Rebuild spec: TypeScript vs Rust for the edge worker, measured

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

REBUILD.md

@@ -99,10 +99,19 @@ dropping garbage; 28 TB of compressed overview traffic. Throughput goal:
    settles).
 
 2. **Edge worker** (node, TypeScript reusing `lib/`). Measured (`edge:bench`,
-   2026-10-09): 73 k headers/s per core on an a.b.ath range from 2013,
+   2026-10-09, after two rounds of work): ~200–215 k headers/s per core on
-   103 k/s on a recent a.b.boneless range (2 M headers each), ~1.2 GB peak
+   recent a.b.boneless, ~320–340 k on a 2013 a.b.ath range, ~0.7 GB peak
-   per process; the 25 B lines/day goal needs ~3–4 cores over all nodes,
+   per process. **Rust** (`tools/edge-bench-rs`, the same RangeIndexer path
-   so no Rust. Per lease: compressed chunks of 10–50 k lines, in order →
+   over the same captures; `check.ts` finds every output identical, the
+   decoded payloads included): 3.5–3.8× less CPU (boneless 2.9 vs 11.1
+   CPU-s per 2 M headers, ath 2.15 vs 7.5) and a third of the memory
+   (0.20–0.24 GB). Its gains: no GC, hand-written matching instead of the
+   regexes, mimalloc, libdeflate (half of zlib's inflate time) and SIMD line
+   scanning. At the full target (~1.25 M/s) that is ~1.5–2 worker cores
+   instead of ~6–7; the speed is still set by the NNTP connections. The
+   cost: a second implementation of lib/headers, lib/release, lib/rules and
+   the chunk format, kept identical (check.ts) for as long as both exist.
+   Per lease: compressed chunks of 10–50 k lines, in order →
    parse subjects → skip crossposts outside their owner group (of the
    indexed groups in the article's Xref, the one with the lowest priority
    number; see the group list) → group articles into files → files