January 2025 FIL Implementer Call

<aside> ‼️

The FIL implementer call notes are public, unpublisized security issues shall not be disclosed in this doc.

</aside>

Aync updates

• FIP discussion: Remove BatchBalancer / Introduce per sector fee
• FIP discussion: Incentivisation of F3. | - Review and merge migration for FIP0097.
• Make F3 “code complete” for nv25. | | Forest | [email protected] | Shipped:
• Forest serverless faucet: https://forest-explorer.chainsafe.dev/faucet
• https://github.com/ChainSafe/forest/discussions/5169
• RPC snapshots testing | - RPC fixes and improvements
• Internal refactoring | | ChainSafe Infra | [email protected] | Shipped:
• automated snapshots after network upgrades

In progress:

• mainnet archival node sync and backfill of diff snapshot archive | | | Venus | [email protected] | | | | Curio | [email protected] | Curio Markets Snap batch | Offline DDO Market 2.0 RFC | | Elliptic Research | [email protected] | Preparing our next feature release | Cut a release when tested and ready | | Filecoin Monitoring & Analysis | [email protected] | | | | Filecoin Spark (Retrieval Check) | [email protected] | Shipped:
• HTTP-only retrieval success score (10%, compare to 20% when Graphsync is included)
• TTFB proof-of-concept (network median: ~1 second)

Live on https://dashboard.filspark.com | Start testing retrieval of DDO (non-f05) deals by the end of January. | | FIPs | [email protected] | FIP0098 in Last Call until Feb 6. | | | FF Security | [email protected] | 1. PDP Audit with zellic. 2. Telemetry aggregation for IR alerting. 3. Smart contract and eth support in Antithesis. | Forest, Lotus rpc api correctness testing in Antithesis.

Adversarial testing in Antithesis.

NV25 testing in Antithesis. |

Sync Agenda & Notes

• (Orjan, Steve) nv25 update

  • Wearing the “F3 lens”

    • Scope:
      • This is a proposal if we ignored onboarding fees FIP(s) and treated it out of scope for nv25. We aren’t saying that’s actually the case; it’s just evaluating an nv25 scope / timeline if that were the case.
      • FIP0097: Add Support for EIP-1153 (Transient Storage) in the FEVM.
      • F3 items
        • Code changes necessary for Mainnet Passive Testing
          • Note: FilOz will update the F3 FIP with what has changed since nv24 for visibility (tracking issue)
        • Functionality to activate with delegated multi-sig approved final F3 parameters ~2-3 weeks after nv25 activation (FIP discussion)
    • Potential Schedule:

    Theme	Item	Date
    FYI	Chinese New Year break	January 25, 2025 → February 3, 2025
    Estimate	F3 implementation work complete	February 7, 2025
    Proposed	FIP Last Call	January 31, 2025
    Proposed	Actor code freeze	February 14, 2025
    Proposed	Lotus RC for Calibnet	February 18, 2025
    Proposed	Calibnet upgrade date	February 25, 2025
    Proposed	nv25 mainnet upgrade date	March 13, 2025
    Proposed	nv25 F3 passive testing	March 14, 2025 → March 28, 2025
    Estimate	F3 Activation	April 2, 2025

  • Potential of “Reduce onboarding gas & align value accrual by removing batch balancer and replacing it with a per-sector fee”

    • FIP discussion: https://github.com/filecoin-project/FIPs/discussions/1105
    • Actively being worked on by multiple parties to try and be included in nv25
    • Work involved:
      • Primarily builtin-actors work that FilOz would plan on taking unless someone else volunteers (at which point FilOz would actively support)
      • Mining software updates (e.g., Venus will do for lotus-miner and Venus, Curio will do for Curio)

  • FIP 0098 is in last call and could be picked up

    • https://github.com/filecoin-project/FIPs/blob/master/FIPS/fip-0098.md
    • This is a easier item to help make lenders happy.
    • Forest: Any interest in taking this? FilOz will support the reviews.
      • 🎬 Forest/Hubert to take a look to see about the amount of work involved and post in #fil-implementers
• Decoupling FIPs / Network Upgrades timelines
  • Pre-Scheduled upgrades
  • Current constraints (at least from Orjans perspective) is that resource constraints is currently a blocker.
  • Implementation capacity / external implementers /
  • Timelines difficult because they often are shifting.
    • Also a burden from implementers teams.
  • Hubert from memory of last time this was discussed: Consensus last time we discussed it was being pragmatic when it comes to network upgrades.
  • Past discussion:
    • https://github.com/filecoin-project/core-devs/discussions/145
    • https://github.com/filecoin-project/core-devs/discussions/177
• DDO <> Spark <> Boost —> Could not be changed in Boost without redoing core Boost.