[arin-ppml] Addressing for other planets

[John Curran]

On Feb 20, 2026, at 11:22 AM, Tony Li <tony.li at tony.li> wrote:

Hi Fernando,

On Feb 20, 2026, at 7:43 AM, Fernando Frediani - fhfrediani at gmail.com <mailforwards at cloudmails.net> wrote:

I am following this and not beleiving this is serious. Forgive me if not but it looks like April's fools day

This is quite serious. Space agencies are already using IP for space missions. That only looks to expand. Today they are using random IPv4 allocations from their own prefixes.  They foresee many additional missions and collaboration between agencies to provide communications backup. If nothing changes, this will become another swamp.

Tony -

By “swamp”, you’re referring to pre-CIDR IPv4 routing swamp??  This is where it gets a little confusing - I guess I’d like to better understand the architectural logic behind the “another swamp” concern.

The historical IPv4 swamp emerged largely because the early allocations were not aligned with provider-based aggregation, which was later addressed through CIDR and provider-aligned issuance.

The draft-li-tiptop-address-space-01 ID draws a parallel to that period, but the proposed solution appears to move toward celestial-region–based allocation rather than provider-based aggregation – I understand the desire to get large dedicated assignments for each deep-space network operator, but that can be done in a very straightforward manner and optimum aggregation.   You suggest that without coordination, space missions would likely receive per-mission allocations from various agencies/operators, but that’s actually a good thing if those allocations are temporary in nature.

I guess that it might be helpful to get some more insight into the expected operating environment, since going by the draft as written it would appear that celestial body based issuance would require deep space operators to carry additional and distinctly non-aggregataable routes for all elements served that have celestial-based allocations…

Can you outline some scenarios where celestial-based allocations (as opposed to deep-space network operator based allocation) make sense?   I would try to create such myself but the draft doesn’t provide any insight into expected number of agencies, operators, missions, etc.

Providing an example scenarios might help folks understand the aggregation model you envision and how it avoids – rather than directly leader to – the non-provider-based allocation dynamics that produced the IPv4 swamp.

Thanks!
/John

John Curran
President and CEO
American Registry for Internet Numbers





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20260221/4f905ed7/attachment.htm>