[ppml] The myth of IPv6-IPv4 interoperation, was: Re: Legacy /24s

[Iljitsch van Beijnum]

On 3-sep-2007, at 17:11, Paul Vixie wrote:

>> It looks like you suffer from the (fairly common) misconception

> what an unnecessarily harsh way to phrase your difference of  
> opinion.  "ouch."

Sorry about that, but I figured you'd set me straight if you  
didn't.  :-)

>> The assumption here of course is that it would have been POSSIBLE to
>> build IPv6 in such a way that an IPv6 host can talk to an IPv4 host.

> that was a design goal.

A few years in the multi6 trenches taught me a thing or two about  
design goals...

> of the IPng candidates being proferred, the one we've
> got now was chosen to become IPv6 on the basis of the features it  
> offered.  if
> it wasn't possible to deliver this, then it's possible a different  
> choice
> might have been made.  but more to the point, if the paragraph i  
> quoted had
> been true then IPv6 would be in wide deployment today and the IPv4  
> runout
> would be a yawner.

The point is that the transition to another address length inherently  
means that hosts only implementing the shorter length can't talk to  
ones only implementing the larger length if those actually get to use  
that length as opposed to the old length and a bunch of 0 bits.