BitDevs DMV
Socratic Seminar 022

BitDevs DMV

Socratic Seminar 022

Wednesday, September 30, 2026
PubKey DC · 410 7th St NW, Washington DC

Chatham House Rule No recordings

bitdevsdmv.com

BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Technical · Protocol

Bitcoin Core 31.1 + v32 Preview

  • v31.1 released July 8, 2026 — bug fixes, performance, translations
  • v32 headline: PR #35295 — parallel prevout fetcher during ConnectBlock
  • Serial disk-read latency eliminated by worker thread pool pre-fetching all block inputs
  • Follow-up PR #36000 adds block read-ahead (prefetch next blocks while connecting current)
  • IBD tracking: PR #32043

bitcoincore.org/en/2026/07/08/release-31.1/ · github.com/bitcoin/bitcoin/pull/35295

Discussion
  1. Does parallel prevout fetching change the node operator cost calculus?
  2. What's a realistic IBD time improvement — and does it matter for decentralization?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Security · Tooling · Postmortem

Coldcard RNG Defect + HWI → Maintenance Mode

  • Coldcard branded "Don't Trust. Verify." — never audited the seed-generation path
  • Since Mar 2021: hardware TRNG silently disabled (MICROPY_HW_ENABLE_RNG=0); Yasmarang software PRNG ran instead
  • Seed looked valid; entropy was not — on-chain losses in the ~$100M+ range
  • Error 1: Source-available ≠ audited (restrictive license ≠ open-source scrutiny)
  • Error 2: Workflow security ≠ seed security — SE/PSBT reviewed; entropy path was not
  • Error 3: Social capital ≠ verification — HWI/Core tooling + endorsements made Coldcard the default
  • HWI → maintenance mode (Optech #418): ecosystem consequence of credentialed trust without seed-path audit

secsov.com/articles/dont-trust-verify · Optech #416 · Optech #418

Discussion
  1. What does "don't trust, verify" require when the marketed verification path never covered seed entropy?
  2. Who failed — vendor, credentialed endorsers, tooling defaults, or the culture that treats reputation as audit?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Research · Spec

Formal Verification of Bitcoin Consensus

Independent efforts, same thesis: Bitcoin needs a machine-checkable spec. Core does not ship one.

btc-verified

ProofOfKeags · Optech #414

Formal Lean/Coq proofs of Bitcoin consensus rules — prove properties of the protocol, not just test an implementation.

Hornet Node

Toby Sharp · hornetnode.org

34 declarative semantic invariants in C++ — one per rule, composable algebra — consensus as explicit, inspectable logic.

Orange Paper / BLVM

Bitcoin Commons · Z3 spec-lock

159 numbered consensus rules; Z3 locking ~251 production functions — a written spec that can reject drift in code.

github.com/ProofOfKeags/btc-verified · hornetnode.org · docs.thebitcoincommons.org

Discussion
  1. What would it take for formal verification to matter in practice — who runs the verifier?
  2. Who decides what the canonical spec says if implementations disagree?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Soft forks · Script

Covenants — CTV, OP_CAT, BIP 448

  • BIP 119 (CTV): most-reviewed covenant primitive; no activation path after 2022 attempt
  • BIP 347 (OP_CAT): re-enabled concatenation; broader expressiveness, more debate
  • BIP 448 (Sanders/Poinsot/Roose, March 2026): "Taproot-native Rebindable Transactions"
  • Saylor "constitutional framing" (July 28 X thread): treats consensus rules as a constitution; any rewrite without near-unanimous buy-in is "economic theft"
    • Lumps BIP-110, covenants (CTV / LNHANCE / etc.), and bigger-block proposals as the same "constitutional offense"
    • Political constraint from the largest corporate holder: soft forks now face an informal near-unanimity bar, not just developer consensus

bips.dev/119 · bips.dev/347 · bips.dev/448 · 446 / 348 / 349

Discussion
  1. Is BIP 448 a meaningful step toward activation or another stall in different packaging?
  2. Does the Saylor constitutional framing help or hurt the technical conversation?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Governance · Soft fork

BIP-110 Failure, Ocean Split, Luke Dashjr Removed

  • BIP-110 (Reduced Data Temporary Softfork): marked Complete June 25, 2026
  • Mined 2 blocks on August 9, then stalled — marked Closed same day
  • 55% threshold via modified BIP9; one-year active duration; UTXO grandfathering
  • One miner rejected BIP-110 while simultaneously mining through Ocean (DATUM)
  • Fork fell 300 blocks behind; mathematically ~6 years from self-correction
  • Knots added undocumented 8th consensus rule (P2A non-empty witness invalid) not in the BIP
  • Luke Dashjr removed as BIP editor August 11

coindesk.com/tech/2026/08/09/… · decrypt.co/375310/…

Discussion
  1. What does BIP-110's failure reveal about Bitcoin's governance model?
  2. Is an undocumented rule in the reference implementation a bug or an inherent feature of implementation-as-spec?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Mining · Infrastructure

Mining — Stratum V2, Poolin Bankruptcy

  • Stratum V2: pools representing 75% of hashrate committed to open block construction standard
  • DATUM (Ocean) is a competing implementation of the same block template concept
  • Commitment ≠ deployment — Job Declaration (miner-side template construction) is the real metric
  • Poolin: Chapter 11 bankruptcy, heading toward $52M asset sale
  • Poolin was once one of Bitcoin's three largest pools

coindesk.com/markets/2026/05/11/… · stratumprotocol.org

Discussion
  1. Does 75% SV2 commitment translate to meaningful decentralization of block construction?
  2. What does continued pool consolidation mean for hashrate decentralization?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Lightning · Security disclosure

Lightning — CLN Embargo Binary + LND / Taproot Assets

  • CLN 26.06.7 (Aug 28): security fixes shipped as signed binaries only — source withheld 14 days so attackers couldn't diff the patch
  • Rationale (Decker et al.): "a patch identifies the code it changes" — operators upgrade first; source later for reproducible verify
  • Embargo failed in hours: release binaries still carried debug symbols; reverse-engineering (helped by LLMs) recovered the patched source and reproduced matching checksums
  • Confirmed on Optech #420 Recap — maintainers: "big learn"; next releases ship source + binaries together
  • Also this cycle: LND v0.21; CLN 26.06 quantum-resistant channels; Taproot Assets v0.8 + Stablecoin SDK

blockstream.com/core-lightning-26-06-7 · Optech #420 · #420 Recap

Discussion
  1. If withholding source fails when debug symbols remain, what disclosure model survives LLMs?
  2. Is "upgrade now, verify later" compatible with "don't trust, verify" — or does it just relocate trust to the signing key?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Policy · DC

CLARITY Act — Cloture Failed (Sept 15)

  • H.R. 3633: digital asset market structure (SEC/CFTC split) — House passed 294–134 (July 2025)
  • Outcome: Senate cloture on the motion to proceed rejected 49–50 (needed 60) — roll call 234. Never reached floor debate on the substance.
  • Why it failed (not the SEC/CFTC split): ethics fights over officials’ crypto profits; illicit-finance / non-custodial developer liability; stablecoin yield vs banks; state AG preemption concerns
  • All voting Democrats opposed; Republican nos included Collins, Hawley, Moran, Tillis (Tillis later filed motion to reconsider)
  • Near-term path: agency rulemaking (SEC/CFTC), not statute — market structure remains interpretive

senate.gov roll call · Reuters · DLA Piper

Discussion
  1. If the fight was ethics and illicit finance — not commodity classification — what does that say about “Bitcoin needs clarity” as a political slogan?
  2. Is agency-driven market structure better or worse for Bitcoin than a compromised statute?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Implementations · 2025/2026 Wave

Alternative Bitcoin Implementations

Core derivatives

  • Knots Classic — non-RDTS/non-BIP-110 Knots line
  • Knots Untangled — bitcoinuntangled.org
  • Bit-Block — hard anti-spam defaults, datacarriersize=0

Longstanding alts

  • btcd — Go; caveat: mostly Lightning / infrastructure today, not a competing full-node culture
  • libbitcoin — impressive IBD times recently; no UTXO model; still no node release

Independent engines (2025/26 wave)

  • Floresta — Rust, Utreexo, ~800MB disk
  • Hornet Node — C++, declarative spec, 11× IBD
  • Bitcoin Echo — pure C, zero deps
  • rbitcoin — Rust, AI-assisted; no UTXO model
  • satd — Rust, Silent Payments
  • Commons / BLVM — 159 rules, Z3 spec-lock
Discussion
  1. What would it take for an independent implementation to reach economic majority?
  2. Is dropping the UTXO set (libbitcoin / rbitcoin) a breakthrough for IBD — or a trade that breaks too many assumptions?
BitDevs DMV
Socratic Seminar 022
bitdevsdmv.com · reading room Sep 30, 2026 · PubKey DC

Privacy · Critical reading · [[alloc] init]

Shielded Bitcoin — Critique

Claim: Zcash-style notes / nullifiers / ZK on L1, “no soft fork,” indexer replay; peg deferred to PIPEs v2. paper · Delving · PIPEs v2

  • Incomplete: transfer layer only; the BTC bridge (theft / inflation surface) is a different paper
  • No consensus enforcement: invalid envelopes can confirm; validity is a software convention
  • PIPEs peg is speculative: experimental witness encryption; historically TB→GB ciphertexts; heavy compute
  • Indexer liveness: can censor / delay / lie; “run your own replay” — most users won’t
  • Edge privacy fails: timing, fees, arity, carrier metadata public; peg edges kill anonymity sets
  • “No soft fork” reframes: hard problem moved into exotic crypto, not solved
Discussion
  1. Is shipping the transfer layer first a research step — or marketing an unfinished bridge?
  2. Is “Bitcoin as bulletin board + indexer” just a sidechain with OP_RETURN journaling?
BitDevs DMV
Socratic Seminar 022

BitDevs DMV

Chatham House Rule applies to everything discussed tonight.