Right now that is the case. How are we to know whether a use case develops somewhere down the road that requires individual allocations rather than netted through someone’s home network?

Nobody in the ARPANET days could have predicted the scale to which things would reach even within their lifetimes.

We should avoid repeating the same mistakes. That “math is on our side” is a somewhat dangerous assumption.  Never mind.  Like I said we’ll all be better off discussing this with the benefit of full 20/20 hindsight.

Meanwhile it is detracting from the main thrust of this discussion, that allocating these ranges to bad actors is something RIPE NCC should be actively working to prevent.  Doing precisely this, handing out /15s at a time that ended up in Spamhaus type blocklists, was a major contributor to the early exhaustion of IPv4, as it is.

—srs

Hi,

On Tue, Aug 04, 2026 at 01:27:36AM +0000, Suresh Ramasubramanian wrote:

> Yes but the conditions at which you calculated IP allocation are set to change if AI comes into the picture for provisioning instances, and IoT catches on more than it already has, to name just a couple of use cases.

400 zillion IoT devices in a single /64 will still be "just a single /64",

and will not affect allocation rate in any way, given that every household

can have a /56 today, or /48.

But that is totally irrelevant - you keep claiming that the way the

NCC is "allocationg space to spammers" would exhaust the IPv6 space,

and we tell you "no, here, math".  Then you bring up something totally

different - unrelated to abuse, spammers, etc.

Gert Doering

        -- NetMaster

--

have you enabled IPv6 on something today...?

SpaceNet AG                      Vorstand: Sebastian v. Bomhard,

                                           Karin Schuler, Sebastian Cler

Joseph-Dollinger-Bogen 14        Aufsichtsratsvors.: Dr. Frank Thiäner

D-80807 Muenchen                 HRB: 136055 (AG Muenchen)

Tel: +49 (0)89/32356-444         USt-IdNr.: DE813185279