New research from the University of Waterloo is making inroads on one of the biggest problems in theoretical computer science ...

And why letting people come up with their own ideas, and actually try those ideas, is almost always the best approach.