By Daniel Smith and Paul Potgieter, with music by Tim Harbour
During LD29 we produced No Hawking, a game we’re both proud of. But it didn’t go exactly to plan, and we fell far short of our initial design. We now examine the process so that we can get closer to our mark in LD30.
The making of No Hawking went like this:
WRONG
Planning:
- We forgot to name our group.
- We aimed too high, and as a result did get core gameplay in place until the last day.
- First release build was attempted after deadline (it failed… continuously)
- Dan forgot to pack a toothbrush.
- We burnt out each day instead of sleeping at regular intervals. This disrupted the flow of development.
Technical:
- We used an interpreted language with horrible garbage collection and no native threading support.
- Custom Engines take a lot of time, which we didn’t have.
- We used a graphics framework that wasn’t actually finished :’|
- We used a release tool that didn’t “just work” and consequently spent the last frantic minutes just getting a distributable build done.
RIGHT
- We had an incredible brain storming session right after the theme was announced and came up with a design really quickly.
- We built a “working” custom engine during LD!
- Despite everything that went wrong, we managed to make a playable game with objectives, obstacles, and win/end states.
- We got enough sleep. (Although it was all over the place)
- We had pizza and fizzers… a lot of fizzers!
- We didn’t give up.
- We had fun!
NEXT TIME
- Use finished engine/framework etc…
- Release early, release often
- Sleep properly.
- Have more fun!
CONCLUSION
- Ludum dare is a time for experimenting with game design but not a time to create a whole engine from scratch
- Ludum Dare is fun!
There will definitely be a more complete Post-LD version.
