On Jan 9, 2025, at 03:32, Joe Abley <jab...@strandkip.nl> wrote: > > ...trying to provide synthetic answers at an authority server is error-prone, > resource-intensive and slow.
(This response isn’t directly related to Joe’s response…just tagging that as a jumping off point.) This thought hit me - the Internet began with the philosophy of smart edges and dumb core, as opposed to the Telephone System it battled for dominance an eon ago. Here the discussion is how and why the server (the core) is doing the work for the edge because the edge (stub) is weak. What that does is upend the strategies for building an Internet-like thing, something with an uncountable population of participants and we end up with discussions ranging from “what would be best for an earnest relying party” and “what can be done to limit abuse of the system.” One option (blue sky again) is to switch to an arrangement where the server, for synthetic or tailored responses, reply with a formula (function). The downsides of this include more resources needed to cross the communications medium (bandwidth, time) and the server has to reveal more than the “need to know.” One high-profile upside is privacy of the querier, no longer needing to provide identifying details or having to reveal intent. Another upside is that the querier might be in a better position to know which service address, when given a choice, is best for its needs. That last is tricky, when faced with a choice of queues and bank tellers, the bank may prefer to spread the load over more, shorter lines to lower service time (hence happier) customers. Customers might prefer to use the one bank teller they like instead, resulting on one longer queue and even if it means longer average/mean service time. A service provider could decide to reveal some but not all service points to still allow queries to make a choice…but at this point, administration of this would be over complicated and probably neglected in operations. _______________________________________________ DNSOP mailing list -- dnsop@ietf.org To unsubscribe send an email to dnsop-le...@ietf.org
