Re: [regext] RDAP questions

[Bernhard Reutner-Fischer <[e-mail]>]

On 23 March 2017 16:38:31 CET, Andrew Newton <[e-mail]> wrote:
>On Tue, Mar 21, 2017 at 3:47 PM, Bernhard Reutner-Fischer
><[e-mail]> wrote:
>>>> 2) entity handles
>>>>
>>>> At least for DNRs the mixture of registrars and contacts seems to
>be
>>>a
>>>> bit unfortunate at first glance.
>>>> For handles (in the DNR / RFC5733 sense) could conflict between
>>>> registrars and contacts. Furthermore there is no (apparent) way to
>>>> output a ROID which would make it easier to distinguish these two,
>>>> disregarding the role of course.
>>>> I could work around this by requiring a -ROIDSFX for contacts and
>>>none
>>>> for registrars, for example.
>>>>
>>>> So i think my question is how to respond to a non-search query for
>an
>>>> ambiguous entity?
>>>
>>>I'm not sure I understand this. You have two entities with the same
>>>handle?
>>
>> Yes. I think in EPP they are unambiguous.
>>
>> thanks,
>
>So my reading of RFC 5730 doesn't say if ROIDs are unique within a
>registry or unique within a namespace. I do have to admit I don't
>understand how section 2.8 fits into it. Hopefully somebody more
>authoritative can shed some light (I'm looking at you, Scott :) ).
>
>That said, if the entity is truly unique but serves different roles,
>then the ROID is good enough. If not, then you can create your own
>namespace in the id.
>
>Does that help? Because I still only have a tenuous grasp of the issue.

Unfortunately it does not since clIDType != roidType

I am talking about
GET /rdap/entity/clIDType

but re-reading the description of "handle" in 7483 maybe it's a non-issue in practice since I can easily workaround it as local convention. 

thanks,