Coin Brief ENDE

Where signing complexity goes: Solana's nonces, Bitcoin's sp(), Ethereum's PQ aggregation

Three technical documents from the past week deal with signatures on three different chains, and each one is about the same choice: which layer carries the complexity.

Solana's draft SIMD-0723 takes it out of the protocol. Durable nonces let a signed transaction stay valid indefinitely, which custodians with air-gapped signing rely on, but the proposal lists bugs from 2019 to 2026, including a 2022 mainnet halt of about 4.5 hours, that came from the special rules nonces need. After activation, not before 5 May 2027, every transaction would need a recent blockhash, and delayed signing would move to onchain programs that verify an offline signature. The proposal is explicit that risk moves with it: those programs must handle replay protection, expiry and revocation themselves, and a failed submission does not revoke an authorisation (SIMD-0723).

Bitcoin's BIP 392 works at the wallet layer. It defines sp() descriptors so silent payment wallets can describe and back up their scan and spend keys in a shared format. This week's fix corrected examples that derived both keys from the same path instead of the 1h/0 and 0h/0 paths from BIP 352, the kind of mismatch that would let a backup restore without finding its payments (bitcoin/bips).

On Ethereum, a post on ethresear.ch measured how to aggregate the post-quantum signatures Lean Ethereum plans to use. Three folding designs lost to leanVM's full recursion inside an aggregator, and the biggest gain came from choosing batch sizes that keep a merge below a step in proving time (ethresear.ch).

Where signing complexity goes: Solana's nonces, Bitcoin's sp(), Ethereum's PQ aggregation
Where signing complexity goes: Solana's nonces, Bitcoin's sp(), Ethereum's PQ aggregation — Coin Brief

What links them

None of the three adds new cryptography. Each decides where existing rules should live and writes down what the decision costs. Solana pays with migration work for custodians, in exchange for one lifetime rule for every transaction. BIP 392 pays with a stricter convention that wallets must follow exactly. The Lean Ethereum measurements show that keeping recursion in one place, with careful sizing, beats moving work into a cleverer scheme.

For users the practical point is the same on all three chains: the details that decide whether funds can be moved, found or verified are increasingly in specifications and examples, not in the user interface. Reading the proposal, including its security section, is part of the job for anyone who holds keys for others.

Written by Victoria Shinder.