Am Mittwoch, 30. Januar 2013 schrieb Ludovic Courtès:

> The attached patch removes that UTF-8 assumption. Can you test whether

> it fixes the problem?

Yes, please commit!

Andreas