July 12, 2017
v2.0.1
In functions postM and postM.ab called with prior = 'uniform', the prior was
incorrectly imputed to be proportional to 0:mmax in v2.0. The prior is now
correctly imputed to be proportional to 1 in v2.0.1. The error was relevant only
when postM or postM.ab was (1) called from the command line in R, and (2) the
default option for the prior ('IbinRef') was manually overridden to 'uniform' by
the user.

July 17, 2017
v2.0.2
The routine for simulating Weibull, lognormal, and log-logistic persistence
distributions has been updated to improve estimation of the variance of the beta
parameter. The change will have negligible effect on estimates of g, M, or
lambda. However, the updated version is more stable when estimating detection
probabilities when persistence times are very long.

An error in the calculation of lognormal persistence distributions has been
corrected. The change will have negligible effect except when estimating g when
persistence times are very long relative to search intervals and a lognormal
persistence model is used, in which case earlier versions of the software would
sometimes crash, but the updated version will not.

In previous versions, the "span" label for custom search schedule in the Single
Class module would sometimes improperly be in "normal" state (black color) when
the "formula" option was selected. In the updated version, the label is
consistently put in "disabled" state (grayed out) when the "formula" option is
selected. The change does not affect any calculations or results.


July 18, 2017
v2.0.3
The routine for simulating Weibull, lognormal, and log-logistic persistence
distributions has been refined to more reliably model carcass persistence when
all (or nearly all) field trial carcasses persisted beyond the duration of
the field trials. The change will have no effect on results except in certain
unusual scenarios when previous versions of the software would occasionally
crash when persistence times extended well beyond the end of the field trials.

In the form for importing and analyzing carcass persistence data (submodule of
Single Class module), a meaningless error message was erroneously generated upon
clicking the table with summary statistics of persistence models. In v2.0.3,
that error message no longer appears.

December 4, 2017
v2.0.4
The label for M* on the figures showing simulation results in the scenario 
explorer has been modified to show the alpha value for which M* was calculated. 
Also, the array of raw simulation results from the scenario explorer can now
be accessed from the command line via getData('res'). 

February 5, 2018
v2.0.5
"Error in quantile(rr, probs = c(0.025, 0.975)): object 'rr' not found." This 
data retrieval error in the analysis of carcass persistence data in the single
class module caused the program to crash in some scenarios but has now been
corrected. 

February 7, 2018
v2.0.6
A more reliable algorithm for displaying EoA version number in window titles
has been adopted.

In previous versions, an error in the processing of carcass persistence data 
caused slight inaccuracy in fitted persistence functions when a large fraction
of carcasses are scavenged before the first carcass check. The error has been
corrected. It is unlikely that the change will cause any noticeable differences
between fitted persistence distributions parameters, estimated detection 
probabilities, or mortality estimates between v2.0.6 and earlier versions. 

February 7, 2019
v2.0.7
In previous versions, the button to "Estmate overall detection probability (g)"
in the multiple class module would properly create a graphical summary of 
results but fail to write a tabular summary, returning an error "object 'g' 
not found" instead. That error has been corrected.