2014-03-21

    o	Improve code for reading all bytes from a connection.

2012-08-31
    o	Use resolved native symbols.

2010-02-17

    o	Added CT n-gram counting.
    o	We now use vectors instead of pairlists to avoid memory
	corruption problems.

2010-01-09

    o	Use perl in word splitting in textcnt. This fixes a problem
	with combining characters when using the default split
	pattern.

2009-10-21

    o	Provide a useBytes argument in textcnt (R >= 2.10.x).
    o	Fixed blank strings in word-sequence counting.
    o	Added reduced n-gram counting.

2009-09-04  0.0-3

    o	Make Unicode latin ligature translation work "portably".
    o	Migrated encoding utilities.

2009-08-31  0.0-2

    o	We now call l10n_info from C to avoid problems on
	non-Linux platforms with langinfo.h

2009-08-24  0.0-2

    o	Migrated core utilities to CRAN package tau.
    o	Fixed to SET_STRING_ELT in C code.

2009-01-15  0.1-2

    o	Added counting of word-sequences.

2008-04-30  0.1-1   Initial Release

    o	Provide efficient C-level counting of n-grams, affixes,
	or words.

    o	Provide utilities for tokenizing, stopword removal, and
	other preprocessing tasks.

