I hope you don't mind replying publicly to this, because I think you did a very good thing describing it as a one-act play. (Seriously.) Kinda like why we parody Harry Callahan at times. (That's "Dirty Harry's" name for those under the age of, oh 50.)
On 7/1/15, 16:59, "Warren Kumari" <war...@kumari.net> wrote: >(Applicant enters stage left) But what if the applicant's name is Godot (and never appears on stage)? [https://en.wikipedia.org/wiki/Waiting_for_Godot] Let's say someone does this. They squat on .foo. They never engage the IETF. Traffic is seen for .foo. A lot of DNS queries to 8.8.8.8 ask for foo. names and data, I mean. The operators of recursive servers have no idea what is generating the ever-growing queries for foo. and so, like other annoying queries, begin to filter them. One day someone else wants to launch a .FOO TLD. (Stay tuned for the sequel "Universal Acceptance") The objective of a registry is to map object to entity for the sake of (among a few services) uniqueness. The latter part of that statement is often forgotten. The uniqueness is not just to protect the entity to which the object is mapped, but to protect all others from running into it. Collisions are not always hi-jacks, where the collider wants to tackle the object, sometime they are like tripping hazards where the collider just didn't see it or have a chance to see it. One red herring here is that there's a domain name market. There is. To the IETF that should simply mean that the likelihood of collisions is higher than it was without a market. And yes, ICANN has processes for dealing in that market. But that's not important here. There's one other factor in all of this, coming in out of the blue. I recall a talk at DNS-OAC in spring of 2013 where someone tracked down a source of query traffic that could be categorized as "annoying." After much effort, the source was a music app using the DNS as a "ping/keep-alive" service, meaning the app, which had not reason to use DNS much, know that by repeatedly sending traffic it kept up tunnels, etc. The point is - the presenter went though a lot of trouble and only with luck, identified the source. This is part of the desire for a registry - so we can figure out where and why traffic is coming into servers. That talk - I just can't find it...I thought it would be here, but I don't see it: https://indico.dns-oarc.net/event/0/contributions Or maybe I'm making it all up. That's why I still think there's a need, even with an alt-TLD, for a process to register special use names, special in a technical way.
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
