[ Charset ISO-8859-1 unsupported, converting... ]
> > >   domain name, which is the correct form?
> >
> > Both are valid.
>
> So, they seem to be equal, but what's the point in having the ':' then?
> RFC 3865 says: "The initial set of solicitation class keywords all begin with
> a domain name with the labels reversed, followed by a colon." Isn't that
> normative?

Um, yes, that would be normative.  :)  That's an artifact of the previous
drafting process for 3865 (we thought we were using an iana registry and
then moved to the reversed domain name issue.  Let me address this
specifically.

>
> > No, and for the ddds app, one does s/:/./g as part of the first well-known
> > rule.
>
> Sure, but if there's no semantic difference a comment to that regard could
> clarify this.

Will do.

>
> > >   from the use of the DNS: label length, overall tag length, 
> > > unavailability
> > >   of ".." (likewise "::", ":.", ".:"). Also, neither RFC 3865 nor this
> > >   draft mention case insensitivity.
> >
> > 3865 does, I believe, restrict each part of a label to be valid, and does
> > deal with overall tag length and stuff.
>
> It says that the tag (or list of tags) may not exceed 1000 characters in 
> length,
> but does not constrain the length to 253 characters or the label length to
> 63. Also it allows "an arbitrary set of characters drawn from the following
> construct", where you can't translate example.com:Foo::Bar into a valid
> domain name.
>

Hmmm ... let me figure out how to deal with it.  A valid point.

> > >   this already happened in RFC 3865.
> >
> > Hmmm ... suggestions at this point on what I should do?  Registrant came 
> > from
> > several folks because it implies "person with admin control and people whom
> > have been delegated that capability by the person with admin control".
>
> Zone administrator is a more neutral term because that applies at every level
> whereas "registrant" is ICANNish vocabulary for someone having registered
> a domain name with a "registry". "dissident.example.com" may be administered
> by an entity applying different policies than the entity responsible for
> "example.com".

Not a problem ... I can clarify this.

>
> > > o solicitation class tags have a very restricted character set, especially
> > >   they don't support non ASCII characters. In what form would IDNs be
> > >   handled? Would "xn--" encoding also apply to the past-colon part?
> >
> > IDN sits in valid domain names, so each part of a label could in fact be
> > encoded in IDN.  The colon is syntactically equivalent to a period, so the
> > rules apply to post-colon parts.
>
> But then I couldn't have "org.?ntisp?m:SP?M" in the header (not even properly
> MIME-Header encoded), but would have to write 
> "org.xn--ntispm-9taf:xn--spm-rla".
> For the part left of the colon an application or library might help, but
> identifying the part after the ':' as a domain label is tough.

OK ... I do see the issue here.  Let me consult with a couple of people who
are more idn literate than I am and propose an answer.  I think the answer is
s/:/./ then apply IDN, but I hear what you're saying on the ambiguity and I
think this may require a sentence explaining what to do if you're going to
use the naptr-based scheme in conjunction with 3865.  Again, I think there
is a clean solution, but let me make sure it works.

>
> > Doesn't really encourage them so much as say (along with other DDDS apps)
> > that they have a potential role and you should look to current best 
> > practices
> >.
> > At least, that's what I meant.  :)  (DDDS requires that I state if they
> > are allowed or not, so I did. )
>
> NAPTR suffers from the subtyping problem, even more if applied at a zone apex
> or with wildcards - that's why I'd prefer them being explicitly discouraged.
> And wildcards are often abused to solve the provisining problem, i.e. to
> save the administrator the work to enter several predictable RRs into the 
> zone.
>

I can add a note explaining that issue, but I'd like to come down in favor
of wild cards.  That said, I think the issue needs to be clearly stated.

Regards,

Carl
.
dnsop resources:_____________________________________________________
web user interface: http://darkwing.uoregon.edu/~llynch/dnsop.html
mhonarc archive: http://darkwing.uoregon.edu/~llynch/dnsop/index.html

Reply via email to