OK, so I guess my point altogether is that I dislike special-ness. There seems to be agreement that DNS anycasting is here to stay so the only point of contention is the size of the prefix. Using longer prefixes does not add much to the conservation and introduces more exceptions. Even if other RIRs are doing /48s already why would introducing an additional chunk of swamp space with the potential for more likely large scale failure be of any benefit? If people agree with assigning the addresses for this purpose, which seems to be the case, just issue a standard allocation and be done. Regarding the issue of flap damp(en)ing parameters, I think we should really revisit the flap damp(en)ing recommendations again at RIPE 50. Joao PS: regarding entries in the routing table and the pollution it causes, keep an eye on RFCs coming soon. PS: regarding the use of "wrong DNS software", the TLD operator is running *SERVERS* and providing services to resolvers/users. The best they can do is provide as good a service as they can to their consumers, not preach to them about them running the wrong client side software.