3 ms·

> If it was a game mechanic I'd call it a cunning way to impose internally consistent limits on what the player can know. But if you assume it's a multi-player

> If it was a game mechanic I'd call it a cunning way to impose internally consistent limits on what the player can know. But if you assume it's a multi-player game, wouldn't it be easier to implement a global space-time? The reality is that you have to keep an internal clock for each player.

throwuwu 3y ago
Partition them by velocity
- vagab0nd 3y ago
  Velocity relative to what?
  - devoutsalsa 3y ago
    Me. I am at the center of my observable universe :)
  - oneshtein 3y ago
    Relative to CMB.
  - throwuwu 3y ago
    If it’s a simulation then there is a fixed point. Any proper game developer would also fake it for the most part e.g. culling things the players aren’t looking at, making distant objects static billboards and they’d also fudge the numbers with something like the cosmological constant.
javajosh 3y ago
In practice you need a global coordinate system. For example, you'd create a "spawn point" for all players, and then, before the game starts, move them to their starting positions at a fraction of c. The game is still keeping track of units within this initial coordinate system, and transform those coordinates for the player. The reverse is done for player decisions. There is effectively an in-game "star network" that mediates between player. Each player/unit has a unique light cone associated with it. Nothing stops us from arbitrarily designating a particular reference frame as "preferred" in a simulation; SR merely says that such a choice is truly arbitrary. Note also that this system could work recursively, such that new units must always spawn at the spacetime location of older units, and do information management (behind the scenes) in a similar way. This would turn the network into a tree - and the transforms are relatively cheap (and we could use a Spacetime algebra library!) Note that I don't think this works for a traditional RTS like C&C or Supreme Commander. But rather a game like Civilization, but galaxy spanning (and spanning very long periods of time), where the smallest objects are something like "suns" and "fleets". The individual lives of humans would come and go as these vast battles took place over thousands of years. You'd only need to add magic like a 1g constant acceleration drive - and a strong rule against using this to directly kill planets - and assume that the fleets are actually large-ish asteroids fitted with the drives and a human and/or robotic crew. The spawn point would then be, fittingly, Sol.
- lucasban 3y ago
  I’d play this
mr_toad 3y ago
> The reality is that you have to keep an internal clock for each player. You can transform between two players clocks easily enough. Time is just another coordinate to maintain. https://en.wikipedia.org/wiki/Lorentz_transformation https://en.wikipedia.org/wiki/Lorentz_transformation

throwuwu 3y ago

Partition them by velocity

vagab0nd 3y ago
Velocity relative to what?
- devoutsalsa 3y ago
  Me. I am at the center of my observable universe :)
- oneshtein 3y ago
  Relative to CMB.
- throwuwu 3y ago
  If it’s a simulation then there is a fixed point. Any proper game developer would also fake it for the most part e.g. culling things the players aren’t looking at, making distant objects static billboards and they’d also fudge the numbers with something like the cosmological constant.

javajosh 3y ago

In practice you need a global coordinate system. For example, you'd create a "spawn point" for all players, and then, before the game starts, move them to their starting positions at a fraction of c. The game is still keeping track of units within this initial coordinate system, and transform those coordinates for the player. The reverse is done for player decisions. There is effectively an in-game "star network" that mediates between player. Each player/unit has a unique light cone associated with it. Nothing stops us from arbitrarily designating a particular reference frame as "preferred" in a simulation; SR merely says that such a choice is truly arbitrary. Note also that this system could work recursively, such that new units must always spawn at the spacetime location of older units, and do information management (behind the scenes) in a similar way. This would turn the network into a tree - and the transforms are relatively cheap (and we could use a Spacetime algebra library!) Note that I don't think this works for a traditional RTS like C&C or Supreme Commander. But rather a game like Civilization, but galaxy spanning (and spanning very long periods of time), where the smallest objects are something like "suns" and "fleets". The individual lives of humans would come and go as these vast battles took place over thousands of years. You'd only need to add magic like a 1g constant acceleration drive - and a strong rule against using this to directly kill planets - and assume that the fleets are actually large-ish asteroids fitted with the drives and a human and/or robotic crew. The spawn point would then be, fittingly, Sol.

lucasban 3y ago
I’d play this

mr_toad 3y ago

> The reality is that you have to keep an internal clock for each player. You can transform between two players clocks easily enough. Time is just another coordinate to maintain. https://en.wikipedia.org/wiki/Lorentz_transformation https://en.wikipedia.org/wiki/Lorentz_transformation