Starting from v4.6.1.119, SpamFilter supports the use of Unicode in the various blacklists/whitelists, which is very useful as it allows to specify keywords in foreign languages for searches in the subjects and body. As a consequence, SpamFilter's logfile and the various blacklist/whitelist files are now encoded in 16bit Unicode (specifically in UTF16 Little Endian) instead of UTF8.

On the day that SpamFilter is upgraded from an older non-Unicode version to the latest unicode one, depending on the types of emails that are received, the new SpamFilter could open the older existing non-unicode logfile and start appending unicode text to it, and then saving it in the existing non-unicode format. This causes unicode characters to be saved as periods, questions marks or squares, but happens only on that one day when the upgrade was being made. From the following day on, all future logfiles will be saved in native unicode format and there would not be any display issues.