[ncc-services-wg] comments on proposal 2012-07

On 02/11/2013 23:07, Sander Steffann wrote:
> I personally find this a bit petty. You seem to have a very different
> idea about the RIPE database than I have. The RIPE database is a service
> to the *community*, not only the NCC members.

I'm genuinely sorry that you find this petty.  It would certainly be petty
if this were a temporary arrangement or if it were something which only
applied to a tiny number of organisations, but it's neither.  This part of
the policy will apply as a permanent arrangement for what will probably be
a sizeable proportion of the existing organisation base which receives
registry services from the RIPE NCC (obviously excluding the responsible
legacy holders who opt for sections 2.1 - 2.4).  From any reasonable
viewpoint, it cannot be written off as petty.

> And the database != the registry.

Indeed - and to be more precise, under the terms of my suggestion, the data
would still be maintained in the registry and the holders can make
modifications as required, but if they decline to pay for services, then
they wouldn't get IRRDB service or the reverse DNS service - which are
slightly separate functions to the address registry.

> But now suddenly an exception must be made for legacy resources?

Third parties can still register data in the ripedb, but they lose out on
stuff which makes it worthwhile, e.g. object grandfathering (giving good
quality object security).  The legacy objects will be equivalently
authorised to native RIPE NCC allocated objects, so it is a subtle but
important fallacy to suggest that an exception would be made for LRHs when
in fact it's the opposite that's true in a difference sense: third party
objects don't get reverse DNS service and don't have an authenticated IRRDB
service.

> Besides, I think that what you suggest will really create a lot of work
> for the database team of the NCC. I can't even imagine the mess of
> business logic necessary to implement what you suggest.

I'll leave it up to the RIPE NCC to comment on this, but the hooks exist
already (mnt-domains: and mnt-routes:) and unless my understanding of the
ripedb is mistaken, I don't see that it would be a huge amount of work.

Sander, it is not unreasonable to expect that organisations which register
data in the RIPE database should contribute towards the running costs of
the service they use.  There is nothing unfair about this and it's not
petty. It is a reasonable and fair approach to creating a permanent, long
term solution for dealing with the requirements of the legacy resource
holder community and the internet community in general.

As a separate issue, there is also no incentive whatsoever in 2012-07 for
legacy resource holders to engage with the RIPE NCC.  When the NCC come
knocking on LRH doors under the current terms of 2012-07, they will present
two options:

1. sign contracts, pay money, service as usual
2. don't bother signing anything, don't bother paying, service as usual

Realistically, most LRHs are not going to bother engaging with the RIPE NCC
under the current terms of this proposal because they do not benefit in any
meaningful way from doing so.

This and the lack of quid-pro-quo are the two main reasons why this
proposal is critically flawed and why it should not become RIPE community
policy in its current form.

Nick