A ledger-carrying reader audit, tested by a skeptic, set what each chapter
fixed. Each chapter then had a narrative pass, an independent review, a
scope and coverage pass against its concept map, a disclosure and
teach-back check, a senior audit, and a check for lost teaching or
overstated claims, which restored what the passes had cut. A volume pass
reviewed seams and first use.
The author then reviewed the 123 high-risk changes. Nine passages where a
caveat or reason was lost, and two where one sentence returns beside the
rewrite, are restored: the ch02 failure-exposure takeaway, three ch03
caveats (sharded gradients, the two oversubscription ratios, rail hop
limits), the ch04 warm-cache naming, the ch09 MFU source, the ch11
consensus correction, two ch13 caveats (attestation, PUF prediction), and
two ch14 reasons (label delay, feature squeezing).
Files: the chapter .qmd for ch01-ch17; concept maps for ch03, ch05, ch06,
ch07, ch10, ch13, ch15 (quotes kept verbatim with the new prose); and
taxonomy/coverage-ledger.yml (one quote follows a ch08 sentence that now
ends in a cross-reference).
METHOD.md marks each applicable clause met, partial, or open, including the
departures: no expert committee, no human indexing trial, and model-assisted
decisions the standard assumes people make.
- Two audits applied the all-and-some test to 127 specializes edges; the 27
both failed are retyped or removed, and specializes now exports as
skos:broader while part-of stays associative under Z39.19 8.3.3.2
- Compiler enforces SKOS S27; tests validate the export
- Concepts record added_in and computed warrant counts
- Release adds an OWL relations vocabulary, DCAT metadata, CITATION.cff,
scope notes, and redirect pages for identifiers
- Agreement reports Hooper and Rolling consistency
878 concepts (+71 net), 2,020 relationships (+326, -25, 25 reworded), 27
redefinitions, 67 alias sets, 33 background tags, 6 moves; guides 16 roles,
24 symptoms, 7 paths; gaps: attention-FFN disaggregation closed, verifiable
ML computation and reasoning-trace monitoring declared. Coverage judgments
updated for the new concepts and a blind held-out checklist added
(OSDI/NSDI/EuroSys/ASPLOS 2026, 202 papers, 22.3% covered, 96.5% homed).
Symptoms now declare the quantity they are measured by (`quantities`), so a
symptom's own metric no longer poses as a cause; the compiler, the StaffML
loader, and the explorer accept a check or lever that touches a cause or the
quantity and reject a quantity listed in another bucket.
README and CHANGELOG record the method, the numbers, and the open author
decisions.
Model swapping and residency (Aegaeon, CoServe) and KV-cache memory
allocation (vAttention, Jenga; paged KV becomes its block-based form),
checked by gpt-6.1-sol, gemini-4-argon-high, and claude-opus-5-5. The
2025 systems-venue set rises to 63.0% covered (85 of 135).
Concepts that point into a specific book section rise from 86 to 190 of
805 (221 links). Each link was proposed by gpt-6.1-sol and claude-opus-5-5
independently and read before recording; four agreed links were dropped
as neighbors rather than what the passage teaches.
Vijay Janapa Reddi