[arin-ppml] Addressing for other planets

[scott]

Hi All,

On Tue, 24 Feb 2026, Tony Li wrote:

Bill said:

>> I watched the TIPTOP presentation at APRICOT a couple weeks ago. It
>> sounded like the idea is to hew as closely as practical to the
>> existing protocol standards and practices that we have now, rather
>> than invent an interplanetary-specific network stack. Relax the timers
>> and change the buffering expectations. Is that about right?
>

To which Tony said:

>
> Yes, that’s about right.
>
> As you well know, the Internet and IP are enormously flexible and 
> powerful. Space agencies would gain many advantages from using as much 
> of our technology as possible rather than re-inventing everything from 
> the ground up (literally :-).
>

I take issue with the notion that there is a need to "invent an 
interplanetary-specific network stack" or for "re-inventing everything 
from the ground up."  To do so ignores the existence of the Bundle 
Protocol networking stack, which is a stable, standardized method for 
delay and disruption tolerant networking suitable for deep space 
environments.  Both CCSDS and IETF have standardized this work, multiple 
implementations exist, and are in production in space environments today.

As such, the assertion that there are not tools which exist to meet the 
needs of deep space networking is an erroneous one, and should be treated 
as such when considering the requirements for interstellar networking.

That said, BP is not a panacea for deep space networking.  It has it's 
limitations as well, particularly a lack of user-facing standardized 
applications.  Tools like NTP, for example, are not really viable in a BP 
only network.  BP excels, however, at delivering payload data across high 
latency or disrupted links, as it was designed to do.

To overcome the lack of end user facing applications, IP networking is 
indeed desirable in locations which can support it locally; i.e. on the 
surface of another world, or on a large spacecraft.  This creates 
localized IP networks.  To make these interoperable with terrestrial IP 
networks via transit of intervening BP networks, application layer 
payloads are encoded as bundles, in a manner which conforms to the 
specific requirements of delayed or disrupted links; i.e. avoid 
handshaking and lookup queries across these network segments.

A full architecture for interoperable interplanetary DNS and SMTP (as a 
demonstrtion application/service) using this method has been developed, 
demonstrated, and spoken upon to both the ISP and Space communities.  To 
ignore the existence of this and other BP related work is akin to burying 
one's head in the sand.

That said, we agree that dedicated IP resources should be 
allocated for/deployed on other worlds, however this is not for route 
aggregation purposes, but route filtering purposes.  In this way, 
robust localized IP services can be enjoyed on other worlds, while 
interoperability with the Internet can be maintained in the application
layer via transit of BP networks, in those cases where said application 
can be made delay or disruption tolerant.  To be clear, it is not enough 
to encapsulate packets in bundles, nor is it sufficient to tweak timeouts 
and buffers.  Specific pragmatic rules exist for delay/disruption tolerant 
application design, for which there exists a textbook, authored by a 
co-chair of the IETF DTN WG.

I urge everyone to make themselves familiar with the rigors of deep space 
networking, and the limitations which are imposed by same.

Sincerely,
Scott Johnson