Re: [DNSOP] Please review and provide feedback -- draft-stw-6761ext

[Vittorio Bertola <[e-mail]>]

> Il 24 agosto 2019 00:35 Warren Kumari <[e-mail]> ha scritto:
> 
> There was also some discussions with Jacob (or perhaps Alec) saying
> that if this had existed when they started, they probably would have
> used onion.alt instead of .onion.
> 
> Whether or not people would *actually* have used it is unknowable, but:
> 1: at least now they *do* have the option and
> 2: in the future we can point at this instead of just having to agree
> that they didn't have an option other than squatting.

I am in favour of trying this: it is simple and it won't do much harm if it fails, but it addresses a few of the problem cases specified in RFC8244 section 3 (I'd say #7, #8, and parts of #5 and #9). Of course it doesn't address the problem of people who do not know or do not care, they will just continue making up TLDs and using them - though some kind of information and peer pressure effort whenever these cases arise could have some effect, as a practical alternative would now exist.

Perhaps, as a guideline either here or somewhere in a future revision of RFC6761 (i.e. here), application developers should be told that before asking for a special use TLD they SHOULD/MUST experiment with a name under .alt, and prove the existence of some running code, adoption and success before moving to a TLD (and plan their implementations since the beginning so that such a move can actually be done). This would act in two ways: it would avoid wasting energy on discussing abstract special use TLD proposals that seem a great idea to some while others claim that they'll never work, and it would encourage people with successful .alt subdomains to move to a special use TLD to get the benefit of guaranteed non-collision, if they see merit in it.

This is also why not having a registry under .alt makes sense. Having one would make .alt second-level domains almost a functional duplicate of special use TLDs, raising the bar to get them and making special use TLDs only better in vanity/shortness, which would lead the IETF to have to deal mostly with vanity TLD applications.

(Also, you have "handing" instead of "handling" once in 4.1.1 and twice in 4.1.2.)

-- 
 
Vittorio Bertola | Head of Policy & Innovation, Open-Xchange
[e-mail] 
Office @ Via Treviso 12, 10144 Torino, Italy