I just tested the new IPv6 support on a VC1S server.
Sadly, the IPv6 support is very badly implemented. Each server or VPS gets a single address (!) inside a /127 subnet. The other address is used as the gateway. Moreover, it looks like a shared /64 prefix is used for multiple customers, or at least multiple servers of one customer. Also, link-local connectivity is broken.
RFC 6177 specifically recommends prefix lengths of /64 or shorter for end nodes in a way suitable for further subnetting (i.e. routed, not on-link). A single address or on-link prefix makes many applications impossible in practice, for instance implementing some kind of VPN. A shared /64 prefix is particularly harmful, too. Blocking nodes (due to abuse etc.) is usually done on /64 granularity.
Seriously, WHAT WERE YOU THINKING? This is not a viable IPv6 setup. It is a mess. It is a good example of how NOT to implement IPv6. Please don’t ignore every best practice regarding IPv6 in existence and then boast about having IPv6 support. It’s just ridiculous.
And as for C1? I remember IPv6 support being planned and promised, but I guess it will never happen now.
(Sorry, just needed to vent a bit.)