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

[Fernando Gont <[e-mail]>]

Hi, Bob,

On 16/2/21 14:14, Bob Hinden wrote:
[...]
>> 
>> FWIW, I don't think this issue has lead to interoperability
>> problems. For the most part, the issue has to do with
>> terminology/architecture, rather than whether things work in
>> practice.
> 
> I am having a hard time seeing value updating a lot of specifications
> (even assuming we could agree about what to change) if there is not
> an interoperability problem. 

There's a mismatch between a protocol spec (RFC4291/RFC4193 saying that 
ULAs are global) and RFC4007 defining scope and global scope.

RFC4007 is an architecture document -- so you want get interoperability 
issues there, since an architecture is mostly abstract.

That said, there are pieces of code (the referenced Python library) that 
do make use of such specifications an architecture.

If one wanted to direct Python's folk to change their library, one would 
need to provide some basis for that. BUt at the time of this writing, 
such basis is contradictory: because RFC4193/RFC4291 says ULAs are 
global, while de "global scope" definition of RFC4007 says it's not.

I ran into this issue when I wanted to offer this explanation myself -- 
but ended up finding conflicting specs.

If the "IPv6 scoped addressing architecture" (RFC4007) is important 
enough to have it -- and assuming we're going to have IPv6 for a long 
time (which I think we all do), it's certainly good to have specs and 
architecture aligned. Otherwise, if the architecture document is not 
important, one might as well fix the issue by simply obsoleting it.

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492