[steveklabnik]

Don't read; do.

Reading is great, but it's a means to an end. Pick a project, work on it. When you get stuck, read to figure out how to proceed.

But doing is worth a lot more than reading.

[wings]

This advice is spot on. I read a lot of books before I realized I wasn't really challenging myself.

Actually putting something into practice will require hours of frustration every day, but it will lead to actual learning.

Pick one project to work on, and no matter what: stick with it.

[sweis]

That depends on what you want to learn and what your goals are. Some of the most fascinating aspects of mathematics and computer science only exist in books and papers. That being said, I think it's perfectly valid to focus strictly on the practical side.