Kevin Higgins and Stephen Roantree
What went right:
It got done.
The music adds a lot and was made easily enough.
The game engine worked reliably, including the guard's field-of-view effect.
Macromedia Flash was a good tool for character and explosion sprites.
The A* pathfinding worked.
The game ended up with a lot of variety and media packed in.
The game was nicely polished.
An enjoyable, puzzle-type gameplay emerged with the first short levels.
The level design tool was functional.
Extra text (song quotes, level names and dialogue) added a lot of polish and enjoyment.
The fixed-width font system worked well and was flexible.
What went wrong:
It had no design.
The gameplay was unforeseen.
The game has little replay value.
It took too long to make - the lack of goals and a schedule slowed down development.
Features were hacked in:
Prisoners at cell doors
Guards standing still
Playing music between levels
The level design tool was slow.
The game was easier and shorter than we would have liked.
The AI was too limited.
The game was released without being tested fully, so two obvious bugs (not enough time to read quotes at the start of levels, no music for the end screen) had to be fixed after release.
The levels weren't designed sensibly all at once, so their size and gameplay and layout changed substantially through the game.
The level design tool required .NET, which meant dialogue writing had to be done in a hurry on the programmer's PC.
This game shows that we're able to make a proper-sized 2D game. There was no problem making the graphics, text and music, and the coding was fast and effective. We didn't hit any obstacles due to the size of the project. However, Jailbreak made obvious what we already knew: design is ESSENTIAL. Having no design document was a stupid mistake, and led to very frustrating situations where features had to be hacked in that could have been elegantly integrated if only the coder had known about them. It also meant that the gameplay was unpredictable - we didn't know what the game would be like. This project was completed in a struggle against the lack of design. If we had made a design doc it would have been a longer, deeper, more feature-rich game. The artist could have been working on characters, levels, dialogue, paintings and sounds long before the game was running. The coder could have focussed on the features needed and completed them faster and tidier. Having no design is looking for trouble. On the plus side, it's encouraging that we're able to bring together a game on instinct, just by knowing what works - but it should never happen again.