Archival Research Report DuckDB Datalake Hydrated Hister Indexed October 02, 2026

Archaeological Cartography:
softwarecraftsmanship.org & The Signatory Ledger

Forensic timeline reconstruction, technical stack decomposition, and archival comparison of the 17-year living monument of the Software Craftsmanship movement (2009–2026).

#106

Mike Hall Canonical ID

March 6, 2009 (7:06 PM CST)

37,842

Latest Lifetime ID

Signed October 2, 2026

27,601

Active Confirmed Ledger

307 API Pages Ingested

28.6%

Link Decay Rate

10 of 35 Curated Links Dead

1. Canonical Signature Verification

While the public web application renders unnumbered pagination tables, the internal search API preserves immutable primary keys and exact ISO-8601 UTC timestamps:

// GET https://manifesto.softwarecraftsmanship.org/signatories/search?name=Hall
{
  "id": 106,
  "number": 106,
  "name": "Michael D. Hall",
  "location": "Crystal Lake, IL",
  "signed_on": "2009-03-07T01:06:59.000Z"
}
    

Local Creation Time: Friday, March 6, 2009 at 7:06:59 PM CST (Central Standard Time).
Canonical Signatory Rank: 106th signatory in the immutable database ledger.
Registered Identity: Signed as Michael D. Hall, situated in Crystal Lake, IL.

2. Forensic Timeline: The Mystery of Positions #106, #102, and #101

A comparative audit between the earliest Internet Archive Wayback snapshot (March 10, 2009) and the 2026 live database answers why different surfaces reported different ranks over time:

The Core Findings:

The Double-Submission Bug of 2009: In the initial March 10, 2009 HTML page, early signatory Cory Foy (Tampa, FL) was accidentally rendered twice (occupying both row #28 and row #32). This shifted all subsequent signatories back by +1. Thus, Mike Hall was visually displayed at row #102.
Asynchronous Email Confirmation Gap: When Mike signed at 7:06 PM on March 6, database auto-increment assigned id: 106. Submissions #101 through #105 were sitting in the database in a pending transactional state awaiting email confirmation tokens.
The SPA Consolidation (2011–2014): When 8th Light rebuilt the site as a Backbone.js SPA, the duplicate Cory Foy entry at #28 was purged and consolidated at #31. The removal of this duplicate shifted all subsequent rows forward by -1 (Joseph Leddy from #29→#28, Nathaniel Talbott from #101→#100, and Michael D. Hall from #102→#101).

Archival Drift Analysis Table

Signatory	2009 Snapshot Rank	2026 Table Rank	Canonical DB ID	Forensic Cause / Notes
Doug Bradbury	#1	#1	1	Primary Author / 8th Light Founder
Robert C. Martin	#5	#5	5	Signed March 6, 2009 at 20:36 UTC
Cory Foy	#28 & #32	#31	31	Double-Submission: Rendered twice in 2009; deduped in Rails rewrite.
Joseph Leddy	#29	#28	28	Shifted forward by -1 after Cory Foy deduplication.
Nathaniel Talbott	#101	#100	100	Shifted forward by -1 after deduplication.
Michael D. Hall	#102	#101	106	Canonical DB ID 106. Rank 102 in 2009 HTML, Rank 101 in SPA table.
Casey Charlton	#120	Purged	120	Withdrawn signature; deleted via confirmation removal workflow.

3. Technical Stack & Infrastructure Archeology

Inspecting static assets (e.g. javascripts/setup.js) and DNS routes identifies the complete runtime environment:

Hosting & Operator

Owner: 8th Light, Inc.
Staging Seam: scmanifesto-staging.8thlight.com
Hosting: AWS Elastic Beanstalk (us-east-1)
DNS: DNSimple Anycast Nameservers

Runtime Stack

Backend: Ruby on Rails JSON API
Frontend: Backbone.js 1.3.3 + Underscore.js
DOM/Ajax: jQuery 3.6.0 (security bump ~2021)
Metrics Visualization: D3.js v3
CSS Template: Dieter Schneider (2006)

4. Curated Reading List: 17-Year Link Decay Audit

The manifesto site curates 35 canonical reference links across foundational literature, essays, and community resources. A live HTTP probing suite reveals an overall 28.6% decay rate:

Resource Name	Section	HTTP Status	Observed State
The Pragmatic Programmer	Background	200 OK	Amazon book listing active
Apprenticeship Patterns	Background	200 OK	O'Reilly Chimera lab active
Unskilled and Unaware (Dunning-Kruger)	Background	404 Not Found	jerwood-no.org.uk domain dead
What's all this nonsense about Katas?	Food for thought	404 Not Found	Uncle Bob's Google Sites page broken
A Call for Apprenticeship (Dave Hoover)	Food for thought	404 Not Found	redsquirrel.com legacy CGI script dead
The Ultimate Code Kata (Coding Horror)	Food for thought	DNS Err	codinghorror.com link unresolvable
The UK Mailing List	Conferences	404 Not Found	groups.google.co.uk redirect dead

5. Local Datalake & Hister Integration

The complete archaeological dataset is persisted locally in the zdots DuckDB datalake and local search engine:

# Query Mike Hall's canonical record in DuckDB
$ duckdb -light-mode ~/.local/state/zdots/datalake.duckdb \
  "SELECT id, signatory_number, name, signed_on FROM scm_signatories WHERE id = 106;"

# Search local hister index
$ hister search "softwarecraftsmanship"
    

6. Panoramic View Investigation Methodologies

This investigation was conducted using the Panoramic View system-cartography discipline, coordinated by Pavel within the Panoramic View Labs (PVL) initiative on the zdots platform. The methodology applies four rigorous forensic phases to reconstruct truth without speculative assumptions:

Phase I: Non-Destructive Probing

Principle: "Snake in a Can" — Shake before opening. Never hypothesize when deterministic interfaces can be interrogated directly.

Targeted the public REST API across all 307 pages without brute-forcing (retrieved 27,601 records).
Queried DNS authoritative nameservers (DNSimple Anycast) and HTTP response headers without intrusive scans.
Audited raw client-side JavaScript assets (setup.js, manifesto.js) to reveal hidden staging seams (scmanifesto-staging.8thlight.com).

Phase II: Archival Stratigraphy

Principle: Historical Stratigraphy — Treat digital systems as sedimentary layers spanning distinct operational eras.

Interrogated Wayback Machine CDX indices to pinpoint the earliest snapshot (March 10, 2009 at 06:55 UTC).
Parsed the raw 2009 table DOM (1,244 static HTML <td> tags) to establish ground truth at launch.
Dissected the transition from 2009 server-rendered Rails views to ~2011–2014 Backbone.js single-page architecture.

Phase III: Diff & Anomaly Resolution

Principle: The Blink Test — Prove the causal mechanism behind every discrepancy rather than accepting surface readings.

Detected the Cory Foy double-render bug (#28 & #32) in 2009, proving why all subsequent rows were shifted by +1.
Uncovered the difference between visual table rank (ROW_NUMBER() over active confirmed rows) and immutable database primary keys (id: 106).
Discovered the internal search API (/signatories/search) that preserves original timestamps stripped by public pagination.

Phase IV: Datalake & Search Grounding

Principle: Durable Knowledge — Don't leave forensic evidence in ephemeral terminal logs; hydrate structured datalakes.

Hydrated DuckDB (scm_signatories, scm_snapshot_2009, scm_rank_drift, scm_reading_links).
Audited link decay with live timeouts and status code logging (28.6% dead link rate).
Indexed public URLs into the local hister search engine for rapid semantic and lexical query.

Architectural Takeaway: A system is not just what its current front-end renders; it is the sum of its database constraints, historical migrations, transactional confirmation semantics, and infrastructure seams. Panoramic View cartography cuts through 17 years of UI abstraction to anchor verifiable engineering history.