Nothing beats experience, you can read all the theory, practice all the exercises until your eyes burn out and your fingers bleed. But experience is key. I'm no expert in the field, heck I'm still new to it, but my personal advice would be to create games in your free time, practice them, study how they work, the mechanics, the structure, and various different implementations of old games, such as Space Invaders, Pong, Breakout. This will help you to understand more about the basic architecture, like states.

But I still suggest you actually get a job, to help build that experience in the field. A few years down the line, you never know, you'll be ready for it.

Game programming from my experience is extremely rewarding and extremely taxing. I know, I'm taking a degree in it. GCC is an excellent book, with a lot of good tips, your fine to keep reading it, because in my opinion it's less about the implementation of the code, and more about the advice, structure and architecture. I can apply what I know from GCC into any other language, like Java, if I wanted too. (Well not ALL of it, but most of the concepts).

Start with small games, building up your knowledge base, get experience, and continue your studies. Nothing else to say on the subject really, your doing fine.

My advice is pretty much the same. I'm the lead AI programmer at Maxis working on The Sims, which means a big part of my job is interviewing candidates and making the recommendation as to whether we should hire that candidate or not. If a candidate comes in without any professional game experience, they had better have a portfolio of personal projects. I've skipped over many candidates simply because they had no experience and had never made a game. The reality is that professional game development is a VERY competitive field. You have to show me something that quickly sets you apart from the dozens of other candidates I'm looking at.

And when I say "make a game", what I really mean is "finish a game". Don't start it, don't show me tech demos with some cool algorithm, show me a complete, working game. It should have levels, user interface, art assets (they don't have to be good, just something to show me that you're loading assets from disk), sound effects, music, input, an options menu, credits, and all the little things that make up a complete, full game. This shows me that you can drive to the finish.

Thedevsykes is spot on: start small and build your way up. My classic first assignment to budding game developers is Tetris. Sit down and make Tetris. It's a trivial game but it requires all the components that go into a full game. Graphics, sound, music, a main loop, etc. There are also one or two tricky parts, so you have to find an elegant solution to a few of the gameplay problems. Breakout is a good second game because it has all of those things but adds collision detection, levels, and power-ups into the mix. The third game should be a simple platformer (like Mario) or over-head action/adventure game (like Zelda). Keep building from there.

If you work at it and dedicate yourself, you'll get there.

-Rez