[jerf]

The fact that it "deserves" its own page is precisely my point. "It has more features" -> "it takes more implementation work than just replacing all addresses". Even ignoring the base work of making things dual-stack you have fundamental rewrites of base assumptions about addresses to take into account.

[scurvy]

Actually, ipv6 takes less implementation work from a network operations standpoint. Every site gets a fixed size network, every building gets a fixed size network, every vlan gets a fixed size network. No more IP budgeting and wrangling nabout justifying IP allocations. "Is that worthy of a /18 or should it be a /20?" That's all gone. Every vlan with people gets a /64 and your router interface to them is fe80::1. Turn on SLAAC and let it run. No more NAT pools and rules. It's great!

[XorNot]

I've been pushing this idea at work lately. We've got a whole network of /28's in IPv4. Provisioning new servers is a nightmare.

[scurvy]

I'd highly suggest you follow the NANOG BCOP. There's a site for it: http://nabcop.org/index.php/IPv6_Subnetting

Here's where I got the idea of using fe80::1 everywhere for the link-local gateway: https://www.edge-cloud.net/2013/08/ipv6-link-local-addresses...