# charlespetrie.com: local roadmap

_Cluster: **fan-tribute-site** (computing history). Generated 2026-09-14. noindex._

## Current state (real signals, 2026-09-14)

- New domain (registered 2026-09-14), new build. Every fact is from his own Stanford pages and CV, his IEEE IC columns, IEEE Spectrum 2018, Stanford News 2007, TED and Wikipedia; see /sources/ and AGENTS.md.
- 17 indexable pages: home, biography (31-entry timeline), 5 idea essays, publications (5 books, 30 papers, the full Peering column list parsed from his archive), /classic/ (three tag-for-tag archives of his original pages with dead links routed to the Wayback Machine), videos (2 verified talks + 7 context), links, FAQ, sources. JSON-LD: Person, WebSite, Article, ItemList of Book and VideoObject, FAQPage, BreadcrumbList, WebPage (archives). Markdown twin of every page, llms.txt, AGENTS.md, api/person.json.
- Not verifiable yet, so not on the site: Scholar citation counts, any award, any recording of him beyond two talks.
- AME: analyzed + activated as a workspace (propose-only; nothing auto-ships).

## Upgrade checklist

- ✅ Designed font system, four themes incl. Burnt Orange dark + day, mobile-first
- ✅ Fan-site disclaimer + cross-link row on every page (footer, and the archive notes on /classic/)
- ✅ /classic/ archive exhibit with link check and Wayback routing
- ✅ /videos/ with verified-vs-context labeling, /sources/ with photo credits
- ⬜ Markdown content negotiation in the vhost (Accept: text/markdown -> index.md) once the vhost is edited (ask first)
- ⬜ A note to Dr. Petrie (via LinkedIn) offering corrections and asking whether he would like anything added or removed
- ⬜ Wayback snapshots of his Stanford pages requested (save-page-now) so the archive links have a 2026 capture too

## Staged, additive upgrades (nothing here changes the live site without a snapshot + approval)

### Stage 1: owner contact and corrections
1. Send Dr. Petrie the link with the disclaimer up front; apply any corrections he sends the same day.
2. Ask whether the Google Scholar counts may be quoted (he can export them) and whether any talk recordings exist beyond the two found.
   Prompt: "Read AGENTS.md and _night/charlespetrie/cp_content.py. Apply the owner's corrections, regenerate with gen-cp.py, deploy, verify every page 200."

### Stage 2: depth
3. One essay per Peering column he wrote himself (about 25), each a short reading guide with a link to his PDF, question-shaped titles ("Is Google evil? What Petrie argued in 2010").
4. A page per project: Proteus, Redux, ProcessLink, JATLite, POEM, the Semantic Web Services Challenge, 88NV, each from his own papers.
5. Archive more of his site: /online/ (the 1997-98 WebWord columns), the Redux overview, the collective-work slides, with the same link-check treatment.

### Stage 3: reach
6. Submit the site to the WWW Virtual Library's successor directories and the Internet Archive; ping IndexNow on each new page.
7. Core Web Vitals pass from a Windows box (never a scanner on the droplet).

_Roadmap = the map. Nothing above is applied automatically; AWE is propose-only for this site._