The Xerox Finite State Toolkit - comprising the c-fsm library and the lexc, twolc and xfst
languages built on top of it - was originally created by Xerox to write spelling checkers,
spelling correctors, tokenizers, morphological analyzers and similar applications for
a small set of commercially interesting European languages. However, to our delight,
and often to our complete surprise, it has also been used successfully to build linguistic
applications for a much larger and growing set of lesser studied, minority and even
endangered languages around the world, thus promoting, we hope, linguistic diversity.
The talk reviews the history of the Toolkit, its successes and failures, and directions for