The first thing that I would do is ensure that all developers at the start of computing agreed to Unicode 2.0, or something else. I don't really care what standard they chose, as long as it works and that they all used it. As time goes past this should lead to me being able to tell my boss that foreign character support in our application is a given and therefore save us several days of testing and head scratching wondering where our Japanese characters have gone.
Character encoding and date formatting are the two things that I hate more than anything in programming. Why can't this be more simple? Why is it that I am restricted to using a specific font to support all the characters when it couldn't be too hard for OS vendors to enable the font subsystem to fall back to a generic Unicode 2.0 full supporting font for individual characters that the desired font doesn't support.