Hi Joe, Thank you for your questions. I should clarify what I meant by "consensus."I'm explicitly not referring to the IETF draft consensus or the formal standards process. You're absolutely right that new RRTypes are assigned based on expert review without requiring broad IETF consensus.What I meant was the practical ecosystem adoption consensus that follows after a new RRType is defined and assigned. This is a kind of "consensus through adoption" where DNS service providers, registrars, hosting companies, and other ecosystem participants effectively vote with their development resources on whether to implement support for a new RRType.When a specification demonstrates widespread real-world usage, it adds significant weight to the likelihood of broader industry adoption. This creates a positive feedback loop - providers are more likely to support RR types that already have demonstrated usage and value in the ecosystem.Consider an analogy from the banking industry: it took banks nearly 20 years to fully adopt chip cards despite the clear security benefits. This wasn't because of technical difficulty in the protocol, but because each institution had to make practical tradeoffs: the actual costs for upgrading software, retraining staff, updating workflows, and considering customer impact.Similarly, for registrars and DNS service providers to support a new RRType, they need to:
- Modify their customer-facing interfaces - Update their API documentation - Enhance their validation logic - Train customer support staff - Allocate engineering resources that might otherwise go elsewhere This economic and practical "market consensus" is what's challenging to achieve, not the technical implementation of the RRType itself or getting it assigned by IANA.Our prefixed TXT approach is intended to provide a transition path that works within existing constraints while building evidence of demand and adoption that might encourage providers to implement proper support for new RRTypes.I hope this clarifies what I meant by consensus - it's specifically about practical industry adoption rather than the standards process itself.Best regards,Victor Zhou On Fri, Mar 7, 2025 at 12:38 AM Joe Abley <jab...@strandkip.nl> wrote: > Hi Victor, > > On 7 Mar 2025, at 00:10, Victor Zhou <zzn=40namefi...@dmarc.ietf.org> > wrote: > > > Our intention was not to claim it's hard to add a new RR type. it is not > hard to "add" it. It's hard to gain consensus without showing adoption > first. > > What consensus do you mean? > > New RRTypes are assigned based on expert review. Consensus is not > required. > > > It's not hard to declare "I" (unilaterally) am going to start treating > RR Type = N as a new type and the client I develop will treat RR Type = N > that way. > > The normal process is that you submit a template and get an assignment. > You don't need to unilaterally declare the existence of a new RRType. > > > It's extremely hard, however, to show the world that some new RR Type = > N has been used by many services and clients for a specific use case and in > a certain way. > > Which parts of the world are you trying to convince in this example? > > > Joe
_______________________________________________ DNSOP mailing list -- dnsop@ietf.org To unsubscribe send an email to dnsop-le...@ietf.org
