[discuss] Leading zeros [Report from the BR meeting local organizing group - Dec 2013]

Tue Dec 31 19:18:25 UTC 2013

On 01/01/2014 05:22, Roland Perry wrote:
> In message <52C1CE65.6080308 at gmail.com>, at 08:49:57 on Tue, 31 Dec
> 2013, Brian E Carpenter <brian.e.carpenter at gmail.com> writes
>>> If IPv6 had been devised as backwards compatible, then my problems (and
>>> those of millions of other business users) would be much less.
>>
>> This meme needs to die. IPv4 is the problem - since it contains no
>> provision for variable length addresses, there is mathematically no
>> such thing as a backwards-compatible design for extended addresses.
>> We didn't invent the dual stack mechanism because we liked it, but
>> because (regardless of all design details of IPv6) there was
>> mathematically
>> no alternative.
> 
> You just need to add a sufficient number of zeros in front of the IPv4
> space, to give it a niche in IPv6 space.

Doesn't help. An IPv4-only host still can't interpret an IPng
packet, which is the fundamental problem. This idea was
investigated in great depth (it was mentioned in RFC 1671, but I
didn't invent it) but it simply doesn't solve the coexistence
and interop problem.

It was actually in the IPv6 addressing architecture from 1995
(RFC 1884) and was only officially obsoleted in 2006 (RFC 4291)
due to having no operational value.

   Brian

> 
> (Or something a bit more sophisticated if there's a good reason to do so).