We Need to Talk About the IPv8 Draft - wolfy https://substack.com/home/post/p-194315405
Quoting from there : IPv4 backward compatibility is genuinely elegant. Setting r.r.r.r to
0.0.0.0 makes an IPv8 address identical to an IPv4 add, processed by the standard rules (#1.5 p3, #3.3). This alone is the best contribution of the draft and should be lauded. It single-handedly kills any further incentive to push IPv6, which has struggled for decades to find adoption due to dual-stack, flag day, and forced migration (#1.2 p3, #2.2 p2). Any draft that replaces this initial iteration of IPv8 should include this design choice.
The IPv8 concept is NOT, under any circumstances, backwards compatible with IPv4. Any IP speaking device reads the first 4 bits of the header to identify the version number. The only 2 valid values today are 0100 (4) and 0110 (6). Just as an IPv4 only device won't know what to do if the field is 6, no IP device will know what to do if the field is 8, as proposed in Sec 6.1 of the draft. The author states in the draft that this would never happen because : IPv4 devices behind IPv8 routers continue to operate because the
router downgrades at the boundary. * IPv4 applications on IPv8 hosts continue to operate because XLATE8 handles version translation on their behalf.
Congrats on re-inventing 6to4! On Thu, Apr 30, 2026 at 4:18 PM Rich Kulawiec via NANOG < nanog@lists.nanog.org> wrote:
When this was floated on various IETF mailing lists, someone took the time to write:
We Need to Talk About the IPv8 Draft - wolfy https://substack.com/home/post/p-194315405
---rsk _______________________________________________ NANOG mailing list
https://lists.nanog.org/archives/list/nanog@lists.nanog.org/message/QD6JBVII...