the way i figure is do it with the knowledge you have and then after you get something working go back and look at your code...once you get something working youll feel like you have accomplished something

then when you go back and look at your code you could learn new ways to optimize your code and/or do it the proper way...by doing stuff this way i have kept myself motivated enough to keep going with my projects and also have learned the correct and more optimized way of writing things...so now when i write to get something going i remember oh i found a better way to write this last time and then i start doing it the correct way the first time

either way i wish you the best of luck on everything you do :)

I generally get bored with examples and small projects for learning, so I will start a large project like a game and implement concepts I learned from my previous projects. I set a goal (finishing the game) and work to that end. It keeps me going and I get satisfaction when I finish a section of it. Most of the time I run into some self-inflicted flaws that might make me quit or start over-- no big deal. I'd say I've started 10 games and only finished 2, but I'm not working for anyone but myself, so whenever I want to start/stop it's my decision.

The point of jumping in with both feet is that you will prod yourself to finish something.

Doing small projects that have only one feature won't show you how it will look/interact with other features. This is also a good reason to come up with a large project and push till your fingers bleed.

The fastest code is the code never written.