Add an option to ignore the Unicode differences in Extended Properties. SQL Compare currently will differentiate between Extended Property text (e.g., for MS_Description) created as Unicode vs Ansi (e.g., N'My text' vs 'My Text'). SQL will apparently maintain this different even though scripts generated from SSMS always show it as Unicode (with N). Thus, this can cause SQL Compare to show differences between two databases where there is none intended (the description text is identical).