[ Register or Signin to view external links. ]

Think of a basic idea.
Write it down.
Think more complexly how you want the program to work.
Write down key methods you think you would need.
Write down key classes.
Design a class diagram.
Formally specify your methods

Start putting your written idea into code following your written structure.

Thats the basic start for it. Follow that and you should be okay.

Just write pseudo code prior to actually writing the code for the software. As a result this enables you to break the code down, instead of diving into the code from direct thought. Which in my opinion is less-efficient.

You need to be certain that you understand the logic involved with the code you are writing and/or copying. If you find that you do not understand the basic concepts, such as while loops etc... Don't be afraid to use Google as your resource.

Think of it this way, if you looked at a trigonometry question and you didn't have a single clue on how you could figure out the hypotenuse/adjacent/opposite line, and you dived straight into someone's answer - you certainly will not know how they managed to get to that final answer. It's the exact same with code, if you do not understand the concept, you will not understand how the programmer achieved his/her final goal.

Examples of pseudo code (if you don't know).

- Set rectangle width to 6.
- Set rectangle height to 4.

- if the rectangles width is greater than 30
print "You cannot exceed the width limit

More examples can be found here: [ Register or Signin to view external links. ]

Anyway, good luck with learning.

Generally what I do is,
- think of a general idea.
(My Mass Effect Guide)
-Think about what I could do, to make this practical
(Live updates from the internet)
-Sit down and start creating basic features
(In this case, basic main screen, pulling information from the internet into proper areas, etc.)
-Thinking about anything else that could be added or tweaked to make the performance or intended goal better
(Added an offline export feature which cut down on loading by 30-40 seconds, Added a notes area, etc)
-Finished up anything else

Now literally at this point I sit back, use my program and think about what could be added or what would be cool.

I find this process works well for me, if I take too long, I get off track and forget too much, so this allows me to develop as I go.