My naming scheme is rather basic - showname.rehearsal.2006.10.09.pdf is rehearsal report, 9th of October, 2006 for showname. My reason behind it is quite logical (at least to me) even though the filename looks ugly. Because of all the data included in the report's file name, I can use regular expressions etc to find almost anything, really quickly (ie the regex hamlet.*.2003.* will return every document from hamlet in 2003, or *.rehearsal.*.2001.7.* will give me every rehearsal report I wrote in July 2001). They also group really well when sorted by title. It is a quirk of mine - and probably not something most others would bother with.