No need to apologize, your not the only one who does this mistake. Ive seen professors do it.
The alternate solution would have been keeping the camera stationary and move eberything else in reverse to the camera And eliminate steps that move too much in the simulation to relative moves that have smaller values. This is not necessarily less work. Splitting cam runs to local si actually very easy, and then render 3 plates is no big deal.*
Anyway ist not less real than it was before this. See the computer when it calculates something flattens the world to stationary camera at the center of the universe anyway at render
The thing is each of your object has a reference frame, and it is accurate in that reference frame (called a trnasform, it would have never been accurate if this was not the case). Now the problem arises because the offset of this reference frame over shadows the numbers inside your transfrom
QUOTE
I've never taken a computer science class
It does not matter, even if you dont know something it still will happen. If you cant learn then theres not much point in doing anything. Anyway you have taken a math class in school. As a consequence of that you have used a pocket calculator at some point i hope. In conjunction with that they did hopefully teach you this (odds are they did, its just that most of what we were teahced slips trough the sieve). So i would expect a 16 year old to know this. Not necessarily where it will appear but the general idea of it. So odds are you knew this phenomenon to start with, you just couldn't think of any application wher it would metter, untill today.
You can even observe google doing this try:
open google
type
30000000000000+1
then type
30000000000000+0
see it issues same answer. you can test same on your pocket calculator. And im allmost sure you knew this for a fact. A computer is nothing more than a much more fancy pocket calculator. Sure a computer could resolve with infinite precicion, but ist slow so slow in fact that you wouldnt survive it. because the processor is optimized for double precision floats, wich is enough for MOST things.
*on this note if you ever render star field you get much better results if you dont move the camera just rotate it, as stars do not really move at all, ever. Not even if you made a simulation where you flew from here to alpha centauri in 5 seconds would you see more than maybe one or 5-10 stars move, very little. Thus for antialiasings sake its best to keep things stationary then at least the magnitude wont flicker because samples shift a bit, whereas the solution should not.
PS: this is why scientists and engineers are obsessed with teh accuracy of a result, not the result itself. Theres not much point in measurement thats saying 1.01 degrees celcius when the inaccuracy of most temperature measuring devices is 1 degree or more.