Does the current decoy selection algorithm single out coinbase (miner reward) transactions?

No, but by factoring in block density when selecting an output, the algorithm is less likely to select coinbase outputs. For example, if you have 3 blocks in a row, block 0 only has a coinbase tx, block 1 is extremely dense, and block 2 only has a coinbase tx again, then the algorithm would pick up more outputs from block 1 than it does from blocks 0 and 2. So coinbase tx's don't carry as much weight with this approach

Anyway, coinbase unlock time is 60 blocks, so if the algorithm selects coinbase younger that this...

I found the relevant piece of code in the output selection: github.com/monero-project/monero/bl…master/src/wallet/wallet2.cpp#L8294

so it looks like it's handled

comment on the line above suggests it's explicitly handled

Also it won't add locked outputs cuz of this: github.com/monero-project/monero/bl…/src/wallet/wallet2.cpp#L7978-L7981

"and the output was originally created in a block that has fewer than 100 outputs total in it"

recent transaction spike kind of fixes this :D

haha yes! it does help was thinking the same

This bug and FloodXMR attack don't work well together too...

hi

are there still regular research meetings?

no

I see :(

however we can set up a meeting if there is something to discuss

cool, yeah perhaps that would be nice based off the issues you sent me

could try my hand at a PoC for this: monero-project/research-lab #84

hmm what do you want to discuss in regards to those? (MRL #70, #84, repo #6456 for context)

ah sweet :)

TheCharlatan: started on a PoC, not sure how for he got

far*

need to familiarize myself with Triptych though

cool, will be nice to touch base with him

#84 is agnostic to the tx protocol, it just spits out a list of on-chain indices of arbitrary size

i see

definitely seems a good amount of research has been done here including thoughts on what binning algo to use

there is code as well so yeah maybe seeing the state of this project would be nice