URLs, IPs and interception

[Peter Fairbrother]

Ian Batten wrote:
> 
> On 2 Mar 2008, at 18:26, Peter Fairbrother wrote:
> 
>> Roland Perry wrote:
>>
>>> ipv6 is simply "more numbers, therefore more addresses".
>>
>> You wish. I wish too, come to think of it.
>>
>> I hereby propose IPv8, which actually _is_ just IPv4 with more numbers.
> 
> A common discussion over lunch in our canteen (when we can get a table: 
> on Friday we had Gordon Brown nabbing a table for himself) is just 
> that.  Why, we ask, do telco contracts so rarely mention IPv6?  The 
> conclusion we came to is that IPv6 suffered from creeping featureism (to 
> be charitable) or OSI complexity envy (to be less charitable).
> 
> ian

Yeah, IPv6 is pure shite. IPs should be just like telephone numbers, 
freely available to all. No further functionality is needed. It's useful 
to have the first portion select eg a geographical area, as is done with 
telephone numbers.


I am also willing to donate a single IPv4 address to IPv8 - probably 
x.x.x.x, but I'd have to check as I am too drunk now - but that alone 
*doubles* the IPv4 address space  ..



I also hereby propose IPvX, where X is variable.

Probably been done before. IPv8 should be enough for the next - 
considering Moore's law about 50 years, and more so historical IP space 
development - about 30 years - ouch.

IPv6 might give 10-30 years, maybe, though I'd bet on around 15 years.


And no, I'm not going to write anything more than some suggestions 
(read: orders and suggestions) about transition to either proposal. No 
RFC's.


-- Peter Fairbrother

[*]((2^32 - 1)/2^32, for those who are as pedantic as I can be when 
drunk.. or being mathematical ..)