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

[Mark Smith <[e-mail]>]

On Wed, 6 Jan 2021 at 23:18, Ted Lemon <[e-mail]> wrote:
>
> On Jan 5, 2021, at 11:38 PM, David Farmer <farmer=[e-mail]> wrote:
>
> I think this is the right direction the previous draft indirectly defined a new scope "non-global", I much prefer explicitly defining a new local scope.
>
>
> Actually, I think you’ve got it right here: the scope is “non-global.”

I







>
> I would add something like the following to better define the relationship between the three scopes;
>
> The boundary of the link-local scope is strongly defined, limiting the extent of the link-local scope to an individual link. However, in contrast, the boundary of the local scope is weakly defined, it is amorphous and imprecise. In some instances, the extent of the local scope can be a single site, in other instances, a group of unrelated sites, a single organization, or even a cooperating group of organizations. Furthermore, the extent of an individual instance of the local scope doesn't necessarily remain constant, it may expand or contract over time as the local situation dictates, for example when two organizations merge. Nevertheless, the extent of the local scope doesn’t encompass the entirety of the Internet which the global scope does.
>
>
> There is at least one obvious problem with this definition: the term “local.” ULAs aren’t really local, despite the name. Using the name “local” is what leads to this confusion. Consider this taxonomy:
>
> GUA: “valid everywhere on the internet scope”
> ULA: “not valid everywhere scope”
> LLA: “valid only on this link scope”
>
> Of course these names are awkward, but I hope they are clarifying. A ULA is “not valid everywhere.” That’s really all you can say about it. You can’t put a ULA prefix in a global routing domain. You can put it in a site routing domain. You can put it in a multi-site routing domain. You can not route it at all. All these uses are valid.
>
> So I don’t really object to your text, but I do object to the name “local.” How about “explicit”? That is, the scope of a ULA is explicit, in the sense that it must be _made_ explicit by the user(s) of the ULA? If that doesn’t work, I’m sure we can come up with a more agreeable term, but please let it not be “local.” Sorry to be a sticky wicket. :)
>

I think "local" is close to, if not the right word.

I consider anything that is non-global or smaller than global to be "local".

I think "local" is a word that has the word's user's perspective and
the context of its use as part of its semantics. The perspective and
context defines what the "local" domain boundary is, and it is
possible that different people might have different definitions,
although they'll probably be pretty similar. People have their own
local perspective on the definition of the word "local" ;-)

When I'm at home, my "local" network is my home network.

When I'm at work, my "local" network is the ISP I work at's network. I
don't consider our customers' home networks "local" though, even
though they're all part of the network with the same AS number.

Some examples of "local" outside of networking.

I go for walks around my local neighbourhood. I ride my bike around
the local neighbourhood. As I can ride further than I can walk, my
local neighbourhood in the context of being on my bike is a greater
geographical area than my local neighbourhood when I walk.



> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------