Re: [DNSOP] back to: Some distinctions and a request

[Mark Andrews <[e-mail]>]

In message <[e-mail]>, Edward Lewis writes:
> On 7/2/15, 13:34, "DNSOP on behalf of Paul Vixie" <[e-mail]
> on behalf of [e-mail]> wrote:
> 
> >manning wrote:
> >> ... STRONGLY suggests that =E2=80=9Cdomain-looking-string=E2=80=9D is , in
>  fact, a
> >> host that is identified using the Internet DNS.
> >
> >i agree with this interpretation, which means, it's the spec itself
> >that's wrong, not hugo's interpretation of it. the internet people
> >didn't love .UUCP addresses either but that didn't stop them from working.
> >
> >what the internet should be doing is defining escape mechanisms for
> >non-internet systems, rather than saying "we are the only thing you can
> >use".
> 
> At the risk of further annoying Andrews ... if there was a definition of
> domain name in contexts external to the DNS, that would be helpful.  Plus,
> in each context, what are the escape rules, if needed?
>
> E.g., At one time, some "funny guy" tried to register ctrl-G as a TLD.
> (He knows who he is.)  How would that be written in a URL?

In a domain name: \007 (RFC 1034 presentation encoding)
In a host name: not possible as it is outside the allowable syntax.
In a url it would depend upon the scheme.  It would not be valid for
http:, https: or mailto: to start with as all three are restricted to
using hostnames.  For those schemes where it is valid input %07.

Mark
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: [phone]                 INTERNET: [e-mail]