Posted: Tue Sep 04, 2007 12:32 am
Eat_The_Path wrote:This FAQ is quite handy in a number of respects, but there are a few answers that are really just wrong, or don't make any sense. So here seems the best place to complain.At this point you may be thinking that's a lot of data and that wouldn't it be a good idea to take snapshots of the game state every so often. Well, the answer is no because of the way the simulation runs, it has to know everything. Ever heard of the butterfly effect? The idea that a butterfly fluttering once can cause a typhoon somewhere in the world. Tiny events building to a big one. Well, a game of Defcon, when run in the simulation, is much like that, a fighter who, say, survives a 1 on 1 with a bomber, might go on to take out a sub who is about to launch nukes. And this is why it is so difficult to cheat in Defcon.
This really makes no sense at all. In the example given, a snapshot would preserve the fact that the fighter survived, and then stream events after that like normal. The butterfly effect is probably present in some degree in Defcon, but a snapshot could contain the intermediate effects of an event and continue on to the same end result, without needing to contain data about the original event. It seems to me that either this is wrong, or needs more expination
Well, if you fall out of sync seconds after syncing, you won't have enough data to jump back on. Syncing everything is the best solution as it also removes the ability for someone to cheat while another player re syncs.
Eat_The_Path wrote:To quote NeoThermic on the matter, "They *are* correct for a spherical world projected on a flat surface. They are incorrect for the fact that launches over the poles isn't supported. Defcon uses Equirectangular projection. Ergo if you draw a line between two points, it'll be curved. Incidnelty, despite the fact that people flag this as incorrect often, it's one of the most correct things in Defcon!"
This is perfectly true... for a certain set of launch points and landing points. The fact of the matter is that how the trajectory curves would be different depending on where you are. If you were launching from one point on the equator to another, it'd be flat. South of the equator, the path would curve 'down'. The way Defcon handles it is great for USA vs USSR scenarios, like Wargames, but breaks down in Africa vs South America. So, not completely wrong, but not completely right either.
I'll stop nitpicking now.
On an equirectangular projection, given enough distance between the two points, you'll never get a straight line. Don't forget, the formula for equirectangular projection details that the x coordinate is longitude from the centre of the projection times the cos of standard parallels where the projection is "true". Not forgetting that the centre of the projection can be totally different from the centre of the map.
NeoThermic