Skip to content

Latest commit

 

History

History
140 lines (82 loc) · 5.79 KB

Meeting 0062.md

File metadata and controls

140 lines (82 loc) · 5.79 KB

Core Dev Meeting #62

Meeting Date & Time: Thursday 5, October 2023 16:00 UTC

Meeting Recording:

Meeting Slides: https://docs.google.com/presentation/d/19r_vnG2F7-8AE6qxfjbyYEnpTA6EsM3MV7xHVb9dh7I/edit#slide=id.g19157ff4f70_0_0


Nemo & Jake (Protocol Labs)

Premise

The discussion focused on the technology and optimization techniques behind Snap Deals and Super Snaps, mechanisms used for efficient data storage and verification on a blockchain network.

Content

Snap Deals

  • Core Function: Enables adding actual user data into an empty sector, known as a CC sector, which is initially committed to the network without user data.
  • Workflow:
    1. Start with an empty CC sector.
    2. Seal it to produce a replica.
    3. Overwrite or mix some empty data in the CC sector with actual data, known as a sector update.
    4. A new replica is generated post-update.
  • Verification: Utilizes Groth16 proofs (also referred to as Snarks) to ensure that data updates in the sector were executed correctly.

Super Snaps

  • Optimization: An extension of Snap Deals designed for efficient proof aggregation.
  • Efficiency Boost: Rather than publishing each Snark proof individually, aggregates them into a single compact proof, making the verification process more efficient.
  • Flexibility: Allows for the aggregation of Snarks across multiple CC sector updates.
  • Gas and Time Efficiency:
    1. Verification time and gas cost scale logarithmically with aggregated proofs.
    2. Crossover point is at 3 proofs, beyond which aggregation becomes more gas-efficient than individual publication.

Multi-Sector Optimization

  • Storage providers with multiple CC sectors can perform data updates across these sectors in a batch operation.
  • Enables the aggregation of all proofs into a single proof that lands on-chain, further enhancing efficiency.
  • Goal is to save on chain bandwidth.

Next Steps

  1. Further technical evaluation of Super Snaps for potential bottlenecks and vulnerabilities.
  2. Benchmark the actual gas and time savings in a live environment.
  3. Explore the feasibility of integrating Super Snaps with existing storage solutions in the blockchain ecosystem.
  4. PR opened for this proposal

Facilitated by Nemo & Jake (Protocol Labs)

BLS Aggregate Signatures to FVM

Context

This FIP proposes the following changes to FVM:

  1. Addition of a syscall for BLS aggregate signature verification (by definition, this also supports non-aggregate BLS signatures).
  2. Removal of the syscall currently used for generic signature (i.e. Secp256k1 and non-aggregate BLS) validation, and refactoring its associated SDK function in terms of existing Secp256k1 signature syscalls and the added aggregate BLS syscall.

Existing Signature Schemes

  • ECDSA: Elliptic Curve Digital Signature Algorithm.
  • BLS: BLS 12-381, another signature algorithm.
  • Both are non-aggregate signatures, requiring single signers and messages.

Aggregate Signatures

  • Enables multiple signers to each sign their own message.
  • Aggregate those individual signatures into one constant-size signature.
  • The aggregate signature retains the same size as individual signatures.

Gas Efficiency

  • Verification of an aggregate signature is more gas-efficient.
  • Estimated to consume half to two-thirds of the gas needed for individual signature verification.

API Changes

  • Additive change to the API, no breaking changes.
  • Doesn't affect current user-facing APIs, adds a new API for aggregate signatures.

Trade-offs and Use-cases

  • No trade-offs, as it doesn't affect the gas costs for non-aggregate signatures.
  • Allows for cheaper multi-signature verification.

Facilitated by Jennifer Wang Lotus, Protocol Labs)

Proposed FIPs for Inclusion in nv21

The following Filecoin Improvement Proposals (FIPs) are being considered for inclusion in the nv21 Watermelon network upgrade:

Updated Timelines

The following key dates are updated timeline for the nv21 Watermelon Upgrade:

  • Last Call for All FIPs: Sept 13
  • Code Freeze: Oct 10
  • Calibnet Upgrade: Oct 19
  • Mainnet Upgrade: Nov 28

Governance Updates

  • Community Governance Call is is October 30- register here
  • Join the Fil-Gov channel here
  • A discussion thread announcing the FIP0001v2 Initiative , a project championed by the Filecoin Foundation on proposed FIP process changes. Read more here.