15:53:12 <CjS77> I hope it's ok if I bounce some ideas related to KES design for Grease in here. The context is https://repo.getmonero.org/monero-project/ccs-proposals/-/merge_requests/651
15:54:21 <CjS77> <rucknium> ■ (eprint.iacr.org/2022/117); ■ (eprint.iacr.org/2021/1445); ■ (eprint.iacr.org/2020/1441); ■ (eprint.iacr.org/2022/744); ■ (ieeexplore.ieee.org/abstract/document/10371398) -- thanks for these. So #1 (AuxChannel) and #4 (Monet) are essentially the same thing. They're different chapters out of Sui's PhD thesis. Monet adds some details and PCN support, basically. #2 
15:54:27 <CjS77> (Sleepy channels) is actually one that I didn't really consider because it uses VTS (verifiable times signatures) which I feel would make it hard to run on mobile devices.. 
15:59:44 <CjS77> #3 (Paymo) also uses time-lock puzzles and is unidirectional, so has similar issues to Sleepy channels imo.
16:26:48 <CjS77> Ideas I've been noodling on wrt trust in KES following Kayaba's comment are here: https://gist.github.com/CjS77/214dd4d63d7abaa93951772a24024193. It's a bit stream-of-consciousness / auto-socratic approach (if that's a thing)
16:28:31 <br-m> <kayabanerve:matrix.org> I think the issue with current possible designs for payment channels is yes, they all require some horrible VDF-esque solution.
20:31:28 <br-m> <torir:unredacted.org> Isn't the entire process of mining Monero effectively a VDF? I don't actually know what properties VDFs are required to have and if mining is a good enough approximation for them.
20:37:33 <br-m> <torir:unredacted.org> Okay, some quick reading informs me that VDFs have the propery of uniqueness: a VDF function cannot not produce two different outputs given the same input. This is obviously not true of mining, where chain splits can occur.
20:38:06 <br-m> <torir:unredacted.org> s/cannot not/cannot/