Changes in 1.0.18 (penultimate version prior to methylumi merge):

 o  Fixed a bug in normexp() (aka mcsv.normexp()) which was causing problems.

Changes in 1.0.17:

 o  Better handling of HumanMethylation450k control probes, along with some 
    suggested improvements from Lavinia Gordon at MCRI Australia (thank you!)
    and implementation of empirical CDF based detection p-values for both chips

Changes in 1.0.14:

 o  Default qcplot method will jitter the negative and normalization controls,
    to give some idea of the degree to which the latter separate from the former
    and thus some indication of how successful the assay was for each sample.

Changes in 1.0.13:

 o  Incorporate bugfix from Andrew Jaffe so that full path names could be used
    for the 'vignette' (450k data import testing and casting to MethyLumiM).
    Thanks Andrew!