Engineering articles on deterministic data processing, standards-grade conformance, and the practices that produce evidence of correctness.
Series
A four-part series on the algorithms, parser, architectural decisions, and evidence-based release process behind a Go implementation of RFC 8785.
View on GitHubThe Burger-Dybvig algorithm in Go: finding the shortest decimal string that round-trips to the original IEEE 754 binary value, with ECMA-262 tie-breaking rules, validated against pinned oracle test vectors.
A strict Go parser that enforces every constraint in RFC 8259. Surrogate pair validation, noncharacter detection, duplicate key rejection after escape decoding, and seven independent resource bounds, all things encoding/json silently accepts.
UTF-16 code-unit key sorting, a thirteen-class failure taxonomy with stable exit codes, and machine-readable ABI contracts. The small decisions that determine whether downstream systems can depend on you.
Unit tests don't prove determinism. An offline replay harness that runs canonical operations across distributions and architectures, captures cryptographic evidence, and gates releases on byte-identical output.
Replacing the Burger-Dybvig multiprecision float formatter with the Schubfach fixed-width algorithm: same ECMA-262 conformance contract, same output bytes, fewer allocations, higher throughput on number-dense workloads. Validated against 286,362 oracle test vectors.
Fused-Multiply-Add instructions produce IEEE 754 compliant results that differ across platforms. What this means for Go programs that depend on identical floating-point output, and what the available approaches to platform-independent digit generation look like.
Lattice Substrate builds foundational infrastructure libraries where correctness is the product. We focus on deterministic data processing, standards-grade conformance tooling, and engineering practices that produce evidence of correctness, not just confidence. github.com/lattice-substrate