Thread: how to develop so many different solutions?

Use those exercises for practice, not for rote learning. With practice, you will remember some things, and much of the rest can just be looked up when necessary. It is more important to understand the concepts so you can apply them in other related situations; there are only so many existing solutions that you can regurgitate.

Originally Posted by c_lady

To memorize all of them doesnt work... they will be gone after one week

Me personally, I like to write short (as short as possible) demo programs for new concepts (K&R appears to encourage this), and keep them all loosely organized in my "C code" directory tree. I know how I think (sort of ) so I put in little helpful comments, for myself. Also, I find code I wrote easier to read than code anyone else wrote.

If you do something a bunch of times, you will start to remember it. Short of that, I wouldn't expect to remember how to write a binary search because I did it once in an exercise. But if I kept my little demo with notes, then it probably won't take me long. The short demo is usually much easier to follow when you need your memory refreshed than going, "Oh, I remember I did that once in that other program which is 1000 lines...". IMO, trying to learn a new technique by coding it straight into your current project is a very very bad idea. Abstract and simplify instead, it saves time and frustration.

As they say, there's no substitute for experience I guess.

Yeah, that is a clever one.

grep is a very great tool in this sense too, if you are looking for a particular command or lib function. Which of course you could pipe head into grep there.