"don''t forget that a float is only accurate to 7 digits, and doubles to 15"
"This mean distance to Pluto is 591,352,000,000cm ... that''s 12 digits already."
Yes and no. The approx limits for a double is +-1.0e+-308 or so. The 15 digit limit is *after* the decimal. Similarly, float offers something like +-1.0e+-38 with 6 digits after the decimal.
However, it can''t store the difference between 591,352,000,000.0002 and 591,352,000,000.0003, which is insignifigant. Floating-point stores values as exponentials. 591,352,000,000 is stored as 5.91352e+11. Once you pass 15 digit numbers, yes, doubles lose accuracy to a unit. To solve that, break up your space into smaller areas.
Anyway, don''t you think a goal of a 1cm resolution is a little unrealistic? After all, all your numbers will be rounded, except for the moon''s distance from earth (earth-sun distance isn''t exact, so this is off too). Also, all orbital mechanics will be off due to you being unable to calcuate all forces on an object. Finally, all masses for gravitational calcs will be, at best, gross estimates.
Good luck though.
To newdeal:
A 5-minute orbit is kinda fast, unless you''re doing sci-fi stuff.
Anyway, you''ll have a lot more than 1/1000 of a second to shoot at the enemy ship, about 1/4 of your orbit time (assuming they ar e headed at each other, can fire in any direction and are moving at about the same speed) Since there is no friction in space, you are free to shoot with anything, from beam to ballistic as soon as he comes over the horizon (or before, but that involves tricky orbital calculations
)To both of you:
How are you going to do the asteroid belt?
