Both sides have good arguments. I am undecisive. However, the discussion is only about the default. Whatever we choose, developers using Lucene will stll have all options.
I briefly thought about integrating David Spencer's ngram-spellchecker into Lucene. It could make FuzyQuery fast AND complete. However, it would make indexing slower since Lucene would have to update the additional index of terms and term ngrams. Furthermore, I don't know a good solution for deleting terms from this additional index if documents get deleted.
Christoph
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
