[Gloucs] IP v6.

[Simon Iremonger (lugs)]

> Personally, I'm skeptical that people won't just keep adding workarounds
> to IPv4 address shortage. I explored this in a blog post, in which I
> suggested that just improving SOCKS a bit would overcome most of the
> downsides of NAT, without the effort of a full IPv6 migration:
> http://www.snell-pym.org.uk/archives/2011/04/11/ipv6-versus-nat/

Consider, however, that, the costs of going through setting up
  gateways and ways to configure them and all the workarounds
  etc etc....     Is comparable to costs of just switching on
  the IPv6 functionality and sorting out the middleware/scripts
  for managing it where applicable...
Then look at the scalability of all the connection-tracking.

A similar story has been looked-at with respect to freeing
  up old huge class-A address ranges.  It will cost lots
  to buy a few moths and this is comparable to the costs
  of just deploying ipv6 once...

> However, since writing, I heard of 6rd, which has given
> me somewhat more hope of IPv6 succeeding eventually:
> http://en.wikipedia.org/wiki/IPv6_rapid_deployment

There are lots of tunneling mechanisms, etc....  Personally
  I think many of these complicate matters.
Certainly, auto6in4 or teredo tunnels being auto-established
  can confuse firewall ACLs and the like.

My College puts out a v6 native router-advertisement prefix
  that works =).  ISP at home connects native v6 over PPPoA,
  etc...  professional network gear has supported v6 for ages,
  consumer stuff is appearing, it'll happen . . .

> ABS

--Simon
(lurker, not from Gloucs lug)