Re: [regext] IANA Considerations in draft-ietf-regext-rdap-reverse-search

[Patrick Mevzek <[e-mail]>]

On Fri, Jul 31, 2020, at 11:21, Hollenbeck, Scott wrote:
> Note "supported extensions". This is why I'm saying that we need to 
> register all extensions with IANA

I agree.

> and include them in the 
> rdapConformance data structure even if they don't describe a response 
> extension. 

I agree, everything should be listed in the reply to an help query.

I am just saying that for any other reply that is a specific one on a specific resource
then the rdapConformance should just list the "extensions" needed to understand this
specific response, and not list absolutely all extensions the server knows about
(and that are irrelevant for this specific response).

The list of what is written in the response should certainly not be just server policy,
especially if there is no automated way to learn about this policy. Otherwise if you include
options like that (the list presented might be the list of all server extensions OR only the subset needed for this specific response) AND there is no way for the client to know which
case he is in, it immediately creates interoperability problems. I prefer no such options
and the protocol clearly defining the content. Or if such options are really needed
(if help response is always all extensions, and any other response is just the specific extensions needed, then nothing more is needed), there should be  a signal to know which
case we are in.

> The help response should include supported 
> extensions that are available to that client.

Yes, the help response allows to "discover" all possible extensions from a specific client.

-- 
  Patrick Mevzek
  [e-mail]