[regext] Re: I-D Action: draft-ietf-regext-rdap-extensions-02.txt

["Andrew Newton (andy)" <[e-mail]>]

On 9/5/24 07:54, Gould, James wrote:
> Andy,
>
> That language looks better.  I believe it would be good for draft-ietf-regext-rdap-extensions to cover how new RDAP JSON Values types are defined.  The RDAP JSON Values registry can be extended by Type and by Value.  The definition of a new RDAP JSON Values type could include the expected format of the values.  Should clients do an exact match of the values or a case insensitive match of the value?  I believe there should be no conflicting values based on case in the registry and clients should implement a case insensitive match instead of an exact match.  Some values may benefit from the use of case and some values may not.  The values for the "notice and remark type" and the "redacted reason" could benefit from the use of mixed case since they're not identifiers but use sentence form.
>
> Thanks,
>
James,

Good additions.

WRT to the expected format for types, the current language says that a 
new type MUST be registered with a stable reference (actually, the 
language has a double negative which needs to be fixed). If we are to 
have the DE's looking at values and evaluating if they meet an expected 
format of a type, I think types should only be added via IETF action. 
That is, the DE's should not be expected to evaluate formats that are 
poorly specified.

Thoughts?

-andy