check semantically wrong, but a trick that prevented an IP lookup in case that the filter was not used did not work. That bugfix causes that crawling gets a huge speed boost for noload urls!