From 41fcdde04d0b9f148290a6ac6ee46abb5239f839 Mon Sep 17 00:00:00 2001 From: Jodson Graves Date: Wed, 16 Sep 2026 12:06:54 -0400 Subject: [PATCH] Carry the residential substrate work into the standard The 2026-09-08 board review raised three physical facts about running substrate on consumer broadband that the document did not address: dynamic addresses behind carrier-grade NAT with acceptable-use policies barring inbound servers, asymmetric and capped uplink, and a host operator who can read the memory of a machine they physically possess. That work was done in the document store and never reached this repository, so the standard has been silent on the residential case while the reference protocol was already built for it. The document gains one paragraph in the Substrate protocol tier: nodes join the market over encrypted overlays and outbound polling, without requiring open inbound ports or a static address. This is the residential case of line 11 rather than a new provision - a substrate that ran only where a provider permits inbound service would carry a chokepoint at every provider, which line 11 forbids. Registered as SUB-no-inbound-requirement, implementation-bound, taking the registry from 25 invariants to 26 and the README counts with it. The seven translations carry the same paragraph and move their source snapshot to 2026-09-08. They do not yet carry the transport fee from #6; the snapshot date says so rather than implying coverage they lack. Bandwidth and confidential execution are not answered by the document. They are argued in P1-004, published here because open question 9 links to it and the link would otherwise dangle - the same reason the concept triage and prior instruments were published in #1. The paper analyzes and never governs; its candidate SUB-redundant-execution is argued, not bound, and reaching the registry is for the board under bylaws 9.16. Open questions gains the residential entry. It lands at 9 by the date order the file has always kept - 2026-09-08 falls between the sybil entry and the transport fee - so the transport fee renumbers to 10. Nothing outside the file's own summary paragraph cited entry 9. Conformance suite: RESULT: all executed checks PASS, exit 0. Signed-off-by: Jodson Graves Co-Authored-By: Claude Opus 5 --- OPEN-QUESTIONS.md | 15 +++- P1-004_Substrate-Constraints_v0.1.md | 123 +++++++++++++++++++++++++++ README.md | 3 +- janus-facing-architecture.ar.md | 4 +- janus-facing-architecture.es.md | 4 +- janus-facing-architecture.fr.md | 4 +- janus-facing-architecture.hi.md | 4 +- janus-facing-architecture.md | 2 + janus-facing-architecture.pt.md | 4 +- janus-facing-architecture.tok.md | 6 +- janus-facing-architecture.zh.md | 4 +- jfa-conformance-suite.py | 5 +- 12 files changed, 165 insertions(+), 13 deletions(-) create mode 100644 P1-004_Substrate-Constraints_v0.1.md diff --git a/OPEN-QUESTIONS.md b/OPEN-QUESTIONS.md index 8cf6e1a..610c0ae 100644 --- a/OPEN-QUESTIONS.md +++ b/OPEN-QUESTIONS.md @@ -2,7 +2,7 @@ The living open-questions document for the Janus Facing Architecture, per the practice carried forward in the [2026-08-24 concept triage](jfa-concept-triage-2026-08-24.md). A stale document here means the project has stopped describing itself honestly. Each entry carries a status and names the constraints it inherits. -Three questions are currently open: readmission after trust suspension (entry 7, resolved in draft pending adoption of the bylaws amendment), sybil resistance in the governance franchise (entry 8), and the pricing of the transport fee (entry 9). The rest were resolved as raised; their record follows. +Four entries carry open business: readmission after trust suspension (entry 7, resolved in draft pending adoption of the bylaws amendment), sybil resistance in the governance franchise (entry 8), residential substrate constraints (entry 9, resolved in draft with its confidential-execution part still open), and the pricing of the transport fee (entry 10). The rest were resolved as raised; their record follows. ## 1. Contestability @@ -75,7 +75,18 @@ Candidate directions, none adopted: witness-attested distinctness at the record What the gap can still reach is the Governance delegate itself, and with it the office of the Vice President and the disposition of expulsion referrals — the powers of Article XII rather than the powers of the membership. The bylaws draft also puts the amendment of the bylaws themselves within prosumer-member reach (bylaws §15.1), which is the one place the gap touches the structure rather than an office; whether that vote is cast directly or channeled through delegates is unsettled in the draft and is the open half of this question. The immediate recall of §6.8 is the standing check, and it is exercised by the same roll a sybil attack would have captured — which is the part that does not resolve itself. **Inherits:** privacy floor (line 7, entry 2); platform-scoped non-portable identifiers (bylaws §9.10(f)); one member, one vote (bylaws §3.4); recallable delegates, one per federation (bylaws §5.3); witness minimum (entry 4). -## 9. Pricing the transport fee + +## 9. Residential substrate constraints + +**Status:** resolved in draft (2026-09-08); part (c) open + +Raised by the Vice President in the board's September 2026 review of the official document: as a governance and system-specification instrument, the document does not address three physical risks of running substrate on consumer broadband — (a) residential connections carry dynamic addresses, sit behind carrier-grade NAT, and come with terms of service that bar inbound servers; (b) home bandwidth is asymmetric and capped, so heavy container layers and large datasets do not move well; (c) a host operator has physical possession of the machine and can read its memory, and consumer hardware carries no confidential-execution feature to prevent it. + +Decided: carrier constraints are treated as a physical, adversarial environment to be routed around in code, not a legal condition to be negotiated away. Lobbying may proceed as a civic matter but is never a dependency of the architecture. Routed at three levels. (a) Resolved in the document — the Substrate protocol tier now states that nodes join the market over encrypted overlays and outbound polling, without requiring open inbound ports or a static address, as the residential case of line 11; registered as SUB-no-inbound-requirement, implementation-bound. The reference protocol already has this shape: its node-side operations poll for work rather than listen. (b) Resolved by design posture in the substrate whitepaper ([P1-004](P1-004_Substrate-Constraints_v0.1.md)): protocol traffic is small by construction under line 7, work ships as sandboxed WebAssembly modules rather than container images, and jobs are spooled locally so a poor uplink delays work rather than losing it. (c) Open: the whitepaper argues for redundant execution across independent hosts with covenant-rated disagreement, plus data minimization, in place of any hardware-enclave requirement, and for attestation as an optional, rated, market-priced capability. Whether that position becomes a registered invariant (a candidate SUB-redundant-execution) is for the board under bylaws §9.16; until then it is argued, not bound. A residual legal question stands with it: whether compensated participation on a residential line engages the non-commercial prong of typical acceptable-use policies, which counsel should review beside the CHC pilot's business-use insurance question. + +**Inherits:** no-chokepoint (line 11); privacy floor (line 7, entry 2); lean, auditable code (P-lean-code); witnessing as compensated substrate work (REC-witness-work, entry 6); document-amendment procedure (bylaws §9.2, §9.16); bootstrap recording (bylaws §16.1). + +## 10. Pricing the transport fee **Status:** open (2026-09-11) diff --git a/P1-004_Substrate-Constraints_v0.1.md b/P1-004_Substrate-Constraints_v0.1.md new file mode 100644 index 0000000..e39ea65 --- /dev/null +++ b/P1-004_Substrate-Constraints_v0.1.md @@ -0,0 +1,123 @@ +# Routed, Not Negotiated: Residential Substrate Under Carrier Constraints + +**Network Theory Applied Research Institute** +Document ID: P1-004 · Version: 0.1 (Draft) · September 2026 + +*Companion to the official document of the Janus Facing Architecture. This paper analyzes; it never governs. Where it recommends a rule, the rule takes effect only by amendment of the official document under bylaws §9.2 and its registry under §9.16.* + +## Abstract + +The Janus Facing Architecture places its substrate layer on consumer hardware in homes, offices and storage. Three physical facts of consumer broadband stand against that: connections carry dynamic addresses behind carrier-grade network address translation, their acceptable-use policies bar inbound servers, and their upstream bandwidth is a fraction of their downstream. A fourth fact stands against the trust model: a host with physical possession of a machine can read its memory, and the confidential-execution features that would prevent this are absent from consumer processors by vendor decision. + +This paper argues that the carrier constraints are best treated as an adversarial physical environment to be routed around in code, not a policy condition to be negotiated legally, and it gives the empirical case for that choice. It then states what routing around them actually requires, separates the parts that technique solves from the parts it does not, and records one legal question that no protocol design can answer. + +## 1. The choice of venue + +There are two ways to answer a constraint imposed by a network carrier. Change the carrier's obligations through law and regulation, or build software that does not need the carrier to change. The architecture's own research says which one to expect results from. + +NTARI's work on democratic information velocity describes a structural mismatch: information and infrastructure move at network speed while democratic synthesis stays locked to electoral cycles (NTARI, 2025a). Acemoglu and Robinson's Red Queen effect names the consequence — when one runner outpaces the other, the corridor is lost (Acemoglu & Robinson, 2019). United States broadband regulation is a clean instance. Two decades of contested rulemaking on network neutrality ended on 2 January 2025, when the Sixth Circuit set aside the Federal Communications Commission's 2024 order in full, holding that the Commission lacked statutory authority to classify broadband as a telecommunications service (Ohio Telecom Association v. FCC, 2025). Relying on Loper Bright, the court removed the deference that had let the rule survive earlier challenges. What remains federally is a disclosure regime: the Commission can require a provider to publish its traffic-management practices and cannot prohibit them. Eight states legislate in the gap, which means a nationally uniform legal answer does not currently exist and a deployment's permissions depend on where its nodes sit. + +That record is not an argument against civic engagement. It is an argument against dependency. A protocol whose viability waits on a favorable rule is a protocol that does not run for years at a time, and the counterparties in that venue are incumbents whose lobbying capacity exceeds this institute's by orders of magnitude. Software written to work under the terms carriers already impose runs today and keeps running whichever way the rule turns. + +**The position.** Carrier constraints are an immutable physical constraint for design purposes. Lobbying may proceed as a civic matter, and this paper takes no position against it, but it is never a dependency of the architecture and no deployment plan may assume its success. + +One caveat belongs here rather than in a footnote. This choice is available for the three constraints below because each has a technical answer. It is not available for every constraint, and the paper's final section names one where it is not. + +## 2. Reachability without a server + +**The constraint.** A residential connection is not a hosting environment. Its address changes, it commonly sits behind carrier-grade network address translation that gives no inbound path at all, and its acceptable-use policy generally bars running servers. Comcast's residential policy is representative: it prohibits equipment that provides "network content or any other services to anyone outside of your Premises LAN, except for your personal and non-commercial residential use," and names web hosting, file sharing and proxy servers as examples (Comcast, 2021). + +**What technique answers.** Everything about reachability. A node that never accepts an inbound connection is unaffected by dynamic addressing, by carrier-grade translation, and by the inbound-server prohibition, because none of those constrain outbound connections. The node opens the connection to the coordinator, holds it or reopens it on a schedule, and asks for work. + +The reference protocol already has this shape and did not need changing. Its node-side operations are `SubmitListing`, `Heartbeat`, `PollJobs`, `Decline`, `ReportJob` and `Fees` — every one of them a request the node initiates. The coordinator implements a pluggable interface and answers; it never dials the node. What was missing was not the mechanism but the commitment. A property that holds by accident of the current implementation can be lost in the next refactor, so the official document now states it and the registry binds it: + +> Nodes join that market over encrypted overlays and outbound polling, without requiring open inbound ports or a static address; the connection as a residential provider ships it is enough. Line 11 depends on this: a substrate that ran only where a provider permits inbound service would carry a chokepoint at every provider. + +Registered as `SUB-no-inbound-requirement`, implementation-bound. A repository binds it by shipping tests that cite the identifier and prove a node completes the full employment loop — register, heartbeat, poll, run, report — with no listening socket, no static address, and a translated, changing address in front of it. Until such tests exist the invariant is reported unbound under bylaws §9.15, and this paper's claim about the reference implementation is exactly the self-attestation §9.6 refuses to recognize. + +The line-11 reasoning is the substantive part, and it is why this belongs in the document rather than in a deployment guide. Line 11 forbids any single host, account or vendor whose removal could stop the network. A substrate that required inbound service would run only where a carrier permits it, making every carrier a veto — a chokepoint per provider, distributed in appearance and centralized in fact. + +**Three notes on mechanism.** First, on transport: ordinary HTTPS and WebSocket connections on port 443 are the right carrier because that is what the network path reliably permits and what every web client already emits. This paper deliberately does not describe that traffic as disguised. It is not camouflage; it is the standard protocol for the job, and the honesty matters because the framing of camouflage invites the belief that a technical measure has answered a permission question, which section 5 shows it has not. + +Second, on internet protocol version six: it removes carrier-grade translation where both ends have it, and adoption crossed half of Google's traffic globally on 28 March 2026, with the United States near 57 per cent (Google, 2026). It is worth using and worth requiring nothing from. Universal addressability is not the same as universal reachability — a globally routable address still sits behind a firewall that drops unsolicited inbound packets, and a node that assumes otherwise fails on the remaining half of connections. Version six is an optimization to the outbound posture, never a replacement for it. + +Third, on encrypted overlay networks. Mesh overlays are a real answer to node-to-node paths that the coordinator should not mediate, and named implementations exist. They cannot enter the protocol module. The lean-code principle holds protocol software to its language's standard library so it stays auditable whole, and the reference protocol currently satisfies this strictly — its module declares no dependencies at all, which is the property that keeps any single coordinator from becoming a hub. A large overlay library inside that module would end it. An overlay therefore belongs below the protocol as deployment-level transport, chosen per deployment and replaceable, or is implemented minimally within the leaf. The document's wording is deliberately generic for the same reason the document carries no product names. + +There is a related tension the institute should not paper over. The reference coordinator deployment currently sits behind a single commercial content-delivery network, and the agricultural stack's transport design assumes that vendor's tunnel. That is convenient, it is not conformant with line 11 in spirit, and it is a chokepoint of exactly the kind this section removes at the carrier layer while leaving in place one layer up. It is out of scope here and belongs on the substrate's open-problem list. + +## 3. Bandwidth asymmetry + +**The constraint.** Consumer connections are asymmetric by design, often ten to one or worse, and increasingly metered. A node cannot serve as a general content origin, and a workload that ships large images to every host will spend its time in transfer rather than compute. + +**What technique answers.** Most of it, by keeping payloads small rather than by moving them faster. Three design choices do the work, and they compound. + +**Coordination traffic is small by construction.** Line 7 keeps narratives and identities out of the shared record — hashes, types, timestamps and references only. A privacy floor adopted for privacy reasons has a bandwidth consequence: the record layer's traffic is bounded by the size of hashes and headers rather than by the size of what was exchanged. Content stays with the parties. The coordination messages that carry the market are a handful of signed structures over a canonical byte encoding. Nothing here strains a home uplink, and this is the sense in which the packages are light: the architecture's own transmissions are light because a line that cannot be crossed forces them to be. + +**Work ships as sandboxed modules, not as machine images.** This is the one recommendation in this paper that asks the reference implementation to change rather than to hold its shape. The current node agent runs jobs through a container executor, which means a first job on a fresh node pulls layers measured in hundreds of megabytes before any work begins. A WebAssembly module for the same task is measured in megabytes or less, starts in milliseconds, carries a deny-by-default capability model rather than an opt-out one, and is portable across the mixed consumer hardware the substrate expects instead of requiring a matching architecture. A runtime with no native dependencies keeps the node agent auditable in the same spirit the protocol module is. Containers should remain available for workloads that genuinely need a full operating environment, on nodes whose connections and operators can carry them, and should stop being the default. The trade is real and should be stated: WebAssembly costs some throughput against native execution and cannot host arbitrary existing software unmodified. For irregular, light, sensor-and-coordination work — the agricultural pattern this architecture serves first — that trade is favorable. + +**Jobs spool locally.** A node holds its queue and its pending reports in a local embedded store and drains them when the connection allows. The consequence is that a poor uplink delays work instead of losing it, and an intermittent connection stops being a disqualification. This matters beyond bandwidth: the agricultural pilot already plans for rural connectivity gaps with offline entry and later synchronization, and the same property makes a node in that setting a participant rather than a liability. + +None of these three is a novel invention and none is registered as an invariant. They are design posture, recorded here so that a deployment can be assessed against them and so the reasoning survives the people who had it. The board may wish to consider whether the module-over-image default should become a registered invariant; this paper does not recommend it yet, because the reference implementation does not satisfy it today and a registry that outruns the code teaches the wrong lesson about what registration means. + +## 4. Execution on hardware the host controls + +**The constraint.** A prosumer hosting a node has physical possession of the machine. They can read its memory, inspect its disk, and observe what it computes. The conventional answer is hardware confidential execution, and it is unavailable at this layer as a matter of vendor product strategy rather than of cost. Intel deprecated Software Guard Extensions on client processors from the eleventh generation of its Core line and retains it on server and cloud parts (Intel, 2021). AMD's Secure Encrypted Virtualization, including the nested-paging generation, is an EPYC server feature and is not present on Ryzen or Threadripper (AMD, 2021). A requirement for either would exclude essentially all consumer hardware and re-admit precisely the data-center gatekeeping the substrate layer exists to displace. Requiring enclaves would not secure the commons substrate; it would abolish it. + +**What technique answers, and how far.** Not confidentiality against a determined host. This section is where the paper's method changes, and saying so plainly is more useful than a workaround presented with more confidence than it earns. + +The architecture's available answer is not to trust the host but to make dishonesty visible and expensive, using machinery the stack already owes itself. Three parts: + +**Redundant execution with rated disagreement.** A job of consequence is dispatched to two or more independent hosts and their results compared. Agreement is evidence; disagreement is an event that enters the covenant, where the scale's lowest rating already means a party was harmed, exploited, or served with malicious intent. The enforcement hook exists: witnessing is compensated substrate work, and the substrate owes a work-type in which witnesses are randomly assigned so that neither party to an exchange can choose its witness. Redundant execution is that same market pattern applied to compute rather than to record-keeping, and random assignment is what makes collusion between a job's executors a matter of chance rather than of choice. The cost is a multiple of the compute, paid deliberately for the class of work that warrants it, and it buys detection rather than prevention — the same posture the record layer already takes toward tampering. + +**Data minimization as the primary control.** A host cannot scrape what never arrives. Line 7 already keeps identities and narratives out of the shared record; the corresponding discipline at the substrate layer is that a job carries the least data that will let it complete, that sensitive inputs are partitioned across hosts where the work allows it, and that a workload requiring a large coherent body of sensitive data about identifiable people is a workload for hardware whose operator is accountable for it. That last clause is a real limit on the substrate's reach and should be stated as one rather than engineered around. + +**Attestation as an optional, priced, rated capability.** Where a buyer genuinely needs hardware-backed confidentiality, the answer is a market, not a mandate. A host with such hardware advertises the capability as part of its listing, buyers who need it pay for it, and the claim is subject to the same covenant rating as any other representation a host makes. This keeps the floor open to consumer hardware while letting the ceiling rise wherever a host has invested, and it locates the decision with the party bearing the risk. + +**Status.** This is the part of the September 2026 review that is not resolved. The position above is argued, not adopted: no invariant is registered, and a candidate identifier for redundant execution is noted in the open-questions record as a decision for the board under §9.16. Registering it would oblige the substrate to build a work-type it has not built. The honest state is that the architecture has a coherent answer to host-operator observation and has not yet committed to it. + +## 5. What technique does not answer + +Outbound polling removes the inbound-server problem completely. It does not remove the acceptable-use problem, and this paper would mislead its readers if it implied otherwise. + +Read the representative policy again. It bars equipment serving anyone outside the premises network "except for your personal and non-commercial residential use." The inbound-server prohibition is a statement about ports and is answered by not using any. The non-commercial prong is a statement about compensation, and compensation is the point: a prosumer hosting substrate is paid, in fiat in the current implementation and in community credit later. No transport choice, port number, or encryption changes that fact, and a design that claims to have routed around it has confused a mechanism with a permission. The question is whether compensated participation on a residential line engages that prong, and the answer is a reading of contract terms in a jurisdiction, not a property of software. + +The institute should have counsel review it, and the natural occasion is at hand: the combined heat-and-compute pilot already has two open questions before counsel about whether a revenue-earning node in a home changes the household's position with its insurer under business-use exclusions. The carrier-terms question is the same question addressed to a different contract, and it should go in the same package rather than wait for its own. Its practical shapes are worth naming now — whether a business-tier connection is required, whether a de-minimis or cost-sharing framing holds, whether the answer varies by carrier enough that node-operator guidance must be regional, and what a deployment tells prospective hosts before they enroll. + +The last of those is a covenant matter as much as a legal one. A prosumer is entitled to know what participation may mean for their own service agreement before they take it up, and the architecture's own commitment to a rated, disclosed relationship between operator and prosumer makes silence on the point the wrong default. + +## 6. Where this landed + +| Level | What changed | Enforcement | +|---|---|---| +| Official document | Substrate protocol tier states the outbound, no-inbound-requirement posture and ties it to line 11 | Suite checks the text; passes at 26 invariants | +| Conformance registry | `SUB-no-inbound-requirement` added, implementation-bound | Unbound until node tests cite the identifier (§9.15) | +| Open-questions record | Entry 9 records the constraint, the resolutions, the open part, and the legal question | §9.3 | +| This paper | The bandwidth and confidential-execution posture, argued and unbound | None; it analyzes and does not govern | + +Two procedural obligations attach to the document change and are not discharged by this paper. An amendment of the official document and its registry runs under bylaws §9.2 and §9.16, and during bootstrap the founder board exercises that power with each act recorded in the governance registry as a bootstrap act under §16.1, open to the membership like any other decision. The suite passes against the amended text, which §9.2 requires before adoption, and the seven-language renderings under P2-002 were updated with the English original. + +## Sources + +Acemoglu, D., & Robinson, J. A. (2019). *The Narrow Corridor: States, Societies, and the Fate of Liberty*. Penguin Press. + +AMD. (2021, March 15). *AMD EPYC 7003 series processors set new standard*. https://www.amd.com/en/newsroom/press-releases/2021-3-15-amd-epyc-7003-series-cpus-set-new-standard-as-hig.html + +Comcast. (2021, February 1). *Acceptable use policy for Xfinity Internet (residential)*. https://www.xfinity.com/corporate/customers/policies/highspeedinternetaup + +Google. (2026). *IPv6 adoption statistics*. https://www.google.com/intl/en/ipv6/statistics.html + +Intel. (2021). *Intel SGX deprecation on client processors* [Intel Community discussion]. https://community.intel.com/t5/Intel-Software-Guard-Extensions/Intel-SGX-deprecated-in-11th-Gen-processors/m-p/1351848 + +Internet Society. (2026, April). *18 years later, IPv6 reaches majority*. https://pulse.internetsociety.org/en/blog/2026/04/18-years-later-ipv6-reaches-majority/ + +Network Theory Applied Research Institute. (2025a, October). *Addressing democratic information velocity* (P1-002). https://www.ntari.org/post/ntari-whitepaper-addressing-democratic-information-velocity + +Network Theory Applied Research Institute. (2025b, June). *The material culture of democratic deliberation*. https://www.ntari.org/post/the-material-culture-of-democratic-deliberation + +*Ohio Telecom Association v. FCC*, Nos. 24-7000 et al. (6th Cir. Jan. 2, 2025). Congressional Research Service analysis: https://www.congress.gov/crs-product/LSB11264 + +--- + +*Network Theory Applied Research Institute, Inc. — 501(c)(3) — EIN 92-3047136 — info@ntari.org* + +*Specification: CC BY-SA 4.0* diff --git a/README.md b/README.md index 14e0e3e..84f5fce 100644 --- a/README.md +++ b/README.md @@ -18,6 +18,7 @@ and protocol. | **Unresolved questions** | [OPEN-QUESTIONS.md](OPEN-QUESTIONS.md) | | **Concepts carried from prior instruments** | [jfa-concept-triage-2026-08-24.md](jfa-concept-triage-2026-08-24.md) | | **Executable conformance suite** | [jfa-conformance-suite.py](jfa-conformance-suite.py) | +| **Substrate under carrier constraints (P1-004, companion paper)** | [P1-004_Substrate-Constraints_v0.1.md](P1-004_Substrate-Constraints_v0.1.md) | | **Prior instruments** | [Historical Docs/](Historical%20Docs/) | The English document is authoritative. Translations are provided for reach, not @@ -77,7 +78,7 @@ python jfa-conformance-suite.py --project PATH # check a repo's open-questions Exit code 0 when every executed check passes, 1 otherwise. Run it after any edit to the official document. -Of the 25 registered invariants, 3 are bound here at the document layer and 22 +Of the 26 registered invariants, 3 are bound here at the document layer and 23 are **delegated** — they bind running software or a governance instrument, and can only be enforced by tests living beside that code or that instrument. They are carried in the registry with stable IDs and reported as delegated and diff --git a/janus-facing-architecture.ar.md b/janus-facing-architecture.ar.md index 9db437b..3bf2b80 100644 --- a/janus-facing-architecture.ar.md +++ b/janus-facing-architecture.ar.md @@ -1,4 +1,4 @@ -> ترجمة مجتمعية (مسودة) — سياسة NTARI رقم P2-002، البث العالمي متعدد اللغات. المصدر: janus-facing-architecture.md (النص الإنجليزي الأصلي، لقطة بتاريخ 2026-08-31). مسودة مجتمعية بمساعدة آلية، في انتظار مراجعة مشرف المنطقة وفقاً للبند §3.1 من P2-002. تبقى المواصفات التقنية الأساسية باللغة الإنجليزية وفقاً للبند §2.2. +> ترجمة مجتمعية (مسودة) — سياسة NTARI رقم P2-002، البث العالمي متعدد اللغات. المصدر: janus-facing-architecture.md (النص الإنجليزي الأصلي، لقطة بتاريخ 2026-09-08). مسودة مجتمعية بمساعدة آلية، في انتظار مراجعة مشرف المنطقة وفقاً للبند §3.1 من P2-002. تبقى المواصفات التقنية الأساسية باللغة الإنجليزية وفقاً للبند §2.2. > > تصحيحات الترجمة مساهمات نرحّب بها ونقدّرها كسائر المساهمات، فإذا لاحظت خطأً > في هذه الترجمة فيمكنك إصلاحه بنفسك عبر إنشاء fork للمستودع وفتح pull @@ -36,6 +36,8 @@ تتبادل التعليمات والأوامر عبر سوق موزّعة للحوسبة والتخزين، تُشغَّل على حواسيب استهلاكية مستضافة في المنازل والمكاتب والمخازن، وكذلك على معدّات صناعية معاد استخدامها. +تنضم العقد إلى تلك السوق عبر شبكات تراكبية مشفّرة (overlays) واستطلاع صادر (outbound polling)، دون الحاجة إلى منافذ واردة مفتوحة أو عنوان ثابت؛ فالاتصال كما يقدّمه مزوّد الخدمة المنزلية يكفي. ويعتمد السطر الحادي عشر (11) على ذلك: فالركيزة التي لا تعمل إلا حيث يسمح المزوّد بالخدمة الواردة تحمل نقطة اختناق عند كل مزوّد. + ### مرتبة المنسّق قدرة حوسبة متّحدة للمنتجين-المستهلكين، تخلق خيارات أوسع على امتداد الجغرافيا. diff --git a/janus-facing-architecture.es.md b/janus-facing-architecture.es.md index 9e6adaf..676f550 100644 --- a/janus-facing-architecture.es.md +++ b/janus-facing-architecture.es.md @@ -1,4 +1,4 @@ -> Traducción comunitaria (borrador) — Política P2-002 de NTARI, Difusión Multilingüe Global. Fuente: janus-facing-architecture.md (original en inglés, instantánea del 2026-08-31). Borrador comunitario asistido por máquina, pendiente de revisión por el mantenedor regional conforme a P2-002 §3.1. Las especificaciones técnicas centrales permanecen en inglés conforme al §2.2. +> Traducción comunitaria (borrador) — Política P2-002 de NTARI, Difusión Multilingüe Global. Fuente: janus-facing-architecture.md (original en inglés, instantánea del 2026-09-08). Borrador comunitario asistido por máquina, pendiente de revisión por el mantenedor regional conforme a P2-002 §3.1. Las especificaciones técnicas centrales permanecen en inglés conforme al §2.2. > > ¿Encontraste un error en esta traducción? Tu corrección es una contribución > bienvenida y valorada: haz un fork del repositorio del proyecto de NTARI y @@ -36,6 +36,8 @@ Es el hardware donde todo ocurre, propiedad de prosumidores de CPU, GPU, impreso Intercambia instrucciones y órdenes a través de un mercado distribuido de cómputo y almacenamiento operado en computadoras de consumo alojadas en hogares, oficinas y depósitos, así como en equipo industrial reacondicionado. +Los nodos se incorporan a ese mercado mediante superposiciones cifradas (overlays) y sondeo saliente (polling), sin requerir puertos de entrada abiertos ni una dirección estática; basta la conexión tal como la entrega un proveedor residencial. La línea 11 depende de ello: un sustrato que solo funcionara donde un proveedor permite el servicio entrante cargaría con un punto de estrangulamiento en cada proveedor. + ### Nivel de orquestador Capacidad de cómputo federada de prosumidores, que crea más opciones a lo largo de la geografía. diff --git a/janus-facing-architecture.fr.md b/janus-facing-architecture.fr.md index 713d6b2..9fe8916 100644 --- a/janus-facing-architecture.fr.md +++ b/janus-facing-architecture.fr.md @@ -1,4 +1,4 @@ -> Traduction communautaire (version préliminaire) — Politique P2-002 de NTARI, Diffusion mondiale multilingue. Source : janus-facing-architecture.md (original en anglais, instantané du 2026-08-31). Version préliminaire communautaire assistée par machine, en attente de révision par le mainteneur régional conformément à P2-002 §3.1. Les spécifications techniques centrales demeurent en anglais conformément au §2.2. +> Traduction communautaire (version préliminaire) — Politique P2-002 de NTARI, Diffusion mondiale multilingue. Source : janus-facing-architecture.md (original en anglais, instantané du 2026-09-08). Version préliminaire communautaire assistée par machine, en attente de révision par le mainteneur régional conformément à P2-002 §3.1. Les spécifications techniques centrales demeurent en anglais conformément au §2.2. > > Vous avez repéré une erreur dans cette traduction ? Votre correction est une > contribution bienvenue et appréciée : créez un fork du dépôt du projet NTARI @@ -36,6 +36,8 @@ C'est le matériel où tout se produit, détenu par des prosommateurs de process Échange instructions et ordres sur un marché distribué de calcul et de stockage, exploité sur des ordinateurs grand public hébergés dans des domiciles, des bureaux et des entrepôts, ainsi que sur du matériel industriel reconverti. +Les nœuds rejoignent ce marché par des réseaux superposés chiffrés (overlays) et une interrogation sortante (polling), sans exiger de ports entrants ouverts ni d'adresse statique ; la connexion telle qu'un fournisseur résidentiel la livre suffit. La ligne 11 en dépend : un substrat qui ne fonctionnerait que là où un fournisseur autorise le service entrant porterait un point d'étranglement chez chaque fournisseur. + ### Niveau orchestrateur Puissance de calcul fédérée de prosommateurs, créant davantage d'options à travers la géographie. diff --git a/janus-facing-architecture.hi.md b/janus-facing-architecture.hi.md index 66fb09e..48c153e 100644 --- a/janus-facing-architecture.hi.md +++ b/janus-facing-architecture.hi.md @@ -1,4 +1,4 @@ -> सामुदायिक अनुवाद (प्रारूप) — NTARI नीति P2-002, वैश्विक बहुभाषी प्रसारण। स्रोत: janus-facing-architecture.md (मूल अंग्रेज़ी पाठ, 2026-08-31 का स्नैपशॉट)। मशीन-सहायित सामुदायिक प्रारूप, P2-002 §3.1 के अनुसार क्षेत्रीय अनुरक्षक की समीक्षा हेतु लंबित। मूल तकनीकी विनिर्देश §2.2 के अनुसार अंग्रेज़ी में ही रहते हैं। +> सामुदायिक अनुवाद (प्रारूप) — NTARI नीति P2-002, वैश्विक बहुभाषी प्रसारण। स्रोत: janus-facing-architecture.md (मूल अंग्रेज़ी पाठ, 2026-09-08 का स्नैपशॉट)। मशीन-सहायित सामुदायिक प्रारूप, P2-002 §3.1 के अनुसार क्षेत्रीय अनुरक्षक की समीक्षा हेतु लंबित। मूल तकनीकी विनिर्देश §2.2 के अनुसार अंग्रेज़ी में ही रहते हैं। > > इस अनुवाद में कोई त्रुटि दिखी? आपका सुधार किसी भी अन्य योगदान की तरह स्वागत-योग्य और > मूल्यवान है: NTARI परियोजना के रिपॉज़िटरी का fork बनाकर pull request खोलें, या @@ -36,6 +36,8 @@ JFA सॉफ़्टवेयर को copyleft परिवेश में एक वितरित संगणन/भंडारण बाज़ार में अनुदेश और आदेश विनिमित करता है, जो घरों, कार्यालयों और भंडार-गृहों में रखे उपभोक्ता-श्रेणी के कंप्यूटरों तथा पुनःप्रयुक्त औद्योगिक उपकरणों पर संचालित होता है। +नोड उस बाज़ार में एन्क्रिप्टेड ओवरले (overlay) और आउटबाउंड पोलिंग (outbound polling) के माध्यम से शामिल होते हैं, जिसके लिए न खुले इनबाउंड पोर्ट चाहिए, न स्थिर पता; आवासीय प्रदाता जैसा कनेक्शन देता है, वही पर्याप्त है। रेखा 11 इसी पर निर्भर है: जो आधार केवल वहीं चल सके जहाँ प्रदाता इनबाउंड सेवा की अनुमति देता है, वह हर प्रदाता पर एक अवरोध-बिंदु (chokepoint) ढोएगा। + ### ऑर्केस्ट्रेटर श्रेणी प्रोज़्यूमरों की संघीकृत संगणन-शक्ति, जो भूगोल भर में अधिक विकल्प उत्पन्न करती है। diff --git a/janus-facing-architecture.md b/janus-facing-architecture.md index 31986f9..ec73c8c 100644 --- a/janus-facing-architecture.md +++ b/janus-facing-architecture.md @@ -30,6 +30,8 @@ This is the hardware where everything happens, owned by prosumers of CPUs, GPUs, Exchanges instructions and orders across a distributed compute/storage market operated on consumer grade computers hosted in homes, offices and storage, as well as repurposed industrial equipment. +Nodes join that market over encrypted overlays and outbound polling, without requiring open inbound ports or a static address; the connection as a residential provider ships it is enough. Line 11 depends on this: a substrate that ran only where a provider permits inbound service would carry a chokepoint at every provider. + ### Orchestrator Tier Federated prosumer compute power creating more options across geography. Orchestrators publish transport offers into the substrate market, each naming a fee, a delivery commitment and a public key; any platform may select any reachable orchestrator, so a dominant carrier is undercut rather than regulated. Transport is delivery, not execution: an orchestrator carries signed spends to the witness set and returns attestations, is never relied upon to determine whether an exchange occurred, and may be lossy and retry-based. diff --git a/janus-facing-architecture.pt.md b/janus-facing-architecture.pt.md index cfafcf3..7d3d387 100644 --- a/janus-facing-architecture.pt.md +++ b/janus-facing-architecture.pt.md @@ -1,4 +1,4 @@ -> Tradução comunitária (rascunho) — Política P2-002 da NTARI, Difusão Multilíngue Global. Fonte: janus-facing-architecture.md (original em inglês, instantâneo de 2026-08-31). Rascunho comunitário assistido por máquina, pendente de revisão pelo mantenedor regional conforme P2-002 §3.1. As especificações técnicas centrais permanecem em inglês conforme o §2.2. +> Tradução comunitária (rascunho) — Política P2-002 da NTARI, Difusão Multilíngue Global. Fonte: janus-facing-architecture.md (original em inglês, instantâneo de 2026-09-08). Rascunho comunitário assistido por máquina, pendente de revisão pelo mantenedor regional conforme P2-002 §3.1. As especificações técnicas centrais permanecem em inglês conforme o §2.2. > > Encontrou um erro nesta tradução? Sua correção é uma contribuição bem-vinda e > valorizada: faça um fork do repositório do projeto da NTARI e abra um pull @@ -36,6 +36,8 @@ Este é o documento oficial, sob a curadoria do Network Theory Applied Research Troca instruções e ordens através de um mercado distribuído de computação e armazenamento, operado em computadores de consumo hospedados em residências, escritórios e depósitos, bem como em equipamentos industriais reaproveitados. +Os nós ingressam nesse mercado por meio de sobreposições cifradas (overlays) e sondagem de saída (polling), sem exigir portas de entrada abertas nem endereço estático; basta a conexão tal como um provedor residencial a entrega. A linha 11 depende disso: um substrato que só funcionasse onde um provedor permite serviço de entrada carregaria um ponto de estrangulamento em cada provedor. + ### Nível de orquestrador Capacidade de computação federada de prossumidores, criando mais opções ao longo da geografia. diff --git a/janus-facing-architecture.tok.md b/janus-facing-architecture.tok.md index 9b7b29c..ef4ef5c 100644 --- a/janus-facing-architecture.tok.md +++ b/janus-facing-architecture.tok.md @@ -1,8 +1,8 @@ # JFA — nasin pi esun kulupu (lipu lili) -> **toki pi lipu ni:** lipu ni li lipu lili tan lipu suli. nasin NTARI "P2-002" li pana e ken tawa ona, tawa kulupu. sina wile e sona ale la, o lukin e lipu suli pi toki Inli: `janus-facing-architecture.md` (tan tenpo suno 2026-08-31). ilo sona li pali e lipu ni. kulupu li lukin ala e ona lon tenpo ni. nasin P2-002 kipisi 3.1 li wile e ni: kulupu o lukin e lipu ni. +> **toki pi lipu ni:** lipu ni li lipu lili tan lipu suli. nasin NTARI "P2-002" li pana e ken tawa ona, tawa kulupu. sina wile e sona ale la, o lukin e lipu suli pi toki Inli: `janus-facing-architecture.md` (tan tenpo suno 2026-09-08). ilo sona li pali e lipu ni. kulupu li lukin ala e ona lon tenpo ni. nasin P2-002 kipisi 3.1 li wile e ni: kulupu o lukin e lipu ni. > -> **Note:** This is a condensed community rendering produced under NTARI policy P2-002. The complete document is the English original `janus-facing-architecture.md` (snapshot 2026-08-31). Machine-assisted draft pending community review per P2-002 section 3.1. Core technical specifications remain in English per section 2.2. +> **Note:** This is a condensed community rendering produced under NTARI policy P2-002. The complete document is the English original `janus-facing-architecture.md` (snapshot 2026-09-08). Machine-assisted draft pending community review per P2-002 section 3.1. Core technical specifications remain in English per section 2.2. > > sina lukin e pakala lon toki ni la o pona e ona: o pana e "pull request" > tawa poki lipu NTARI, anu o toki tawa info@ntari.org. pana sina li pona tawa @@ -20,7 +20,7 @@ JFA li wile pona e ni tan insa: **kulupu pi pali nasin li kulupu pi lukin nasin* JFA li jo e kipisi luka: -1. **kipisi ilo** (Substrate) — ilo ale. jan li jo e ona lon tomo ona. +1. **kipisi ilo** (Substrate) — ilo ale. jan li jo e ona lon tomo ona. ilo li toki tawa esun kepeken nasin len; ona li wile ala e lupa open lon tomo. (nasin 11) 2. **kipisi sona** (Record) — sona pi ijo pini. ona li awen. 3. **kipisi pi toki awen** (Covenant) — jan li toki e ni tawa jan ante: "mi ike ala e sina." 4. **kipisi lawa** (Governance) — jan li kama kulupu, li pali e nasin. diff --git a/janus-facing-architecture.zh.md b/janus-facing-architecture.zh.md index aca9675..11f2c25 100644 --- a/janus-facing-architecture.zh.md +++ b/janus-facing-architecture.zh.md @@ -1,4 +1,4 @@ -> 社区翻译（草稿）——NTARI 政策 P2-002《全球多语种广播政策》。来源：janus-facing-architecture.md（英文原件，2026-08-31 快照）。机器辅助的社区草稿，按 P2-002 §3.1 尚待区域维护者审校。核心技术规范按 §2.2 保留英文。 +> 社区翻译（草稿）——NTARI 政策 P2-002《全球多语种广播政策》。来源：janus-facing-architecture.md（英文原件，2026-09-08 快照）。机器辅助的社区草稿，按 P2-002 §3.1 尚待区域维护者审校。核心技术规范按 §2.2 保留英文。 > > 发现译文有误？您的更正与其他贡献同样受欢迎、同样宝贵：请 fork NTARI 项目仓库并提交 > pull request，或写信至 info@ntari.org。 @@ -35,6 +35,8 @@ JFA 软件的设计以 copyleft 环境发布与管理，通常采用 GNU Affero 在一个分布式的算力/存储市场中交换指令与订单，该市场运行于家庭、办公室与仓储中的消费级计算机，以及经再利用的工业设备之上。 +节点通过加密覆盖网络（overlay）与出向轮询（outbound polling）加入该市场，无需开放入向端口，也无需静态地址；住宅宽带提供商交付的原样连接即已足够。第 11 条正依赖于此：若基质只能运行在提供商允许入向服务之处，那么每一家提供商都会成为一个咽喉要道。 + ### 编排器层级 产消者的联邦化算力，在地理上创造更多选择。 diff --git a/jfa-conformance-suite.py b/jfa-conformance-suite.py index fd0107d..b7cf0b6 100644 --- a/jfa-conformance-suite.py +++ b/jfa-conformance-suite.py @@ -6,7 +6,7 @@ The prior instrument and its suite are preserved in Historical Docs as *-pre-rebuild-2026-08-24.*. Their clause-numbered IDs are retired; this suite's registry cites the new document's sections by name and its lines -by number (L1-L11). Where an invariant descends from the prior registry, +by number (L1-L12). Where an invariant descends from the prior registry, the old ID is noted for lineage. Which concepts carried, changed, or retired is recorded in jfa-concept-triage-2026-08-24.md. @@ -94,6 +94,9 @@ "instrument", "governance instrument assigns each layer's discipline by exit cost (was 2.3)"), Invariant("P-lean-code", "principles", "auditable whole", "implementation", "protocol repo: dependency audit in CI - standard library only; copyleft check (was 8.2)"), + # Substrate layer — the residential posture + Invariant("SUB-no-inbound-requirement", "substrate", "without requiring open inbound ports or a static address", + "implementation", "node tests: a node registers, heartbeats, polls for work and reports over outbound connections only - no listening port, no static IP, behind NAT/CGNAT and dynamic addressing; the residential case of L11 (new 2026-09-08)"), # Record layer — the topology Invariant("REC-six-holders", "record", "held six ways", "implementation", "record tests: each transactor, the operator, and both witnesses keep records beside the chain (was REC-four-holders; recounted 2026-08-27)"),