Changes:
1) from version 1.0 to 1.1:
  - bin2raw and raw2bin files: endian variable was added
  - raw2bin: additional error checking
  - runmean, runmax, runmin, runquantile, runmad: additional error checking and
    corrected error checking
  - EndRule: added error checking
  - runmean, runmax, runmin: additional parameter "alg" was added
  - runmean - new C code can survive round-off errors
  - several generic functions were moved to a new "caTools" package
  - msc.split.data renamed to split.data and moved to "caTools". Shell function
    "msc.split.data" still provided which works with the warning.
  - read.mzXML and write.mzXML functions were added

2) version 1.2
  - following functions were added:
       msc.peaks.write.mzXML
       msc.peaks.read.mzXML
       msc.rawMS.write.mzXML
       msc.rawMS.read.mzXML
   - function 'msc.msfiles.read.csv' was renamed to 'msc.rawMS.read.csv'
   - in functions 'msc.peaks.write.csv' and 'msc.biomarkers.write.csv'
     order of arguments was swappped from '(fname, X)' to '(X, fname)' in order  
     to be consistent with other functions in the package
   - functions 'msc.peaks.write.csv' and 'msc.peaks.read.csv' got better error 
     checking
     
 3) version 1.3
  - better integration of mzXML data in the flow of the functions
  
 4) version 1.4
  - INTERFACE CHANGE: msc.*.read.mzXML functions switched from outputting list 
    with 'scans' and 'mzXML' fields to outputting scans with mzXML attached as
    attribute. That was done to make it consistent with other functions in the 
    package. 
  - Improved automatic self-checking through use of 'stopifnot' function 
    throughout the examples.
  - msc.*.read.csv functions got a new argument "mzXML.record=FALSE" that allows
    for creation and storing of meta-data
  - msc.*.read.mzXML functions got a new argument "wipe=TRUE" that controls if 
    samples returned from the function should be removed from mzXML data.
  - better integration of mzXML data in the flow of the functions
  - added more documentation to caMassClass/inst/doc directory
  
 5) version 1.5
  - Bug in 'msc.peaks.find' function found by Phingzhao was corrected
  - Modification were made to 'write.mzXML' function in order to make resulting
    files compatible with 'readmzXML' software 
    (http://tools.proteomecenter.org/readmzXML.php) which were particularly 
    sensitive to XML formatting used. Problem and solution was provided by 
    Daniel Lorenzini.
  - Daniel Lorenzini proposed an approach to store additional scan attributes 
    from mzXML files. I liked the idea but did not liked the approach, so as a 
    result each mzXML.scan gained 'scanAttr' field storing all scan attributes
    not captured otherwise in a string form. At the same time 
    mzXML.scan[].header field was retired.
  - Bug was fixed 'read.mzXML' function that was causing examples to crash. 
    The bug was probably triggered by changes to XML package.
  - Examples of 'msc.rawMS.read.mzXML' function were modified in order pass 
    'R CMD check' with R-2.3.0 alpha. Seems like 'all.equal' function have changed.
  - changes to 'mcs.biomarker.fill'  and 'msc.project.read' function to handle 
    bad input data and make functions more robust.
    
 6 ) version 1.6 (Feb 1 2007)
 - fixed a line (#239) in mzXML.R which was causing wrong xml terminator to be 
   used in the xlm header.
====================================================\==========================                 
 Jarek Tuszynski, PhD.                           o / \ 
 Science Applications International Corporation  <\__,|  
 (703) 676-4192                                   ">   \
 Jaroslaw.W.Tuszynski@saic.com                     `    \

