2004-05-26 Update this NEWS files again. We already gone through 2 
	   sub-version during this period of time. From 3.2 to 3.3, 
	   a live mode recognzier is included, from 3.3 to 3.4, the
	   bugs in live mode recognizer is large exterminated and 
	   a lot of facilities such as fast GMM computation and phoneme
	   look-ahead are included. 
	
2002-11-30 Release of Sphinx-3.

2001-12-07 Dropped "naming" conditions from LICENSE.

           Use SphinxTrain to train acoustic models:
              http://www.speech.cs.cmu.edu/SphinxTrain

2000-01-27 CMU Sphinx is going Open Source!