The good news is that many publishers are making headway in this direction. The bad news, however, is that in some ways localization is going to get more difficult. Why? Because in the ongoing drive for increased market share, developers are looking at ways of making software smarter. And one of the many ways of doing this is making programs more sensitive and sophisticated concerning language. While much of this core linguistic technology will be licensed from your trusted OEM suppliers, who are specialists in this field, there is still going to be tremendous challenges for developers to design word processors and other applications which can exploit these powerful but complex techniques.

Consider, for example, the challenge of building a search-and-replace function which can handle morphological reduction for all the languages you support. Or take grammar checkers. Having a parser for a given language is only part of the equation. You also need an elaborate structure around the parser for handling unrecognizable phenomena, for diagnosing problems, and for providing meaningful feedback to your user. As programs get more helpful, they also get more verbal. And most developers will want to have some control over the output. All this means there will be some tough localization nuts ahead to crack.