[db-wg] 2022-01 Resource holders identity

Hi Leo

On Wed, 1 Jun 2022 at 16:27, Leo Vegoda <leo at vegoda.org> wrote:
>
> Hi Denis,
>
> On Tue, May 31, 2022 at 5:12 PM denis walker <ripedenis at gmail.com> wrote:
>
> [...]
>
> > That might be interesting to know but unless the RIPE NCC's internal
> > database has a flag to indicate a natural person they would have the
> > same problem.
>
> I expect they do. Or a reasonable proxy, like whether they validated
> that the member exists via a company registry or a registry of natural
> people.
>
> > But the numbers are not the real question. Anyone can
> > see by flicking through the split file there is a 'significant' number
> > of objects with full name and address details of people. And if one of
> > them is you it is very significant. We also know there are people who
> > won't become a resource holder or even get an assignment because of
> > the risk of their personal details being entered into the RIPE
> > Database. So we always come back to the fundamental question, for what
> > purpose do we publish the name and address of a natural person who is
> > a resource holder or who has an assignment from an LIR?
> >
> > We can also differentiate here between publishing the details and
> > storing the details not for public access. If and when we establish
> > the purpose of this information we can look at technical solutions. We
> > should also keep in mind that this information is completely separate
> > from being able to contact the operator of a network. So we are trying
> > to establish why anyone needs to know who is responsible, accountable,
> > liable for a block of IP addresses...and who needs to know that.
> > Technical solutions may address how they can access this information.
>
> I think we should try to establish if the proportion of natural people
> that are LIRs is so low

It is not only LIRs. So many assignment objects also contain name and
address in the "descr:" attributes. We know one large telco has about
700k customer assignments documented in the RIPE Database.
Unfortunately I can't easily find one in the 4.2m entries in the
INETNUM split file to see if they have entered the customer name and
address in the INETNUM as well as in referenced PERSON objects. These
types of issues also need to be addressed.

> that any policy will always be handling corner
> cases. If that is the case, it might be better to establish policy
> objectives and just delegate the specific implementation to the RIPE
> NCC.

I do see this policy as determining the principles. If approved then
there would need to be further community discussion on a migration
plan. This is not going to be a quick, overnight fix. Ultimately
technical details will drop down to the RIPE NCC.

cheers
denis
proposal author

>
> Kind regards,
>
> Leo