RFC 8799: Limited Domains and Internet Protocols

HTML is more "legible" than text.

Is this a widely-held opinion?

They point out that different networks have different needs: home network, deep space network, iot network, in-car network, etc. They call these "domains" (bad name, as it can be confused with DNS).

These networks can be isolated, or connected to the Internet trough a boundary. But each of these has specific needs, therefore they argue that interoperability should not not be strictly necessary between new protocols that are designed for a domain-specific use-case (up to now, I think every protocol had to work everywhere and be able to reach, ie, from one home network to another trough the internet[1])

In part 6, they argue that a protocol is needed to (securely/cryptographically) discover what machine belongs to which domain (this is a bit unclear, but likely to determine what domain a new machine is supposed to join, and where the domains boundaries are).

[1] Which obviously wasn't the case in practice: you want to filter some packets to avoid someone on the internet playing whatever they want on your TV, or print something. You could allow anyone on your home network to do so, but not necessarily in a company network either. Such filtering has been done with firewalls, deep packet inspection, etc. until now.

This time, I agree - this is very abstract, and reading the summary, intro and some details, I don't know what concrete problem is being solved.

I think its china saying we have special requirements for our internet and we should standardize that some networks should have different requirements then the rest of the internet (Like we want to require nodes of our network to be authenticated, so that they cant be part of our network unless we say so). Its them wanting to be able to fork the internet, and still call their lockdown version of the internet.

See the section: "3. Examples of Limited Domain Requirements"

...the problem statement starts by saying that home networks might optimize for "compatibility and security by default", whereas a giant CDN might optimize internal traffic for performance and violate some of the general "Internet Protocol" recommendations. That giant CDN's internal traffic might use MANY of traditional IP protocol components, but might not be fully compatible with "the open internet" for a variety of reasons.

"""Nevertheless, because of the diversity of limited domains with specific requirements that is now emerging, specific standards (and ad hoc standards) will probably emerge for different types of domains. There will be attempts to capture each market sector, but the market will demand standardized solutions within each sector. In addition, operational choices will be made that can in fact only work within a limited domain. The history of RSVP [RFC2205] illustrates that a standard defined as if it could work over the open Internet might not in fact do so. In general, we can no longer assume that a protocol designed according to classical Internet guidelines will in fact work reliably across the network as a whole. However, the "open Internet" must remain as the universal method of interconnection. Reconciling these two aspects is a major challenge."""

...the underlying point is that the internet is fragmenting (home != open, cdn != open, home != cdn), and they're calling it out as a RFC... stating the problem.

Their recommendation (towards the end) alludes to tagging / nesting of networking groups. (see: "With this perspective, we can list some general functional requirements.")

Whether a node "IN" or "OUT" of a "LIMITED DOMAIN" is a signal that is useful to know (probably either to restrict or allow traffic in and out of a particular domain).

There are definitely some valid points in there, but it's also possible to take a very dystopian/pessimistic view as to what this allows.

The Good? A local wifi in your car with special rules, hardware, or protocols might be super cool! ...especially if it's compatible with other "CarWiFi(TM)" networks. Same goes for "CDN-Net9000gb(TM)".

The Bad? Those same tools could be very effectively used at political, not just technical boundaries. "US-Net", "China-Net", etc. with varying degrees of compatibility or access between them.

Just my own impression after lightly reading the RFC.

It appears to be the idea that something akin to Microsoft's Active Directory is needed for the Internet instead of it essentially being reimplemented in many different contexts. That such domains would control membership, access, and discoverability for services and applications, based on some standardized protocols.