- Create a file without BOM
- Compare the file with some file
- Change compare mode to UTF8<>UTF8
- Edit the file without BOM
- Save
- Check for BOM
E.g. if you have a PHP library and some program will add BOM without your knowledge, then you will get "Headers already sent" error or something like that and you will spend hours searching for this (if you don't know about this BOM)... I got the same with Notepad some time ago.
BTW. Programmers Notepad 2 uses some algorithm to check if a file is in UTF-8 or not (if BOM is not present). It works more or less fine, but you can always change the encoding, plus you can use some tricks for it to work (add some non-ANSI characters near the beginning of the file).
One more thing - it's get even stranger if you compare two files which are almost exact copies: At first you will start with ANSI<>ANSI; change one so the files become the same; compare again and you get ANSI<>UTF8 and the files seems different at first (but you just synch. them, right?).