Re: draft-bourbaki-6man-classless-ipv6-00

[David Farmer <[e-mail]>]

Sent from my iPhone

> On Jun 16, 2017, at 19:00, Brian E Carpenter <[e-mail]> wrote:
> 
> Because, I think, the word 'prefix' has three distinct aspects in
> IPv6: a prefix used by a routing protocol, a prefix used by a
> node to determine if another node is connected to the same link,
> and a prefix used to construct the complete address of a node.
> We have generally been a bit careless in distinguishing these
> three aspects.
> 
>   Brian

Also, in the last two aspects of an IPv6 prefix, the part of the address leftover, the righthand side, can quite reasonably be referred to as an Interface Identifier (IID) in both cases. And, of those two aspects of IID, only the last one is currently defined to be 64 bits, the other can have any length from 0 to 128 but it Is 64 bit as well in most cases. 

This is all exasperated, by using the term subnet prefix in RFC4291 and it's predecessors, because in IPv4 the last to aspects of a IPv6 prefix, and the two aspects of IID, are bound together and called a subnet prefix, and in IPv6 those two aspects are supposed to be separate.

While I could support a variable length addressing prefix and IID that is no small change and is not something to be included in RFC4291bis, but clearly resolving the ambiguity of the three aspects of an IPv6 prefix and the two aspects of an IPv6 IID is required to move RFC4291bis forward to Internet Standard.

Thanks.

David Farmer