Right, this is the reason. I missed the implications of your reference to 
RFC8552 in my previous response.

Names like this however don't actually work with DNS-SD anyway, so it's weird 
that someone would ask for one. DNS-SD uses 
<subname>._sub._service._transport.<domain> to do something similar. subnames 
depend from the service name, so aren't registered.

On Fri, Oct 2, 2026, at 8:47 PM, John Levine wrote:
> It appears that Paul Wouters  <[email protected]> said:
> >
> >Dear DNS and DNS-SD experts :)
> >
> >The Designated Experts for the Underscored and Globally Scoped DNS Node
> >Names IANA Registry (RFC 8552) received a request to register a new
> >entry to be used with DNS-SD. See:
> >
> >https://www.iana.org/assignments/dns-parameters#underscored-globally-scoped-dns-node-names
> >https://www.rfc-editor.org/info/rfc8552/
> >
> >
> >The question is, I believe that all DNS-SD entries are under a
> >parental underscore domain that is not global (eg _xmpp._tcp.local or
> >_printer._udp.local). As such, I believe entries meant for DNS-SD
> >are unrelated to the Underscored and Globally Scoped DNS Node
> >Names IANA Registry and should not (need not?) be registered there.
> 
> I agree for two-ish reasons. You're right, these don't go in the public DNS, 
> but
> names are cheap and it's nice to avoid needless confusion, BUT ...
> 
> The important reason is that they are all sub-labels. We had the same 
> confusion
> with the names in SRV records. The transport (usually _tcp or _udp) which goes
> immediately before the base name goes in the registry, the service names which
> only go before another underscore name do not. It looks to me like the service
> names are already registerered so there's nothing for you to do.
> 
> R's,
> John
> _______________________________________________
> DNSOP mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
> 
_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to