Dev Update - January 27, 2026

News and Announcements
1

Dev Update - January 27, 2026

1. Last Week by the Numbers

Activity for the last seven days across our core repositories:

  • neptune-core
    • Issues: 1 closed
    • Pull Requests: 1 merged
  • triton-vm
    • Issues: 1 closed
    • Pull Requests: 1 opened, 1 merged
    • Commits: 6 merged into master

Note: tasm-lib, twenty-first, neptune-types, neptune-explorer, and neptune-proton showed no activity this week and have been omitted.

2. Stand-up Summary

Alan Szepieniec

  • Last Week: Focused on the lib-p2p branch, specifically investigating peer connection restrictions. Completed a comprehensive presentation on Neptune’s transaction architecture (mutator sets and pipelines).
  • Coming Week: Plans to convert the presentation into a video. Will continue office discussions with Ferdinand regarding the porting of the proof system to WASM 32-bit.

Ferdinand Sauer

  • Last Week: Patched a critical soundness issue in Triton VM and managed the downstream integration across tasm-lib, triton-tui, and triton-cli. Successfully established Triton VM compatibility with 32-bit/WebAssembly (WASM) architectures.
  • Coming Week: Leading benchmarking and technical discussions on the potential adoption of the STIR proof system. Will also investigate performance optimizations via additional instructions and univariate batching.

Thorkil Værge

  • Last Week: Merged PR for light client and wallet endpoints, including new mempool query functionality. Conducted a performance review of mempool linear searches, determining them sufficient for current needs.
  • Coming Week: Once the lib-p2p networking branch is stabilized and merged, focus will shift to the high-level architecture of data structures for the “reboot.”

3. Technical Discussion

Networking: Dual Stack Strategy

A significant portion of the discussion centered on the networking layer. Alan proposed maintaining a dual networking stack. By keeping the legacy stack alongside the new lib-p2p implementation, we provide a “security-in-depth” design.

  • Integration: Both stacks will serve as entry points into the blockchain-peer-loop.
  • Cleanup: To reduce bloat, legacy components like call-peer, answer, and older discovery mechanisms are being phased out where they overlap with lib-p2p’s automatic connection paradigm.

Triton VM Soundness

A recently found bug in the Triton VM verifier allowed arbitrary claims to be “proven.” While technically, the proof system was sound regardless, all verifiers we know about were vulnerable. The flaw has been reported to the Rustsec advisory and all affected versions have been yanked from crates.io.

Proof Systems: FRI vs. STIR

The team is evaluating the transition from the current low-degree test, FRI (Fast Reed-Solomon Interactive Oracle Proof of Proximity), to STIR (Shift To Improve Rate). STIR opens the path to improved concrete efficiency, which could facilitate the development of high-performance web wallets. It also has a simpler security proof. However, it is a more complex system, and might bring performance degradation in the short term. We will make sure that the impact to Neptune Cash is as small as possible.

Triton VM WASM compatibility

A draft PR to make Triton VM compatible with 32-bit architectures, most notably WASM32, is in working state. Since it builds against the version of Triton VM that contains STIR, merging it is contingent on the merger of STIR for the time being.

4. Updates and Announcements

2 Likes