Re: [v6ops] IPv6 addressing: Gaps? (draft-gont-v6ops-ipv6-addressing-considerations)

[Ted Lemon <[e-mail]>]

Neither of those are good options. We’ve discussed this ad nauseum. RFC 4007 is correct. We do not need a new scope. A document that addressed this problem would be almost completely different. This is not a good starting place. 

> On Feb 18, 2021, at 17:10, Fernando Gont <[e-mail]> wrote:
> 
> Hi, Ted,
> 
>> On 18/2/21 18:53, Ted Lemon wrote:
>>> On Feb 18, 2021, at 4:32 PM, Fernando Gont <[e-mail] <mailto:[e-mail]>> wrote:
>>> This text predates draft-gont-6man-ipv6-ula-scope which effectively does try to formally update the scope of ULAs.
>> I know you haven’t asked for a call for adoption yet, but it’s pretty clear that you’re in the rough on adding a “local” scope
> 
> The intent was not to add a local scope, but to fix the inconsistency among RFC4291, RFC4193, and RFC4007.
> 
> It could very well mean adding a local scope (even if it's ULA specific), obsoleting RFC4007, or any other option.
> 
> Right now, the "IPv6 Scoped Addressing Architecture" (RFC4007) disagress with the definition of ULAs as "global scope" (as per RFC4193/RFC4291).
> 
> Yes, networks do not break down as a result of this (it's unlikely that that might happen as a result of an architecture document which, by definition, it's going to be rather abstract).
> 
> But we already have existing cases (e.g. Python's library) where they solve our spec inconsistencies in their own way. And, certainly, we cannot really offer authoritative responses, because we have different specs arguing in different ways.
> 
> Given that we'll likely have IPv6 for a while, I would understand if people have different ideas regarding how inconsistencies can be solved (e.g. "new scope" vs "obsolete RFC4007"). But IMHO it would be awkward for folks to argue in favor of keeping conflicting specs.
> 
> Thanks,
> -- 
> Fernando Gont
> SI6 Networks
> e-mail: [e-mail]
> PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
> 
> 
> 
>