Decentralized DNS Monday — August 3 2026 Weekly Update

TL;DR — This Week

• ENS Governance Overhaul — "Next Era of ENS DAO": Temperature check proposal to move operational, treasury, grants, and capital management responsibilities from token-weighted DAO voting into a strengthened ENS Foundation (full-time Executive Director, 5-member board). Protocol governance (contract upgrades, registry, fees, constitution) stays with tokenholders. Inspired by Mozilla, Signal, Linux Foundation, ISRG/Let's Encrypt models. COO Katherine Wu defending publicly. [Source: EtherWorld, ENS DAO forum]
• ICANN 2026 gTLD Window Closes August 12: ENS has applied for .ens as a .brand TLD — closed to public registrations initially, potential open TLD transition later. Earliest delegation 2028. Strategic pivot: ENS works within ICANN rather than outside it. Application fee: $227,000. [Source: ENS Blog, ICANN]
• Blockchain Domain Reckoning Intensifies: Namecheap exited all Handshake TLD support on June 10 with no migration path. Unstoppable Domains CEO Matthew Gould publicly called blockchain domains a "craze" with limited mainstream penetration — 90%+ of UD's active business is now conventional DNS. HNS token down 99% from ATH ($0.85 → ~$0.002). Namespace collision risk: ICANN may delegate strings (.wallet, .crypto, .nft) that blockchain registries already sold. [Source: MonstaDomains, webhosting.today]
• HNS Post-Halving Bounce: 62% weekly price surge from ATL ($0.001139 on Jul 8) to ~$0.001935. Low volume ($6.8K/24h). CoinGecko notes "no clear narrative" — likely technical bounce post-halving rather than fundamental shift. [Source: CoinGecko]
• Pubky Core v0.10-alpha (Jun 29) + Active July Development: Major breaking changes (grant-based auth, deeplink refactor). v0.9.x bugfixes through June. Migration guide published Jul 31. Homeserver improvements, TypeDoc setup, strict capabilities refactor. SDK maturing toward stable v0.10. [Source: GitHub releases, commits]
• GNUnet 0.28.0 Released (Jul 17): New release of the experimental GNUnet framework including GNS. Continues as the strongest privacy-preserving naming system (petname-based, zone delegation, RFC 9498). [Source: gnunet.org, Wikipedia]
• ICANN Root KSK Rollover — October 11, 2026: ICANN published detailed preparation guide (Jul 27). KSK-2024 replaces current root trust anchor. DNSSEC-validating resolvers must update or face total DNS resolution failure. First rollover was 2018; this is only the second. [Source: ICANN blog, PDF guide]
• DNS-AID (Linux Foundation) — Agent Discovery via DNS: Launched May 27, 2026. Open-source project for AI agent discovery and MCP server lookup via DNS records. Backed by Cloudflare, GoDaddy, Equinix, Infoblox, ISC, CSC. Companion academic paper (arxiv 2606.02314). Growing ecosystem with Agent Name Service (ANS). IETF standardization underway. [Source: Linux Foundation, Infoblox, arXiv]
• Freedom-tech takeaway: The naming layer is bifurcating. ENS is integrating into the establishment (ICANN, Foundation governance). PKARR/Pubky and GNS remain the sovereignty-maximalist paths. The agentic web's naming layer is being standardized at IETF NOW — if cypherpunk voices don't participate, agent identity will be locked into DNS-based establishment frameworks.

The Week in Detail

1. ENS — The Great Governance Restructuring

The most consequential ENS governance event since the DAO's inception is now in temperature check: "Next Era of ENS DAO: Empowering the ENS Foundation."

The Problem: ENS DAO was designed to steward credibly neutral infrastructure — tokenholders should control protocol-level decisions (contract upgrades, registry, fees, constitution). But the DAO gradually accumulated operational responsibilities: budget allocation, grants, working groups, treasury oversight (~$65M Endowment Safe in ETH/stables), ecosystem planning. Token voting is slow, binary, and lacks operational context. Running an organization requires constant judgment, clear accountability, and execution speed that DAO governance cannot provide.

The Proposal: Create a fully-resourced ENS Foundation with: - Full-time Executive Director - 5-member Foundation Board - Responsibilities: treasury stewardship, endowment oversight, public goods grants, ecosystem coordination, mission-driven advocacy, ENS trademark/brand management - Model: nonprofit open-source foundation (Mozilla, Signal, Linux Foundation, Wikimedia, ISRG/Let's Encrypt) - ENS Labs remains independent engineering/product org, funded by Foundation grants + trademark license

What Stays with Tokenholders: Protocol governance only — contract upgrades, registry control, fee structures, constitutional changes. The token's role narrows back to what it was designed for: safeguarding the protocol.

Key Quote (Katherine Wu, COO of ENS): ENS never held an ICO or venture-backed token sale. The token was created for governance, not as a revenue claim. The proposal corrects a structural mismatch between token-weighted governance and organizational operations.

Why This Matters for Freedom Tech: ENS is maturing from a crypto-native experiment into institutional infrastructure. The Foundation model mirrors how the internet's core infrastructure actually gets governed (IETF, ISOC, ISRG). Whether this is a positive (credibility, sustainability, longevity) or negative (bureaucratization, reduced cypherpunk edge) depends on execution. The 8-member Security Council activated July 21 adds another governance layer.

• Links: Proposal on ENS DAO forum | EtherWorld analysis

Relevance to Ungovernable.tech: ENS is the most production-deployed decentralized naming system. Its governance evolution sets precedents for how all decentralized infrastructure matures. See ENS for deeper technical details.

2. ICANN 2026 gTLD Round — The Window Closes August 12

ICANN opened the 2026 New gTLD application window on April 30 — the first opportunity in 14 years (last round was 2012, which delegated 1,200 new gTLDs). The window closes August 12, 2026.

ENS's .ens Application: ENS has applied for .ens as a .brand TLD: - Closed to public registrations initially (brand protection) - Could support names like wallet.ens for experimentation - Leaves door open to transition to open TLD in future - Earliest possible delegation: 2028 (multi-year process) - Application fee: $227,000 (up from $185,000 in 2012) - .eth remains in ICANN administrative "no-man's-land" (ISO 3166-1 alpha-3 reserved for Ethiopia)

ENS's ICANN Engagement: ENS is the only blockchain naming protocol that has consistently played by ICANN's rules: - ENS DAO Constitution: "ENS aims to integrate with the legacy DNS naming system to the greatest extent possible without sacrificing decentralization" - DNSSEC integration (ENSIP-6, 2018): import DNS names into ENS via TXT record verification - D3/Doma partnership: tokenize DNS domains on-chain - Registry integrations: .box, .art, .locker (Orange Domains) now ENS-enabled - Active engagement at ICANN 84 (Dublin), WIPO UDRP 25th Anniversary, DNS-OARC

Namespace Collision Risk: When Handshake, Unstoppable Domains, and others issued extensions like .wallet, .crypto, .nft unilaterally, they assumed those strings would remain outside ICANN's jurisdiction. The 2026 round makes that assumption untenable. ICANN's current procedures have no mechanism for assessing collision risk with pre-existing on-chain registrations. If ICANN delegates .wallet to a traditional registry, two separate authorities govern the same string simultaneously. Blockchain registrations may stop resolving correctly.

• Links: ENS ICANN 84 Blog | ICANN 2026 Round | Applicant Guidebook

Relevance to Ungovernable.tech: The collision problem is existential for blockchain-anchored naming. ENS's integrationist approach (working within ICANN) may prove more durable than Handshake's/UD's separatist approach. But it comes at the cost of submitting to ICANN governance. PKARR and GNS sidestep this entirely — they don't claim TLD strings at all.

3. Blockchain Domain Market — The Reckoning

The first generation of blockchain domain names is being publicly assessed as a commercial failure at scale.

Namecheap Exits Handshake (June 10, 2026): - Suspended all registrations, renewals, transfers, management functions — permanently - No migration path, no refunds, no recommended alternatives - Cited unnamed upstream provider winding down - Sequence: Jan 2026 Namecheap sold Namebase → Feb 1 Namebase went "offline for migration" → Jun 10 full exit - Namecheap manages tens of millions of domains globally; their exit is a market verdict

Unstoppable Domains CEO Concedes: - Matthew Gould publicly acknowledged blockchain domains were a temporary "craze" (March 2026) - 4M+ blockchain domain names sold, $70M VC raised — but 90%+ of active UD business is now conventional DNS - UD has become "primarily a conventional registrar that also sells NFT-backed names"

HNS Token Reality: - ATH: $0.8519 (May 5, 2021) - ATL: $0.001139 (July 8, 2026) — 99.8% below ATH - Current: ~$0.001935 (62% weekly bounce from ATL) - Market cap: ~$1.3M (smaller than most seed-stage companies) - 24h volume: $6,884

Structural Problems Identified: 1. The Browser Wall: Chrome, Safari, Firefox, Edge each declined native resolution for alternative roots. Required extensions or custom DNS — invisible to non-developers. 2. Censorship-Resistant but Unreachable: On-chain ownership means no registrar can be pressured. But censorship resistance only has value if people can visit your site. A domain requiring custom setup is already inaccessible to most of the internet. 3. No Commercial Incentive at Scale: At $1.3M market cap, no registrar can justify maintaining Handshake integrations.

Counter-Narrative: The impulse behind blockchain domains remains valid — register and operate online without identification, tracking, or traceable payments. The technology failed to deliver in a form the mainstream internet would support. But PKARR (sovereign naming without blockchain), ENS (integration with existing DNS), and private registrars (crypto payments, WHOIS protection) offer alternative paths.

• Links: MonstaDomains analysis | webhosting.today reporting

Relevance to Ungovernable.tech: This is a cautionary tale about the gap between cryptographic soundness and practical adoption. See Handshake, Unstoppable Domains for deep dives. The lesson: browser integration is the make-or-break choke point for any naming system.

4. Handshake (HNS) — Post-Halving Landscape

The 2nd halving completed ~July 22-23, 2026 (block reward 2000 → 1000 HNS). Key developments since:

• Price Action: 62% weekly surge from ATL, but this is from a very low base ($0.001139 → $0.001935). CoinGecko notes "no clear narrative" — likely technical bounce, not fundamental shift. Still 99.8% below ATH.
• Mining Decentralization: ViaBTC pool closure (Jun 1, ~50% hashrate) and Namebase shutdown (Jun 2) preceded the halving. Community views this as opportunity for further decentralization, though hashrate distribution data is thin.
• Development: hnsd (light resolver) last commit March 2026. Core protocol in maintenance mode. Community-led alternatives being discussed.
• Domains Remain On-Chain: Namecheap exit does not delete registrations. Names exist on the Handshake blockchain regardless of registrar support. But resolution now depends on running independent nodes or specialist resolver services.

Freedom-Tech Assessment: Handshake's cryptographic architecture remains sound. The challenge is purely adoption and infrastructure. Post-halving supply reduction (1000 HNS/block) combined with near-zero commercial interest creates an interesting long-term dynamic for patient sovereignty-minded operators.

• Links: CoinGecko HNS | Handshake.org | See Handshake

5. PKARR / Pubky Ecosystem — v0.10 Alpha and Active Development

The most censorship-resistant naming approach (DHT + public keys, no blockchain) continues steady technical progress:

Pubky Core v0.10.0-alpha.0 (June 29, 2026): - BREAKING: Grant-based authentication system (PR #373) — new auth flow replacing older mechanisms - BREAKING: Deeplink refactor (PR #392) — updated URI scheme - Migration Guide: Published July 31, 2026 (PR #493) - First alpha toward stable v0.10; significant API stabilization

v0.9.x Bugfixes (June): - v0.9.3 (Jun 24): TLS revocation hard-fail disabled on ICANN HTTP client (fixes false "Revoked" errors on Android) - v0.9.2 (Jun 23): Transport resolution with full _pubky.<key> qname so ICANN fallback applies correctly (fixes #422) - v0.9.1 (May 27): Various SDK fixes

July Development Activity (commits through Jul 31): - TypeDoc documentation setup (PR #523) - Homeserver: Skip PKARR republishing for migrated users (PR #515) - Strict capabilities and shared storage paths refactor (PR #514, breaking) - Consolidated stateful OpenDAL layers (homeserver fix) - Static testnet setup documentation

PKARR (Core Library): - Dependency bumps (Jul 22-26): cargo-dependencies group updates - CI: dependabot interval changed to weekly - crates.io: pkarr v6.x actively maintained

pkdns: - Stable, no new releases since March 2026 - Public pkdns.net resolvers operational - 191 GitHub stars, 17 forks - Supports A, AAAA, TXT, CNAME, MX records

SDK Maturity: Pubky SDK available for Rust (cargo add pubky) and JavaScript (npm install @synonymdev/pubky). 31 releases on crates.io, 251 downloads/month, used in 7 crates. pubky crate v0.10.0-alpha.0 published Jun 29.

Freedom-Tech Strength: Records live on Mainline DHT (millions of nodes). No blockchain dependency, no ICANN dependency, no registrar. Your keypair IS your name. Ideal for pseudonymous identity, mesh networks, Nostr integration, fully offline-capable naming. Long cache TTLs provide resilience against network disruption.

• Links: pubky-core releases | pkarr GitHub | pkdns GitHub | Pubky.org | Awesome Pubky | See PKARR

6. GNS / GNUnet — 0.28.0 Release

GNUnet 0.28.0 released July 17, 2026 — the second major release of 2026 (0.27.0 was March 19).

GNS (GNU Name System) remains the strongest lookup-privacy story in decentralized naming: - Petname-based: users assign local names to zones, no global namespace required - Zone delegation: chain of trust through signed zone delegations - RFC 9498 published (formal IETF documentation) - R5N DHT underneath (randomized recursive routing for restricted-route networks) - No blockchain, no consensus mechanism — ecologically sustainable - Go implementation (GoGNS) can participate in GNUnet DHT - Ascension tool: migrate DNS zones to GNS via incremental AXFR/IXFR - reclaimID: self-sovereign identity management on top of GNS

Why GNS Matters: Unlike blockchain naming systems, GNS doesn't just decentralize the registry — it decentralizes the resolution itself. Queries are opaque to the network. No central authority, no blockchain, no mining. The privacy model is fundamentally different and stronger for lookups.

Maturity: Academic roots (Christian Grothoff, Martin Schanzenbach). NLnet funded. Active since 2001. GNS is the most theoretically grounded alternative to DNS, but suffers from the same adoption problem as all alternative roots — requires client-side installation.

• Links: GNUnet.org | RFC 9498 | NLnet GNS project | See GNS

7. ICANN Root KSK Rollover — October 11, 2026

Not decentralized DNS, but critical DNS infrastructure that affects every resolver worldwide:

• What: Root zone Key Signing Key rollover. KSK-2024 replaces the current trust anchor.
• When: October 11, 2026. KSK-2024 was pre-published in January 2025 (21 months prep).
• Impact: DNSSEC-validating resolvers without KSK-2024 configured will experience total DNS resolution failure — cutting off internet access for users.
• History: First-ever KSK rollover was October 2018 (originally planned 2017, delayed for outreach). This is only the second.
• Preparation: ICANN published detailed guide (Jul 27, 2026) + webinar series (EMEA Aug 26).

• Freedom-Tech Relevance: Operators running their own resolvers (pkdns, GNS proxies, Handshake resolvers) must ensure DNSSEC compatibility. Self-sovereign infrastructure that breaks on KSK rollover defeats the purpose.

• Links: ICANN KSK Rollover Blog | What to Expect PDF | root-dnssec-announce mailing list

8. DNS-AID — Naming Layer for the Agentic Web

A new frontier in naming that directly intersects with decentralized DNS philosophy:

DNS-AID (Linux Foundation, launched May 27, 2026): Open-source project enabling AI agents to discover and communicate via DNS records. Initially developed by Infoblox.

What It Does: Standardizes how AI agents and MCP (Model Context Protocol) servers are published, discovered, and verified using existing DNS infrastructure. Python SDK, CLI, and MCP server reference implementation.

Why It Matters: - The agentic web needs a discovery layer. Currently fragmented with hardcoded URLs and centralized registries. - DNS-AID bets on the internet's existing DNS hierarchy — the only discovery system "universally deployed, universally interoperable, and proven at planetary scale." - Companion: Agent Name Service (ANS) — adds verification and identification layer. - Academic foundation: arxiv paper 2606.02314 ("Discovering Agents for Discovery: The Case for DNS") synthesizes DNS-AID and ANS approaches. - IETF DAWN working list active on standardization.

Backers: Cloudflare, GoDaddy, Equinix, Infoblox, ISC, CSC, Indeed, WWT.

Freedom-Tech Implication: Agent identity and discovery is being standardized NOW. If cypherpunk/privacy voices don't participate in IETF standardization, agent naming will be locked into DNS-based establishment frameworks with built-in surveillance potential. This is analogous to the early DNS standardization decisions that embedded certain trust assumptions. PKARR-based agent discovery (self-sovereign, no DNS dependency) is technically possible but not represented in current standardization efforts.

• Links: Linux Foundation announcement | GitHub dns-aid | Infoblox analysis | arXiv 2606.02314 | Help Net Security coverage

Evidence at a Glance

Freedom-Tech Takeaways & Recommendations

1. The Bifurcation Is Real: ENS is integrating into the establishment (ICANN gTLD, Foundation governance, institutional credibility). PKARR/Pubky and GNS remain sovereignty-maximalist paths. Handshake is wounded but alive. Choose your path based on threat model: ENS for discoverable public identity with institutional backing; PKARR/GNS for seizure-proof pseudonymous naming.

2. Browser Integration Is Everything: The blockchain domain market's failure had one root cause — no native browser resolution. Any naming system that requires extensions or custom resolvers has a hard ceiling. PKARR's DoH endpoints (pkdns.net) and ENS's eth.limo gateway are partial workarounds, but true sovereignty requires running your own resolver.

3. Agent Naming Is Being Standardized NOW: DNS-AID at the Linux Foundation with IETF standardization is defining how AI agents discover and verify each other. This is the naming layer for the next internet era. If freedom-tech doesn't engage, agent identity will be DNS-anchored with built-in surveillance. PKARR-based agent discovery should be proposed as an alternative/complement.

4. ICANN KSK Rollover (Oct 11) Affects Everyone: Self-sovereign resolver operators must prepare. Test your DNSSEC validation against KSK-2024 before October. This is operational hygiene for anyone running their own naming infrastructure.

5. The Collision Problem Is Existential: ICANN's 2026 gTLD round can delegate strings (.wallet, .crypto, .nft) that blockchain registries already sold. There is no recourse. Holders of affected blockchain domains face resolution risk. This validates key-rooted approaches (PKARR, Tor .onion) that don't claim TLD strings.

6. Monitor ENS Foundation Governance Vote: The "Next Era" proposal is a temperature check now. If it passes, ENS transitions from DAO-operated to Foundation-operated for everything except protocol governance. This affects the most widely-deployed decentralized naming system. Participate in the discussion if you hold ENS tokens.

7. PKARR Remains the Best Freedom-Tech Bet: No blockchain dependency, no ICANN dependency, no browser dependency (with local resolver). Active development (v0.10-alpha). Pubky ecosystem maturing. Run pkdns locally for maximum sovereignty. The "your keypair = your name" vision is closest to realization here.

Primary sources

• GNUnet: early-alpha software status