I'm sorry your thread was so badly derailed as to be basically unreadable.
This egotistical shit isn't going to help you.
Without an expert on cryptographic hashes commenting here on the possible breakage, I don't think you can claim omniscience.
As I explained in private, the output of a Random Oracle is assumed to have no topological structure, i.e. just random points in any multi-dimensional space you choose. We don't have to have a preimage of the solution in order to potentially have a multi-dimensional
curvepartial-order appear due to structure found in the hash function and be able to reduce the entropy. I understand that even Cryptonote has the same reliance on a Random Oracle to prevent undetected inflation. Thus I also understand that concern about hash functions may be out-of-scope of discussing or orthogonal to the innovations of this paper. My point was essentially wouldn't it be safer to use a 256-bit curve.
Recommendations are for 256-bit ECC and hash security to be protected until 2032:
http://www.keylength.com/en/compare/Heck after I opinioned that his original sum of two squares was lacking entropy, johoe proceeded to prove my intuition was correct.
As for the other idea I introduced about combining mixing (e.g. CN) and two curves, I was attempting to brainstorm simplifications. I asked 3 times whether it was possible to find multiple uncommitted values that could map to the same committed value, which I understood my idea hinged on. When I finally got the answer from Mixles, I dropped that idea.
Actually guys like me who toy around with simplifying ideas, in spite of not studying ECC extensively, sometimes do achieve simplifications.
So watch your ad hominem derogatory pompous mouth. Because you are going to eat the words you spout.
If you are Andrew Poelstra, I am about to make mincemeat of your whitepaper on PoS.
Narrow, unenlightened, closed-mindedness doesn't impress me. What impresses me are humble, open-minded, patient experts. Johoe appears to have this enviable quality.