James Thain, from One Limited, submitted Internet-Draft draft-thain-ipv8-00 on April 14, 2026, proposing IPv8 as a new network protocol suite. This individual submission carries no IETF endorsement and remains in early “Active” status, expiring October 16, 2026. It claims to secure and manage networks from home setups to the global internet without disrupting existing IPv4 infrastructure.
The draft positions IPv8 as a drop-in evolution. IPv4 addresses become a subset: set the routing prefix to zero, and it works as IPv4. No hardware, software, or application changes required, per the authors. This sidesteps IPv6’s adoption woes—global IPv6 traffic hovers around 40% as of 2024, per APNIC data, leaving billions of devices on IPv4.
Addressing exhaustion gets a direct fix. Each Autonomous System Number (ASN) holder claims 4,294,967,296 (2^32) host addresses—one full IPv4 space per ASN. With roughly 130,000 public ASNs today, this unlocks over 500 trillion addresses without address sharing tricks like CGNAT. The global BGP table shrinks structurally: one entry per ASN, capping it at ASN count versus today’s 1 million prefixes and climbing.
Unified Management via Zone Servers
IPv8 centralizes everything under “Zone Servers.” These handle authentication (OAuth2 JWT tokens from local caches), service discovery (single DHCP8 lease delivers all needs), packet validation (egress checks against DNS8 and WHOIS8 routes), plus telemetry, NTP, and NAT. Every network element authenticates via tokens; every internet-bound packet verifies against registered routes.
Companion drafts flesh this out: BGP8, OSPF8, and others replace legacy protocols; WiFi8 for wireless; NetLog8 for monitoring; Update8 for NIC certification. IPv8 MIB enables SNMPv8 management. Regional Inter-Network Exchanges (RINE) connect zones.
This setup enforces top-down control. Operators monitor and authorize at the edge, reducing rogue traffic. Egress validation blocks hijacks if WHOIS8 confirms routes—potentially slashing BGP leaks, which hit 10,000+ incidents yearly per BGPStream data.
Skeptical Lens: Promises vs. Reality
Backward compatibility sounds ideal, but verify. IPv4 as prefix-zero subset implies larger addresses—likely 128+ bits, embedding ASNs for hierarchy. Routers need IPv8 stacks; DHCP8 replaces DHCP. “No modification” holds only if networks upgrade gateways first, creating hybrid pain points. Full rollout demands new routing daemons (BGP8 et al.), WiFi firmware, and Zone Server deployments.
Security trade-offs emerge. OAuth2/JWT centralizes auth—convenient, but Zone Servers become juicy targets. Compromise one, and you own the network. Unified logging (NetLog8) aids forensics but raises privacy flags under GDPR or similar. WHOIS8 mandates route registration, curbing anonymity but exposing AS paths.
Adoption hurdles loom large. IETF standards favor incrementalism; IPv6 took 25 years for partial uptake. IPv8’s suite—10+ interdependent drafts—risks chicken-egg problems. ASNs cost $500-$3000 yearly; smaller players might balk at managing 4B addresses. Routing table compression helps backbone routers (current tables stress 1-2M IPv4 routes), but ISPs must buy in.
Why this matters: IPv4’s 4.3 billion addresses strain under 5 billion+ internet users. Carrier-grade NAT masks it, but breaks peer-to-peer apps and complicates security. IPv8 offers IPv4’s simplicity with expansion, potentially accelerating IoT (50B devices by 2030, per Statista) without dual-stack nightmares. If Thain’s team prototypes and demos—say, on a testbed with real ASNs—it could gain traction. Absent that, it joins the draft graveyard.
For security pros: Watch for egress filtering’s impact on DDoS mitigation and prefix hijacks. Crypto implications? None direct—focuses on L3/L4, but JWTs invite token replay attacks if not tuned. Finance nets might adopt for compliance telemetry. Track updates via IETF datatracker; latest version active as of now.
