03:20:30 <m-relay> <r​ottenwheel:unredacted.org> > Monero's next hard fork is expected to deploy Full Chain Membership Proofs, which will eliminate the on-chain ring signature privacy model. However, in certain situations, decoy-based privacy will still be used to provide protection to users' wallets from a potentially malicious spying remote node. Therefore, the OSPEAD-derived decoy distribution can be used in those circumstanc<clipped mes
03:20:32 <m-relay> <r​ottenwheel:unredacted.org> es. For more details, read "Initial Probability Density Function for OSPEAD".
03:20:34 <m-relay> <r​ottenwheel:unredacted.org> Rucknium Does this mean OSPEAD's new decoy distribution you mention few lines above will get added to FCMP++ future network upgrade?
03:20:36 <m-relay> <r​ottenwheel:unredacted.org> > The OSPEAD techniques suggest a new decoy distribution, which would reduce the average attack success probability to 7.6 percent, corresponding to an effective ring size of 13.2.
04:00:48 <m-relay> <e​longated:matrix.org> Fcmp needs decoy selection?
04:17:19 <m-relay> <g​ingeropolous:monero.social> not directly elongated . I think its for instances pulling outputs from a remote server, so the remote server will have a hard time figuring out which output is own by the requestor.
04:17:29 <m-relay> <g​ingeropolous:monero.social> or I could be way off.
04:33:36 <m-relay> <a​remor:matrix.org> Yeah something like that.
15:37:24 <m-relay> <r​ucknium:monero.social> rottenwheel: Yes[*]. The default way for wallets to construct FCMP++ txs will be to build the tree locally in the wallet cache file (based on data queries to nodes), then pull from that. The default method does not require decoy requests. In the event that local tree building is not possible, a backup method is for the wallet to request a set of decoys (plus the real spend) from a<clipped message
15:37:26 <m-relay> <r​ucknium:monero.social>  node at the time of tx construction. The OSPEAD-derived decoy selection algorithm can be used for that.
15:38:26 <m-relay> <r​ucknium:monero.social> Also, decoys would be selected for the light wallet "filter-assist" key tier. This won't be implemented in the first FCMP++ hard fork, but it is likely that it will be added later.
15:39:58 <m-relay> <r​ucknium:monero.social> [*]: The distribution currently posted is a draft, based on data up to October 2024. We would want to update the data and possibly adjust a few choices in the estimation for the final version. The final version is supposed to be delivered for Milestone 3 of OSPEAD.
15:51:22 <plowsof> tw