13:29:59 <br-m> <gingeropolous> hrm, at some point I should build a thing that runs simulations using a mixed network of current main and a tagged release. I'd probably need to spec out a proper gauntlet though .... 
13:35:40 <br-m> <gingeropolous> im currently meandering about scale maxing . 3k node sim finished a 16 hr sim run in 36.5 hrs. 2k node 16 hr run at 20.4 hrs, 1k nodes 16hrs finishes in 9.4 hrs. KSM ftw - all the randomx memory thingers are identical, so we only need 1 copy! apparently the slowdown at higher scale is from logging all the going through the nodes....
13:36:41 <br-m> <gingeropolous> 1k run has 200 users, 2k run has 400 users, 3k run has 600 users (nodes count is just relay nodes. users include a wallet and accompanying sending and receiving of txs). 
13:39:35 <br-m> <gingeropolous> logging all the txs obvi
13:40:28 <br-m> <gingeropolous> and im not sure what to do with the SoP data and sims