On a side note, i think hes actually trolling you, so troll him back by becoming an anti vac person, and refuse to take the next vaccine if you still go to school 
is there any reason why he wouldn’t believe Google Earth? you can zoom in from a whole globe view right down to the level you need to prove the curvature issues.
A thought.
If we split a scene into 2 scenes …
one for small scale models, and one for huge scale models …
both scenes could render independently. After, they are combined in composite.
Could this be a way to bridge giant stellar scenes and smaller objects?
1 Like
That could work, but only if the two parts do not interact. In the case here, you need the small objects to cast shadows on the large object, and I’m not sure how to do that with two scenes.
Also, if you use the z-depth for merging, you’ll again have issues with floating-point numbers.
I can’t believe it, the discussions on the flat earth have come this far … hehehe 
But I **** you all by raising the bet, whe are living in a fractal universe that could also be simulated, does not have too much importance at certain levels … apparently tied to the sequential time in a quantum pattern as something in continuous growth as a plant or any other living creature in a cycle of birth and death of its cells, the alchemists called it Oroborus, where everything in motion from the zero point to the infinite, but cyclical …
So it does not matter if the earth is flat or round, if the universe is limited to what we perceive and see or it is great, infinite galaxies, if in a dimension or parallel space-time we are on Jupiter or on Mars or Venus or in your room or on the street to make any crap you can think of, we are all connected, and this is the only fact that we are a collective consciousness called god or universe that is totally chaotic, whether real, artificial, or dreamed of by someone. …
^ ___ ^
It all comes down to the individual shots. And, if you are dealing with truly-huge objects, you might need to (for example) construct some kind of programmed script which selects those portions of the “huge object” (which is actually composed of many objects …) that need to appear in that particular shot. (So that you don’t waste time “culling through” them all at render time.) You might also choose based on level-of-detail.
“The Death Star(s)” from the Star Wars franchise are, of course, “huge objects,” but you never get an architectural-visualization of them that you can somehow troll through. Instead, you saw shots.
All shows consist of multiple shots, prepared individually and edited together, and the “gigantic world of gigantic things” really exists only in the viewer’s mind. When it comes down to rendering each one of those scenes, the renderer should only be presented with the minimal amount of “work to do” that is required to make each shot work.
don’t do that! we might end with a caveman and a dead kid!
1 Like
Okay, thanks for the ideas. I think I will do some tests today and gather some arguments. I think I will meet him after few weeks.