Re: [weirds] Fwd: I-D Action: draft-blanchet-weirds-bootstrap-ianaregistries-01.txt

[Ernie Dainow <[e-mail]>]

Some comments.

"the longest match" as used in several places is not fully descriptive, 
and different.

Section 2. Domain Name RDAP Registry
     domain names are found by doing the longest match
This is a string search of the longest match starting from the end of 
the target name and the end of each value in the Domain column.

3.1.  IPv4 Address Space RDAP Registry
     IPv4 address is found by doing a longest match of the target 
address with the values of the Prefix column
The example query for "192.0.2.0/24" matches the "192/8" entry does not 
follow. There is no string search that will return this match. This 
search requires some semantics, which should be explained.

3.2.  IPv6 Address Space RDAP Registry
same issue as 3.1

4.  Nameserver
Isn't the lookup the same as for a domain name? Just refer back to 
Section 2.

5.  Entity
Add recommendation to use links.
     if the entity identifier was received from a previous query, it 
will generally contain a link to retrieve the complete entity. If it 
does not, the same RDAP server could be queried for that entity.

7.  Deployment Considerations
While the suggestions for doing a sequence of DNS checks if there is no 
lookup match could be useful to some clients, I think a much simpler 
recommendation should also be made.
1. The IANA registries are fetched by an off-line process (such as a 
nightly cron job) and stored in files.
2. The client reads the registries from the stored files.
3. If the RDAP client does not find an entry in the IANA file, it sends 
the request to an RDAP redirector (ref: draft-ietf-weirds-redirects).


-Ernie