So I think I might've made a break in my dimensional equation. I'm working on....twelve dimensions, now, and it's finally allowing me to really get that correct, almost-randomized shape of the City. I really need to calibrate my detection device, but I think I've accounted for the shift in p-dimensional subspace due to the relative warping of
(
Read more... )
Reply
Reply
Reply
Reply
Reply
Leave a comment