Creating demand for IPv6

Wed Oct 3 23:14:24 UTC 2007

On 3-Oct-2007, at 1907, <michael.dillon at bt.com> wrote:

> I see that you did not change anything on that page. Specifically what
> is wrong with the wording below?
> ---
> This is a transition mechanism in which the user configures a 6to4
> client in their PC or home gateway. The 6to4 client requests dynamic
> tunnels

(not quite right; the client doesn't actually request anything)

> from a 6to4 server which is found via the anycast address prefix
> 192.88.99.0/24 allocated in RFC 3068.

(most 6to4 implementations allow a relay router to be configured as  
an alternative to the RFC 3068 well-known relay router address. That  
address is exactly 192.88.99.1, incidentally; it's not something that  
needs to be found)

> This tunnel then attaches the IPv4
> host to the IPv6 network using the IPv6 address 2002:V4ADDR::/48. The
> mechanism is documented in RFC 3056.
>
> ISPs can improve connectivity for their customers who are currently
> running IPv6 on their PCs by setting up a 6to4 relay. This avoids the
> increased network latency caused by a trombone path to the IPv6
> destination through a distant 6to4 relay.

(for an ISP's customers to find that relay it either needs to be  
explicitly configured in their client stacks, or it needs to be  
numbered 192.88.99.1 and the clients need to use the RFC 3068 address)

> In addition, a content provider can also add IPv6 access to their
> services by configuring 6to4 on their network

(... and configuring all the servers and related infrastructure  
responsible for those services to use IPv6, using a 6to4 prefix. Note  
that this is not particularly different from any other kind of IPv6  
transit a content provider might decide to arrange.)

> . Again, by shortening the
> routing taken by one of the protocols,

(shortening the IPv4 path over which the tunnel is provisioned is  
clearer; I'm not sure in general what "shortening the routing" means)

> you ensure that there is no
> tromboning of the path and network latency is close to the minimum
> possible.

I did not change anything on that page, either. For the record,  
that's because I have a screaming two-year-old trying to use me as a  
climbing wall right now.

Joe