Internet 2.0 — multi-scale connectivity infrastructure analysis and conceptual design.
Coverage is not connectivity.
An address can be marked served and still be unaffordable, fragile, slow, or one cut away from isolation. PACKET scores networks at international, national, regional, or local scale across coverage, capacity, latency, availability, affordability, competition, and path diversity.
Series: Applied Systems
| Audience | Path | Time |
|---|---|---|
| Planner / researcher | SHOWCASE.md → broadband adoption divide | 15–25 min |
| CLI implementer | SHOWCASE.md → cargo run -p packet-cli -- gap --scale regional --corpus corpus |
10–20 min |
Research lab only — not RF design, regulatory filing, carrier build plan, or FCC/NTIA endorsement. Coverage is not connectivity.
PACKET is one domain implementation of a shared evidence-first method:
PUBLIC SOURCES → CORPUS → SCORE → SERVICE PROMISE → GAP MAP
↓
CONCEPT → REVIEW → DESIGN
| Lane | Repositories |
|---|---|
| Movement | ROUTE, GAUGE, TARMAC, HARBOR |
| Lifelines | PYLON, PACKET, BASIN, DRAIN |
| Public access | SHIELD, SLATE |
| Civic boundaries | ZONES |
The family shares evidence labels, explicit scale and demand bases, T1–T4 service promises where meaningful, adversarial review, and acceptance of a rigorous null result. Each repository owns its domain semantics and safety boundary.
PACKET is a research and conceptual-design project. It is not an engineering plan, RF or optical design, regulatory filing, network build plan, or advocacy brief, and it claims no FCC, NTIA, carrier, or standards-body endorsement.
PACKET is public and open to use as a reference model, cited connectivity finding, diagnostic pattern, review discipline, or local adaptation starting point.
PACKET is currently a specialist connectivity-infrastructure analysis product,
not a supported cross-repository library. Its network model, evidence taxonomy,
DIM-01..13 scoring, service tiers, gap policy, and CLI remain product-local; no
portfolio repository pins a packet-* crate or owns compatibility proof.
Infrastructure 2.0 siblings share and adapt an evidence-first method, not a versioned PACKET dependency. Family source similarities, findings, worksheets, and local adaptations are not stable provider contracts. Direct reuse requires a named downstream consumer, a bounded versioned surface, and consumer-owned compatibility tests.
If you want to apply it to a region, broadband market, middle-mile question,
resilience problem, affordability gap, or connectivity service question, start
with docs/adoption/README.md. It lays out safe
reuse, first adaptation steps, contribution targets, and claim boundaries.
Broadband programs can count passings while missing price, reliability, middle-mile, competition, and route diversity. PACKET makes those promises explicit and evaluates them at the scale where the decision is actually being made.
The transferable principle is: availability is a bundle of service promises, not a binary coverage flag.
| Crate | Responsibility |
|---|---|
packet-network |
Connectivity elements, paths, markets, and scale-aware relationships. |
packet-corpus |
Evidence-labelled corpus parsing and validation. |
packet-score |
DIM-01..13 score artifacts. |
packet-tier |
Tier-SLA classification and shortfall reporting. |
packet-gap |
Scale-filtered gap analysis and null-result reporting. |
packet-cli |
CLI front door for corpus, score, tier-SLA, and gap commands. |
The first cited broadband adoption divide analysis reports both tested dimensions with 50% of observations below the declared bar.
That result is scoped to the cited run, not a universal claim about every network or market.
cargo run -p packet-cli -- corpus corpus/us-ca.md
cargo run -p packet-cli -- gap --scale regional --corpus corpus
cargo test --workspaceCORPUS -> SCORE -> TIER-SLA -> GAP -> CONCEPT -> REVIEW -> DESIGN
Every element carries a scale and market. Cross-scale comparisons must say so, and a rigorous null result remains valid.
PRODUCT_PLAN.md— scope, product shape, and next work.docs/adoption/— open reuse, local adaptation, and review path.docs/vtrace/— VTRACE requirements, architecture, trace, and verification.context/waves/— repo-local execution history..roles/ROLE.md— adversarial review panel.
MIT. See LICENSE.