Re: [art] draft-ietf-dnsop-attrleaf

[I sent this directly to the document-related personnel a few days ago, 
but I'm re-sending it to the ART list as I think it could benefit from 
additional input from the ART community.]

I have some tiny editorial comments, but I'm leaving them aside for now 
to bring up what appears to be a major structural defect in the document.

The document is pretty clear about its intended scope, containing 
statements like:

> Only the right-most names are registered in the IANA Underscore table

and

> The current definition of a global underscore registry attends only to 
> the "upper-level" names used for these RRs, that is the "_proto" names.

But then the table includes things like "_ldap" and "_certificates". The 
only SRV records for (e.g.) LDAP will look like "_ldap._tcp...", which 
(according to the cited statements) would mean that "_ldap" doesn't get 
an entry in this table: it is not the "upper-level" name, and it is not 
the "right-most" name. It is clearly under "_tcp". On a casual check, 
the vast majority of the SRV records in the table shouldn't be in here 
by the rules I cite above.

So, what I *think* this document probably wants to do is define a 
top-level registry containing things like "_tcp", "_udp", and 
"_domainkey", and then create two additional sub-tables (one for "_tcp" 
and one for "_udp"), which register all the "_service" types that can 
appear under these two "_proto" types (e.g., "_sip", "_certificates", 
"_xmpp-client", "_crls"). You might want to also add a table for 
"_sctp", although it would appear to only have two entries at the moment 
("_diameter" and "_diameters" -- cf RFC6733).

The alternative approach to making the document internally consistent 
would be to remove almost all of the SRV entries presently in Table 1, 
and leave them to fend for themselves (which would make this document a 
kind of weak half-measure).

/a

On 7/25/17 9:56 AM, tjw ietf wrote:
>
> Application Folks
>
> We in DNSOP have a draft that is formalizing the use of _underscore 
> labels in DNS, and the creation of a "DNS Global Underscore Scoped 
> Entry Registry".   We want the application folks to give us some 
> comments, as we feel we are fairly  close to Working Group Last Call.
>
> The draft is here:
>
> https://datatracker.ietf.org/doc/draft-ietf-dnsop-attrleaf/
>
> and I'll be glad to shepherd feedback toward the author, or bring them 
> here if it becomes quite contentious.
>
> thanks
>
> tim/suzanne
> DNSOP chairs
>
>
>
>
>
> _______________________________________________
> art mailing list
> art@ietf.org
> https://www.ietf.org/mailman/listinfo/art