DNS Attrleaf Changes: Fixing Specifications That Use Underscored Node Names
draft-ietf-dnsop-attrleaf-fix-07

Echoing comments from my review of draft-ietf-dnsop-attrleaf: I believe this
document needs to also include RFC 6763 and RFC 4386; and that it should not
include RFC 6733. Please see that review for details.

---------------------------------------------------------------------------

§1:

>  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
>  "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
>  document are to be interpreted as described in [RFC2119].

Minor nit: please consider using the boilerplate from RFC 8174.

---------------------------------------------------------------------------

§§2.1 and 2.2:

>  An effort has been made to locate existing drafts that
>  do this, register the global underscored names, and list them in this
>  document.

I think this text ("list them in this document") is left over from before this
document was split from draft-ietf-dnsop-attrleaf.

---------------------------------------------------------------------------

§2.3:

This ties back to my discuss on draft-ietf-dnsop-attrleaf, and needs to be
changed in a way that is consistent with however that issue is resolved. The
current list of entries added by draft-ietf-dnsop-attrleaf strongly implies that
the contents of https://www.iana.org/assignments/enum-services were
automatically imported into the namespace created by the Underscore Global
Registry by the simple existence of RFC 7553. If that's the case, it seems that
the following text in this document...

>  For any document that specifies the use of a "URI" RRset

...doesn't capture the real process here. As RFC 7553 will continue to exist
into the future, it seems that the trigger is addition of new Enumservice
entries, rather than the explicit specification of a new URI RRset.

---------------------------------------------------------------------------

§3.1:

>  The specification for a domain name under, which an SRV [RFC2782]
>  resource record appears, provides a template for use of underscored

Nit: "...a domain name, under which..."

---------------------------------------------------------------------------

§3.2:

As a very minor nit, the cited original text for RFC 6117 §4.1 kind of blends in
with the text of this document. I would propose indenting it as was done with
the rest of the quoted content in this section.

---------------------------------------------------------------------------

§3.3:

>  consumes all names beginning with the string "_ta-", when using the
>  NUL RR in the query.

Nit: I believe the record type is called "NULL" rather than "NUL".

Thank you for addressing my DISCUSS and COMMENT.

I support Alissa's Discuss points (and in particular would think that this document
would be fine as Proposed Standard).

One other comment, in  Section 3.1:

Why is the new text only placing a "SHOULD be registered" requirement, as
opposed to MUST?

I don't quite understand why it was seen as beneficial by the group that this doc has been split up, however, BCP does not seems adequate for this part of the doc anymore (maybe PS instead as it updates some PS docs?).

Also, I think that the OLD/NEW style is not really necessary in most cases as simply a sentence/reference to the registry is added.

My previous position was a "plea for clue Discuss" about three MUSTs that appeared in the document, and Dave pointed out to me that those MUSTs will be removed in the next version of this document, because there is already a MUST in the underlying specification that this one is providing usage guidance for. 

Thank you for the speedy feedback!