docs: expand feature/go coverage
- Architecture: full feature/go section — system layering, package
layout, data-structure footprint at ft12 MF=3 scale, 10 key design
decisions (strong refs, uintptr slots, seqlock op cache, fingerprint
validation, 64-bit hash as full key, open-addressed shards, lock-free
slab, global SatCount memo, bdd.Diff fusion, goroutine spawn model)
- Performance: add n-queens head-to-head table (feature/go 14-22 % faster
than feature/c at N=10..13)
- Building-and-Running: expand feature/go section with Go-native
nqueens workflow, pprof recipe, .so swap procedure for sre-ndd A/B,
pointer to in-repo benchmarking.md
- JNI-and-Java-API: document feature/go JNI auto-tuning
(SpawnPairThreshold=1024, CacheClearInterval off, ignored slab
fields), handle representation, pprof tag
docs: label experiments by branch; add feature/go coverage
- Home: split feature/c and feature/go status
- Performance: add 2026-04-24 feature/go v1.18 vs feature/c SRE sweep
(full topology x MF table, observations); tag existing nqueens data
with feature/c
- Architecture / Building-and-Running / JNI-and-Java-API /
Configuration-Reference / Visualization: scope notes distinguishing
the two branches; feature/go build section; feature/go field mapping
- ft12 MF=3 C baseline refreshed (3-run median 578.29 s) — Go 2.92x
faster
docs: rewrite wiki for submodule layout + 2026-04-22 perf data