v.6.23.01

Moderate update that fixes two bugs:

DISCUS
output Convolution with profile function was erroneous
for very short wavelengths at low Q

KUPLOT
load Data with large arrays with the y-dimension larger than the x-dimension caused a crash

v.6.23.00

Moderate update
Fixed several small internal issues

GENERAL
Revised WINDOWS WSL installation script

DISCUS
save/read Added a field to HDF5 files to specify if coordinates are in range
of basecell or supercell fractional coordinates
export Added a 'radiation' and 'qmax' to refine preparation
EXPERI
Improved plots in interactive mode, added wave length input

v.6.22.01

Minor update; needed for CALGARY workshop
added a "coordinate_unit" field to HDF5 unified file format

v.6.22.00

Important new version!

GENERAL
Introduced two common file formats for:
disordered structures and
diffuse data

DISCUS
mmc Improved convergence of 'cn' coordination number target
mmc Modernized 'apply_symmetry' output
import Modernized preparation of refine macros
output
read Files with ending ".hdf5" ".h5" are interpreted as
unified common structure format
save Added format 'nexus' for unified common structure file format

KUPLOT
calc Extending calculations to 3D
'val' qualifier substitutes old 'wz'
'wz' refers to coordinates for 3D data
load Added 'rmc' for RMCprofile 3D data
load Added 'nexus' for DAunified common diffuse data file format
match Extended to 3D data
variables val[data_set, h,k,l] returns value at pixel for 3D data
val[data_set, h,k ] returns value at pixel for 2D data
val[data_set, h ] returns value at pixel for 1D data

v.6.21.03

Moderate update,
Predominantly a bug fix regarding the 2D convolution in DISCUS
Ongoing development in DISCUS/mmc not yet ready for a general release

v.6.21.02

Moderate update
Most importantly, a bug related to the Lanczos filter was fixed

GENERAL
do/if Improved error handling while reading macros inside a loop
SUITE
parallel BUG FIX Parallel command at Windows did not work
DISCUS
GENERAL Revised internal calculation to recognize molecules at read
Molecules far from the origin were not detected properly
fourier Lanczos filter did not work properly for 3D data
import Corrected lambda and qmin in refine/diffev setup
wyckoff Improved output to signal if the position is on a
secondary symmetry operation like a mirror plane at x=1/2
KUPLOT
smooth Lanczos filter did not work properly for 3D data

REFINE
newpara Added a 'move' parameter to the 'newpara' command to calculate
derivatives via an additive parameter change.

v.6.21.01

Minor update,
predominantly one bug was fixed.

DISCUS
read BUG FIX Molecule recognition was faulty

v.6.21.00

Note January 23, 2026
The Windows installation guide has been updated.

Small update,
adapted the DIFFEV newparam command to use identical format to REFINE

DIFFEV
newparam Changed style to be identical with REFINE
release BUG FIX, did not work without reading the GENERATION file,
did not work for Parameter names
DISCUS
save Improved storage of anisotropic values to avoid rounding errors

v.6.20.00

Fairly big update, several new features, some BUG Fixes

GENERAL
Changed warnings into blue color
Variables i[] and r[] can be used with a range of indices as in
r[1:10] = 0.0
r[ :10] = 0.0
r[1: ] = 0.0
Added system variables 'F_XMIN', 'F_XMAX', 'F_XSTP' and equivalent
for 'Y' and 'Z'
Their values are set to the data limits by the 'data' commands
in REFINE and DIFFEV
functions Added a lognormal function
DIFFEV
GENERAL Adapted diffev_best.mac to include 'data' command
data Added a 'data' command that creates variable F_XMIN, F_XMAX etc
purge Added a command to purge the initial generations from the logfiles.
DISCUS
import Added options to prepare a refine / diffev setup
save Added a unified HDF5 type structure format
DISCUS
GENERAL REPLACE command with several new atom types
GENERAL Most variables can be used with a range of indices as in
x[1:10] = 0.0
occ[ :10] = 0.0
b[1: ] = 0.0
export Added option to shift origin to "-1" at SHELX export
Added option to include wave length at SHELXL export
mmc Fixed "displacement correlation"
Added atom value correlation
output BUG FIX Powder output modified limits given in powder menu
powder Improved internal workings for convolution
BUG FIX, Preferred orientation and non-00l stacks needed to
be aligned as in 6.19.00
show Added display of "-1" location
super Added multiple waves from separate input files
Added interpretation of atomic value as phase
surface Extended options of 'set' to replace a shell by other atom types

v.6.19.00

An important update related to powder diffraction
Possibilities to calculate the powder pattern with stacking sequences along [100] or [001] have been improved and made much faster.
If a stack consists of N-unit cells, the optimum calculation is to use steps of 1/N along the diffuse rods. At smaller steps, the weighting scheme between diffuse and sharp rods was off. This bug has been corrected in this release

DISCUS
chem Improved aver command if exact mode is set
perioditize Improved algorithm
powder/stack Major BUG fix. Stacks along axed other than [001] did not work well.
Corrected relative weight of sharp/diffuse rods