I have a problem with a customized Windows based application which is developed in .NET.
It is a conversion application used to transliterate text in romanized format into native languages viz. Chinese, Urdu and the like; using the font Arial Unicode MS which follows the Unicode Standard. The conversion used to take place correctly until recently when we changed our operating system from Windows Server 2003 to Windows Server 2008 r2.
Interestingly, a bizzare event is happening. Suppose we want to translitereate Romanized script into two languages A and B. Using the application If we transliterate language A first then it gets converted correctly. After that language B converts incorrectly. Following it in second attempt after re-running the application if we transliterate language B first, it gets converted correctly following it with Language A, conversion gets inccorect.