Coding practices

[clayemore]

Chime in...I am interested on your thoughts...

Which coding method do you prefer? Put every little logical routine into a sub function (example a), or write code into the main (example b)

Comments

[jlipeles]

As I code more and more, especially in Java, I do A. In fact, now I always do A. Had to get away from the PHP (scripting) mentality. Hahaha... Greg's gonna come after me for that one. Muhahahaha!

[toddito]

Just to be a troublemaker, I'm gonna go with B.

[arichi]

A. It's not only easier to debug and read, it's also easier to write - if you need a logical subroutine, you just assume it's there and go back to fill it in. If you forget to do so, the compiler will remind you of your obligation.

[clayemore]

I actually write the code as B...then I take what I just wrote, cut and paste and create the subroutine. Usually, every once ina while I know to make the routine before writing what it actually does.

[arichi]

This works too. At least you go through the trouble of making it a different routine. On behalf of anyone following you on the project, thank you.

I guess a PSA to anyone reading this - if you find yourself copying and pasting code, possibly changing one or two things, this screams "make it a function"

[maniakes]

I rarely put any serious code directly in main unless the entire program is only a few dozen lines.

In real-world production code, the most common model I've seen is to use main only to instantiate a few objects. The constructors of the object perform initialization tasks, and wire the objects up to receive input.

If you're asking a more general question about how fine to chop up functions, I'll echo arichi's comment about the copy-and-paste rule of thumb.