Files
dlcspecs/meeting/meeting_august_3_2021.md
Nadav Kohen b16070cf6f Tuesday August 3 2021 Meeting (#172)
* Added August 2021 meeting doc

* Added meeting notes
2021-08-13 08:33:29 -06:00

5.7 KiB

August 3rd (7 PM CST)/6th (9 AM JST) Meeting 2021

Individual Updates (Sync)

  • #Query(Nadav)
    • Not much, just rebased secp-zkp
  • #Query(Tibo)
    • Serialization work (back and forth with Matt Corallo)
      • New proposal that more heavily uses TLVs
      • Implementation work on this
      • Working on custom messaging with LDK folks
  • #Query(Lloyd)
  • #Query(Chris)
    • Working with Atomic Finance
      • Automating oracles
        • Deployed last week
  • #Query(Jesse)
    • Working on Shamir & FROST
      • WIP PR pushed to secp256k1-zkp
        • Includes DKG
  • #Query(Matt)
    • Created proposal for DLC mutual close protocol
    • Troubleshooting and improvements on node-dlc
    • cfd-dlc serial-id stuff
  • #Query(Ben)
    • Working on networking P2P over TOR
  • #Query(all)

Implementation Updates (Sync)

  • bitcoin-s
    • P2P work beginning to set up DLCs over TOR without copy/paste
  • rust-dlc
    • Began work on networking as well
      • Hopefully will work with bitcoin-s in the future
    • Serialization stuff
  • atomic finance
    • cfd-dlc serial-id stuff
  • others?
    • Binary outcome DLC implementation from Lloyd

Specification Writing

#Status_Update_Interrupt

  • If anyone is interested it would be nice to document loss-of-fund vectors
  • Draft Updating P2P message serialization to include sub-types, optional fields, tlv streams
    • https://github.com/discreetlogcontracts/dlcspecs/pull/171
    • Closer to TLVs than the previous draft
    • #Discussion
      • Every tiny thing had a big type associated with it and this will simplify things and make them more clear and extensible, current LN is too restrictive
      • Open call for review!
    • Also adds protocol versioning to offer message
      • Anyone have new thoughts on this?
  • Mutual Close Message
    • https://github.com/discreetlogcontracts/dlcspecs/pull/170
    • Open for review!
    • Adds a way to request a close with mutual cooperation
    • Ben commented that it is very restrictive as is to have a fixed template closing transaction and it would be nice to have it be more flexible (like LN dual funding).
      • General response was that since this is more of an application level standard it is fine to have multiple iterations including the very simple one and later more general ones.
      • Thibaut also notes that the trust assumptions behind this piece of the spec should be as explicit as possible.
  • Multi-oracle support for oracles with different num_digits algorithm proposed
  • bLIPs
    • E.g. premium_fee for fee-bumping, TLV extensions, etc.
    • #Discussion
  • P2P discussion (TOR?)
    • #Discussion
    • Thibaut is looking for a Rust library for doing TOR stuff
    • bitcoin-s currently just sends binary blobs without all the extra LN stuff
      • e.g. init message and ping/pong
        • These things will be needed in the future to get cross-implementation communication working
    • Things moving along, need serialization finalization from earlier
  • Path to Lightning
    • #Discussion
    • It's gonna take a lot of work
    • Hard part seems to be figuring out the separation of abstraction between the LN specific and DLC specific stuff
    • Once Thibaut is done working on networking stuff he hopes to formalize the protocol required for DLC-related updates inside of a LN channel
      • Thibaut plans on implementing some form of sub-channels
    • New virtual channels paper!
      • Virtual channels in theory could be used for "in-channel" DLCs on LN between people not connected.
  • Time to start thinking about Taproot!
    • #Discussion
    • Next time. We don't have enough time.

Oracle Specifications

#Status_Update

  • Breaking Oracle Changes for v0
    • https://github.com/discreetlogcontracts/dlcspecs/pull/167
    • Splits up oracle key into announcement and attestation key
      • Lloyd notes that there should be a separate attestation keys for each kind of attestation scheme
    • Adds oracle_keys message which is static and separate from announcements
    • Adds nonce proof-of-knowledge signatures to announcements
      • Thibaut and Lloyd suggest getting rid of this
    • Adds option for ranged timestamps on oracle events
    • Current review is mostly small things, are people okay with the general structure?
      • #Discussion

Oracle Attestation Research

#Status_Update

  • #Query(Lloyd)
    • paper: https://eprint.iacr.org/2021/350.pdf
      • Compress s values in signatures with R values present
      • This would fix our aggregation woes
      • TODO: Get feature branch from Jonas to tinker with
        • Nadav - ask Jonas for this code
    • Coblox
      • https://coblox.tech/
      • They are working on building loans on Liquid
      • Use an oblivious oracle scheme to replace escrow in hodl.hodl kind of scheme
      • They require liquidation conditionals
      • Follow-ups will happen
    • In the future perhaps covenants can be used to replace signature anticipation?