZ format without the -X flag
		- zoom in/out will proceed untill the button is released

Fri May 10	- Better encapsulated postscript
		- Better hydrogen bond display for PDB files
		- support for hydrogens on RNA/DNA pdb.
		- molden -r fname: read file with per line;
			atom color(1-15) VandeWaalsRadius, (- = skip)
		- molden -w opt: write all points of a movie to a file:
			opt specifies format; xyz(=1) zmat(=2,mopac)

Thu Apr 25	- Increased support for CSD .fdat files.

Wed Apr 17	- Support for mopac93/6.0 irc output
		- gamess-uk inputfiles with no variables/constants oke now
		- added command line flag for switching off automatic
		  cartesian -> zmatrix conversion; "molden -a"

Tue Apr 09	- Some platforms dont support the "logical*1"
		- Enhanced cartesian -> Z-matrix conversion

Tue Mar 29      - Support for mopac6.0/7.0 force runs.
		- Allow to set the background color, via palette icon.

Tue Mar 28      - Support for CSD .fdat files, molden -S will generate
                  symmetry positions and a unit cell.
                - xmol files that used scientific notation (1.0e05)
                  are now also handled well

Tue Mar 19	Mopac input files didnt show up in the z-matrix editor
		fixed it. You read mopac input files as fragment in the
		z-matrix editor.

Fri Mar 15	Added built in GIF writing capability, dropping the use
		of the external "convert" program from ImageMagick

Wed Mar 13	Extended and improved the rotation of the molecule,
		holding the mouse button down in the center of the graph
		will invoke rotation around a z-axis

Tue Mar  5	A VERY serious bug effected the calculation of the
		electrostatic potential and esp charges with  Gamess-US/UK
		output. (It was introduced with Gamess-US support nov. 95)
		esp's from Gaussian were unaffected !!

		- ESP charges: introduced keywords AXIS-X,AXIS-Y,AXIS-Z
		  They the define the principle axis used in constructing 
		  the atomspheres which make up the connolly surface.
		  Sometimes necessary when you want to reproduce symmetry.

Wed Mar  4	- support now for dih, -dih variable use when reading
		  Gamess/Gaussian inputfiles.
		- With Gamess-UK optimise jobs molden would show
		  orbital energies of the first point, instead of the last
		  point, fixed.

Mon Feb 26	A number of bug fixes and new features, see also
		The history of molden2.9:

		- Added color postscript for Density Display.
		  Triggered by commandline flag -C

		- Removed bug that would mess up the EUCLID display
		  when clicking the Postscript button.

		- considerably enhanced reading and displaying of PDB
		  structures.

		  Helix, beta secondary structure information, will
		  by default be read from the PDB file, if not available
		  Molden will calculate secondary structure information.
		  Use "molden -g" to allways force the latter.
		  From the "Residue Command" window this information can
		  be edited.
		  The color postscript rendering of PDB structures has
		  also been dramatically improved.
		  Molden also lets you display parts of the molecule as
		  a connolly surface.

		- molden now supports pure F/D-functions (7f,5d)

		- GAUSSIAN: molden now also read orbitals from dft
		  calculations.

		- Added commandline flag -h, which prints the available
		  options.




		Molden2.9:

Thu Feb 22	- Added color postscript for Density Display.
		  Triggered by commandline flag -C

		- Removed bug that would mess up the EUCLID display
		  when clicking the Postscript button.

		- Made postscript file more suitable for customizing.

Tue Feb 20	GAUSSIAN: molden now supports pure F-functions (7f)
			  molden only used to read cartesian F (10f)
			  (As with pure D, not yet for the electrostatic
			   potential)
			  With Gaussian92 and older in combination with
			  ROHF all orbitals were printed, Molden now
			  reads all the orbitals.

Fri Feb 16	GAUSSIAN: the use of nosymm could lead to problems.
		          Think I've tackled it now.
			  Also the SCF convergence reading when
			  "SCF Done:  E(RB+HF-LYP)" was on the output
			  could cause problems.

Thu Feb 15	Changed default to Color Postscript for the molecular
		display. (Color postscript for the density display not
		yet implemented) Mono postscript can still be generated
		by use of the commandline flag -M (molden -M)

Thu Feb  8      - Removed bug in postscript printing.
                - PDB: default is now to read secondary structure info
		       from the PDB file, if there is non the info is
		       calculated. Use "molden -g" to allways force the 
		       latter.

Wed Feb  7	The postscript (color) rendering of protein structures from 
		the PDB has considerably been enhanced

Fri Feb  2	Removed small bug affecting only Ultrix versions of molden
		when displaying dummy atoms in the solid mode.
		Removed small bug affecting Ultrix,Solaris when adding
		a fragment in the Z-matrix editor when no line is selected

Thu Feb 1	On the rs6000 (ibm,aix) molden sometimes gave annoying
		run time warnings, such as "C invalid decimal digit"

Wed Jan 31	- GAMESS-US:
		  With geometry optimisations, for each geometry point
		  the z-matrix is read from the output. It used to be
		  generated by Molden. To update the coordinates on screen
		  with the coordinates belonging to the z-matrix,
		  push "Aplly current changes to Z-matrix"
		  Supplying "molden -u" will restore the old behaviour.

		- Removed small bug which caused problems on
		  a Gaussian optimise job on an already optimised
		  geometry (vectors are still a problem in this case)

Tue Jan 30	- Added connolly surface's for protein (PDB) viewing.
		- Added commandline flag to scale normal motions
		  molden -s4
		  will increase the amplitude of the vibrations times 4
		
		Molden2.8:

Tue Jan 25	- Removed small bug in the automatic z-matrix generation
		- GAMESS-US: not converged optimisations showed a last
		  invalid point, rendering de energy plot useless.
		- Problems encountered when having more than 200 points
		  in an optimisation or IRC have been dealt with.

Tue Jan 23	Gaussian: molden used to generate its own z-matrix
		for the z-matrix editor (except gaussian input files)
		Now it retains the z-matrix on the outputfile.
		(except for "force jobs")

Tue Jan 22	Removed small bug in displaying SCF convergence, SCF
		convergence of last point sometimes not displayed.

Thu Jan 18	Provisionally supported gaussian 5d functions.
		(density only, electrostatic potentials not yet supported)

Wed Jan 17	Removed a bug affecting the reading of gamess input files.

Wed Jan 16	Added beta sheet solid rendering.

Fri Jan 12	Added numbering to atoms on screen for the convenience
		of the Z-matrix editor (switch on label when in the
		zmat editor.

		Added "substitute atom by fragment" capability in the
		zmat editor, you can also read a gamess/gaussian/mopac
		input file as fragment.


1995


Thu Dec 21      Added nucleotide support when reading PDB files

Thu Dec 20	Removed bug in z-matrix editor which yielded
		wrong guesses for bond distances for rare atoms
		Removed bug in z-matrix editor, which in case of a wrongly
		user supplied value and when the # of zmat lines was less
		than 9, would garble the zmat-editor screen.

Thu Dec  7	Removed a bug which only affected the Sun solaris version.
		(In the z-matrix editor whilst adding the second atom)

Tue Dec  5	MAJOR BUG FIXES:
	
		- molden with GAUSSIAN94 UHF showed only alpha orbitals
		- GAMESS-US UHF geometry optimisation showed no orbitals

Thu Nov 30	removed sleep call in inferr.f, caused problems on AIX,linux
		and openvms.
		Zmatrix editor now supplies the first fragments

Wed Nov 29	On AIX molden hangs after pushing the "space" button,
		it was caused by a call to sleep, which should return
		after 2 seconds but it never returns, you have to
		deactivate this line in file inferr.f

		GAMESS-US IMPORTANT BUGS:

		- Scf-Convergence of GAMESS-US outputs would be empty
		  when "* * *   INITIATING DIIS PROCEDURE   * * *"
		  was on the output. Taken care of.

		- fixed inproper reading of orbital occupancies.
	
		- fixed incorrect handling of normal vibrations > 9

Tue Nov 21	When entering the z-matix editor without having read a file
		first, an error occured while adding the second atom of a
		new z-matrix, removed it.
		Cleaned up z-matrix editor code for monochrome displays.
		Removed a bug in routine hidedr, only OpenVMS affected.
`		Z-mat editor is a little more intelligent now.

Fri Nov 10	Protein viewing: removed inconsistency in displaying hetatm
		records

Mon Nov  6	Gaussian-94 users NOW should use iop(6/7=3) instead of
		iop(6/7=1), the latter option under g94 nolonger prints all
		occupied orbitals, the first prints all orbitals.

Fri Oct 20	First beta support for Gamess-US
		And both alpha and beta orbitals (UHF runs) are now displayable.



		Molden2.7:

Mon Oct  9 	Made minor changes to let the code compile on Linux,
		the user has to edit file inferr.f before compilation.

Thu Sep 28	The distance,angle and dihedral button now display their
		output in a dedicated window.

Wed Sep 27	Support of Gaussian94 output.
		Fixed bug introduced on Sep 11, which made the gamess
		vibration animation routines use the wrong set of coordinates.


		Molden2.6:

Tue Sep 26	Support for pdb ambiguity amino acids, asx,glx,his
		and hydroxyproline (hyp).
		Also HETATM A.. are now recognised.
		Added popup box that lets you choose the color of your HETATM
		molecule.

Fri Sep 22	Fixed an important bug introduced on Mon Sep 18.

Mon Sep 18	The cartesian coordinates to z-mat converter now WILL
		produce a z-matrix for molecules containing 3 atoms on a line.

Fri Sep 15	Added the capability to write cartesian coordinates from
		the Z-matrix editor.

Thu Sep 14	Fixed a bug that lets molden crash on some systems, in the
		case of a molecule containing 3 atoms on a line.
		Fixed a few other unclarities in the z-mat editor, however
		the z-mat editor still needs a lot of work.
		The cartesian coordinates to z-mat converter will not
		produce a z-matrix for molecules containing 3 atoms on a line

Mon Sep 11	Single point gamess calculations retain their user specified
		z-matrix in the z-matrix editor

Thu Sep  7	For pdb files only:
		Made second display type for the solid helix default
		Added "neigh=r" command in the residu command window.
		Neigh looks for neighbours at distance r
		Changed "substrate" to "hetatm" in the molden control
		window, clicking it will lets you choose which hetatm molecules
		you want to display

Tue Aug  1	Added second display type for the solid helix

Mon Jul 24	Display of vibrations for mopac (tested on mopac93)

Fri Jul 14	Incorporated calculation of charges fitted to the
		ElectroStatic Potential (ESP) on a series of connolly
		surfaces. (Only for gamess/gaussian) (see test/esp.txt)


		Molden2.5:

Fri Jun 30	The compilation of xwin.c caused some problems on non-ansi
		compiler systems (hp,sun), fixed it.

Thu Jan 26	Removed redundant code which made molden eat cpu when it
		should be idle.
		Apparently molden plotted inverted molecules, changed it

Jan		Improved lighting (At least for protein viewing)

1994

Fri Oct 21	Made support for XMOL cartesian format. Use -X flag
		
Thu Oct 20	Changed resolution of postscript molecule plotting,
		depending on number of atoms. (Plotting proteins was
		very slow)
		Eliminated trailing in rotating the molecule.
		Molden can now generate color postscript for the molecule
		display by using the -C flag.

Tue Sep 22	ZMatrix editor implemented, incorporating a very rudimentary
		builder.



		Molden2.4:

Tue Aug 16	Scf/Geometry convergence graphs improved and will now
		display function value when cursor is over the graph point.
		clicking in the graph will bring up the current point for
		display

Tue Aug 10	Added support for reading Gamess/Gaussian/Mopac inputfiles
		and Mopac/Ampac arc files.

Mon Aug  1	Added listbox to select orbitals, made internal rearrangements
		for systematic handeling of queryboxes, listboxes
		Some query boxes now remain on the screen untill a toggle
		button is pressed

Mon Jul 11	Gamess/Gaussian frequencies/Norm. Modes can now be displayed.


		Molden2.3:

Mon Jun 27	Improved shaded plotmode & switched to fully 3D controls

Tue May 17	Removed bugs on commandline parsing on some systems

Tue Feb 1	handle gaussian92 scf convergence
		(turned on by #P in gaussian inputfile)

Jan 20		Made molden pc X compatible

Thu Dec 23	-c0.2 define depth of shading 0.0 is max, 1.0 is no shades

Tue Dec 16	removed bug in the H-bond on screen display routine

Wed Dec 15	-f with PDB files, forces connectivity to be build by molden
		by default connectivity between amino acid atoms are read
		from the PDB file and for all other atoms molden determines
		the connectivity

Fri Dec 3	better handling hbond when more than one chain
		better spheres, and beter coloring with shade

Thu Dec 2	Better atom selecting by cursor.
		perspective is now uniform from large to small molecules
		-p10.0  perspective flag

Tue Nov 30	Added reading of binary mopac/ampac *.gpt files

Fri Nov 26	-ofilename change the default plotfilename='plot'
		trying to put in spheres with real depth, not satisf. yet
		-d debug flag

Thu Nov 25	bug out of getmul.f, by Dirk Huckriede

Nov 10		Dec Alpha version, corrected problems in wrzmat.f

=========================================================================


TODO:

- improve zmat editor

- Density Functional orbitals of Gaussian94 ?

- Clean up handling of dummies

- Third rotation axis.

- Mopac IRC

