File: CHANGES.old.4

package info (click to toggle)
horae 071~svn536-1
  • links: PTS
  • area: contrib
  • in suites: wheezy
  • size: 12,996 kB
  • sloc: perl: 67,215; lisp: 744; sh: 78; makefile: 76; ansic: 35
file content (259 lines) | stat: -rw-r--r-- 13,330 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259

   +--------------------------------------------------------------+
   |                    Change log for Artemis                    |
   |   http://leonardo.phys.washington.edu/~ravel/software/exafs  |
   +--------------------------------------------------------------+
   |                         Bruce Ravel                          |
   |                  ravel@phys.washington.edu                   |
   +--------------------------------------------------------------+


Begin change log for version numbers starting at 0.6.  See CHANGES.2
for the 0.5 releases.  See CHANGES.1 for earlier versions whose
version numbers were the same as release dates.

December 2 2003 -- December ?? 2003 0.6.011  (never released, go to 0.7.000)
  * make all toplevels in the same class (and common with athena) so
    that they group in the taskbar
  * correct a situation where spurious path parameter math expressions
    may be left in Ifeffit's memory, leading to incorrect evaluation
    of a path (one way this might have happened was to define a
    pathparam math expression, evaluate the path, delete the math
    expression entirely, then reevaluate the path.  in that case, the
    previous math expression would have been left in Ifeffit's memory,
    leading to an incorrect evaluation of that path)
  * added several Echo messages specific to certain actions
  * cloning a path now halves the degeneracy of the current and cloned
    paths
  * using the right click menus on a path parameter math expression to
    define a new variable on the GDS page now inserts "0" as the math
    expression if it is a newly created variable
  * the "Add a new variable" button now focusses on the name of the
    new variable
  * change the page layouts on the GDS, feff, and path views to use a
    frame and the packer.  this gives more consistent layout on
    different screens/font sizes
  * added a Messages tab so that errors and warnings issued by Artemis
    have a place to go without overwirting the contents of the Files
    tab.
  * several improvements to the Edit math expression dialog
  * write initial guesses to the log file
  * made the window a bit narrower and made size of views a global
    parameter so it's easier for me to change
  * found several places where the flag for inquiring about saving the
    project was not being set when it should have been
  * when importing feff paths, the choice for importing the first N
    paths is now configurable
  * added Scott's ZnO example to the distro.  removed the Pu and
    NiC4N4 examples
  * fixed a logic problem in one of the comments in the fitting script
    generation which misidentified a normal fit as a difference fit

November 28 2003 -- December 3 2003 0.6.010
  * fixed the color of the scrollbar of the prefs description
  * added line styles for data, fit, and parts
  * disable log viewer when prefs dialog is shown

October 30 2003 -- November 25 2003 0.6.009
  * bound &quit_artemis as the delete protocol for the main window
  * prompt for verification when clearing variables or converting all
    guesses to sets
  * clear the Files tab before every fit (so that warnings from
    previous fits aren't still reported there)
  * disable plot and include data buttons if discarding data brings
    the project down to a single data set, also toggle including of
    first data set if it is excluded from fit
  * several appearence problems uncovered in Tk 804.025
  * fixed a problem contructing command string for an fft using a feff
    path for the phase correction
  * fixed a problem constructing suggested file names for res and diff
    output files
  * ask ifeffit about its limits on number of paths and such

September 4 2003 -- October 27  2003 -- version 0.6.008
  * fixed a problem reporting fit results when a path parameter
    evaluates to NaN, resulting in a string of *'s in ifeffit path
    report.  artemis now handles this gracefully and issues an error
    message useful for diagnosing the problem.
  * when saving results to a log file, default file name is based on
    project file name, if it exists
  * made the intrp box in the APA a bit smaller
  * added an echo bar message to the "Reset all variables" option in
    the GDS menu
  * added an echo area message for mouse-over in intrp box
  * Clicking the "include path" button now issues an echo area
    message.
  * Can now specify a path to be ifeffit's current path so that after
    the fit the def parameters are evaluated in terms of that path.
    This info is saved to the project file.
  * Moved the "Show" submenu from Edit to it's own menu.
  * Plot running r-factors in k, R, or q.  Put a cascade in the Data
    menu.
  * Log viewing utility, with reports and plots + averages and
    Einstein fits.  In Edit menu.
  * added an xbm icon so the unix windowmanager toolbars/panels/icon
    boxes/docks/whatever have something pretty to display
  * post explanation of problem when a fit is aborted due to
    parenthesis mismatch
  * can now convert all guesses to sets in GDS menu
  * disable right click menus for the column heads on the GDS view
  * modified many keyboard shortcuts to align more closely with common
    UI choices, e.g. C-o for open data, C-p for print, and so on.
    zoom is now C-= and unzoom is now C--.  The key chains for
    plotting in k, R, or q are now C-; k, C-; r, and C-; q
    (i.e. control-semicolon) 
  * grab previous and grab next in GDS right click menu
  * projectbar underneath menu bar for displaying path+name of current
    project + configuration option for displaying file, project title,
    or nothing
  * closing project now asks if you want to save before closing.
  * right click to define a new param with Stay option, vhanges view
    on GDS grid without showing GDS grid.  also do an update before
    moving when clicking the "Add a parameter" button.
  * split path() commands into separate lines for each path
    parameter.  this helps to avoid very long lines caused by long
    file names and/or long math expressions
  * include error checking for
     -- number of paths used
     -- number of variables used
  * deal with upper/lower case feffnnnn.dat file names.  this is an
    issue when transferring a project between windows and unix
  * fixed bug where a null paths entry would be created if the pcpath
    parameter for a data group points at a nonexistant path
  * fixed a bug where a fit would abort if a path for a data set NOT
    included in the fit does not exist
  * added the ability to delete data sets (but not the first one --
    use change data, instead)


June 20 2003 -- August 20 2003 -- version 0.6.007
  * save and restore plot features in project file
  * handle Mac terminated files transparently
  * handle long filenames transparently
  * added a right click menu to the numbers on the GDS view with the
    following two features + a way to clear that variable
  * added a way to move GDS params up and down the list more flexibly
    than the little up and down buttons
  * added a way to ask Artemis where a parameter is used
  * can now choose which data set to plot once fit is finished
  * fixed a problem managing guess variables that lead to inconsistent
    fit results.  the problem was an erase() line that was too long
    for ifeffit.  the solution was to break it up into many shorter
    erase()'s
  * difference fitting is implemented.  the diff spectrum can be made
    from an arbitrary set of paths.
  * preferences implemented
  * can now restore default values for operational params
  * added a feature for automated generation of parameters.  it's not
    super-brilliant, but it should give a decent starting point,
    particularly for first shell fits.
  * fixed a bug counting data sets in an MDS fit and some data sets
    are excluded from the fit.
  * figured out how to explicitly set backpage and inactive colors in
    the NoteBook widget
  * added a dialog at read data time asking if this is to be a
    multiple data set fit or a change of data sets.  renew data in the
    Data menu circumvents this
  * fixed a bug where title lines get lost when a data set was changed
  * when changing paths to feff, update the headers, reff, etc for
    each path + fix a bug when on feff view when changing paths

June 18 2003 -- June 19 2003 -- version 0.6.006
  * Fixed a bug handling the is_rec flag
  * perl/Tk 804.024 is pickier that earlier versions about minus signs
    in front of method arguments, such as width and height, so put in
    the dash wherever needed

May 24 2003 -- June 16 2003 -- version 0.6.005
  * Fixed bugs in save_all_paths and save_selected
  * better column labels (short_descriptor method) and title lines
    for save_selected. also use plotting k-weight for r and q
  * fixed major problems in saving column data for individual groups.
    also, data in r or q is now saved using the current k-weight for
    plotting
  * also use plotting k-weight for save_all_paths
  * fixed a bug identifying the parent in the situation where the
    include path menu is used while the feff calcualtion is current
  * added dphase to the path parameters list
  * added $config{meta}{extpp} and the "extended path parameters"
    (dpahse and (k|amp|phase)_array).  by default, the extended pps
    are hidden.  when the config variable is true, they are unhidden
    and added to the end of the list.  a checkbutton is in the paths
    menu for unhiding them.  when hidden, they are not used in the fit
  * Now give the user the option of carrying on even if certain errors
    are found in the math expressions.  This is a modal dialog box:
    Continue or Abort.
  * fixed problem in Project Assistant where the same feff calc could
    not be imported for more than one data set (which is useful for a
    multiple-temperature fit)
  * removed Control-c keyseq, replaced it with Control-period
  * allow restraints to be defined on the GDS page and use them as
    such in the fit.  (all restraints are applied to the final
    data_set)  Added an example of using a restraint to the tarball.
  * put epsilon_k on the data view.  put an item in the data menu for
    reporting epsilon_k
  * allow saving selected as k-weighted chi(k)
  * erase delta_ scalars before the fit
  * added some more options to the right click menu for words in the
    path parameter math expressions.  added skip + the option not to
    jump to the GDS view
  * verify that an data file for import is actually a data file
  * fixed a bug where athena chi(k) records were only read correctly
    as the first data set
  * try to make sure that the plot options have sensible values
  * constrain the fitting space variable to be the same for every data
    set (like cormin)
  * added a right click menu for most op params
  * added a "Save as ..." to the file menu.  Made Control-s save
    without prompting if a default project filename already exists.
    Project saving now writes a .bak backup file.

May 2 2003 -- May 18 2003 -- version 0.6.004
  * use features of ifeffit 1.0077
  * refuse to run without Ifeffit 1.0077 or higher installed
  * new syntax for multiple k weight fits
  * fix a minor bug in project saving (only save titles for data
    groups)
  * make use of the list of ill-posed variables reported by ifeffit
    when writing results of a fit
  * fixed a bug displaying keys containing commas (by enclosing the
    key string in quotes)
  * report reff+dr for each path in the results
  * fixed a bug where skip variables were not saved to the project
    file
  * made foreground text of anchored, unincluded paths the brownish
    color rather than the black
  * do a simple check of math expressions to verify that parens match
  * reporsition Data and Paths list when a new data file is imported

April 24 2003 -- May 2 2003 -- version 0.6.003
  * add a confirmation dialog to project deletion. also move the
    delete project choice farther away from the save project in the
    File menu
  * check that files are writable before writing
  * fixed a bug in path selection where the selection rule was not
    sticking to the current feff calc in situations where that makes
    sense.

April 22 2003 -- April 23 2003 -- version 0.6.002
  * move rc and mru files to the .horae directory
  * fixed a bug whereby the first path may be assigned index=0
  * fixed a bug involving the C-n (rename_this) keyboard shortcut when
    the shortcut is done on a feff calculation
  * fixed a bug regarding how titles get saved to project files
  * exclude a few data-group parameters from the project file, one of
    which was causing trouble in the Safe box evaluation of the
    parameter array when the project was read in

April 21 2003 -- version 0.6.001
  * Multiple data sets + many changes to accommodate this
  * multiple k-weights
  * Artemis Project Assistent
  * removed templates (see version 2002.08.19)
  * report Nidp in echo area whenever kmin,kmax,rmin,rmax are changed
  * fixed a bug in up/down buttons and skip variables
  * always compute residual in k-space
  * fixed a bug running ff2chi which resulted in bad low-r Fourier
    components
  * changed layout of widgets controlling the plotting options, also
    make widgets for plotting k-weight separate from fitting k-weight