[karterk]

You will have to prove that you can "ship it". I suggest that you start off by picking a domain that you're interested in and building a body of work that demonstrates your ability. Be sure you pick some thing that's neither too wide nor too narrow. For e.g. picking something as broad as "web development" is too difficult - these days that involves everything from knowing semantic HTML to CSS to JavaScript to CSS to Rails.

Picking a practical "project" will help you stay motivated in learning things and you will also have the benefit of "publishing" it as something you "shipped".

Once you start feeling a little confident about the code you write, start putting your projects up on github. Also, start contributing to other projects in github (e.g. libraries you have used in your own project). There are many projects where you can easily begin by fixing small bugs. Then you can slowly learn the code base enough to contribute features.

[Leepic]

Would things like an eShop, knock-offs like game reviewing websites where a user can register to recommend and post reviews for videogames and a tumblr clone would qualify as practical?

[karterk]

Sure, but wouldn't it be better if it's actually used? A tumblr clone is probably too much work for an eventual ghost town. You can pick more focussed projects that solve a small problem really well - in that way, you have a chance that's actually used. For e.g. a chrome extension for something that has always bothered you.