Hmm, see maybe: http://bugs.archlinux.org/task/18339 I do not know which ghostscript bug may be related to this one, but would be nice to fix this and print again...