I've been reading IETF draft-martin-retry-over-ipv6-02 which proposes serving a HTTP 566 error code over IPv4 to indicate that a site is only available over IPv6.

I've thought about this before since I've been looking for a way to signal to IPv4-only clients that a site is not "broken", but that their endpoint is missing connectivity.

Needing to maintain a whole IPv4 stack to do it means it's good for transition, but maybe less useful longer-term.

I don't really see why browsers don't already gracefully handle the IPv6-only-site-on-IPv4-only-network situation. I wonder if it'd need a standard to exist before any would adopt it. And where that standard would live.

#IPv6

Bitte, sperre ich eben die halbe Türkei aus.

Nein, ich schreibe keine Abuses, ich blockiere Providernetze.
Notfalls alle 2^92 Adressen.

#ipv6 #spam

Finally! My 2a06:9801:7c9::/48 subnet is online via AS219366

https://bgp.tools/prefix/2a06:9801:7c9::/48#

and can be pinged! 💪

Next: Setting up multiple POPs worldwide and creating the anycast network.  

#network #bgp #ipv6 #routing #asn

https://tf.nist.gov/tf-cgi/servers.cgi

> The generic name time.nist.gov will continue to point to all of our servers on a round-robin basis, and users are encouraged to access the service using this name.

Except that it doesn't any more. That name now points to a load balancer that is only reachable via legacy IP.

#ipv6

Saw-horses, or IPv6? You be the judge!

#woodworking #ipv6

If you use #MySQL, watch out as the INET6_NTOA function is not compliant with #IPv6 RFC 5952 §4.2.2.

In particular:

SELECT INET6_NTOA(UNHEX('00010002000300040005000600070000'));
yields "1:2:3:4:5:6:7::" rather than "1:2:3:4:5:6:7:0".

(For added fun, perl's Net::IP rejects the former as a corner-case due to a bug.)