Re: "External" hosts in EPP

[Daniel Manley <dmanley@tucows.com>]

I like the flexibility this change would offer.  Should the spec should 
say that the two forms cannot mix?  ie. if host mapping is used (as 
stated in the server's greeting) then "in-zone" host objects cannot be 
implicitly created through domain create or update commands.

And for out-of-zone hosts... after being created in a domain create or 
update command, should a <host:info> work these hosts?  I personally 
would still like the registry to create host objects implicitly for 
these out-of-zone hosts -- mostly to facilitate mass updates of domains 
using a particular host.  Beside, it wouldn't surprise me if a registry 
that is already handing domain-host relationships via RDMS normalized 
table relations would implicitly create host records and relations for 
out-of-zone hosts.

Dan


Klaus Malorny wrote:

>> Klaus,
>>
>> Actually, while looking at a way to modify the domain schema to 
>> accommodate
>> the "out of zone" name servers I think I've figured out a way to make 
>> both
>> the host object proponents and opponents happy.  What do you think about
>> this?:
>>
>> I modify the domain mapping so that delegations are made using one of 
>> two
>> name server identification forms:
>>
>> 1. As is described in the drafts right now using an existing host 
>> object, or
>>
>> 2. A new, second form that allows the client to provide a server name 
>> and
>> optional IP address, but this form doesn't refer to a host object.
>>
>> The second form can be used by registries that support objects for 
>> in-zone
>> hosts to fix the out-of-zone host problem (leaving off the IP 
>> address), but
>> it can also be used by registries that don't want to support host 
>> objects.
>> Such registries wouldn't advertise a host mapping in the <greeting>, and
>> they would require delegations to be done using the new form with IP 
>> address
>> info only as appropriate.
>>
>> -Scott-
>>
>
>
> Hi Scott,
>
> this would be a similar host-less model as DENIC/.de has. So far, I 
> don't see any problems with it. I wouldn't say that I would implement 
> this kind of model if I had a choice. As you know, I don't have a 
> problem with host objects in general, but only with the way they are 
> implemented in EPP. Nevertheless, at the very end it is only a 
> question of personal preferences.
>
>
> regards,
>
> Klaus Malorny
>
>
> ___________________________________________________________________________ 
>
>      |       |
>      | knipp |                   Knipp  Medien und Kommunikation GmbH
>       -------                           Technologiepark
>                                         Martin-Schmeißer-Weg 9
>      Dipl. Inf. Klaus Malorny           44227 Dortmund
>      Klaus.Malorny@knipp.de             Tel. +49 231 9703 0
>
>
>