guidspace.arpa or similar seems like a good potential solution to any of
the usecases around such "magic" negotiated systems where the user isn't
going to need to type in or see the actual name.  By generating a guid, any
system could easily generate unique names without any registration.

But it doesn't do anything for usecases that want nice names for users to
directly see or enter.  Anything with a guid is unlikely to ever be
considered a "nice" name.  Is this a usecase that also needs to be solved,
or are nice names even more into the territory where the system should be
registering a name first?

On Tue, Nov 17, 2020 at 1:57 AM Brian Dickson <[email protected]>
wrote:

> Comments made in the chat, about the private-use presentation/draft:
> Me:
> One potential approach is to say (in the RFC) that one of the two-letter
> reserved codes should avoid name collision by putting a collision-resistant
> second-level label, below .zz and above the private use usage (and use that
> particular two-letter code in that manner exclusively).
>
> E.g. whatever-i-want.guid-as-label-to-prevent-collisions.zz rather than
> whatever-i-want.zz
>
> Warren:
> @Brian: Yes, but as you know (being a registrar), people really want
> semantically meaningful names... People have seen using www.corp, not
> www.dfads3e4r34324rwefe.corp..
>
> Me:
> Correct, and I think possibly providing guidance that this private use
> SHOULD be generally limited to "magic" things like automation or
> automated/negotiated systems. The UI can/should hide the GUID component and
> possibly also the zz TLD
>
> Ben:
> That could be $guid.guidspace.arpa, no need for .zz
> "guidspace.arpa" is not a thing. I'm just suggesting an adjustment to
> Brian's proposal.
> Me:
> That's a good idea, can be pursued independently of the private-use TLDs
> like .zz
> (possibly addressing different use cases?)
>
> Brian
> _______________________________________________
> DNSOP mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/dnsop
>
_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to