0.4.0a (27.03.2005)
---------
    * First public version

0.5.0a (11.05.2005)
---------
    * pyExcelerator now can import Excel 5/95/97/2000/XP/2003 files.
      With imported information you can do what you want.
      See ./examples/xls2txt.py for reference.
    * fixed bug in MS OLE2 compound document dumper (not in writer :))
      Bug causes dumper to combine links in sectors' allocation chain sectors
      contained in additional MSAT with SAT data.
    * Excel files dumper supports now Excel 5/95/97/2000/XP/2003 files
      and produces more informative output.
      
0.5.1a (14.05.2005)
---------
    * improved floating point decoding code

0.5.2a (18.05.2005)
---------
    * improved perfomance (especially on nonfragmented streams)
    * now pyExcelerator can process XLS files with short streams
    * OLE2 compound document dumper supports short streams
    * reading UTF-16-LE and Latin-1 with more accuracy
    * in DSF (double stream file) XLS data extracted from stream "Workbook"
    * improved floating point/integer values decoding code

0.5.3a (19.05.2005)
---------
    * typos in biff-dumper.py fixed :(
    * now pyExcelerator can correctly import XLS files with charts
    * added ability to override default file's encoding
    * new examples (or tools :)) -- ./examples/xls2csv.py, ./examples/xls2html.py

0.6.0a (19.07.2005)
---------
    * blanks, numbers, currency (use number formats!) and dates (use number formats!)
    * pyExcelerator uses 1.5-2 times less memory (thanks to __brains__ and __slots__)
    * fixes for Big Endian CPUs
    * some refactorings
    * new examples: blanks.py, merged1.py, dates.py, numbers.py, num_formats.py, formulas.py
    * most important change: formulas support

0.6.1a (29.09.2005)
---------
    * fixed: exception when reading OLE2 files with incorrect MSAT
      (sector ids points to nonexistense sectors)

0.6.2a (30.09.2005)
---------
    * fixed: exception when reading OLE2 files with incorrect SSAT
      (sector ids points to nonexistense sectors)

0.6.3a (25.10.2005)
---------
    * slightly new algorithm for reading OLE2 files. I hope it is
      more robust
    * splitting and frozing
    * worksheet protection
    * protection password
    * workbook protection
    * new example: protection.py, hyperlinks.py, panes.py
    * extracting formula results
    * speed optmizations(for example, benchmark big-35Mb.py runs about 
      20-30% faster)
    * updated THANKS file
    * xls2csv, xls2txt, xls2html now resides in ./tools