About the author of this article
My name is David Mekersa and I manage the French studio Casual Box.
I started creating my first games on Amstrad CPC, then on the Amiga…
I am a veteran in the world of new technologies with over 20 years of experience. At times developer, creator, designer, technical director, and head of international projects, I have explored all the intricacies of IT. In 2006, I created the site www.casualgames.fr and in 2009 I launched Casual Box, the French video game studio.
How to create a game in 2 months with Corona SDK: my method
1 - Familiarize yourself
Read some "How to start", open some samples, play with the code, and try to understand it. Don't try to understand every line!
2 days is my average learning time for a new programming language. At this point, I know only basics... Then I start coding!
2 - Learn when you need
You want to work fast. So, you'll have to accept to start programming with just some basic understanding of the language...
My method: I start coding, then I learn only when I need something I don't know.
Example: For Chicken Deep, I needed to use the accelerometer. I coded everything until I needed to use the accelerometer. At this point, I had no idea how to use it. I looked at the Corona SDK documentation, and I played with some samples, and eureka!
I use this method for everything in my games: code, learn, code, learn... It is practice makes perfect.
3 - Decide
What is your game? Try to sketch it on paper, dividing it in chunks.
I like to use wireframe templates, like:
Start simple: a menu scene and a gameplay scene (use the storyboard functions of Corona).
4 - Think simple
You'll add new scene when you'll need. Don't try to plan everything, because you don't really know how your game will look like before the 1st prototype (and not even sure)!
I hate detailled Game Design Document... Planning is guessing. I prefer action.
5 - Limit yourself
Decide how many weeks you'll work on the game.
The schedule will decide of the features, not the opposite!
Check your progress every week and take decisions.
6 - Create with nothing
You don't need art, sound and music right now. Don't be shy to create with nothing. Use sounds from other samples or from the web.
Same for art, just use placeholders. My method is to search images with google, adding "png" to the search terms.
7 - Prototype
A game is a gameplay. Prototype the gameplay in a minimum of time (few hours). It's like a homemade game jam!
For Chicken Deep, I prototyped the gameplay in 2 hours. If the prototype sucks, try to enhance it. If you cannot, go to sleep and try again the day after.
8 - Create by successive refinement
A game is like a painting. You have to have something complete at every stage of its creation.
Create everything, even if it's brief, then refine each part until it's good enough.
9 - Integrate
When you have something playable (and demonstrable to your wife...), it's time to integrate real art, music and sounds.
But don't be naive, your game will change, until the last version before the Appstore submission. Your artist need to be aware of that...
10 - Play
Play your game at every stage. A lot.
11 - Plan
Divide your process in 4 steps:
- Alpha (playable, with lot of bugs and missing features)
- Beta (few bugs, feature complete)
- Final Beta (only your wife can ask for some changes...)
- Release Candidate (this is the girl...)
12 - Kill you babies
Less is more. Remove features if you feel it will impact your schedule too much, or if its make the game too complicated.
I know, it's hard to remove lovely lines of codes... but you have to, man!
13 - Stop
You have to stop coding at some point and decide it's the Release Candidate (fear!!).
You did it!
You can try the game I created here: http://bit.ly/PSEcjz, it's called Chicken Deep.
And for more about my studio: http://bit.ly/RDLitJ ...
Corona SDK official website: http://www.coronalabs.com
Talk to you soon! Email me: david at casualbox.fr.