250M-plus reserved IPv4 addresses could be released

Original Article ↗ Hacker News Discussion ↗

A long-reserved block of IPv4 addresses (240.0.0.0/4) is being floated for reclassification and potential use, sparking debate over whether this would meaningfully ease address shortages or just create chaos. Network engineers point out that many routers, operating systems, and consumer devices either block or mishandle this range, so enabling it would fragment connectivity and be far harder than simply deploying IPv6 and carrier-grade NAT where needed. Others note that large organizations and governments still sit on underused IPv4 space and argue that reclaiming or better managing those allocations — alongside serious IPv6 rollout — would be more effective than trying to retrofit 240/4 into today’s internet.

Scope of the 240/4 Proposal

  • 240/4 is currently “reserved”; proposal is to reclassify it as unicast, not (in the main proposal) as private RFC1918 space.
  • It’s already used in some VPNs, SDNs, and cloud/internal networks as non-routable hop space.
  • Advocates argue that simply updating the IANA/IETF status would “match reality” and is mostly a policy/ledger change.

Feasibility and Compatibility Concerns

  • Many commenters argue it’s effectively impossible to deploy 240/4 on the public internet:
    • OSes (notably Windows) refuse to assign or use it.
    • Routers and middleboxes often hard‑block it.
    • Debugging partial reachability would be a nightmare (some paths work, some silently drop).
  • Others counter that:
    • Many routers (e.g., OpenWRT, various IoT OSes) already accept it.
    • Windows support could be flipped “in a patch Tuesday,” and code paths that special‑case 240/4 could simply be removed.
  • There is disagreement whether existing internal “squatters” on 240/4 should block public use (some say “their problem,” others say this makes it de facto private space).

Private Space vs Public Unicast

  • Some want 240/4 added to RFC1918 for large internal networks that have exhausted 10/8, especially in mergers and very large enterprises.
  • Others think using ~6% of IPv4 as private space is wasteful; reserved space should become public if reclassified.
  • Comparisons are made to organizations squatting on DoD /8s and to the federal government holding many idle /8s.

IPv6 vs Squeezing IPv4

  • Large contingent: releasing 240/4 is “too little, too late” and distracts from IPv6. A /4 would be exhausted quickly and add fragmentation.
  • Counterpoint: reclaiming 240/4, 0/8, and “.0 broadcast” addresses could free meaningful IPv4, and patches already deployed for 0/8 didn’t break the internet.
  • IPv6 deployment is uneven:
    • Some ISPs (including multiple fiber providers) offer no IPv6; some that do give only dynamic /64s or small, unstable prefixes.
    • Consumer gear often has buggy or opaque IPv6 implementations; support ranges from “works great” to “6 hours of pain and custom daemons.”
    • Users are frequently told to “just disable IPv6,” which some say is ignorance and others say is a pragmatic workaround for broken devices.

CGNAT, Security, and Policy

  • CGNAT is seen as inevitable for residential IPv4; some suggest using 240/4 for CGNAT external pools.
  • Concerns raised that routable IPv6 exposes insecure IoT devices; others reply that firewalls and the vast IPv6 space make blind scanning far harder than with IPv4.
  • If 240/4 became public, several expect RIRs to tightly control allocations and possibly favor networks that already deploy IPv6, limiting big players’ ability to hoard.