Re: [DNSOP] additional special names Fwd: I-D Action: draft-chapin-additional-reserved-tlds-00.txt

[Stuart Cheshire <[e-mail]>]

On 28 Jan, 2014, at 07:52, Paul Hoffman <[e-mail]> wrote:

> It does, but it doesn't call it out very well. It the middle of section 3, it says that the list is “names that may not be used for top-level domains".

That doesn’t describe what the labels are used for. It describes what they may *not* be used for. What they *are* used for remains unspecified.

If the user types “www.host” into their web browser, the name should *not* be resolved in the global DNS. I get that. But what *should* happen with that name? Should it result in NXDOMAIN, like “www.invalid”? Should it result in 127.0.0.1 like “localhost” does? Resolved via mDNS, like “www.local”? Something else? I have no idea. If it’s the same as one of the other existing special-use TLDs, then an argument needs to be made as to why we need another reserved special-use TLD that duplicates the functionality of an existing one. These names are not supposed to be vanity names. The special-use names are there to trigger special behavior by software, and as such we probably don’t need more than one way to trigger each particular special behavior.

The current use of various de facto reserved names like “.onion” results from there being no formal IETF mechanism for documenting and discussing such uses.

The goal of RFC 6761 was to remedy this omission, and give people who feel they need such names a process to apply for such names and initiate discussion about whether such use is appropriate. That way the IETF community can be involved with these decisions about how names are used, instead of it happening outside the IETF with no IETF scrutiny or input.

I think it would be fairly easy to produce a draft documenting what “.onion” is for, how it works, and why resolving those names via the conventional DNS is not appropriate. I’d love to see a draft like that from one of the people who understands the details.

For some of the other names I don’t know what those documents would say. If people in the IETF community do know what those names are used for, having those people write and submit a quick two-page draft describing the usage would be a wonderful contribution to greater IETF understanding of what’s going on. Observing that certain weird names are hitting the root name servers is a useful first step. Understanding *why* that’s happening would be even better.

Stuart Cheshire