I made the same next-year prediction yesterday (and safe to use in 2006), but everyone seemed to think I was crazy.
Perl 6 has already been worked on for years. If a computer language takes more than three years to create, it's being developed incorrectly, except in some cases if it's a special purpose language, which Perl is not supposed to be. The longer it takes, the the more likely it is that the niche will be filled by another product by the time it comes out. That's disastrous for development of a commercial product, and a huge waste of time for a free product created by volunteers.
Ok, I just made up the 3 year time frame, but it feels right.