﻿April 2006, ver. 4.8.3
-----------------------------------

x) An updated *.par file not read (?). Another problem in windows: *.par file behaves as opened permanently and cannot be modified
FIX in ver. 4.8.3: both OPEN and CLOSE operations are done by the calling DLL

x) FIT/INIT: if the number of parameters changes, the command line interpreter does not reflect it 
FIXED: ver. 4.8.3

x) FIT/PLOT: dispersion branch is plotted at EN, not at omexc(1). (it is correct for MFIT)
FIXED: ver. 4.8.3

x) when statring MFIT> submenu, the prompt shows FIT>, similarly to the FIT> submenu. That was misleading at some point.
FIXED: ver. 4.8.3

x) The "plot" made in the FIT> submenu:
shows khi2 (within the plot frame) whithout the last digit of the exponent:
0.aaaE+0N, the symbol N is not visible. 
FIXED: ver. 4.8.3

x) after >MPHON the MC cloud of the Res. Function is positioned (on the plot)
not at the (qh qk ql) point which is given last, but at the point taken from the data file. The energy en is plotted correctly (that is the cloud is centred at the energy value given by en, bot not one taken form the file).
On the plot it is seen as a shift of the cloud with respect to the central point of the plot. This effect does hot happen after >PHON
(probably visualization problem)
Similarly, when working with the exci module, after >FIT and FIT>plot
the ellipse from TRAX is centered on the selected point on the dispersion curve. But after >MFIT and FIT>plot the same effect is seen:
the dispersion curve is plotted around the point given by the command >qh ...
but the MC cloud is plotted at another point, as if it is taken from the data file (but correct energy en). If your point (qh qk ql) is far form the point taken from the file, you may not see the MC cloud at all. 
FIXED: ver. 4.8.3

*) funkce OMEXC v submenu fit ma jako argumenty q(1..3), bylo by potreba
pridat i q(4), protoze v nekterych pripadech je ho potreba (kdyz ma ta cara
vlastni sirku)


*) na elvire, kdyz se prekompiluje procedura exci, pokud se nezmeni nazev
knihovny, dynamicky link se neupdatuje - pri opetovnem spusteni tam zustane
stara verze - jsem  presvedcen, ze na HP to nebylo - tam ten system zrejme
hlidal verzi nebo datum/cas - nevim, jestli se s tim da neco udelat?
??

x) Partial fitted curves for individual branches in the data output have wrong scale
FIXED, ver. 4.8.3

x) Both FIT and MFIT use ray-tracing for resolution function 
FIXED, ver. 4.8.3


December 2004, version 4.8.0
----------------------------------------------------------
1) Runtime error in EXCI compiled with g77: cannot read text file.
(FIXED)
Since this version, EXCI is loaded dynamically at runtime. 
Module for loading EXCI is written in C. 
The library can be compiled with any fortran, e.g. g77, but it must be linked 
with some additional files: reclat.f, exci_io.f
Use configure.EXCI script to generate makefile, e.g.
% configure.EXCI g77
% make -f exci/makefile
 
2) Error when loading multiple files using range of names:
> file 31440,31443
should read files 031440 to 031443, but it didn´t. 
There was a problem in the subroutine SpecFileName.
(FIXED)



September 2004, version 4.7.7
----------------------------------------------------------
In TRAX only: Guide section, if used, gave always m=1 reflectivity in TRAX
(FIXED)

In TRAX only: Outer dimensions of SOLLER collimators were ignored (i.e.
for ALPHA<500), giving slightly (few %) larger resolution 
(FIXED)

MFIT: The fitted curve is not correctly plotted after return (QUIT),
unless it was plotted already during the MFIT session.
(FIXED)

September 2003, version 4.7.6
----------------------------
1) Error in taking number as string argument with the command FILE, e.g.
>FILE 28230 
didn't work. 
(FIXED) 

2) Incorrect conversion of a number to the data filename, e.g. from
28230 to 028230 (FIXED) 

September 2002, version 4.7.3
----------------------------
1) In command RESOL 4:
   uninitialized matrix values lead to incorrect result (or even crash)
   (FIXED)   
   
2) problems with correct x-axis scale on the scan plot, if scan centre
doesn't agree with the QH,QK,QL,En values (FIXED)


April 2002, version 4.7.2
--------------------------

1) Sample mosaicity has not been taken into account correctly. (FIXED)

2) PROF 2 and FIT 2 commands: conversion of (Q,E) events to the
reciprocal lattice coordinates was wrong. Transformatin matrix wasn't
updated when needed, resulting in nosiy scattering curvea. (FIXED) 

3) Corrected script Build.EXC, generating the makefile for EXCI library.
It didn't generate correct dependencies and system-dependent linker
options. (FIXED)   

   
