2021-10-07 Stephane Germain  <stephane@psychometricon.net>

	* Add the frequencies table to the classical test theory statistics output

	* Now using libirt 1.4.1

	* version 2.0.3

2021-09-08  Stephane Germain  <stephane@psychometricon.net>

	* Now the stdcall/mrtd verion of the 32 bits dll use explicits __stdcall prefix for each functions

	* version 2.0.2

2021-09-07  Stephane Germain  <stephane@psychometricon.net>

	* Now using libirt 1.4.0

	* version 2.0.1

2016-10-26  Stephane Germain  <stephane@psychometricon.net>

	* Uninstall the old eirt.xla

	* Update the selection on sheet change

	* version 2.0.0

2016-10-24  Stephane Germain  <stephane@psychometricon.net>

	* Convert xla/xls files to xlam/xlsx/xlsm

	* Replace MS RefEdit by a more robust and usable custom solution
	
	* Stop using named range for translation and load all translations into a collection

	* version 1.3.1

2014-04-06  Stephane Germain  <germste@gmail.com>

	* Now using libirt 1.3.0

	* Added support for mixed models, graded and nominal

	* version 1.3.0

2013-07-29  Stephane Germain  <germste@gmail.com>

	* Create the menu each time the xla is open

	* version 1.2.1

2013-02-22  Stephane Germain  <germste@gmail.com>

	* Now using libirt 1.2.0

	* Added the biserial/polyserial correlations in the "CTT" output

	* version 1.2.0

2012-06-13  Stephane Germain  <germste@gmail.com>

	*  Now using libirt 1.0.3

	* version 1.0.9

2011-11-07  Stephane Germain  <germste@gmail.com>

	* Add the possiblility in step 4 to have the abscisse of the
	graphics on the score scale instead of the normal scale.

	* Reflect the change in irt_wrapper.c from the rirt project.
	
	* version 1.0.8

2011-04-15  Stephane Germain  <germste@gmail.com>

	* Add a parameter to change the behavior on errors in the GSL
	library and set the default to continue on error.

	* version 1.0.6

	* Now using libirt 1.0.2

	* version 1.0.7

2011-04-06  Stephane Germain  <germste@gmail.com>

	* remove a name insertion of the model in the worksheet (was
          there for future use by other plugins) it was causing a bug
          in excel 2007

	* release 1.0.5
	
2010-12-20  Stephane Germain  <germste@gmail.com>

	* restablish correct error handling
	
	* add a 64 bit version of the dll eirt64.dll and add conditional declare statement
	(I can't test if this work since I do not have a 64 bit version of office, let me know)

	* release 1.0.3

	* add the ptrsafe keyword in the declare for 64bit

	* release 1.0.4
	
2010-12-16  Stephane Germain  <germste@gmail.com>

	* sort the options for the MC models
	
	* change the name of the latent variable from "ability" to "z" by default
	
	* now possible to change the latent variable name in the settings

	* release 1.0.2
	
2010-12-15  Stephane Germain  <germste@gmail.com>

	* correct a bug in the slopes display with the graded model and the normal ogive

	* change the names of the models (Dichotomous, Bock, Samejima)

	* release 1.0.1
	
2010-01-18  Stephane Germain  <germste@gmail.com>

	* correct a bug in "all data points" when using the normal ogive

	* print the assistant version in the summary

	* print the correlations in red when over 0.8 and under -0.1

	* show the good option in the parameters (in bold)
	and in the plots (bold and dash) when using MC

	* add the test caracteristic curve

	* release 1.0.0

2009-11-26  Stephane Germain  <germste@gmail.com>

	* correct a bug in the graphics points of the information functions

2009-11-26  Stephane Germain  <germste@gmail.com>

	* add the score of each subject as optional output

	* add new examples with theirs reports

	* use libirt version 0.9.9

	* new version 0.9.9

2009-11-24  Stephane Germain  <germste@gmail.com>

	* add Warm's weighted maximum likelihood estimator (WMLE) for the
	abilities, and the option to change the estimator int the algo tab
	of the settings

2009-11-23  Stephane Germain  <germste@gmail.com>

	* change the degrees of freedom used in the local independance test
	for the mc models
	
	* add the correlation between each item pairs as optional output

2009-06-14  Stephane Germain  <germste@gmail.com>

	* fix an error in the test of fit for the multiple-choice, the
	statistic was sometime negative when some options had a zero
	frequency in some class

	* use libirt version 0.9.8

	* new version 0.9.8

2009-04-08  Stephane Germain  <germste@gmail.com>

	* Compute the standard errors at each iterations of the mc and graded models

	* Trim the spaces in the responses cells before processing

	* Show each items information functions in its own chart

	* use libirt version 0.8.5

	* new version 0.9.7

2008-09-12  Stephane Germain  <germste@gmail.com>

	* The minimums are now auto-detected for the graded model
	hence if the minimum or maximum option is missing it's ignored

	* Add the option in the settings to have the results
	with the normal ogive

	* Add a way in the settings to change the slope of the 1PLM

2008-01-31  Stephane Germain  <germste@gmail.com>

	* check for the convergence of the likelihood in all the EM

	* change the default for max.em.iter from 20 to 100
	
	* change the default for precision from 0.001 to 1e-5
	
2007-12-12  Stephane Germain  <germste@gmail.com>

	* Correct typo on the Back button in french

	* Check the settings with locale dependant functions

	* Always read and write the settings with a point for decimal

	* In the selection step, change the text for the check boxes

2007-09-17  Stephane Germain  <germste@gmail.com>

	* add the test information function

2007-09-14  Stephane Germain  <germste@gmail.com>

	* permit to choose the one, two or three PLM when the data
	is multiple choice

	* open the online html help in the user define browser

2007-09-06  Stephane Germain  <germste@gmail.com>

	* release version 0.9.0
	this version uses libirt version 0.8.2

	* add a user guide in english

2007-09-05  Stephane Germain  <germste@gmail.com>

	* validate the change in the settings

	* correct a bug in the selection, the selection
	was wrong when the data wasn't at the top left of the sheet

2007-09-04  Stephane Germain  <germste@gmail.com>

	* change the behavior of installeirt.xls
	now it quit the application only if there is
	no other workbooks open else it just close the workbook

	* add a workbook to uninstall from Excel's menu

	* create subdirs in the project
	(bin, xls, doc, setup, examples, vba)

	* add a user guide in french

2007-08-30  Stephane Germain  <germste@gmail.com>

	* check if there is at least two options

	* add installation instruction and various notes in README.txt

2007-08-29  Stephane Germain  <germste@gmail.com>

	* let the nominal model be selected when the data is graded

	* select the faulty cell when there is an error in the data

2007-08-28  Stephane Germain  <germste@gmail.com>

	* complete rewrite of the add-in

	* now in a "wizard" (assistant) style with 4 steps,
	instructions and tooltips

	* generate a nicely formatted report

	* separate advanced settings in another form

2007-07-20  Stephane Germain  <germste@gmail.com>

	* compute the log likelihood ratio test for local dependance

	* compute the log likelihood ratio goodness of fit test

	* add the information functions

2007-07-05  Stephane Germain  <germste@gmail.com>

	* add the graded model

2007-06-28  Stephane Germain  <germste@gmail.com>

	* add the multivariate model

2007-05-18  Stephane Germain  <germste@gmail.com>

	* add the estimation of abilities

2006-04-28  Stephane Germain  <germste@gmail.com>

	* uninstall the previous add-in installation before installing

2006-04-24  Stephane Germain  <germste@gmail.com>

	* disable screen updating during processing

2006-04-13  Stephane Germain  <germste@gmail.com>

	* change the default bandwidth by a factor of 1.1 in accordance with
	testgraf

	* add an option to have the parameters of the normal ogive instead of the
	logistic (divide the slope by 1.702)

	* add internationalization support

	* translate into french

2006-04-07  Stephane Germain  <germste@gmail.com>

	* disable the automatic recalculation during processing

	* change the cursor to "hourglass" during processing

	* improve the error handling by adding the normalEnd and
	retryEnd exit points

2006-04-06  Stephane Germain  <germste@gmail.com>

	* initial release 0.1
