Changelog

You can also check the full svn changelog

1.7.6.2 - April 2008, 28th (Revision #996)

NEW FEATURES

• Support Fourier maps for single crystal data
• Display Fourier maps with the original resolution used for the extraction, no matter what resolution is used afterwards (e.g. during the optimisation)

IMPROVEMENTS

• More efficient random moves for molecules, particularly for large molecules or with flexible cycles.
• Better automatic changes of temperature for parallel tempering
• Linux: enable compilation using "make -jN" (j2, j4...)
• Nicer atom drawing. Force double buffering

BUG FIXES

• Fixed incorrect formula for sample displacement
• Linux: avoid lock of optimisation when 3D view is opened
• Correct monoclinic indexing
• Fix issue with manual auto-indexing
• Disable auto-check for updates (could crash under Linux when no network can be reached)

1.7.5.1 - February 2008 (Revision #962)

NEW FEATURES

• Add support to compute Fourier maps (observed, difference), with a "live" update during optimization

• Profile fitting and Le Bail extraction is now directly available from a right-click on the powder pattern graph.
• Support for Intel macs
• Indexing triclinic lattices
• Explore centered lattices during auto-indexing
• Le Bail-extracted structure factors are now saved in the xml file

IMPROVEMENTS

• More stable profile fitting
• Linux: keyboard shortcuts are working again
• Correctly compute reflections beyond the supplied max(sin(theta)/lambda), but whose profile extend below the limit.

BUG FIXES

• Fixed incorrect computation of dynamical occupancy correction in some asymmetric unit configurations
• Fixed crash when switching labels on/off during optimization
• Fixed windows crash when removing atoms in a Molecule/Polyhedron

1.7.1 - December 2007 (Revision #903)

NEW FEATURES

• Add peak detection & auto-indexing (still beta for triclinic)

• Add Le Bail fitting & profile fitting (including asymmetry parameters)

• Add graph representing Iobs and Icalc as a function of 1/d for single crystal data.
• Add user-controlled rotation of atoms around any existing bond

IMPROVEMENTS

• Better support for CIF files import
• More responsive UI (refresh display even when no new solution has been found, every 30s at least, and no more than once every second)
• Always report *real* time instead of process-time for the first 100s, and then real time

BUG FIXES

• Correct distance table calculations in some cases, and therefore dynamical occupancy correction...

1.7.0 - February 2007

NEW FEATURES

• Support for multiple optimizations - store all the solutions obtained, then allow 'browsing' the solutions.
• Support for neutron time-of-flight data
• Support for reflection profiles other than symmetrical pseudo-Voigt
• Add bond-valence computation
• CIF import (powder patterns / crystal structures)
• Background can be automaticaly estimated using a Bayesian approach (David-Sivia method)
• Background calculation can now use either a linear or a cubic spline interpolation
• Support for Mac OS X (using XCode /Apple Developper tools)
• Add a 'tracker' window to display the progress of the optimization, including all the Chi^2, antibump costs, ...
• Add preferences (remembered between sessions) to automatically open crystal 3D view, powder pattern graph, display names / reflections Miller indices,...
• Enable saving files in a compressed (gzip) format (*.xml.gz)
• Add graph display for Single Crystal data (Iobs and Icalc vs 1/d)

IMPROVEMENTS

• Add choices for scaling of powder pattern graphs: x as 2theta/tof, 1/d or 2pi/d ; y as I, sqrt(I) or log10(I)
• Compilation for windows is now done using the free (as in 'free as a beer') compiler Visual C++ 2005. Large speed improvement.
• Fox can be compiled without GUI under *nix, using "make Fox-nogui" removing the requirement for wxGTK on *nix servers.
• Support for large Molecules (1000's of atoms) - these *must* be treated as rigid bodies
• Better graphical interface to a Molecule's parameters, including rigid sub-groups
• Better graphical interface to access scattering power parameters (Biso, antibump, bond valence, colour...)
• Significant speed improvements for powder patterns when using the default integrated Chi^2
• Rotation amplitudes for torsion angles and overall molecule rotation is now tuned so that the average displacament per atom is now set to 0.1 Angstroem.
• Faster refresh of graphical user interface
• Using cctbx rather than atominfo + sglite, enabling using either Hall, Hermann-Mauguin or spacegroup number for spacegroup description. Fox will revert to the Hermann-Mauguin symbol.
• freeglut and wxWidgets are statically linked under Linux as well to avoid compilation issues for user without root access.

BUG FIXES

• Avoid creating H-H bonds when converting ZScatterer to a Molecule

1.6.0.2 - december 2003

NEW FEATURES

• Added the export to a POV-Ray file to create a Ray-traced view of the 3D Crystal window, with support for both Crystalstructure and Fourier maps.
• Added more user control on the flexibility of a Molecule:
  • - Added "Rigid Body" and "User-Chosen Free Torsion" choices for the flexibility model of the Molecule. - Also added an option to disable the optimization of the orientation of the Molecule (only useful for "Rigid Body" mode) - Give access to the delta and sigma parameters to calculate restraints, although users are HIGHLY advised not to change the default values unless they really know what they are doing.
    • Also, a Molecule will by default be less flexible than before
• Added import of DSN6 Fourier maps, which can be created by forplot in GSAS.
• Reverted to using GLUT under Linux to display atom labels in the OpenGL window, as some graphical drivers did not fare well with the new method.
• Ability to remove several atoms/bonds/angles at the same time. When removing a bond, offer to remove the bond angles and dihedral angles using that bond.
• Added reflection labels on the powder pattern graphical display.
• Speed improvements by avoiding (more) recomputations.

• Fox now uses wxWindows 2.4.x (no support for versions <2.4).

• Added support for Cobalt X-Ray tube. Also, correctly save X-Ray tube parameters when they have been entered manually.
• Fourier maps with negative contour values will now be highlighted towards negative values, for Fobs-Fcalc maps display.

BUG FIXES

• The position of peaks on a powder pattern was randomly displaced by up to +/-2 pixels for each reflection. This could be serious for patterns recorded with a low resolution relatively to the peak widths, although the use of "integrated" Chi^2 and R-factors reduced the gravity of this bug. (thanks Michal Husak for serious testing and reporting)
• In the CIF output, some atoms were wrongly labeled as being overlapping with another identical element.
• In the graphical display of powder patterns, improve the precision of the display.
• Correct center of view coordinates when the center of the displayed structure is not at (0.5,0.5,0.5)

1.5 - June 2003

NEW FEATURES-MAJOR CHANGES

• The optimization will now use log(likelihood) as a criterion: in practice, this means that the Chi^2 statistics are used rather than the user-chosen cost functions. The likelihood can sometimes be parametrized in the objects (crystal, pattern), but not any more in the Global Optimization objects.
• Added a Molecule class, which supersedes the ZScatterer approach. The scatterer is defined by a list of atoms, and restraints (bond lengths, angles and dihedral angles) to specify the geometry. This gives more flexibility, allows a better convergence and allows to correctly define cyclic molecules.

• Added the ability to define a statistical positional error for atoms, which will be taken into account following Maximum Likelihood principles (ML field in ScatteringPowers), of course with a significant amount of approximations to be computationally friendly.

• CIF output (with the help of Brian Toby).

OTHER NEW FEATURES

• Faster computation when using integrated profiles for powder pattern objects
• Error bars are now displayed on the Powder Pattern graphical display.
• Further improvements of the Fourier maps display (mostly from Brian Toby)
• A cursor can be displayed in the center of the 3D Crystal view, its coordinates are displayed at the bottom of the window, with the value of Fourier maps (if any)
• Added atom labels in 3D Crystal view (right-click menu to deactivate)
• In OpenGL view of a Crystal structure, shift-dragging with left mouse button now correctly changes the center of the view without changing the rotation center.
• Changed lighting model in 3D Crystal view

• Added the ability to save structure factors from a DiffractionSingleCrystal object, including detailed contributions from different elements (useful for anomalous diffraction experiments)

BUG FIXES

• Fixed bug when exporting to a Fenske-Hall z-matrix, the 4th atom was always repeated twice
• Overall ZScatterer occupancy is now correctly taken into account.
• Auto-save of xml files will now use local time rather than gmt.

1.3 - december 2002

NEW FEATURES

• Added a contribution from Michael Y. Polyakov, (working with Brian Toby) to display Fourier maps exported from GSAS in Fox (use right-click when displaying 3D structure to import).
• Added scattering power for a sphere, e.g. as a model for a disordered fullerene.
• Added the option to use pseudo-crystallographic symmetry in a Crystal structure (i.e. use a spacegroup which is not allowed by the unit cell). Could be useful for phase transitions.
• Added data format for Fulprof formats 4 (synchrotron) and 6 (multi-detector, LLB G42)
• Now more than 100 atoms can be displayed in a ZScatterer unit.
• Added the ability to set relative limits on all scatterer translations from their current positions.
• Added the ability to run Fox without the graphical interface, once all objects to be optimized have been setup and saved in a xml file.
• shift-dragging with left mouse button allows to change center of the 3D Crystal view.

BUG FIXES

• Fix serious bug in the determination of systematic extinctions, which lead to missing reflections for (body|face)-centered spacegroups with a center of symmetry not on the origin (e.g. for Fd3m (Si,...))
• Fixed various bugs in the display and update of powder patterns

1.2 - august 2002

NEW FEATURES

• FOX now supports Preferred Orientation, using the March-Dollase model. The texture parameters can be searched ab initio.
• The 3D display of the crystal structure is now also refreshed automatically under WinNT and above (still much less responsive than under Linux -hint,hint).
• The default size of crystal structure and powder pattern are more reasonable under windows.

• New atoms in ZScatterer are now given a sensible default name, ScattPowName+number of the atom.

• The wavelength is now directly displayed in the window, rather than exclusively through a menu.
• Added ability to duplicate scatterers in Crystal.
• Powder diffraction pattern keeps zoomed during optimization.

BUG FIXES

• The atom names in ZScatterer are now validated without the 'enter' key being hit.
• Adding atoms in a z-matrix scatterer does not crash anymore.
• removing a scatterer (or any object...) after running an optimization, does not crash any more when optimization is relaunched (the parameter list is now correctly rebuilt).
• Now correctly zooming when the dragged area goes beyond 2theta limits.

1.1.2 - 2002/02/18

License changed to the GNU General Public License. NEW FEATURES

• Added a maximum value of sin(theta)/lambda for diffraction data objects. All data above is ignored.
• Added an *experimental* option in single crystal data for twinned data, to work on the sum of metrically-equivalent reflections. Entirely untested (contact me to use it !)

• Some parameter are automatically fixed for global optimizations (unit cell, background & profile parameters).

• Fox will write a few messages for some key events (importing data, creating spacegroup,...)

BUG FIXES

• The dynamical occupancy correction now begins for interatomic distances below 1A, and proceeds to full correction for .1A (before that 2 and .8A where used (why?))
• Spacegroup and Unit Cell values can now be changed even if scattering data has already been computed, without crash.
• Importing a Z-Matrix would give the 3rd and 4th atoms the same number, and if atoms were of the same type, both atoms would have the same name in the .xml saved file, and would crash on reloading. Molecules thus created need to be re-imported.

1.1.1 - 2002/01/17

BUG FIXES

• Under windows, the global optimization lead to a crash after some random time.
• Under windows, clicking on some part of a powder pattern graph would make Fox crash

1.1 - 2002/01/8

NEW FEATURES

• Now it is possible to import Z-matrices to build a molecule structure (eg a Z-matrix file can be generated from a .pdb file using babel, and then imported into Fox)
• Added integrated R and RW factors as new cost functions (thus less sensitive to badly defined profiles). Unless you have *very* good profile parameters, I recommend using the integrated Rw as a cost function.
• Now it is recommended to choose an "exponential" schedule for the amplitude, from 8 to .125, and a "smart" schedule for the temperature. This should be optimized for *all* experiments, so that users don't have to guess the right choices.
• It is no longer necessary to validate all input by hitting the 'return' key (which was *very* annoying)
• Ability to zoom on the powder pattern graph, using the left button (double-click to unzoom)
• Possibility to reduce the part of the displayed unit cell in 3D view (use the popup menu with right-click)
• Significant speed improvement.
• The Parallel Tempering algorithm has been tuned to yield a higher success ratio.
• Possibility to used single-crystal data (eg extracted intensities) for crystal structure optimization (honest crystallographer's note: if you have high quality extracted intensities, try direct methods first...)
• Added interface for 2theta zero/displacement/transparency parameters in a Powder Pattern object
• Added an overall temperature factor (Biso) for each crystalline phase in powder diffraction (also available in single crystal).

• new examples for organic molecules (potassium tartrate and cimetidine with powder & extracted pseudo-single crystal intensities)

• The program now forbids to check lattice or profile parameters for a global optimization.

LINUX:

• Now using a distribution from source using the wxWindows RPM, which
• should make things easier under Linux.

BUG FIX:

• under linux, the antibump cost function was not properly restored from an xml file (preventing merging of identical atoms)
• Changing the spacegroup (and thus extinction rules) will now will re-generate the list of reflections for powder diffraction. Same if the lattice parameters or the wavelength has changed.
• a number of minor bugfixes...

1.0.1 - 2001/10/15

BUG FIX:

• corrected minimum FWHM for profiles from .001 to .00001 radians

1.0 - 2001/10/15

NEW FEATURES

• support for multiple crystalline phases in Powder Diffraction (you can refine several phases from one pattern ...provided you can index them)
• the Windows version now displays a console to give a better idea about what the program is doing.
• Import of Sietronics data files (.cpi)
• Linux binary version also available for glibc 2.1

BUG FIXES:

• removed another bug in ZMatrix interpretation, which prevented the correct coordinates to be generated when using mono/triclinic unit cells (thanks Mark Edgar).
• the program will no longer crash when importing data with null or negative points.
• U,V and W parameters leading to negative profile width will no longer crash the program (profile forced to be strictly positive)
• various cosmetic bugs

0.9.1 Beta - 2001/09/20

NEW FEATURES

• Added menus to change limits globally in a ZScatterer(bond,angles)
• Now the pivot atom (around which the ZScatterer is rotated, and which corresponds to the displayed x,y,z position of the ZScatterer) is saved in the .xml file. It can also be changed from a menu.
• Limits can be changed individually for all parameters, using a popup menu: right-click on the parameter name to get the local menu

BUG FIXES:

• removed a nasty bug in ZMatrix interpretation, which prevented the correct coordinates to be generated for chains longer than 3 atoms. Did not affect polyhedra... (thanks Yuri Andreev)

0.9 Beta - 2001/09/18

• Initial Public release

0.5-0.9

• Internal release only... Thanks to Laure Guenee, Guillaume Renaudin and Radovan Cerny for testing !