:orphan:

.. include:: /_templates/include-file/page-prefix.rst

.. meta::
   :description: LLog of the MMv3 revision session for b12-syseng (systems engineering paper), documenting all changes made in response to the adversarial review.
   :keywords: e7Day, llog, systems engineering, MMv3, revision, adversarial review, OSCR, OKO, Jubilee, WoLC

.. note:: **LLog: MMv3 Revision of b12-syseng (2026m04d05).**
   Session by: Claude Opus 4.6 (``dv_ClaOp46_MMv3_syseng_llog_2026m04d05``).
   Date: 2026m04d05.
   This llog documents the revision from MMv2 to MMv3 of the systems
   engineering paper (b12-syseng), integrating all 12 issues from the
   adversarial review.


*****************************************************************************
LLog: b12-syseng MMv3 Revision (2026m04d05)
*****************************************************************************


.. contents:: LLog Contents
   :depth: 2
   :local:


----


.. container:: verbatim-prompt

   Prompt (2026m04d05): Revise the b12-syseng paper from MMv2 to MMv3 by
   integrating all accepted feedback from the adversarial review and author
   reply. All DISCUSS items have been resolved by LLoL.

   Step 1: Read These Files
   =========================

   Read in this order:

   1. ``.claude/CLAUDE.md`` --- project rules, language rules, EDEN system.
   2. ``source/matheology/hell/ll/study/b/12/review_b12-syseng_2026m04d05.rst``
      --- the adversarial review (12 issues, 3 S3 + 5 S2 + 4 S1).
   3. ``source/matheology/hell/ll/study/b/12/reply_b12-syseng_2026m04d05.rst``
      --- the author reply with all decisions resolved. **This is your
      primary instruction set.**
   4. ``source/matheology/hell/mm/b/12/mmv2/b12-syseng_2026m04d05.rst``
      --- the current MMv2 paper you are revising.
   5. ``source/matheology/hell/mm/b/12/mmv3/b12-math_mmv3_2026m04d05.rst``
      --- the MMv3 math paper (reference for updated axiom numbering).

   Step 2: Revision Actions (by priority)
   ========================================

   **HIGH PRIORITY (the 3 Serious issues):**

   1. **S3-1: Sharpen the OKO pattern.**
      - Replace "never declare OK" framing with the three-level distinction:
        Operational OK, Architectural OKO, Existential OK.
      - Add a paragraph acknowledging regulatory/contractual contexts.
      - Reframe ADR-with-review-date as meta-level OKO layered on operational OK.
        Key distinction: whether an operational OK is then treated as the final
        word and data collection stops.

   2. **S3-2: Fix the 30% UMP threshold.**
      - Drop the Shannon derivation claim.
      - Keep Shannon for qualitative insight.
      - Present 30% as conservative engineering heuristic with industry benchmarks.

   3. **S3-3: State OSCR domain boundaries.**
      - Add "Domain and Boundaries" subsection.
      - Include Boeing 737 MAX, Knight Capital (fitting cases).
      - Include Therac-25 (non-fitting case).
      - Acknowledge CrowdStrike product-vs-pipeline distinction.

   **MEDIUM PRIORITY (the 5 Moderate issues):**

   4. **S2-1:** Relabel 6:1 as Schelling point.
   5. **S2-2:** Restructure OSCR detection indicators by measurability.
   6. **S2-3:** Develop maturity model assessment questionnaire.
   7. **S2-4:** Add case studies (subsumed by S3-3).
   8. **S2-5:** Honest labeling throughout.

   **LOW PRIORITY (the 4 Minor issues):**

   9.  **S1-1:** Expand Dual-Nothing to 2--3 paragraphs.
   10. **S1-2:** Focus Tuckman on Storming = EQUAL.
   11. **S1-3:** Add references: Meadows, Perrow, Senge, Leveson.
   12. **S1-4:** Develop Luhmann/autopoiesis connection.

   Step 3: Constraints
   ====================
   - MMv3 axiom numbering (m0.ax0, not m0.ax1).
   - Citations: [Matheo-2-m]_ for math paper.
   - Guarded sections: do not modify.
   - RST quality: clean, mmv3- prefixed labels.
   - Audience: systems engineers, software architects, organizational designers.
   - Assess audience and word counts BEFORE writing.

   Step 4: Output
   ===============
   Save at: ``source/matheology/hell/mm/b/12/mmv3/b12-syseng_mmv3_2026m04d05.rst``
   Create llog at: ``source/matheology/hell/ll/study/b/12/study_ll_2026m04d05_b12-syseng-mmv3-revision-llog.rst``


----


Pre-Writing Assessment
========================

**Audience:** Systems engineers, software architects, organizational
designers. Engineering literacy assumed. Mathematical and theological
concepts defined on first use.

**Word count assessment:** The MMv2 paper is approximately 5,000 words
(748 lines of RST). The MMv3 revision adds:

- S3-1 (OKO sharpening): ~300 words added (three-level distinction +
  regulatory paragraph + data-collection insight)
- S3-2 (UMP threshold): ~200 words net change (relabeling + benchmarks)
- S3-3 (OSCR boundaries + case studies): ~800 words added (new subsection
  with Boeing, Knight Capital, Therac-25, CrowdStrike)
- S2-1 (Schelling point): ~200 words added (benchmark table + rationale)
- S2-2 (OSCR indicators restructured): ~400 words added (three tiers +
  seven alternatives + false-positive discussion)
- S2-3 (assessment questionnaire): ~600 words added (two-test framework +
  7-level questionnaire table)
- S1-1 (Dual-Nothing expanded): ~200 words added
- S1-2 (Tuckman focused): ~50 words net change
- S1-3 (references): ~100 words added (5 new references)
- S1-4 (Luhmann expanded): ~150 words added

**Estimated total:** ~7,500--8,000 words. This is ~50--60% growth over
MMv2. The growth is justified by the review's demand for case studies,
assessment instruments, and honest grounding --- all of which require
text. The paper remains within reasonable bounds for an engineering
audience (comparable to a long conference paper or short journal article).

**Flag:** The paper is at the upper end of comfortable length. If a
further revision adds more case studies or develops the questionnaire
further, consider splitting the assessment framework into a companion
document.


----


Revision Actions Taken
========================


S3-1: OKO Pattern Sharpened
------------------------------

**Section affected:** 4.1 (The OKO Pattern)

**Changes:**

1. Replaced "Never declare OK" framing entirely with the three-level
   distinction:

   - **Operational OK:** Cleared for production within bounded scope.
     Necessary. Compatible with OKO.
   - **Architectural OKO:** Adequate for current requirements, review
     scheduled. The useful pattern.
   - **Existential OK:** Fine, no further review needed. The BABL
     pattern.

2. Added explicit acknowledgment of regulatory contexts (FDA 510(k),
   SOX Section 302, ISO 9001) where operational OK is required.

3. Added the meta-level insight: the relationship between operational OK
   and architectural OKO is *layering*, not contradiction.

4. Added the data-collection distinction per LLoL's instruction: a
   pharmaceutical company that gets FDA clearance and stops collecting
   data has converted operational OK into existential OK; one that
   continues post-market surveillance has layered architectural OKO on
   top. The regulatory approval is the same; the epistemic commitment is
   different.

5. Retained the ADR-with-review-date implementation as the primary
   example, now explicitly framed as meta-level OKO.


S3-2: UMP Threshold Relabeled
---------------------------------

**Sections affected:** 2.6 (CARE), 4.4 (UMP Monitoring Pattern)

**Changes:**

1. Dropped the Shannon derivation claim for the 30% number throughout.
2. Kept Shannon for the qualitative insight: alert fatigue is
   information-theoretic, not a people problem.
3. Presented 30% as a **conservative engineering heuristic**, citing:

   - Google SRE: <50% non-actionable
   - PagerDuty: degradation measurable above 50%
   - Paper's 30% is more conservative: "act before reaching the 50%
     level"

4. Acknowledged that the exact threshold requires per-organization
   calibration.
5. Added the Beyer et al. (2016) SRE book to the reference list.


S3-3: OSCR Domain Boundaries Stated
---------------------------------------

**Section added:** 3.2 (OSCR: Domain and Boundaries)

**Changes:**

1. Added new Section 3.2 with explicit boundary statement: "OSCR models
   progressive systemic degradation through self-assessment failure. It
   does not model design-time defects, latent single-point bugs, or
   acute update failures."

2. Included two fitting case studies:

   - **Boeing 737 MAX:** Textbook OSCR through all three stages.
   - **Knight Capital:** OSCR in deployment infrastructure.

3. Included one non-fitting case:

   - **Therac-25:** Single design-time defect, not progressive drift.
     Pointed to Leveson (2011) as the appropriate framework.

4. Acknowledged the CrowdStrike product-vs-pipeline distinction:
   OSCR may apply to the update pipeline even when the product was
   healthy.

5. Updated the Abstract to include the boundary statement.


S2-1: 6:1 Relabeled as Schelling Point
------------------------------------------

**Sections affected:** 1.1 (WoLC table), 2.8 (TRUST), 4.2 (Jubilee
Pattern), 6 (Conclusion)

**Changes:**

1. Dropped "constrained optimum" language entirely throughout.
2. Relabeled 6:1 as a Schelling point with three properties: culturally
   resonant, memorable, resistant to erosion under feature pressure.
3. Acknowledged the Biblical origin explicitly and intentionally.
4. Added the industry benchmark table (Google 20%, Spotify hack weeks,
   typical tech-debt sprints).
5. Noted empirical calibration is future work.


S2-2: OSCR Indicators Restructured
--------------------------------------

**Section affected:** 4.3 (OSCR Detection Pattern)

**Changes:**

1. Restructured indicators into three tiers:

   - **Primary (measurable):** configuration surface growth, onboarding
     time.
   - **Secondary (partially measurable):** hotfix-to-feature ratio,
     exception handler trends, cyclomatic complexity trends.
   - **Cultural signals (qualitative):** "works on my machine,"
     scope creep language, fossilized workarounds.

2. Added the reviewer's seven alternative indicators from Grey Meadow #2
   as "additional indicators for future testing."
3. Acknowledged the false-positive problem explicitly with estimated
   rates.
4. Added baseline guidance.
5. Framed OSCR as a general pattern; detection as a toolkit of
   analytical + heuristic methods (per LLoL's resolution).


S2-3: Assessment Questionnaire Added
----------------------------------------

**Section added:** 5.3 (Assessment Questionnaire)

**Changes:**

1. Acknowledged the three gaps (no assessment instrument, no transition
   guidance, no empirical calibration) in Section 5.2.

2. Added a preliminary assessment questionnaire based on the
   life-trifecta / death-trifecta framework:

   - Life-trifecta: "Is this ultimately gentle, kind, and reasonable
     --- all three, for all sides, for all time?"
   - Death-trifecta: "Does this over-Simplify, over-Complicate, or
     over-Reach?"

3. Applied both tests at each WoLC level (Levels 1--7) in a table
   format.

4. Added the reviewer's comparison table (CMMI, DORA, Westrum, Spotify)
   as Section 5.2.

5. Added scoring guidance: diagnostic instrument, not certification.
   Death-trifecta on Levels 6--7 = highest risk.


S2-4: Case Studies (Subsumed by S3-3)
-----------------------------------------

Handled entirely within S3-3 above.


S2-5: Honest Labeling Throughout
------------------------------------

**Sections affected:** Throughout

**Changes:**

1. All quantitative claims are now either:
   (a) derived from cited theorems with explicit derivation, or
   (b) labeled as engineering heuristics with empirical grounding cited.

2. Specifically: 30% UMP threshold labeled as heuristic (S3-2), 6:1
   ratio labeled as Schelling point (S2-1), Shannon connection labeled
   as qualitative (throughout).


S1-1: Dual-Nothing Expanded
-------------------------------

**Section affected:** 3.5 (Dual-Nothing)

**Changes:**

Expanded from 1 paragraph to 3 paragraphs with concrete engineering
examples:

- VOID-nothing: maximally dangerous. No defined requirements, undirected
  action, the greenfield project with no spec.
- TRUST-nothing: maximally stable. Null aggregation, directed
  non-action, the maintenance window.
- The asymmetry: VOID-nothing is the problem; TRUST-nothing is part
  of the solution. Start by scoping (m1), finish by resting (m7).


S1-2: Tuckman Focused on Storming = EQUAL
---------------------------------------------

**Sections affected:** 1.2 (Connections), 2.3 (EQUAL)

**Changes:**

1. Focused the Tuckman discussion on the Storming = EQUAL parallel as
   the central (and strongest) claim.
2. Acknowledged explicitly that Tuckman's "performing" has no WoLC
   equivalent and WoLC Levels 3--5 have no Tuckman equivalents.
3. Characterized the broader mapping as approximate.


S1-3: References Expanded
-----------------------------

**Section affected:** References

**Changes:**

Added 5 new references (total now 10, up from 5):

- Beyer et al. (2016). *Site Reliability Engineering.*
- Meadows (2008). *Thinking in Systems.*
- Perrow (1984). *Normal Accidents.*
- Senge (1990). *The Fifth Discipline.*
- Leveson (2011). *Engineering a Safer World.*

All are engaged substantively in the text, not just listed.


S1-4: Luhmann/Autopoiesis Developed
---------------------------------------

**Section affected:** 1.2 (Connections)

**Changes:**

Expanded from 1 sentence to a full paragraph connecting Luhmann's
framework to m5.ax1:

- Luhmann's operationally closed / structurally coupled systems map
  to m5.ax1's conditional-data machines (self-manage = operational
  closure; UMP noise threshold = structural coupling).
- Key shared insight: self-reproduction is necessary but not sufficient;
  without the self-correction mechanism (m6), the system can reproduce
  itself into a corner.


----


Technical Notes
=================

**Label convention:** All labels use ``mmv3-se-`` prefix (e.g.,
``mmv3-se-abstract``, ``mmv3-se-sec3-2``).

**Axiom numbering:** Uses MMv3 numbering from b12-math_mmv3: m0.ax0
(not m0.ax1), m6.ax5 (Environmental Novelty), th5 derivation chain
(m2.ax2 + m6.ax5 + m5.ax2 + th3).

**Citation convention:** References use version-suffixed keys
(``Ashby1956-sev3``, etc.) to avoid cross-paper collisions.
Cross-references to the math paper use ``:ref:`` with display name
``[Matheo-2-math]``.

**No guarded sections** in the MMv2 paper, so no guarded-content
constraints applied.

**RST quality:** Clean RST throughout. No indentation errors. All
list-tables properly formatted. All cross-references use ``:ref:``
with descriptive display names.


----


Summary and Recommendations
==============================

**All 12 review issues addressed in MMv3:**

- 3 S3 (Serious): OKO sharpened with three-level distinction, UMP
  threshold relabeled as heuristic, OSCR domain boundaries stated with
  case studies.
- 5 S2 (Moderate): 6:1 relabeled as Schelling point, OSCR indicators
  restructured by measurability, assessment questionnaire added,
  case studies integrated (via S3-3), honest labeling applied throughout.
- 4 S1 (Minor): Dual-Nothing expanded, Tuckman focused, references
  expanded, Luhmann developed.

**Paper growth:** ~5,000 words (MMv2) |rarr| ~7,500--8,000 words (MMv3).
Growth justified by case studies, assessment questionnaire, and industry
benchmarks.

**EDEN assessment of revision:** Green Meadow (count = 3 remaining
improvement paths):

1. Develop the assessment questionnaire into a standalone instrument
   with scoring rubric and worked examples.
2. Add 1--2 more case studies (Healthcare.gov fits OSCR well; Heartbleed
   illustrates boundaries).
3. Empirical calibration: pilot-test the WoLC assessment with willing
   organizations.

**Recommendation for next steps:**

- LLoL should review the three-level OKO distinction (Section 4.1)
  carefully --- this is the highest-impact change and the most
  conceptually delicate.
- The data-collection distinction (operational OK as waypoint vs.
  destination) is new material from LLoL's prompt that was not in the
  review or reply. LLoL should check that the formulation captures the
  intended insight.
- The assessment questionnaire (Section 5.3) is preliminary. LLoL should
  evaluate whether the life-trifecta / death-trifecta questions capture
  real organizational dynamics or are too abstract for the target
  audience.

**No BABL dangers detected in this revision.** The primary risk (dressing
heuristics as derivations) was the review's central finding and has been
addressed by honest labeling throughout.


----


MMv3r1 Revision (2026m04d06)
================================

.. container:: verbatim-prompt

   Prompt (2026m04d06): [LLoL provides extensive structural feedback on
   MMv3, addressing multiple issues. Full prompt preserved in conversation
   context. Key instructions summarized below.]

   - **BABL-before-ZION ordering:** ALWAYS mention BABL first, then ZION.
     Death-trifecta first, life-trifecta second. (0) BABL default, (1)
     ZION escape. Within life-trifecta: reasonable first, kind second,
     gentle third.

   - **Shabbat not Jubilee:** The 6:1 cycle is Shabbat. Jubilee is the
     larger 7×7+1=50 multi-scale reset. Fix throughout.

   - **Section 4.1 rewrite:** Replace static three-level OKO distinction
     with dynamic ZION cycle (seed-feed-grow-reap). Use OK vs NOT OK, not
     OK vs OKO. Frame as a living system (crop in a field), not a static
     artifact with a certificate. Each cycle yields operational adequacy
     ("good enough for this harvest"); stopping the cycle is BABL.

   - **Section 5.3 table restructure:**
     - Add Level 0 (VOID) --- the null hypothesis, wicked problems, past
       failures. Ignoring the structure of VOID is itself over-simplification.
     - Death-trifecta column FIRST (gravest concerns first).
     - Within life-trifecta: reasonable → kind → gentle.
     - Each cell: 3 rows, each starting with keyword spelled out.
     - 3-fold table header: (0) death / (1) life; Tree of Knowledge-faking
       / Tree of Life-giving; BABL / ZION acronyms.
     - TYPE overreach: address systemic risk --- the assumption that
       "doesn't eventually destroy the world" is being tested for when
       systemic risks are routinely forgotten.
     - Add wicked problems (Rittel & Webber 1973) and super-wicked problems
       (Levin et al. 2012) references.

   - **Paper length:** Up to 10K OK if needed for clarity. Cut fluff even
     below 5K. Clarity and Truth over brevity.

   - **Word changes:** Batch them; tell Claude at end of review or save in
     a text file.


MMv3r1 Revision Actions Taken
--------------------------------


R1-1: BABL-Before-ZION Ordering Enforced
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Sections affected:** Throughout entire paper.

**Changes:**

1. All BABL/ZION pairs now present BABL first, ZION second.
2. Abstract rewritten: "The default state is BABL... The narrow escape
   is ZION..."
3. Section 2.7 (HOPE): governance options now numbered (0) BABL, (1) ZION.
4. Section 5.3: death-trifecta column moved to the left (first), life-
   trifecta column to the right (second).
5. Section 6 (Conclusion): point 3 now reads "(0) OK = BABL (death by
   default). (1) NOT OK = ZION prerequisite (narrow escape through active
   cycling)."
6. New Language Rule 7 added to CLAUDE.md codifying this ordering.


R1-2: Shabbat/Jubilee Distinction Corrected
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Sections affected:** 1.1 WoLC table, 2.3, 2.8, 4.2, 5.3, 5.4, 6.

**Changes:**

1. "Jubilee Pattern" renamed to "Shabbat Pattern" throughout.
2. Section 4.2 title changed to "The Shabbat Pattern."
3. All "e7Day Jubilee" references changed to "e7Day Shabbat."
4. Clarified: "The Shabbat cycle is the smallest unit in the Jubilee
   System's multi-scale framework" (Section 2.8, 4.2).
5. New Language Rule 8 added to CLAUDE.md codifying this distinction.


R1-3: Section 4.1 Rewritten (Dynamic ZION Cycle)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Section affected:** 4.1 (renamed from "The OKO Pattern" to "The OK vs
NOT OK Pattern").

**Changes:**

1. Replaced static three-level OKO distinction entirely.
2. New framing: engineered system as living crop, not static artifact.
3. Introduced ZION cycle structure (seed/feed/grow/reap =
   Zoning/Investigating/Organizing/Navigating) as the correction
   mechanism.
4. Each cycle yields "operational adequacy" (good enough for this harvest).
   This is the legitimate go decision that regulations require.
5. BABL pattern = stopping the cycle. Treating the harvest as permanent.
6. ZION pattern = keep cycling. Treat each harvest as waypoint, not
   destination.
7. Pharmaceutical example retained: the distinction is whether data
   collection continues after regulatory clearance.
8. ADR implementation retained, now framed as "seed of the next ZION
   cycle."
9. OK vs NOT OK distinction replaces OK vs OKO throughout.


R1-4: Section 5.3 Table Restructured
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Section affected:** 5.3 (renamed to "Death-Trifecta / Life-Trifecta").

**Changes:**

1. **Level 0 (VOID) added.** Death-trifecta: treating wicked/super-wicked
   problems as tame; analysis paralysis; claiming full understanding.
   Life-trifecta: mapping unknowns before scoping; hearing all parties;
   pacing the VOID→TYPE transition.

2. **Death-trifecta column moved to first position** (left column).
   Life-trifecta column second (right column).

3. **3-fold table header:** Row 1: (0) Death-trifecta / (1) Life-trifecta.
   Row 2: Tree of Knowledge-faking / Tree of Life-giving. Row 3: BABL
   acronym / ZION acronym.

4. **Each cell restructured to 3 keyword rows:** over-Simplify,
   over-Complicate, over-Reach (death-trifecta) and Reasonable, Kind,
   Gentle (life-trifecta).

5. **Life-trifecta order corrected:** reasonable → kind → gentle (not
   gentle → kind → reasonable).

6. **TYPE overreach sharpened:** "Claiming the scope covers systemic risks
   it was never tested against. This is among the gravest patterns in
   existence: the implicit assumption that 'doesn't eventually destroy
   the world' is being tested for, when in fact systemic risks are
   routinely forgotten until they become system-threatening."

7. **Wicked problems references added:** Rittel & Webber (1973) for wicked
   problems, Levin et al. (2012) for super-wicked problems.

8. **Scoring guidance updated:** Most dangerous findings at Level 0 (VOID),
   Level 1 (TYPE), and Level 6 (HOPE).


R1-5: Wicked Problems and VOID Expanded
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Sections affected:** Abstract, 1.2 (Connections), 2.1 (VOID), 6
(Conclusion).

**Changes:**

1. VOID (Section 2.1) expanded with wicked/super-wicked problem framing.
2. New connection added in Section 1.2: Rittel & Webber (1973) and Levin
   et al. (2012).
3. Conclusion point 7 added: "VOID is not empty."
4. Two new references added.


R1-6: OK vs NOT OK Language Throughout
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

**Sections affected:** Throughout.

**Changes:**

1. "OKO self-assessment" replaced with "NOT OK self-assessment" where
   it referred to the dynamic correction state.
2. "OK/OKO bifurcation" replaced with "OK vs NOT OK bifurcation."
3. Senge paragraph updated: "NOT-OK self-assessment" replaces "OKO
   self-assessment."
4. Westrum comparison updated: "Generative ≈ NOT OK (actively
   correcting)."


----


Summary After MMv3r1
======================

**All structural feedback from LLoL integrated.** Key changes:

- Dynamic ZION cycle replaces static three-level OKO distinction
- BABL-before-ZION ordering enforced throughout
- Shabbat/Jubilee distinction corrected
- Assessment table restructured with VOID, death-first, 3-keyword rows
- Wicked problems and super-wicked problems integrated into VOID
- OK vs NOT OK replaces OK vs OKO
- 2 new references (Rittel & Webber 1973, Levin et al. 2012)

**Paper length:** Estimated ~8,500 words. Within LLoL's 10K ceiling.

**EDEN assessment of MMv3r1:** Green Meadow (count = 3 remaining paths):

1. Further develop the e7Ch connection (ZION cycle stages) once that
   paper exists.
2. Add Healthcare.gov as a third fitting case study if space permits.
3. Empirical pilot-test of the death-trifecta / life-trifecta
   questionnaire.

**BABL self-check on this revision:** The MMv3 version presented
death-trifecta/life-trifecta in the wrong order (life first), which
subtly biased toward comfort over safety. LLoL correctly identified this
as a BABL pattern: "gentle for the powerful, forget BABL over-reach
against the unconsidered." The correction (death-trifecta first) is not
just a formatting preference --- it is a structural requirement of the
framework itself. The framework that warns about over-simplification
must not over-simplify its own presentation by putting comfort before
safety.