Rapid development

[louisg]

I was thinking the other day some more about rapid development, and was thinking how an oldschool approach which produces harder-to-maintain code is actually kind of a benefit when it comes to just getting a game together quickly.

For example, I made a simple Monaco GP clone once in a week. The code was bad, non-object oriented, and I don't think I'd reuse any of it. But, it was simple and to the user there was no difference. Then, some time later, I decided to produce a sequel. The code was very meticulous, clean, modular, and highly object-oriented, but after a month of spare time I was still not done. Part of this I can say was because I opted to use SDL instead of Allegro which needs a little bit more coding anyway, but I think much of the time was spent organizing the program well, and generalizing the problems into modular chunks for subclasses.

What do you guys think? Is a non-OOP approach generally quicker?

[Ex-Cyber]

Good programming (which is not synonymous with OOP) typically takes longer than sloppy programming for small projects. The benefits come with code reuse and lack of obnoxious interdependencies in large projects or across multiple projects. If you can basically keep the whole game in your head anyway, having it written down neatly is not nearly as important.

[null1024]

I'm doing the messy approach for a simple Java shmup I'm starting to work on, and I at least got some stuff to display, move [I didn't have the eEvogun sources [formatted the drive to put OSX 10.4 on], my previous attempt to do a shmup in Java, and in the same rapid, messy attempt].

The clean approach, seems like it'd be too much of a pain. I've always been a messy coder, but I'd document about everything that wasn't obvious with large amounts of comments and descriptive function names. Variable names, they ended up being sort of terse, but usually not unintelligable [like a,b,r,i,etc.].

[the2bears]

What do you guys think? Is a non-OOP approach generally quicker?

Generally not, though it can seem so at first. I've done 10 or so rapid prototypes in the last year, each one in "2 weeks", and having an OO framework to test and prototype off of has been of enormous benefit. But I had to spend some time getting that framework to a point where it helped.

Bill