11/06/99 |
First public release of GUI version |
15/06/99 |
Files packed into an installer. No need to edit autoexec.bat etc.
Resize-able windows now have minimize and maximize buttons.
This version no longer relies on any environment settings. It's
location is set at install time in c:\windows\wincrys.ini
The code for the model window has been re-written to make it more
stable. A view menu has been added to set flags to do with drawing
the model.
Windows (optionally) store their sizes on closing so that they will
open in the same place. The file used is script\winsizes.ini (in
case something goes wrong!) The only windows using this feature are
the main window and the Cameron window.
Some scripts have been corrected where errors have been found by
user testing.
CrysLoad executable added for choosing a directory to start
CRYSTALS in. This is the file that any desktop and startup icons
should point at.
"Tip of the day" script added, with lots of useful hints for users.
This may be disabled from the window that pops up. (This fact is stored
in script\winsizes.ini) |
18/06/99 | Fix: Text output was always just off the bottom of the screen.
Feature: You can now select output with mouse and copy with CTRL-C.
Fix: Startup tips no longer flickers nor forgets the state of
the "Show tips at startup" box when you choose More...
Fix: The check for old DSC files gave an invalid default choice.
Feature: Status bar is much more infomative about what CRYSTALS is
up to while you are waiting.
Feature: Split data menu into X-ray data (cell, scattering, space group,
reflections) and Input Structure. Added Shelx file to the
Input structure menu.
Feature: Added perhydro to the refinement menu.
Fix: Made default sizes of CRYSTALS and CAMERON window smaller
so that they startup first-time on 800x600 screens.
Feature: Reduced space used by buttons in Cameron window. Added a
Hide Qs/Show Qs button.
|
22/07/99 | Fix: CAMERON: Torsion angle output to progress bar now fits.
Fix: CAMERON: Dotted bonds and HBONDS can now be seen even
when very small.
Fix: CAMERON: Added dummy atoms type L and M to the properties
file, so that the VOIDS code works better.
Feature: Routines #SLANT and #FOURIER will now output ASCII files
containing Fourier maps, aswell as the original binary files.
The direcive is MAP FILE=YES for both instructions.
Fix: Any element types that didn't have a default colour in
propwin.dat, were given a default (small) radius in the model
manipulator which would then fail to find bonds to them.
Fix: Reflection scaling for SHELXS output was broken.
Feature: #STORE CSYS MAP 'filename' sets the default output name for
ASCII fourier maps.
Feature: Informative status bar text shows whether CRYSTALS is
waiting for commands, directives or script input.
Fix: Fixed cursors.
Feature: Re-organised some of the menus. Put a link to the web page
in the help menu. Added some more types of Fourier map for
completeness. Added inlist14 to the Fourier menu.
Removed all menu items that didn't do anything.
Added an input section to the Structure menu. Added several
commands to the structure menu: Resequence atoms, collect
atoms, distances, add hydrogen, remove hydrogen, remove peaks.
Added a new refine dialog with a few more options than just
'How many cycles'. Added a shift limiting restraints dialog
to the refinement menu.
Added scripts for running RC93 and DIPIN from within CRYSTALS.
Changed the default editor to notepad for edlist12, edlist16
and edlist5 scripts.
Added a publish option to XAUTO12 (guided refinement) script.
Added output to several scripts to keep users informed of progress.
Feature: Cameron window is now a better size, compared to its button panel.
Feature: Overhauled XCELLIN, XSPACE, XINLIST3, XINLIST6 and XSHELXIN,
so that they work in harmony, and in the right order.
Feature: Click on a few atoms , then right-click for a context-sensitive
menu. Choose "Slant Fourier Map", to get a dialog to compute a
slant map through the plane defined by the chosen atoms. You
have the option to run either CONTOUR or Michal Husak's Marching-
Cubes 3D electron density viewer.
Fix: When scale was fixed in Cameron, postscript scale would be 5 times
too big. Fixed.
|
30/07/99 | Feature: Check for H atoms in LIST29 to decide whether to run FINDH in the
XAUTO12 script.
Feature: #SCRIPT ACTA makes sure LIST30 has acceptable values for publication.
Fix: Obscure progress bar related bug fixed.
Feature: CIF routine outputs summary table to screen.
Feature: #script DEMOP runs through a demo of some features
Fix: VOIDS no longer brings the model manipulator to its knees.
Feature/Fix: Accidentaly typing #CAM no longer does anything.
|
03/08/99 | Glasgow IUCr release. Demo menu added for aiding demonstrations of the
software to the conference delegates. This might stay as it provides a
very rapid introduction to the package.
|
15/08/99 | There is a problem under NT. As a workaround download the file
"crystals_nt.srt", rename it to "crystals.srt", and replace your
existing copy of "crystals.srt" in c:\wincrys\ (or your installed
location).
The workaround means that you will NOT be able to use the Open or
Save As options on the File menu. For the time being under NT, your
datafile must be called crfilev2.dsc. If you have an old file by another
name (e.g. CRFILE.DSC), you must copy or rename it to crfilev2.dsc to use
it. If you double-click on any DSC file, CRYSTALS will start in the
correct directory, but will use the file CRFILEV2.DSC as its datafile.
Fix pending...
|
10/09/99 | Fixed a bug in #MOLAX, when calculating an angle between two planes,
the angle would not come to the screen, and could even crash the program.
|
17/09/99 | Added link to SIR97.
Better weighting for weak reflections.
|
21/09/99 | v11.41 release
Added script for writing and reading Sir97 files. Though can't run Sir97
automatically yet.
Added link to each manual section to the Help menu.
Added links to new sir97 scripts to the menus.
Fixed a bug in Cameron where distances between symmetry related atoms were
omitted from the calculation. Also fixed a old bug where the translation
part of the output of symmetry operators during a dist calculation was
going missing. Thanks Bruce!
|
22/09/99 | v11.41a release
Changed bond tolerances used by the model viewer to match Cameron to some
extent: The bond must be at least 1.21*(sum of covalent radii) to be drawn.
Unlike Cameron the model viewer will not draw bonds which are shorter than
0.5*(sum of covalent radii), because that's just too short.
|
08/10/99 | Correct input of atom nunbers from SHELXS.CRY files.
|
13/10/99 | RC93 - Add new input format for .DAT files from French machine
(3 spaces at start of line, 2 blank lines between data lines)
|
26/10/99 | Fix SCRIPTS so that Fsq refinement is stable even if the user
forgets the weighting scheme
|
15/11/99 | Patch link between SCRIPT SLANT and CONTOUR/HUSAK.
Add DIFABS TEST to Analyse menu
|
09/12/99 | Changed colour-scales in Contour, compute number of expected
atoms for SHELXS input, add PRINT command to CAMERON menus.
|
24/01/00 | Add directive RESET to edit to facilitate strucutre editing
with graphics window.
Syntax: RESET 'parameter_name' 'parameter_value' 'atom_list'
The named parameter is set to the given value for all the
specified atoms.
'PUNCH' Directive in MOLAX causes orthogonal coordinates to be
send to the 'punch' file (*.PCH)
|
23/02/00 | '#PUNCH 5 D' creates a Chime .XYZ file called publish.xyz
(Chime is a strucure viewer downloadable from the Web)
'OUTPUT CHIME' in CAMERON creates a file chime.xyz containing
everything in the CAMERON window.
Help added to 'Guided Refinement'
Extinction reset if it goes negative (physically unreasonable)
CAMERON layout changed to give full width text line.
|
18/04/00 | Bugfix in Chime output from CAMERON, attempt to stop 'guided
refinement' trashing user-defined LIST12, more bug fixes for
weird CAD4 output.
|
04/07/00 | Allow continuation lines in a #COMPOSITION instruction.
|
11/07/00 | List 29 extended to include COLOUR.
Fix bug where NT version would not start without NT_patch.
|
17/07/00 | Check for unreasonable Uequiv.
Output special position info to the screen during refinement.
|
19/07/00 | Fixed bug where GUI hangs for hours if the atomic parameters shift
to very unreasonable values.
Warn about invalid elements in a #COMPOSITION instruction.
|
20/09/00 | Re-vamped colour output. Added colour highlighting
to important pieces of information.
Re-arranged the menus slightly.
|
21/09/00 | Added new text output control. Faster, better colours, and hyperlinks.
|
03/10/00 | Perhydro can add H to nitrogen atoms. H can be manually placed along
likely H-bonds.
H atoms are no longer implicitly included in a refinement by LIST12.
ie. FULL X'S U'S will not refine hydrogen.
FULL FIRST(X'S,U'S) UNTIL LAST will.
|
04/10/00 | Benchmarking code added. #BENCH.
|
10/10/00 | #PURGE DATE produces a fresh dsc file with date and time for a filename.
|
11/10/00 | Generation numbers are now added to the file root rather than extension.
|
30/10/00 | File root for output files has changed to BFILE
|
31/10/00 | Added Split Atom option to context menus.
|
03/11/00 | Bitmap control for displaying pictures in the GUI windows.
|
17/11/00 | List 5 extended to 18 parameters per atom.
|
04/12/00 | Shared some things between the popup menus and the Appearance menu.
Added links to all chapters of all manuals from the Help menu.
|
11/12/00 | Change type popup menu now has access to XOTHER script for other element
types.
|
12/12/00 | Voids map option in context menu gives figure-field output compatible with
contour programs. Height is distance to nearest VdW surface.
|
13/12/00 | Script for manually adding H to atoms added to context-popup-menu.
|
19/12/00 | Added Kappa CCD input option to the menus.
|
22/12/00 | List 6 can now be re-indexed as it is read-in.
Twin-handling improved.
|
11/01/01 | Analyse against FC as well as FO.
Cameron mode=special starts cameron without writing new files, so that you
can view old data. Added this option to the menus.
|
12/01/01 | Default PERHYDRO action is now 'BOTH'.
|
17/01/01 | Overhauled log file. The log files are now useable, with no dependency on
external programs - e.g. #SCRIPT command isn't logged, but anything the script
does is logged. e.g. #USE isn't logged, but the commands in the file are logged.
$ and spawn commands are not logged. The only problem with the logs is that
they fill up with #generaledit commands issued by the scripts. These don't do
any harm, but make it difficult to browse the log.
|
18/01/01 | #DISTANCES - new OUTPUT PUNCH option:
OUTPUT PUNCH=SCRIPT - outputs ride instruction for each pivot atom
found in the following calculation.
Fix: EDIT directive to add new atom (ATOM) did not match format of new
extended list 5.
Input editbox for CRYSTALS and CAMERON limited to 80 characters.
|
19/01/01 | Punched list 6 files now only contains keys that are actually present. (Save
space)
|
22/01/01 | Peak height passed to GUI molecule viewer. It shows up in popup text over the
atom.
|
25/01/01 | New TAB control for the GUI for organising more information than will fit
in one window comfortably.
|
14/02/01 | Added new instruments, programs etc to LIST 30 Output. Correct Occ
if the user has set LIST 23 to suppress its computation.
New cif generator, including alphabetically ordered references,
proper accout taken of Occ in M, Z and mu
INTWIN script used for inputting twin law for TLS and TLQS twins. Reflection
list is updated automatically. For two component twins the twin element scales
are automatically added to the parameter list and can be refined from the
guided refinement scripts.
|
26/02/01 | Resize bar control added. Contains two panes laid out horizontally or vertically.
If contents of both panes can resize, then you can drag the bar inbetween them
to adjust the sizes. If only one pane can resize, then the other may be collapsed
away by clicking on the bar. (If neither can resize, you don't need a resize bar).
Toolbar control added. Same sort of syntax as menu items.
|
02/03/01 | Added exponentiation operator EVALUATE X = X ** 3.5 to scripts.
|
08/03/01 | #DISTANCES - two new directives:
PIVOT - atom can be a pivot in the calculation. Disables the
INCLUDE directive. All non-specified atoms will be type BONDED,
unless there is a BONDED directive, in which case non-specified atoms
are excluded.
BONDED - atom can be bonded to in the calculation. Disables the
INCLUDE directive. All non-specified atoms will be type PIVOT, unless
there is a PIVOT directive, in which case non-specified atoms are
excluded.
Changed directive:
ONLY - no longer takes any atom specifications, use it in conjuncition with
any other directives to make the calculation use the LIMIT card rather than
the covalent radii of the atoms.
GUI - new info tab below text output window is updated with many details
of the analysis as it proceeds. Just so you know what CRYSTALS knows.
The "Files" tab has buttons that let you open the current log, list and punch
files for reading. Very handy.
GUI - matrix defining thermal ellipsoid is sent to gui model viewer. You can
now choose between Ball and Ellipse views.
GUI - non modal windows (i.e. windows that let you carry on interacting with
the program) now have a thinner title bar to distinguish them from modal windows.
SCRIPTS - added random number function EVALUATE X = RANDOM ( 0.0 )
the 0.0 doesn't do anything yet, but has potential to be a seed.
New GUIDE set of scripts improving on the Guided-Automatic refinement set
of scripts.
Important Changes:
GUIDE stays open, but you can still use menus, type commands, use model etc.
GUIDE window is much smaller so that it doesn't hide so much information.
GUIDE works for whole analysis, (import and solve added).
Refinement options include various H treatment methods, including automatic RIDE.
GUIDE is much better at remembering what has been done so far.
|
21/03/01 | New #SPAWN ($) method. You can now type "$file.lis" and CRYSTALS will
open the document using the Windows associated program (for .lis files).
Dos window no longer pops up for external programs unless it has to. You
can still say "$notepad file.lis" to specify a program. Spawn (and $) no
longer wait for the called program to finish, unless you include a + symbol
before the program or document name (i.e. straight after the $)
e.g. "#SPAWN +calc" will wait for calculator to be closed.
|
23/03/01 | Fourier scripts updated to use the /FOT/ key for scaling and phasing before
a map calculation on a twinned data set.
|
27/03/01 | Fix: Prevent CRYSTALS from occasionally "freezing" until mouse moves.
XLIST28 updated to provide many more options for filtering reflections.
Fix: If you say no to six lsq cycles before publication, a cycle with no shifts
is carried out so that atom list matches current lsq matrix.
|
28/03/01 | Fixed site occupancies in LIST 5 part of cif output so that the
effects of crystallogrpahic occupancy are explicitly removed. This is
not explained in any CIFDIC, but is stated in current Notes for Authors.
|
30/03/01 | Script XSPLIT now checks that an atom is actually aniso before attempting
to split it.
|
11/04/01 | Fix: Bill Harrison reported a mis-match between the list of symmetry operators
and the symmetry codes on distance/angles for bonds hanging out of the +1 side
of the cell. Fixed.
|
18/04/01 | Echo Cameron input lines to output window.
Several improvements to Cameron interface.
GUIDE - Reset flack parameter if no-longer being refined.
Manual H addition - generate sensible serial numbers.
Reorganised UI so that it fits in 800x600 screen.
SHELXIN - added some more previously unencountered keywords.
XSPACE - limit input to 20 chars and correct a few space groups in the
drop-down list which were missing spaces. (Commonest problem - P21/c)
Auto H addition - removed that annoying "Do you want to remove H" question,
when there aren't any H's in the model.
|
01/05/01 | New IUCr CIF checking web address.
|
16/05/01 | SCRIPT XLIST28 - Fo/sigma(Fo) should say I/sigma(I).
SCRIPT XSIR92 - enable rerunning of edited Sir92 input files.
|
18/05/01 | SCRIPT EDLIST5 - now always uppercases element type before searching for it.
SCRIPT XPARAM - now always uppercases element type before modifying it.
|
23/05/01 | Patches to LIST 6 to improve twin handling.
|
31/05/01 | Fixed: bug in listing file of #SUM L 28.
Enhanced: New processing of SLICES and CONDITIONS in LIST28. See the
distributed version of the manual for details.
|
04/06/01 | Fix: s.u's of overall parameters now displayed when you type
#PARA / END.
|
06/06/01 | Fix a colour postscript bug in Cameron when drawing unfilled circles.
Also fixed save postscript dialog - it was overriding users colour choices
if printing in black and white.
Added an alternative B&W colour scheme to the Cameron main window options.
Make sure Refinement dialog is closed if there is an error due to an
invalid LIST 12.
|
08/06/01 | #REGULARISE RENAME option for matching and labelling structures.
Fix: updating F/F2 status between L23 and L30.
|
11/06/01 | Fix: Script ED30 would only show mol weights up to 999.
Update: Molecular weights for common atoms brought into line with CIF checking software.
Update: Anomalous scattering factors for common atoms.
Fix: Script XINLIST3 (Edit Atomic Properties) would cause Mo scattering factors to be read
regardless of the radiation type chosen.
Update: SHELXIN: ZERR card used to get Z (for list 30) and cell errors (for list 31)
Update: SHELXIN: Support for separate SFAC cards.
Update: SHELXIN: Support for = continuation character at any point in atom card.
Update: SHELXIN: Wavelength read into list 13 from CELL card.
|
12/06/01 | Printing of SU's of bond lengths and angles to non-refined
H atoms to the CIF is supressed by default in scripts XCIF and CIF in accordance
with IUCr's notes for authors.
Warning on startup if > 150 atoms and option to turn off updating of model at this point.
|
14/06/01 | Added SHELX to list of files to create in publication tables dialog.
|
19/06/01 | Size of internal store doubled to about 1Mb. (On request, versions with much more store
can be specially compiled - current record is about 32Mb).
|
03/07/01 | Added ROTAX algorithm for identifying non-merohedral twin laws. (written by Bob Gould and
Simon Parsons, Univeristy of Edinburgh)
If reading Fo^2 from a file, then assume any Fc values (if being read) are also squared.
|
11/07/01 | Enable Flack parameter to refine to negative values. (It may be negative within bounds
of experimental error.)
|
13/07/01 | Display number of outliers during chebychev weighting analysis.
Add menu option to Results for helping to find H-bonds.
When reading full SHELX files (as opposed to SHELXS files) don't ask how many atoms to
keep. Just get them all.
|
16/07/01 | Ensure F/F^2 status info is update as soon as CRYSTALS starts to avoid confusion.
|
19/07/01 | Torsion - fix crash when outputting some angles to CIF.
Added an _atom_site_type_symbol to the CIF output, as some programs like this info.
Output /Fo/ and delta**2 column headers when showing weighting scheme analysis.
Changed the GUI update logic so that #SET UPDATE OFF only stops the model from being updated
(this makes sense as it is the model that really slows things down for big strs).
Autochange model style back to "ball" when doing the geometric/fourier H comparison.
Added dialog to help packing "windows" in Cameron. |
RELEASE b1165
|
24/07/01 | Fix bug where Cameron fails to display drawing on initial start up.
|
25/07/01 | Rotax improvements. XROTAX script added. UNTWIN script added for reverting to no twinning.
|
RELEASE b1165a
|
08/08/01 | Ignore Q atoms when generating VDW VOID maps.
|
09/08/01 | Added an _atom_site_type_symbol to the CIF
|
14/08/01 | After applying twin laws, round indices to nearest integers.
#FOREIGN SHELX SPECIAL now allows serials for one character elements to run up
to 999, instead of just 99.
Holding down CTRL and clicking a "resize" bar will swap the panes either side of it.
Holding down SHIFT and clicking a "resize" bar will change its orientation from
VERTICAL to HORIZONTAL or vice versa.
|
RELEASE b1166
|
15/08/01 | Some new SCAN types for LIST 30. (e.g. PHI and OMEGA)
New option to "view list6" from the X-ray data menu
New dialog for updating List23 (sfls control).
|
16/08/01 | Twinning. Backup original list 6 to a file so that recovery is possible.
|
RELEASE b1166a
|
20/08/01 | Print information about validation tests as they are carried out by script ACTA.
|
23/08/01 | Avoid singularities in certain situations in #HYDROGENS
|
03/09/01 | If importing a SHELX data (but not from SHELXS), read in all the atoms regardless of
how many the UNIT instruction say are present.
|
04/09/01 | Scripts can write to the PUNCH file!
|
07/09/01 | New GUI feature allows nice bitmap backgrounds behind the view of the molecule. Looks nice, but
slows down drawing, so don't use on slow machines.
Early Win95 fix. The Toolbar control isn't fully implemented under early version of Win95. This
fix works around a GetMaxSize() bug, but the result is ugly. Recommend upgrading to the latest
Internet Explorer as this includes several service packs which fix this problem.
|
09/09/01 | Tidied up CIF output. All data are lined up in column 35 of the CIF if possible, to make
reading and editing easier.
|
11/09/01 | New option to step back and forth through any list 5's (parameters) stored on the disk. Can
be used as an UNDO option.
New date format for info in the DSC file. Lists now have time included (as well as date). This
change is backwards compatible with old DSC files. Old dates will be displayed as before.
Stop #GENERALEDIT commands from filling up the log file. The log records if a change has been
made to a particular list.
New script to produce summary of all data (accessible from "Results" menu).
New script (XFRIEDEL) called by all the major import scripts, for non-centro structures, it gives
an option on whether to merge friedel opposites, and makes a recommendation based on the heaviest
atoms present and the type of radiation used.
|
RELEASE b1167
|
12/09/01 | Fix: Include missing "sum" script.
|
RELEASE b1167a
|
08/10/01 | Reorganised calling structure of program, now uses one single command processing loop at the top.
Increase H-bond upper limit from 2.0A to 2.2A (Foxman).
|
17/10/01 | Added Lists 40 and 41 to hold bond lists, and commands for generating bond lists. Watch this space.
Added #BONDING command to generate a L41 from L5 and L40.
|
18/10/01 | Bond labels in the GUI now display "fragment number" as well as bond length.
|
06/11/01 | In #REGULARISE add new facility for propogating atomic typing and numbering
from a 'master' fragment into other fragments. Add corresponding SCRIPT to
Development
In #REGROUP add GROUPNUMBER (=yes/no) to add a group number to each isolated
molecular fragment. This may become a way of references 'groups' in future
releases. Instructions like SORT PART already work in #EDIT.
In Cameron, produce correct bounding box for .EPS output.
Added 'quiz' scripts to Development menu.
|
08/11/01 | Output of graphical analyses from weighting/analyse code. Still in development, but running.
|
13/11/01 | Fix free format reads of reflections. Fix bug when no file name given. Fix default format when reading shelx files.
|
16/11/01 | New commands for outputing graphs from #WEIGHT and #ANALYSE (GUI GRAPH=ON) and (PLOT
CLASS=ON PARITY=ON etc) respectively. By default all are off.
Streamline the guided refinement dialog. 2nd dialog (ncycles/stop on converge/edit restraints) is not displayed unless
user clicks advanced checkbox.
Added support for SHELX ins/hkl import to the import guide.
Streamline the H-placing dialogs. Now contains checkboxes for 'all atoms found' and 'geometric or fourier'.
|
20/11/01 | Allow SHELX input scripts to read a .ins with no atoms without complaining, and without creating an empty list 5.
|
RELEASE b1169
|
23/11/01 | Link to Platon's squeeze - new output style, and program convplat.exe to work platon.hkp back into a list 6.
More improvements to agreement analysis plots.
Call GUI 'filter reflections' dialog from xinlist6 (import hkl) script, rather than assuming no filters.
|
02/12/01 | Fix auto-riding-H bug, when thing H is bonded to is on a sym-op (e.g. 2-fold). The sym-related H was also
being included in the instructions causing the same parameter to be referenced twice in L12. (Which would halt
refinement).
|
04/12/01 | New button type. Enables users to abort long forced cycles of refinement. Added to guided refinement script.
|
RELEASE b1169a
|
14/12/01 | TLS and MOLAX code combined into new GEOMETRY module. Enables angles between plane normals
(etc) and libration axes to be computed. Add code to enable TLS restaints to be generated,
and to enable the user to modify the T and L tensors.
|
18/12/01 | IF environment variable CRYUSEFILE is set to a filename, that file will be used upon start
up. This is for WinGX integration.
|
14/01/02 | Added scatter plot of Fo vs Fc for checking extinction and finding outliers.
|
24/01/02 | Reflection output now goes to separate CIF, publish.fcf.
|
01/02/02 | #SFLS/CALC now outputs additional R-factors with data cut at -10I/sigma and 4I/sigma.
|
13/02/02 | Extinction su now stored in List 30.
|
15/02/02 | Update ROTAX to SP's latest version. Added option to view listing file to results dialog.
New XWRITE5 code works without sorting L5. List 12 generated may be longer, use of until
is made to shorten slightly.
|
20/02/02 | Avoid drawing that blue box at the bottom of the Cameron display. Also Cameron: improved
colour tables so that background is actually white, not off-white.
|
21/02/02 | Add calls to VALID12 script all over the place, so that SFLS is never attempted without
a valid list12.
|
27/02/02 | Enable long path names and path/file names with spaces in. Filenames containing spaces
must be "double quoted". CRYSTALS should now install in "Program Files", and refine
structures in "My Documents"!
|
28/02/02 | Added code for calculating and plotting Wilson plot.
Added #SIGMADIST command for outputting I/sigma(I) histogram.
|
01/03/02 | Correct and improve presentation of weighting scheme in CIF file.
|
04/03/02 | Absences and merging-R data output to named files during data reduction. Can be picked
up later for user inspetion.
|
05/03/02 | Added output of N(Z) graphs.
|
06/03/02 | On startup, use CRYSTART.DAT if present in working directory. Differs from CRYSINIT.DAT,
which gets read before the GUI is initialised.
Added "initial analyses" script to the GUIDE. Displays Wilson, merging R, absences, so
you can spot problems early on.
Added wmf output and printing support to Cameron.
Improved the demo so it no longer overwrites your current structure.
|
08/03/02 | Option to filter reflections using LIST 28 when outputing for Sir92.
|
11/03/02 | Enable refinement of extinction for twinned CRYSTALS. Considered unwise (see manual), so
now a warning is printed if you attempt it.
Add histogram output to ROTAX to show where the rejection cutoff is.
|
12/03/02 | More analysis graphs - added R-factor analysis and analysis by h, k and l layers.
Improve spawning of external files/programs to handle URL's with #'s in them.
Eased updating from old DSC files - now no need to reenter lists 29 or 30.
Changed extinction analysis to display Fo vs Fc scatter. Much better.
|
13/03/02 | Preliminary support for output of MOL2 and CSD connectivity file formats.
Overhauled distangl code (again). Now much more robust and flexible.
Rewritten code for updating GUI. Bond types are now calculated and displayed in the
model window.
Spring clean of GUI-menus. Attempted one help link for each menu, explaining what each
menu item does. Added explorer icon to toolbar to open current working folder.
Improved directory open dialogs - they now remember the last working directory.
New popup menu for the model window: context-menu for clicking on a bond.
|
RELEASE b1180
|
17/03/02 | Ensure final diff map before CIF production - if min max density values really aren't
present, use ? in CIF rather that 0.00
|
18/03/02 | Minor bug in new #SFLS/CALC threshold fixed.
|
21/03/02 | Clear error flag before bond calculations (previously bond calc would echo
previous command error)
Extend I/sigma(I) graph down to -5.
Ouput B and scale below Wilson plot.
Output |E**2-1| below N(Z) plots.
|
28/03/02 | Fixed occasional race condition (in b1180) in code that starts Cameron. (more likely
to affect slower machines).
|
29/03/02 | Reduce number of steps in progress bar output for speed increase on slow machines.
|
RELEASE b1180b/c
|
11/04/02 | Remove "optimal" weighting option - default is now "Chebychev" (which was what "optimal"
did anyway. Also, for unit weights, use quasi-unit weights of 1/2F for F**2 refinement.
|
15/04/02 | Fixed bug where formula was not displayed in GUI window info tabs until after
the structure had solved.
|
16/04/02 | Ensure mixed case output of atom types in CIF - for compatibility with some
other program.
|
19/04/02 | Try representing size of Q peaks using thermal parameters - click Ellipse to test.
|
30/04/02 | Swap red and blue bars in Fc and rho agreement plots. Now in chebychev weighting
analysis, the unweighted residual and the parameterised fit are both red for easier
comparison.
Font sizes in Cameron: Font is now proportional to the window/output medium size so
that the picture scales properly. Should fix problems with metafile output. To change
font size type "FONT 18 view". Default font size is 12.
Trial implementation of "sticky refinement modes" for atoms: Right-click an atom and
set a refinement mode which will override the general settings used during "Setup
Refinement and directives".
Added workshop manual volume. Guides user through the solution, refinement and analysis
of two structures, which are included in the release. (With pictures!)
|
RELEASE b1181
|
07/05/02 | Cameron: Fix font size in direct output to printer.
Cameron: Allow output of emf or wmf metafiles. (emf doesn't work on Win'95).
|
RELEASE b1181a
|
15/05/02 | New directive in #EDIT: GUISELECT, works like SELECT, but matching atoms are
'selected' in the GUI window.
Output R(sigma) during #SUM L 6.
|
16/05/02 | Check for updates added to help menu. Currently just loads the download page.
|
31/05/02 | List 31 is automatically created if it is required but doesn't exist.
Update SHELX output so that occupancies are fixed. (+10).
Export sub-menu added to X-ray data menu.
|
05/06/02 | Corrected slight error in calculation of S (GooF) during refinement.
|
07/06/02 | If no sigma cutoff, or I<-10u(I), display 'none' in refinement info tab.
R-factors for all and cutoff reflections recalculated during publication
table production. R and wR factors are quoted for _all, _ref, and _gt. The
_gt threshold is the same as the one in L28 if present, otherwise I>4u(I).
CIF output tidied a bit.
If non-zero, the Z value from list 30 is used to calculate Z', which is
then used to modify molecular formula in the CIF.
|
14/06/02 | Replace CONVPLAT.EXE with working version, you should now be able to run
Platon's squeeze/bypass successfully!
|
21/06/02 | Fix missing esd's on bonds across a symmetry operator between atoms on
special positions.
|
24/06/02 | New Auto-statistical wieghting scheme (#17) Automatically determins optimum
a and b parameters for scheme 16 and applies them. (Optimum: flat wrt Fc ranges and GooF=1)
|
25/06/02 | During Chebychev weighting (using GUI) output wDelSq and number of outliers to the GUI.
Updated all F' and F'' values to match Vol C. Int tables.
|
26/06/02 | Remove debugging output from bond calculation routines. Treat QH atoms as H when calculating
bonding.
Cameron: Fix mouse labelling. Positions are now correct, and labels easier to click onto.
|
28/06/02 | NEW field in atom list can now hold characters. (These will be used for an alternative
atom name at some point in the future.)
Representations of the special shapes SPHERE and RING now appear in the GUI window. Note
that special shape flag is lost if changes are applied after Cameron.
Cameron: try to fix postscript line thickness problems during consecutive output of ps files.
|
03/07/02 | Allow polygon selection of atoms in the model window.
|
08/07/02 | Removed text from toolbar buttons in order to save screen space. Popup 'tooltips' tell you
instead. Let me know what you think.
|
11/07/02 | Work out theta min and max during publication. Work out completeness at theta max, and also
best value for theta_full (currently highest theta with completeness > .995, or failing that,
the theta of the highest completeness possible)
Added new required absolute structure options to CIF. Set to 'unk' by default.
For calculation of F000, don't use anomolous bits if anomolous scattering
was not used in the refinement.
|
RELEASE b1185 - NB changed to MSVC7.0 compiler, so new DLL's etc. Any probs, let me know.
|
16/07/02 | Fix format overflow in CIF if cell was bigger than 10,000 Angstroms.
|
17/07/02 | In #DISTANCE added option: SEL RANGE=L41
This will use the bond list (41) to determine what is connected.
In #DISTANCE added option: OUT PUNCH=DELU D1DEV=x D2DEV=y
This will output VIBR restraints to the punch file with an esd
of x for 1,2 distances and y for 1,3 distances. You probably want an
EXCL H if using the DELU option.
In #DISTANCE added option: OUT PUNCH=SIMU S1DEV=x S2DEV=y
This will output U(IJ) restraints to the punch file with an esd
of x for normal bonds and y for terminal bonds. A bond is terminal
if it connects to an atom with no other connections except hydrogen.
This option uses LIST 41 to determine whether the atom is 'terminal',
but the RANGE can still be set at user's discretion. You probably want an
EXCL H if using the SIMU option.
In #EDIT added option: INSERT NCONN
This will put the number of connections to other atoms (according to L41)
into the SPARE slot in List 5.
In #EDIT added option: INSERT RELAX
This will start with the electron count of each atom in SPARE, and
keep cycling through the bond list, adding the current value of
SPARE to all the neighbouring atoms. It will cycle until the number
of unique values of SPARE stops increasing.
|
18/07/02 | Two new DIST directives - NOTBONDED and NOTPIVOT. This brings the total list to:
PIVOT, INCLUDE, BONDED, EXCLUDE, NOTPIVOT, NOTBONDED. Issue directive
OUTPUT LIST=DEBUG to see on screen what each of these directives thinks it is doing
in terms of including/excluding atoms from certain positions.
|
22/07/02 | Added new #EDIT directive: CLASH. MODE=REPORT, FIXLATTER or FIXFORMER. Either
just reports clashes, or fixes them (FIXLATTER changing serials lower down
the list, and FIXFORMER changing those higher up the list.) The new serial is
chosen to be one higher than ANY existing serial in L5.
|
23/07/02 | Minor bug in interprocess (GUI<->Crystals) queue locking behaviour. Should work better now.
|
29/07/02 | Fix in-memory list 29 when doing updates to GUI model from in #EDIT.
Added CIF to L30 items to store completeness. Theta full is set to an optimal value
automatically unless you set it negative, in which case its absolute value is used for
the CIF and the corresponding completeness is computed.
|
23/08/02 | Fix ALLDIST=YES option in #DIST. Now prints all distances.
Internal storage for the least-squares matrix has been increased eight-fold to 32Mb. This
is enough for approximately 400 anisotropic atoms full matrix.
Display script name in status bar when running.
|
29/08/02 | Overwrite list 41 when updating if possible (bonds unchanged). Overwrite L5 when trivial
changes are made e.g. refine scale factor, setting refinement flags etc.
|
30/08/02 | Added pressure to L30. Only appears in CIF if non-zero.
|
02/10/02 | Support for MAPVIEW (WinGX) fourier map output from #FOURIER (Slant not supported yet.)
MOL2 format output (#FOREIGN MOL2 / END)
GUI now includes a list of atoms, co-ordinates and uequiv.
New analyse menu between refine and results to better group some of the tools.
|
RELEASE b1187
|
07/10/02 | Minor CIF changes (quotes and spaces) for EnCIFer compatibility.
|
31/10/02 | Changed default I/u(I) threshold for #SFLS / CALC from 4.0 to 2.0 - 4.0 generates warnings
from cifcheck EVEN IF THE THRESHOLD WASN'T USED FOR REFINEMENT. The CRYSTALS cif now
records as a comment whether the reported I/u(I) cutoff was used or is just for info.
The calculation of R_gt and R_all now respect all L28 cutoffs EXCEPT
the I/u(I) minimum value. This is more like the IUCr definition (which
states that these values should respect the theta limits.) The IUCr don't
mention OMITted reflections, but it's unlikely anyone wants these in
the R_gt and R_all values.
Fix bug editing the z-coordinate using XPARAM (right-click an atom and choose "Edit").
|
04/11/02 | Output SPARE value in L5 consistent with input.
Fixed bug in some scripts dealing with charged atom type (e.g. SIP4).
In XPRUNE script set "Delete remaining peaks on exit" to OFF following constructive criticism.
|
06/11/02 | Regularise was failing to fill out data structure for new atoms correctly.
|
07/11/02 | Fix to completeness calculation so that it works with data that hasn't been SYST'd by
CRYSTALS.
|
16/11/02 | The GUIDE: Ensure some aniso refinement is done before further options are recommended.
|
24/11/02 | Three new buttons on toolbar accessing: EDLIST12, EDLIST16 and XREFINE.
Fix bug that was causing winsizes.ini to lose stored values (e.g. the PLATON directory etc.)
Side effect: TIP of the DAY now works again!
Preliminary support for using the USERPROFILE directory to store local per-user copies
of winsizes.ini. If you're running under XP or 2000 and don't have administrator privs then
the program won't rememeber fonts, window sizes, locations of programs etc. Email me for
details. (david.watkin@chem.ox.ac.uk) - should make it into a release soon.
|
29/11/02 | #SFLS / REFINE PUNCH=MATLAB / END output the design matrix (multipled by the Cholesky
factor of the weights) and a vector of weighted delta F's, suitable for reading into
the MATLAB package. Restraints are tagged onto the end of the design matrix and wDF vector.
NB The design matrix can be very big.
Some attempts at reducing output to essential info, especially during SFLS. Complaints/ideas
welcome - the job isn't finished yet.
|
03/12/02 | Set defuault font size in Cameron to 20.
Add CUT to end of PACK WINDOW instruction in the right-hand toolpane.
When UNFILL is on, outline the bonds on the screen (this already happens in the postscript).
Added help menu to Cameron.
|
05/12/02 | Option in "Set map ranges" dialog to specify the orientation of the Fourier map, if desired.
Accuracy of values used in the dialog box has been increased.
Bug in Fourier script stored location of MAPVIEW in MCE slot in winsizes.ini - problem: MAPVIEW opens
instead of MCE. Now fixed, but if affected, open winsizes.ini and delete the line beginning "MCEDIR"
|
11/12/02 | Fo vs Fc graph now also displays omitted reflections in blue.
|
16/12/02 | New #PERTURB command. Allows per parameter shift multipliers, and a general
multiplier to apply to all shifts.
The applied shifts respect weights setup in List 22 so that atoms do not move
from special positions, and competing occupancies continue to add up to the same
number. By default it creates new list 5s, but can be forced not to by directing:
"WRITE OVER=YES". Shifts for XYZ are given in Angstrom, for OCC as a fraction,
for overall (scale) and temperature factors as a fraction of the existing value.
Fix problem where Cameron menus cease to work on the third and subsequent times that
you use Cameron during one session in CRYSTALS.
|
17/12/02 | For mol2 output, only output atoms with lowest part number in each group.
|
RELEASE b1188 |
17/12/02 |
For mol2 output, only output atoms with lowest part number in each group. |
14/01/03 |
This is a useful one and I'm quite pleased with it.
Added a new lexical token PART to the scanner. The following
examples show what you might do with it:
----
Get distances between all atoms in part 1:
#DIST
INCLUDE PART(1)
END
----
Competitively refine the occupancy of two parts, e.g. in disorder:
#LIST 12
FULL X'S U'S
EQUIV PART(1,OCC) PART(2,OCC)
WEIGHT -1 PART(1,OCC)
END
----
Quickly id whole fragments in regularise:
#REGUL
target c(1) to c(6)
ideal c(11) to c(16)
cameron
map part(2)
onto part(1)
END
----
Move a part of the structure:
#EDIT
ADD 0.2 PART(3,X)
END
----
Have different regimes for refining e.g. a host and a guest molecule:
#LIST 12
FULL PART(0,X'S,U'S) PART(1,X'S,U[ISO])
END
----
So it's quite handy. Part numbers are stored in the 15th slot of LIST 5, and
are defined as follows:
( 1000 * GROUPNUMBER ) + COMPONENTNUMBER.
e.g. 123456 is group 123 and component 456.
The same syntax is used to select a part in the lexical input:
e.g. PART(123456) selects atoms in group 123, component 456.
BUT, if the value of the group or component is 999, all atoms in that
group or component are selected.
e.g. PART(999003) selects all atoms with component 3 in their partnumber and
PART(4999) selects all atoms in group 4.
The reason for having groups and parts is to enable logical grouping of, for example
a disorder where all affected atoms are group 1, with each component being part 0,1,2 etc.
NB. #script xparts runs a useful tool for defining part numbers.
|
15/01/03 |
Output F^2 to fcf file if requested. |
RELEASE b1190 |
19/02/03 |
Added list 7 support to lots of instructions. Note (programmers)
that sfls, calc and refine have the parameter ityp06, which corresponds
to the types picked up for lists 6 and 7 from the command file.
|
20/02/03 |
Overhauled weights. Main additions are application
of Dunitz-Seiler or Tukey type Robust-Resistant weight modifiers to any
other weighting scheme. Use e.g. #LIST 4/SCHEME 9 ROBUST=YES/END for unit
weights with robust modification. Note that downweighting only occurs when
the weights are initially calculated. To recalculate (e.g. after some refinement)
use #WEIGHT/END. |
27/02/03 |
Two BRAND NEW lexical things:
You can now use, wherever appropriate, FRAG(1,X's) syntax to
select all atoms whose fragment number (slot 17 in L5) matches. This
works in EDIT, LIST 12, DIST, REGULARISE etc.
You can now use, wherever appropriate, TYPE(H,X's) syntax to
select all atoms whose type matches that given. This
works in EDIT, LIST 12, DIST, REGULARISE etc. Main use I can think
of is to quickly FIX all H positions (especially if using FRAG to specify
bits to refine) or to special case a few heavy atoms: TYPE(AU,U'S).
Still problems overriding implicit temperature factor directives: If you say
FULL U[ISO] TYPE(PB,U's) - it causes an error - it would be nice
if it just overrode the iso setting. Maybe one day. |
06/03/03 |
Changed gui a bit so that the ! prompt appears next to the input line, and not
in the text output - it changes to > when expecting directives and disappears
while CRYSTALS is processing things. |
20/03/03 |
Added more information tabs: Constraints, Restraints, Twin laws etc. Added useful
buttons for changing many of the things in the info tabs. |
20/02/03 |
Added HTML output option - nice as it contains all the fancy greek symbols and you
can copy and paste bits into word documents. Requires fairly modern browsers to view
all the symbols correctly.
The output of weighting schemes has been overhauled to be clearer. |
25/03/03 |
Include the deepest hole in the list of Fourier peaks that are found.
It is labelled QN. |
26/03/03 |
#MATCH/MAP FRAG(1)/ONTO FRAG(2)/END - works a bit like regularise, but
no need to get atoms in the correct order (indeed, you can just use the
new FRAG keyword to reference all the atoms, provided you have already set up
the fragments). It doesn't have much of regularise's functionality, currently
it will only output the fitting matrix and deviations and a cameron model
of the two fragments superimposed. Use Graphics->Special->Use exisiting input files to view it.
Changed some of the C++ so that all application settings are now stored in the
registry. Crystals should now work better for people without admin rights on their
machines, since previously they probably couldn't store setting such as window sizes,
whether to turn tip of day off, whether to offer help files at start up, where programs
such as PLATON and MCE were stored. These settings will now all be stored on a per-user
basis. You might even be able to install CRYSTALS without admin privileges, but I
haven't tried it yet.
New icons, splash screen in anticipation of 12.0 release.
Streamlined the guide especially for the case of SHELX format input.
Added Open Shelx option to the FILE menu. Horst pointed out that that is
probably the most obvious place to put it.
New dialog for adding H's. Should be easier to understand for first
time users - though the whole pink/white Fourier/geometric thing
still throws people most of the time.
Added MCE to the distribution. Thanks Michal! New MCE options in the
Slant Fourier dialog include options for special 3d hardware, and the
simultaneous output of a VOIDS map to complement the Fourier map, and
take advantage of MCE's new feature of displaying two maps at once.
Added buttons to edit atoms, constraints, restraints to the toolbar.
Enhanced these scripts so that they are more user-error friendly i.e. they
spot errors as the files are read in, and give the user an option to re-edit or
recover the previous file.
Made the Shelx input script find the space group, if it is provided on the
TITL card after the word 'in'. e.g. "TITL cyclo in P 21 21 21". It still
must contain the proper spacing, so the script presents it to the user
for checking before continuing.
|
RELEASE b1199 |
13/05/03 |
CRYSTALS is now much more likely to generate valid
mogul queries for extended structures and structures with atoms on special
positions. Analysis of the MOGUL output is also faster in these cases.
The MOGUL results dialog now remains open after you choose "Examine in Mogul" to
take a closer look at a result.
Update to SFLS control code: command input can now specify up to 10
least squares cycles in one instruction. More will generate a warning,
rather than crashing the program. This addition will not affect users of
the menus and scripts.
Final difference Fourier maps for twinned structures when using the GUIDE were
not optimal.
User may now abort reading in the systematic absence graph during initial
analyses if it is taking too long.
|
RELEASE b1201 |
10/07/03 |
SUMFIX constraint: Use "SUMFIX c(1,occ) c(2,occ) c(3,occ)" to constrain
the total of a list of parameters to be fixed. More subtly, use
"SUMFIX c(1,occ) h(11,occ) AND c(2,occ) h(21,occ) h(22,occ) AND
c(3,occ) h(31,occ)" to constrain the shift of each group to be identical,
and the sum of all groups to be constant.
SAME restaint: Use "SAME 0.04, 0.02 FOR C(1) C(2) C(3) AND C(2) C(1) C(3)
AND C(4) C(5) C(6)" to restrain all of the 1,2 and 1,3 distances in the
first group to be the same as those distances mapped onto the following
groups.
Better completeness graph, now shows all data down to theta = 0.
New twin law editing dialog, used by ROTAX, and for straight input
of twin laws. It automatically normalises element scales to add up
to unity.
Optional transform of data as it is input by script XINLIST6
Put number of restraints in the CIF.
Changed I/u(I) vs resolution graph to be more concise.
Allow punching of twinned data with more than 3 elements.
Filter rotax results against the current point group.
Put rotax graph, twintol parameter and new ed25 graph on same scale (A-1).
Use \PUNCH 6 F to output a standard HKL file with no headers.
GUI speed ups by removing redundant code.
Store new information in the REFINE slot in the atom list (riding
H, special positions, rigid bond restraint, dist/angle restraint,
occ constraint etc) to be used to set the _atom_site_refinement_flag
in CIF and reduce your checkcif problems.
Don't output esd on angles involving riding H's to the CIF - they are
deemed to be 'senseless'.
Improved extraction of space group symbols from the INS file TITL
card.
Fixed (we hope) mysterious occasional 'crash on exit'.
Speeded up dialog and window creation.
Fixed context menu use on a sorted parameter list (below the 3D model
view)
GUIDE script can now handle twins with more than two domains.
Fix problem which prevented impatient people from entering commands
while CRYSTALS was starting up.
When changing min. func. from F to F^2: offer reflection filter before
reweighting.
CIF distances and angles now use the bonds in list (41), rather than the
cov radii - this means that you can make and break bonds in the GUI and
they will appear (or not) in the CIF.
Cell su's are included by default in your CIF distance calculation (makes
no detectable difference for routine data as cell su's are so tiny).
Improved Mogul support (bond types - 'un' and 'pi'), better treatment
across sym ops (again) and improved Mogul result processing.
|
RELEASE b1209 |
17/07/03 |
Fixed bug: New twin law editing dialog failed if there were no twin
laws already entered.
Added an SGPLOT instruction, for plotting intensity distribution
of various classes of reflection. Not for general use yet, but the
interested may see spga.scp for an example.
Extended the \REGROUP code, so that if you pass a -ve value as a
MOVE parameter, it uses covalent radii to compute bonding rather than
using a fixed range. The absolute value of the MOVE parameter is used as
a tolerance for allowing distances longer than the sum of covalent radii.
The atom renumbering script now uses this method (with MOVE=-0.2), and
should avoid pulling some 'connected sets' to bits...
Avoid potential crash when displaying 7 membered aromatic rings. Still
not quite right.
\MATCH outputs a file regular.dat containing lists of the atoms in each
fragment in identical order.
Added tools for automatically adding thermal vibration and thermal
similarity restraints to List 16. Also moved this method into the
shift limiting restraints adding tool.
Fixed bug which caused infinite script loop when using 'manually add H'
script on atoms with high serial numbers.
Corrected the covalent radii of calcium. All radii may be updated
by users by editing script\propwin.dat.
Squeeze interface - recompute scale after reading A and B parts
back from Platon.
Speeded up reading of systematic absence file during 'initial analyses'
by moving the LOOP into a sub-script.
Fixed bug in 1209, writing a mol2 file could crash the program if performed
immediately after refinement (shadowing of non existent
List 12 was carried out in KDIST4).
|
RELEASE b1210 |
31/07/03 |
Fix bug in 12.10 when adding Hydrogens using the ADDH wizard (script GFINDH).
Fix bug in contour.exe (Ashley Ibbett's Fourier map viewer), that caused it
to grind and then crash on some systems.
|
RELEASE b1211 |
14/08/03 |
Improvements to Mogul interface.
|
18/08/03 |
Fix \PERHYDRO on C atoms lying on symmetry operators.
Fix overflow in output of Shelx HKL format files.
|
05/09/03 |
Theta absorption correction added.
|
10/09/03 |
Fix formatting of Sheldrick type weighting scheme in CIF.
|
16/09/03 |
Thin out labels on x axis of histogram plots if the font gets
too small. Much better.
|
22/09/03 |
Fix \PURGE code that was supposed to allow one list can be purged
at a time. E.g. useful for clearing out normal matrices after use.
|
24/09/03 |
Upgraded matrix operations in \EDIT and \MATCH and \REGULARISE so
that they also apply the transformations to the Uij tensors of each
atom. Previously these data were just translated to the new location.
|
22/10/03 |
Fixed minor CIF output bug. SUs greater than 1 were correct but not
formatted to Acta standards. Now angles and distances involving H but
without SUs are quoted with less precision at request of Acta editor.
|
23/10/03 |
Move all \DEFINE functionality into the main crystals executable. This
won't affect any users, but simplifies the build.
|
24/10/03 |
Fix intermitent crash in #DISTANCES because required data not loaded.
|
RELEASE b1214 |
31/10/03 |
Add option to output CIF without ESDs on parameters. Allows CIF to be
produced without access to reflection data, or inverse normal matrix.
Use #SET BONDTYPE OFF, to suppress calculation and display of bond types.
Output Hydrogen atom restraints from \DIST.
Don't append reflections to end of 'special' SHELX files (ie. with atoms).
Change old REFINE column in interface to RESIDUE.
Fix bug where halogens appear grey in CRYSTALS, but green in CAMERON. Now
they will be green everywhere. (Existing data will require re-entry of L29).
In the \perhydro + \fourier dialog, remove peaks that are coincident with
the existing atoms, but leave peaks that might be coincident with the
recently added H atoms.
When an atom is selected by clicking on it, the associated field in
the list of atoms scrolls into view, if not already in view.
|
03/11/03 |
Make \PARAM output a list of twin element scales (with su's), if present.
|
04/11/03 |
If F000 (non-dispersive) is integer, then output as integer in the CIF.
Fix CIF 'rule of 19' treatment of (1), now given as (10) with extra digit
of precision.
Move Sheldrick geometric definitions of T_min and T_max to more appropriate
place in CIF.
Lower threshold for su's being treated as zero.
|
05/11/03 |
Split _atom_site_refinement_flags into three separate data names,
as required by new CIF dictionary.
|
06/11/03 |
Fixed long-standing bug where crystal colour and shape strings (stored in
List 30) would have spaces removed if they fell at 4 byte boundaries. E.g.
'Pale pink' became 'Palepink'.
Added atom_site_disorder_assembly and _group data names to the CIF. These
values are taken from the PART flag in list 5. The rightmost three digits
together with the sign form the _group number, and the next three digits
form the _assembly number. E.g. If an atom is given a part number
of -123456, it is in assembly 123 and group -456.
Atoms in the same assembly, but different groups will not be bonded
together for display and CIF purposes. Atoms in groups with negative values
will not bond to symmetry generated atoms in the same group. I expect
to see the following constructions in list 12:
- SUMFIX PART(1001,OCC) AND PART(1002,OCC)
|
07/11/03 |
Added restrain C-H option to the routine refinement dialog
box. This uses the original option (fix,refine,ride) for all non C-H
hydrogens, but restrains the C-H ones. (Best bit: allows methyl groups
to rotate).
Assembly and group numbers are now displayed in the list of atoms in
the GUI, instead of the REFINE flag, which is now for internal use.
Added some ACTA checks to the Tabbed initial analyses. Now you can tell
if you won't be able to publish a structure even before it is solved.
|
10/11/03 |
Added a CIF2CRY dialog, and put it into the menus. This enables direct
import of structural data from CIF.
|
13/11/03 |
Tweak font creation in Cameron graphics - might fix JCD's font
problems. Now use Arial in on-screen and postscript display to make
the labels as similar as possible on the two media.
|
17/11/03 |
Manual updated to explain PART syntax. CRYSTALS now has a proper
citable reference:
Betteridge, P.W., Carruthers, J.R., Cooper, R.I.,
Prout, K., Watkin, D.J. (2003). J. Appl. Cryst. 36, 1487.
|
RELEASE b1215 |
01/12/03 |
Fix truncated cell parameters in .fcf file.
CIF2CRY now works in directories with spaces.
Increased detail level of aromatic rings when rotating the model. Those
pentagons were freaking people out.
Ensure L41 is used for bonding when creating automatic RIDE instructions
(affects disordered fragments).
Atoms no longer automatically 'collected' by symmetry on input from
solution program. New 'Collect' button in re-typing dialog.
Setup refinement dialog no works if there is no existing constraint
list (list 12).
Use 'propwin.dat' not deprecated 'properti.dat' in output from CIF2CRY.
Restrain C-H option in 'Setup Refinement' dialog wasn't remembering its
last state properly.
Output of MapView format data from generalised Fourier section routines
(\SLANT).
Slightly changed output format of file, e.g. 'Map.fou', for MCE from normal
\Fourier routines. Definition is now identical to \Slant output.
|
RELEASE b1216 |
02/12/03-04/02/04 |
Fix bug in L41 which missed symmetry related bonds in one
direction where the only operation was a cell translation.
E.g. Ga(1)-O(1,1,1,1,1,1) would show up, but O(1)-Ga(1,1,-1,-1,-1) would
not.
Distance routines: Loop backwards through L41 (which is constructed upside-down in
the first place) so that SEL RAN=L41 doesn't have (much) effect on the
order of distances output compared to SEL RAN=LIM.
Change VIB to REST for hydrogen U[iso] restraints output from #DIST.
Enables a scale factor to be used.
Correct CIF data names for disordered parts. (_refinement_ shouldn't have
been in the data name).
Fix colours of data series in plots (was wrong if >6 data series).
Moved popup atom labels in model window to prevent cluttering view of
atom of interest.
Made mouse pointer more accurate in selecting objects in model window (it
was 3 pixels out in the vertical).
Changed Chem Cryst address in the default CIF template, refcif.dat
Reduce tolerance for 'snapping' atoms to special positions when reading in
a Shelx .ins or .res file to .01 angstroms.
Added a CONTINUE button next to the CAMERON command input box, this
allows long command strings to be input over several lines, like the old days.
Fix internal storage of PLATON executable location if you've installed it in
a directory with a space in.
"Restrain C-H" box in the refinement setup dialog now remembers its unchecked state
between runs.
Added some index entries for the CAMERON manual. Might be helpful.
|
RELEASE b1217 |
06/02/04 |
Change Mogul syntax slightly ready for Mogul Alpha 2 release.
|
RELEASE b1218 |
07/02/04-18/02/04 |
Add option to tabbed initial analysis to use List 28 reflection filters
when computing Wilson plot and E statistics. The Wilson plot should use
all data, but the filter allows investigation of problems with
very poor data.
Experimental: \SCRIPT xphase plots a distribution of the phase angles. Not
very useful in centrosymmetric space groups.
Include a list of missing reflections in the tabbed initial analysis, next
to the completeness graph.
Bug fix: Versions 12.17 onwards had trouble producing Fourier map files
for MCE, contour and MapView - the file names were wrong.
Catch and fix some old atom lists which may have invalid data in the
slot now used for ASSEMBLY and GROUP numbers. (Symptom: **** in CIF output).
Computed TMIN and TMAX values (used when no other values are available)
were being set to zero in the CIF. Now fixed.
Reduce number of decimal places in bond lengths that popup over
the model window. These distances are not always recomputed at every step
of the refinement, so may be slightly out-of-date. (Output from \DIST and
into CIF is unaffected).
|
RELEASE b1219 |
18/02/04-10/03/04 |
Added a 'ROTATE' directive to #EDIT. Rotates specified atoms a certain
number of degrees around a bond or vector while keeping a point or atom
fixed. The rotation is in an orthogonal system so the geometry of the
rotated group is not changed. E.g. Rotate the hydrogens of a methyl group by thirty degrees:
#EDIT / ROTATE 30 C(1) C(2) H(20) H(21) H(22) / END
Latest version of Cif2Cry released - now handles reflection CIFs and has better handling of
non standard atom names (apostrophes, stars, no numbers etc.)
\EDIT mathematical operations (ADD, SUBT, DIV, MULT) now work on the integer
fields of the atom list (PART,REFINE,RESIDUE).
Fix date style in the CIF (month and day were reversed).
Running SHELXS from the menu now works again.
Choosing 'Input/Edit formula/radiation' from the X-ray Data menu (SCRIPT XINLIST3)
now allows input of alternative wavelengths into LIST 13. It also runs a new dialog
(SCRIPT EDLIST3) automatically for non Cu or Mo wavelengths and allows you to type
in the F' and F'' for each element. NB. Re-inputting data from almost anywhere (CIF,
SHELX, KCCDIN) will overwrite these changes. Ror now, check the values are still correct
in the final CIF - this is being worked on.
Cameron: Fix size of medium resolution meta-file output.
Allow the 'File'->'Save As' option to work in directories with spaces.
CIF output from menus now recomputes bonding at the last minute, just in case
you've just changed the covalent radii in LIST 29 for some reason.
|
RELEASE b1220 |
11/03/04-02/06/04 |
Fourier maps: when using data with unmerged
Friedel pairs in a non-centro space group,
the pairs should be merged prior to
computing a Fourier map. This is now handled
automatically by all SCRIPTS (e.g. if you
choose items from the Fourier menu.) For
manual merging, do this:
#COPY 6 7
#SYST 7 FRIEDEL=YES FILTER=YES
#SORT 7
#MERGE 7
#SFLS 7
CALC
#FOURIER 7
....
END
As you can see, it uses something called
LIST 7 to store and work with the merged
data set. The scripts automatically delete
and purge list 7 afterwards.
Bugfix: If GUI updates are turned off, make
sure #EDIT creates a List 40 automatically
if it needs one.
Fix use of UNTIL in ROTATE directive.
Report error for zero length vector in
ROTATE directive.
Wilson scale and B factors are stored in
L30, and output in the CIF for info.
The publish script uses a new switch to
inhibit the addition of su's to parameters
of riding hydrogen atoms.
Added a normal probability plot of the
residual.
More robust H13 directive in hydrogen
placing.
Major CIF reorganisation and tidy.
Fix bug when purging large dsc files.
Maximum disk address is now 2^31.
Fix some obscure L16 error handling.
Don't crash if 'ALL' is too near the end of
the command line during lexical
scanning.
Fix bug in \CHECK for AVERAGE restraint on
LAYER SCALE factors. Very obscure.
Allow more that 9999 directives per
instruction. There is no longer any upper
limit.
Fix a resource leak on Windows version.
This affected some machines, when
CRYSTALS was used for long periods.
Added another option to the filter
reflections dialog. For twinned data set
maximum ELEMENT value to, say, 1.1 to only
include data from the primary domain, that
isn't overlapped.
Added buttons and scripts to start mercury
and encifer, if installed.
New CIF producing script gives more options,
and ability to open CIF in more
programs.
New refcif.dat to go with new CIF generation
routine.
Fixed some typos in reftab.dat.
Fixed bug in retype.scp (change atom types),
where nothing worked after clicking the
Collect button.
In general: all path names can now be given
with forward slashes instead of backward
slashes. (If you like).
|
RELEASE
b1224 |
02/06/04-08/07/04 |
The latest version: Replaced old NAG and
Harwell libraries with BLAS and LAPACK.
Automatically determine the CIF value
for Hydrogen atom treatment.
Storage of absorbtion correction details brought
into line with CIF.
Normal probability plot: axes swapped to match literature, fix bug
where k&l indices on labels were swapped, added ability to omit reflections by
right-clicking.
Change calculation of theta_full so that
it is always at least 25 degrees if theta_max is > 25.
Cameron: When packing cell,
operators are written to the listing file so that you can match the pack number
(e.g. 23 in label C1_23) to the symmetry applied.
Reviewed and tidied a lot of the
interface code, tests show ~20% speed increase.
Fix new Fourier scripts so that
merging is not done unneccesarily, and twin flag is set again if required after
returning to original data.
Cameron: fixed outline of dotted bonds in postscript (was too thick).
|
RELEASE
b1239 |
09/07/04-12/10/04 |
SFLS: code reorgansised and optimised. Test show approx 25% speed increase
for structure factor calculation, but matrix accumulation is unchanged.
Output enantiomer sensitive reflections to the listing file.
Compressed output from both scripts and refinement routines.
Removed List 22 dependency from SFLS (in English: you can now do
#SFLS/CALC or #SFLS/SCALE without having any valid refinement directives.
(Likely) H-bonds are now appended to the CIF in an appropriate format. Go
through the list and remove the ones you don't want.
CIF output overhauled in response to Editor's comments
Output of the SHELX-like weighting scheme to the CIF has been changed to
suit the Editors.
Cameron: bonds are now outlined in black! Atoms too, unless they are very small.
(Makes on screen rendering more like the postscript).
Fixed intermittent 'freeze and crash' when choosing 'Find H-bonds' in Cameron.
Use new CheckCIF web address.
Fixed and improved KCCD input script. Now waits and block, and also re-enables
import button after KCCDIN window closes.
Some scripts attempt to keep working directory cleaner by removing their scratch
files after use.
V1239 didn't work with Mogul. Fixed.
Option on Fourier dialog to choose weighting scheme.
|
RELEASE
b1250 |
RELEASE
1282 |
12/10/04-01/06/05 |
HYDROGENS: After hydrogen atoms have been found by the using the GUIDE,
they can be refined by them selves to optimise their posititions using
restraints on geometrty and ISO adps.. After
this, they are refined as RIDING.
NOTE: If you reselect the hydrogen option in the guide and don't delete
the existing hydrogens, they will be re-refined with restraints. This
might be a useful penultimate tweak before the final refinement.
CIF: There are more samples if 'useful texts' in the refinemnt section.
Don't use them all!
The hydrogen bond loop is only added if there are H-bonds. Note that the
D-A bond has no esd. For the moment, patch in the value from checkcif.
The formula for Sheldrick weigting has been streamlined.
SFLS: If the scale factor is EXACLTY unity, several scripts assume it has not
been calculated, and so carry out a SCALE. At any time, if the scale
factor from least squares deviates substantially from the ratio
Sum(Fo)/Sum(Fc), a red warning is out put. The Wilson plot scale is
also displayed (if a plot has been calculated). the LSQ scale can go
very wrong if there are a few really bad reflections, or a very poor
model, inhibiting convergence. Atoms that go NPD are listed in red.
SFLS: SCALE, REFINE and CALC can now all be given in a single SFLS
block.
BONDTABLE: A script assists in the creation of a list
of bonds to be added or removed from those created automatically
STRUCTURE: Items have been added to the STRUCTURE menu to assist in
allocating RESIDUE, ASSEMBLY and PART values to atoms. Useful for
Z'.GT. 1 structures or disorder.
|
RELEASE
1282-b |
12/10/04-02/07/05 |
In release 1282, the hyrogen Uiso were restrained to a value 1.2 times
the documented value. This was because a constant was applied twice by
mistake.
|
RELEASE
1283 |
02/07/05-01/02/06 |
Absorption
Fix error in calculation of theta in THETABS, the theta dependant
absorption correction
Weights
Stabilise Sheldrick-type weighting if parameters start to go out of
range. Include observed sigma to protect against zero division if the users
sets QUASI to 0.0
SFLS
Improve conditioning of Normal Matrix by using a softer restraint for floating
origins
Add code to permit use of data from multi-batch samples
fix a weird (possibly compiler) error in XCOMPL. The dynamic scratch area causes the program
to die if there are more than about 90,000 reflections
Enable REF and CALC to be given in the same SFLS instruction. CALC must be last
command
REGULARISE & MATCH
New code to control near-singularities for planar/linear/circular
groups
CIF
Skip LIST 6 items for cif if there is no LIST 6
Lots of small changes, including more comments to help authors
|
RELEASE
1284 |
02/02/06-01/06/06 |
adp Principal Axes
Compute some axes statistics in AXES
Hydrogen Atoms
Tune Sp3 hydrogen detection.
Output a "delete file" for hydrogens created by perhydro.
This can ue USED to delete the hydrogen atoms created by
the last PERHYDRO command. Hydrogen found by other means
are not affected.
WEIGHTS
Another patch to Sheldrick Weighting. e-mail me
for more details
HTML output
Compress the HTML Publish file for binding into reports
TORSION
Fixe format overflow in cif output of torsion angles
Tabbed Analysis
Enable filtering in Tabbed Analyse
|