...

...

...

According to the agreement between RIPE NCC and DENIC we plan Has this agreement been published? No, this is a working agreement about technical details. It is based on the community consensus that domain related information (excluding reverse delegations) should be moved from the RIPE Database to the respective ccTLD databases. Fair enough, but community consensus must be recorded.

i am confused here. i presume the communities involved, ripe ncc and denic, reached a consensus and one or the other would be whining if not. why should this be a concern of the other lirs or the wg or you or me?

What ever happened to the auto-referral mechanism, whereby a query about a moved object could be referred onward to the appropriate whois server?

is this not just for domain:s? randy

...

No, this is a working agreement about technical details. It is based on the community consensus that domain related information (excluding reverse delegations) should be moved from the RIPE Database to the respective ccTLD databases.

Fair enough, but community consensus must be recorded. Was this consensus achieved at RIPE 36 and we're still waiting for the relevant minutes?

Frankly, I don't see what the great fuss is about, Mike. I think this has been discussed in the CENTR arena, and if I'm not mistaken there's been developed a consensus between the RIPE NCC and the CENTR folks for the preferred way forward. The RIPE DB's primary purpose isn't to serve as a repository for domain objects and related person objects -- I have for a long time had the impression that this has been rather obvious, and that a better solution for all parties would be preferable (less work for the CC registries to maintain the RIPE DB data, tighter coupling to the underlying registry database, costs for running the system where they can be accounted for, etc. etc.) Since the data published is in some sense being published by the relevant CC registry, they're free to do as they see fit with the data, including moving it to a separate whois server. That's as a matter of fact what has happened to the .NO domain objects -- the deletion happened shortly after the last RIPE meeting (we still have a little cleanup left to do). The .NO registry started using its own registry database late last year, and after that time, the .NO domain objects had not been maintained in the RIPE database. Thus, in order to not continue publishing increasingly stale data, it was seen as better to remove the objects from the RIPE database, and instead insert a pointer to the .NO registry's own whois database. No, this wasn't published to the RIPE lists, but instead published to those who have dealt directly with the .NO CC registry previously.

...

You are right, we are moving only domain objects. The person objects will remain in the RIPE Database for several months.

What ever happened to the auto-referral mechanism, whereby a query about a moved object could be referred onward to the appropriate whois server?

With the reservation that we're talking about the same thing, that works quite nicely as is for the domain objects. It will however by default only trigger if the corresponding domain object does not exist in the RIPE database, thus the push to remove the stale data, so that the otherwise-masked but maintained objects could become visible. As for the person objects referred by the deleted .NO objects, we just went ahead and (tried to) delete the referred person objects, and the RIPE database reference integrity check made sure that we didn't actually remove objects which were referenced from other remaining objects in the database. If the RIPE database doesn't have as primary purpose to serve as a domain name registry database, it certainly didn't have as its primary purpose to serve as a phone book (serving otherwise-orphaned person objects). Best regards, - Håvard