The water proved to be a difficult task to simulate. When a comet would hit the planet, a crater would form, or a bulge would rise up, a certain amount of water would be displaced and these changes would be propagated across the entire globe. Richard agreed that the optimal way would be particle effects, but it would be possible to “fudge” it by simply using static slices that moved up and down. The calculations would be processor-intensive, so the whole thing would be done in the cloud and then sent back to the user. In an ultimate version of the game, the entire planet could be procedurally generated and simulated, but then it would stop being arcadey and become incredibly complicated.
Some more ideas about the game’s structure were thrown around during that time, as what it was originally designed to be was a full “life simulation” (except that you control the planet, not the life). The player goes through different “stages”, each with a different objective to complete.
- Create life
- Grow plant-life
- Grow animal life
- Grow humans
- Protect civilization
Of course, the player would have to contend with space-borne obstacles such as meteors and the like.
As for the player character, I just wanted it to be a tremendous, superintelligent robot. That would open up some great multiplayer opportunities when someone else tries to wreck your carefully-maintained planet. The voiceovers will be funny, in a deadpan way. And, of course, if central command gave you orders to do insane things for the purpose of observation, you wouldn’t really be able to disagree.