Hi Mario,

From: Mario Loffredo <[email protected]>
Date: Wednesday, August 20, 2025 at 3:30 AM
To: Jasdip Singh <[email protected]>, [email protected] 
<[email protected]>, Mario Loffredo 
<[email protected]>, [email protected] <[email protected]>
Subject: [regext] Re: Fwd: I-D Action: draft-ietf-regext-rdap-jscontact-22.txt


<snip>


On 18.08.25 14:45, Mario Loffredo wrote:

[ML] I think a reasonable compromise is to allow both formats, but we could 
require that they NOT be used simultaneously. This could avoid confusion for 
clients. The same could apply to the full name and name components.

[JS] Glad to see that we won’t be excluding unstructured addresses in RDAP 
JSContact. Though, not sure about "they NOT be used simultaneously”. Even if 
structured components are available, unstructured data is typically intended 
for human-friendly presentation of an address or a name per some economy or 
culture. IMO, the guidance should be to encourage supplying structured 
components but not discourage additionally providing unstructured data, be it 
for an address or a name.
[ML2] Clients don't need to get the unstructured format if they can get the 
structured one and present the address as they want, instead of how servers 
chose.  It could happen, perhaps it is unlikely but we cannot completely rule 
it out, that there could be inconsistencies between the two formats if both 
were presented.

[JS2] That’s an interesting point -- whether a client should honor the 
unstructured address’ presentation when provided from the server side or not. 
RFC 9083 discusses structured versus unstructured addresses [2] but not 
sufficiently to help with our discussion point.

[2] https://www.rfc-editor.org/rfc/rfc9083.html#name-structured-vs-unstructured-


PK> I agree. I propose to add SHOULD for components, unless there is no 
technical mean to provider structured data.

Regarding JSContact address components, I think that we should only allow 
"name", "locality", "region" and "postcode". The "country" component is 
redundant because countryCode is already allowed.

PK agree

[JS] Per RFC 9553 [1], “country” is country name and “countryCode” its 2-letter 
code. Should we not keep both?
[1] https://www.rfc-editor.org/rfc/rfc9553.html#name-addresses


[ML2] I believe it's always better to manage a code rather than its decoding. 
Furthermore, RFC5733 uses the country code rather than the country name. 
Therefore, I imagine that RDAP servers using registration data based on the EPP 
data model derive the country name from the country code to comply with the 
jCard format and this is also the reason why the vCard CC parameter was 
introduced by RFC8605. The JSContact format does not require this conversion.

[JS2] This makes sense. Come to think of it, even RDAP’s IP Network and 
Autonomous System Number object classes have the “country” member as a 
2-character country code.

 Considering the EPP data model, the name will be presented primarily in the 
unstructured format, while the postal address will be presented in the 
structured format.


Jasdip
_______________________________________________
regext mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to