A fix to cleanly support non-ASCII characters in file names (i. e. on systems using unicode character sets) has been committed to SVN recently, so if you have a need for this, we recommend to upgrade to the latest code. However, it’s a good idea for everyone, because using such characters in file names causes SVS not to scan those files…

A new (stable) version will thus likely be released soon, but there’s another open issue we have to track down and fix before we release the new version (a possible rare race condition / deadlock which is hard to reproduce) — stay tuned.

Update (December 10, 2010): we’ve just committed a few additional patches which should avoid the deadlocks. However, we’ll do a number of intensive tests before we officially release it.