Contributed Post:
Hot topic · builder-side · non-market · drawn from X and related sources
Retiring chain scans, counting subtrees
The practical edge since the last check is about what has to change before high throughput is usable day to day — not only at the node, but in how wallets and apps recover coins and verify data.
Project Babbage has proposed BRC-234 as a path to retire reliance on zero-cost full-chain scanning and indexing. The argument is that private keys alone are no longer treated as enough for recovery once scale bites, and that applications still built around continuous chain scans will need a different model. That sits next to a concrete inventory of remaining WhatsOnChain / scanning dependencies across Wallet and Overlay Services, with a public GitHub issue tracking the gaps.
Teranode’s capacity framing this week is subtree arithmetic rather than a single TPS headline: about 1 million transactions per subtree, one subtree per second at baseline, on the order of 600 subtrees in a 10-minute block — with no hard block-size cap. Separate “by the numbers” posts continue to cite sustained million-plus tx/sec capacity, multi-year development, and a predominantly Go codebase.
On the application layer, builders are still filling infrastructure that the base chain does not provide out of the box. DXS has described STAS-side wallet, indexer, and contract work needed to bridge account-based stablecoins onto BSV’s UTXO model. BSVKey continues to push pay-per-call and per-token inference settled as BSV micropayments from a BRC-100 wallet. Elsewhere, a Metanet-side token design claims lineage verification without indexing, contingent on protocol-level opcode support.
The current pulse is the same tension as recent editions: node capacity is described in concrete units, while wallets and apps are still being forced to stop depending on scan-everything indexers.
Source Links
Leave a Reply