libera #monero-dev

17 Apr 2023

Official repo: github.com/monero-project/monero || Don’t ask to ask, just ask:) || Monero GUI dev is at #monero-gui || Monero website dev is at #monero-site || Non-dev questions -> #monero, please || No fork help, you’re on your own

• 00:03
  plowsof11
  i dont know what c++ is, and wasnt aware of the latest commits (only seen the 1st one) my apologies sir
• 00:15
  ofrnxmr[m]
  i really dont see what problem this pr hopes to solve
• 00:16
  ofrnxmr[m]
  torrc has options of its own for reconnects and circuit dirtiness.
• 00:16
  ofrnxmr[m]
  this reimplementation on monero isnt a monero problem
• 00:17
  M53a45b263745298
  Tor doesn't actively drop long-lived connections
• 00:18
  M53a45b263745298
  So circuit reuse ensues
• 00:19
  M53a45b263745298
  Tor doesn't actively drop any connections actually, not just long-lived
• 00:19
  M53a45b263745298
  Circuit dirtiness is for new streams
• 00:20
  ofrnxmr[m]
  NewCircuitPeriod 30
• 00:20
  ofrnxmr[m]
  MaxCircuitDirtiness 600
• 00:20
  ofrnxmr[m]
  EnforceDistinctSubnets 1
• 00:20
  ofrnxmr[m]
  CircuitsAvailableTimeout 86400
• 00:20
  M53a45b263745298
  Which one of those drops connecitons?
• 00:24
  M53a45b263745298
  I don't know what that reaction means?
• 05:32
  moneromooo
  Expecting users to tweak their tor config this way is not good. Much better to ensure monero uses it safely in the first place.
• 05:33
  M53a45b263745298
  That torrc doesn't do what he thinks it does, there is no way to modify torrc to actively drop connections
• 05:33
  moneromooo
  Also, historically vtnerd has been around only occasionally, so unobtrusively pinging from time to time is a good way to ensure this doesn't get forgot.
• 05:35
  M53a45b263745298
  Thanks, he responded above
• 10:39
  ofrnxmr[m]
  moneromooo: im not expecting users to tweak their tor configs, but "monero" cant build new tor circuits
• 10:40
  ofrnxmr[m]
  And onions are different than "circuits" that use exit nodes
• 10:41
  M53a45b263745298
  SOCKS clients (monerod) can force Tor to build new circuits
• 10:42
  M53a45b263745298
  IsolateSocksAuth flag is set by default on all Tor SocksPorts
• 10:42
  M53a45b263745298
  It is standard practice to use this for stream isolation
• 10:43
  M53a45b263745298
  Your configuration above does nothing to prevent circuit reuse whatsoever
• 17:04
  vtnerd
  M53a45b I thought only tor control connections could request new circuits, but socks 5 perhaps?
• 17:39
  fedorafan
  what makes it impossible to fake buyin
• 17:45
  moneromooo
  Your question was cut off.
• 17:46
  moneromooo
  If it's about markets (got cut off at "buyin") then try #monero-markets. Or #monero if not about monero development. Otherwise ask again here.
• 17:49
  fedorafan
  ah no its about development but fine
• 17:50
  fedorafan
  ah you dont provode every question to this channel?
• 17:51
  moneromooo
  Monero dev chat in #monero-dev. Market chat in #monero-market. Research in another. Catch all in #monero.
• 17:51
  fedorafan
  ok
• 20:58
  M53a45b263745298
  <vtnerd> "M53a45b I thought only tor..." <- SOCKS5 auth is the standard way for a stream (connection) to get its own circuit
• 21:01
  M53a45b263745298
  It's not "requesting a new circuit", it's "don't put this stream on a circuit with any other streams that have different auth"
• 21:01
  M53a45b263745298
  Which is the same effect
• 21:08
  M53a45b263745298
  I reference this in the PR, the host blocking won't be necessary once SOCKS5 exists
• 21:08
  M53a45b263745298
  I'm going to put it as one of the items for my CCS when I get around to it