Monero Research Lab meeting - October 21, 2026Monero Research Lab meeting - October 21, 2026

The Monero Research Lab met on Wednesday, October 21, chaired by Rucknium. In updates, Rucknium is doing further analysis for his eclipse-attack fix (#11527); tevador is working on Share or Perish and relative time locks; vtnerd is adding tests to the p2p-ssl patch for ping-backs and SOCKS, since the normal connect flow already has functional tests; thomasbuilds is on more audits; gingeropolous reported monerosim work on Share or Perish, early antiproxy work and that stressnet.moneroworld.com is up; jeffro256 is handling #11586 and opened a PR to the Carrot spec (#16); and jberman said the upstream FCMP++ integration PRs are ready for review with stressnet ongoing. On FCMP++, jberman said all LLM comments were addressed, the RFQ for the Phase 2 & 3 integration audits has gone out and quotes are being fielded, the Least Authority circuit + gadgets audit is ongoing, multisig is in progress, and Carrot tx proofs are also in progress. jeffro256 noted that mx25519 got merged, and jberman planned a v3.1 stressnet release that day with the already-merged fixes (including a GUI wallet-reopening fix), and wants to check the spam-period logs to confirm tx relay v2 holds up before the upstream tx relay PR is merged.

The main discussion was issue #11586: get_tx_key/check_tx_key doxes the sender in 2-out transactions. jeffro256 said the Carrot spec PR (#16) is up and a carrot_core PR should follow within days, and that the issue affects the Carrot schedule. Only 2-out transactions are at risk because of the cryptographic construction of the shared ephemeral tx pubkey, and in Carrot every recipient learns d_e, making the impact wider than the old "be careful sharing your tx key" warning; selsta found the change-output leak was already documented in the CLI wallet user guide, though the sender-tracing implication was not. ofrnxmr argued the RPC endpoints need not be deprecated - the real problem is that wallets and services (Bisq, explorers) ask for tx keys instead of proofs - and that wallet providers should be approached to support proofs. The proposed fix would make InProofs impossible for self-sends (tevador: only those); alternatives floated were two ephemeral pubkeys per 2-out tx, doubling scan time, or removing Janus protection, which tevador wants to keep as a 16-byte mitigation that fits Jamtis well. No decision was reached, and the item returns next week. After the meeting, UkoeHB said he did not see the D_e sharing issue because Carrot has a separate path for change derivation, and ofrnxmr noted the problem only affects legacy wallets.

On relative time locks, tevador's PR #507 should target whichever branch takes the consensus changes (fcmp++-stage per jeffro256), jberman will review it and fold it into the upstream PRs, and tevador raised the lock time to 3 days so a Friday channel force-close gives users until Monday to react. On Share or Perish, Rucknium reviewed the researcher's shared code, found it "OK, but not amazing" with heavy LLM usage, and argued it is premature to endorse SoP without a more exhaustive analysis such as a Markov Decision Process; tevador has frozen the specs for now unless a major issue appears and named temporary network splits from the subjective fork-choice rule (limited to 1 block + 32 workshares) as the biggest potential problem, while gingeropolous plans to run new sims and then point an LLM at attacking them. On the Shi et al. eclipse paper, Rucknium ran performance tests written by thomasbuilds' LLM and judged the impact of growing the white_list to 45,000 peers acceptable, especially if shared peerlists are cached rather than reshuffled; ofrnxmr's complementary work is blocked by the 101-second peer drop, which he says drops a good peer and selects a likely attacker, and Rucknium's 100-hour simulation suggests his patch reduces the effectiveness of the related Moros attack on newborn nodes.

News roundupNews roundup

• Monero Observer clearnet site experiencing intermittent outages, onion operational - the clearnet site has had intermittent outages when accessed through Tor over the past ~72 hours, while direct clearnet access and the onionsite remain reachable; the operator is investigating and testing new name servers, with feedback welcome via email/XMPP.
• Biscuit Wallet v1.2.1 released with support for swaps in Tor mode - Biscuit adds swaps in Tor mode via PegasusSwap, ETZ-Swap and Stereoswap, and now holds SOL plus Dogecoin on the same light wallet as Bitcoin and Litecoin, with ETH, USDT and USDC on Arbitrum alongside Ethereum; verify the download before installing.
• XMRPod v0.1.24 released with new features and fixes - the open source AntennaPod fork for sending XMR superchats to XMRChat podcasters adds live-stream listening while browsing, pause/resume/stop from Android media controls and playback preservation when leaving fullscreen; some live streams may still stall with the screen off depending on the provider and Android WebView version.
• Monero CCS website temporarily 'disabled' after disclosure of critical PHP vulnerabilities - Core Team member binaryFate disabled the CCS website/app as an emergency measure after the public disclosure of two critical PHP vulnerabilities that can allow unauthenticated remote code execution; PHP has released 8.5.12, 8.4.27, 8.3.36 and 8.2.35 to fix them, and anyone running PHP should update now.
• Monero Research Lab meeting scheduled for 28 October 2026 1700 UTC - the next MRL meeting is set for Wednesday, October 28 at 17:00 UTC in #monero-research-lab, chaired by Rucknium, with topics including get_tx_key/check_tx_key sender doxing in 2-out txs, Share or Perish and the eclipse-attack paper; the full topic list will be posted on GitHub soon.
• Cuprate Meeting scheduled for 27 October 2026 1800 UTC - the next Cuprate meeting, for the effort to build an alternative Monero node implementation, is set for Tuesday, October 27 at 18:00 UTC in #cuprate, moderated by Boog900, with updates, the project's next steps and other business on the agenda.
• Monero Tech meeting scheduled for 26 October 2026 1800 UTC - the next Monero Tech meeting should take place Monday, October 26 at 18:00 UTC in #no-wallet-left-behind, probably chaired by rbrunner7, with Seraphis/JAMTIS, FCMPs and other long-term development issues on the agenda.
• woodser releases monero-java v0.8.63, monero-cpp v0.9.10 - monero-java now throws MoneroError on malformed base58 addresses, while monero-cpp updates monero-project to v0.18.5.3, supports disabling TLS certificate verification, validates the daemon URI before setting the connection, persists a random wallet restore height at creation and pins hidapi and libusb for Windows builds.
• Haveno v1.11.0 released with security updates, fixes and UI improvements - Haveno ships Tor 0.4.9.14 (Tor Browser 15.0.25) with high-severity security fixes, Jackson 2.21.7 and Logback 1.5.38, plus a fix for arbitrators dropping simultaneous offer-signing requests from the same maker, IPv6 Monero nodes without Tor (Tails still uses IPv4), a restored themed chat background, ticket filters and restyled context menus.
• Community Workgroup Meeting scheduled for 31 October 2026 1600 UTC - the next Community Workgroup meeting should take place Saturday, October 31 at 16:00 UTC in #monero-community, moderated by ofrnxmr, focused on community highlights, CCS updates and workgroup reports.
• Monero v0.18.5.4 'Fluorine Fermi' released - the minor bug-fix release fixes DNSSEC validation for checkpoints, blocklists and update checks, restores string-to-integer conversion in JSON RPC requests, and fixes reopening testnet and stagenet wallets in the GUI plus other minor bugs; CLI and GUI binaries are available on Getmonero.org.
• silverpill submits CCS proposal to migrate 'monero-rs' packages to 'monero-oxide' — silverpill's first CCS proposal requests 2.96 XMR for a roughly one-month migration of monero-rs packages to monero-oxide, aiming to eliminate the split in the Rust ecosystem and help users move to the better-maintained library; silverpill co-maintains monero-rpc and develops Mitra.

Sources: Monero Observer · MRL meeting logs