Re: [v6ops] Scope of Unique Local IPv6 Unicast Addresses (Fwd: New Version Notification for draft-gont-6man-ipv6-ula-scope-00.txt)

[Brian E Carpenter <[e-mail]>]

So, my thought (and it belongs on this thread OR the 'IPv6 addressing: Gaps?' one) is something like:

We should abolish, delete, expunge and deprecate the word "scope" from all IPv6 documents. It clearly doesn't have an agreed meaning, so it is worse than useless.

All addresses have a region of reachability. This may be confined to a single "link" (whatever a "link" means), some type of limited domain (such as, but not limited to, a "site" (whatever a "site" means)), or to a large part of the Internet as a whole (knowing that there is in reality no such thing as "the" Internet.)

LL addresses MUST NOT be used off a given L2 link.
ULAs MUST NOT be routed outside a given limited domain.
GUAs MAY be routed anywhere.

(MANETs and other mesh networks don't fit in there very well.)

Regards
   Brian

On 19-Feb-21 12:58, Fernando Gont wrote:
> On 18/2/21 20:39, Manfredi (US), Albert E wrote:
>> -----Original Message-----
>> From: ipv6 <[e-mail]> On Behalf Of Fernando Gont
>>
>>> Well, this is a spec inconsistency. You have one spec (RFC4007) defining
>> "scope" and "global scope", and another specs:
>>>
>>> a) making use of the same terms in an incorrect way, or,
>>>
>>> b) employing same terms but with a different definition.
>>>
>>> i.e., either the definition in RFC4007 is incorrect, or the use in
>> RFC4193 and implicit use in RFC4291 is incorrect.
>>
>> You can also argue, if there are prefix bits sent in the clear, and those prefix bits are used to send the packets to a pre-determined gateway, and that gateway is then used to decrypt all of the remaining address bits, then route packets through a walled garden intranet with global span, then global scope could still apply.
> 
> "global span" is defined as "Internet-wide" span. i.e., if an address 
> does not unambiguously specify an interface Internet-wide, it's not 
> global scope as per RFC4007.
> 
> 
> 
>> Just sayin'. These still aren't like RFC 1918.
> 
> The only practical differences I see with respect to rfc1918 are:
> 
> 1) ULAs are not intended to be used with NAT.
> However, were RFC1918 strictly specified to be employed along with NAT? 
> Besides "not indended" != "won't be".
> 
> 2) ULAs are intended to have a small probability of collision when a 
> subset of ULA-based networks are interconnected.
> 
> This is the product of mandating that some bits are generated from a 
> PRNG, plus the fact that ULAs have more bits than their RFC1918 counterpart.
> 
> If I have missed any other differences, please enlighten me. :-)
> 
> Thanks,
>