This will teach you a lot of things... I can try to come up with a short list of what I've observed of my own development as a programmer.
You will recurrently come up with particular patterns, you start to identify which approaches will work, which will be dead ends and so on.
Try to separate your understanding and model of a problem to at least data, model and view domains. You will have certain data - you should keep this as simple as can be for the problem you are solving. There will be the model - how is the data mutated, which rules will guide it - and there will be a one or more view to this model and data. Note, that as a pattern the view encompasses both the possibility of an API to your system, as well as the actual display of data.
Try to keep you code as simple as can be. Don't go for the more complex solution unless you know you really need it.
'Pragmatic programmer' and 'code complete 2' are pretty good books about software development in general.
I would suggest you figure out what interests you and then try to implement simple programs in that ___domain. If something feels too complex, then first try a simplified version... more simplified etc. until you understand how to solve it.
One of the best knacks to learn is how to approach something from a direction which makes the problem easier to solve. There are usually an infinite amount of ways to write something and the challenge is not to just start writing code but actually figuring out which solution method leads to the most understandable and easy solution.
I would suggest you familiarize yourself with the practically important datastructures: the list, the binary tree, the map and the graph, etc. "Algorithm design manual" is a pretty good book for this. Aho's foundations of computer science is a pretty solid and free reference: http://infolab.stanford.edu/~ullman/focs.html
It's probably most motivating to mix the computer science with practical exercises unless you find you get a great kick out of it.
This will teach you a lot of things... I can try to come up with a short list of what I've observed of my own development as a programmer.
You will recurrently come up with particular patterns, you start to identify which approaches will work, which will be dead ends and so on.
Try to separate your understanding and model of a problem to at least data, model and view domains. You will have certain data - you should keep this as simple as can be for the problem you are solving. There will be the model - how is the data mutated, which rules will guide it - and there will be a one or more view to this model and data. Note, that as a pattern the view encompasses both the possibility of an API to your system, as well as the actual display of data.
Try to keep you code as simple as can be. Don't go for the more complex solution unless you know you really need it.
'Pragmatic programmer' and 'code complete 2' are pretty good books about software development in general.
I would suggest you figure out what interests you and then try to implement simple programs in that ___domain. If something feels too complex, then first try a simplified version... more simplified etc. until you understand how to solve it.
One of the best knacks to learn is how to approach something from a direction which makes the problem easier to solve. There are usually an infinite amount of ways to write something and the challenge is not to just start writing code but actually figuring out which solution method leads to the most understandable and easy solution.
I would suggest you familiarize yourself with the practically important datastructures: the list, the binary tree, the map and the graph, etc. "Algorithm design manual" is a pretty good book for this. Aho's foundations of computer science is a pretty solid and free reference: http://infolab.stanford.edu/~ullman/focs.html
It's probably most motivating to mix the computer science with practical exercises unless you find you get a great kick out of it.
Good luck!