20:23:16 <Rucknium[m]> I guess I should have posted this here, since it is more Research Lab than Dev:
20:23:23 <Rucknium[m]> Status report: I brainstormed and have a really rough draft for what an overhauled mixin selection algorithm may look like and I shared it with jberman. This is distinct from the small fix that jberman is working on. It doesn't incorporate binning as of yet; I'm not sure whether it can incorporate binning, but I think it might.
20:47:04 <UkoeHB> re: binning, I originally got my ideas by looking at this presentation for Maxwell's How to Squeeze a Crowd paper (which is far less intuitive) http://ieee-sb2018.cs.ucl.ac.uk/presentations/4-Chator.pdf
20:51:50 <UkoeHB> IMO the polynomial interpolation to fit multiple real spends into one bundle of samples was overkill (better to have a separate ring member set per input).
20:54:31 <UkoeHB> ah I guess Chator is the main author, my bad