[anti-abuse-wg] Draft Anti-Abuse Working Group Minutes - RIPE 60

On 10 Aug 2010, at 3:44, Richard Cox wrote:
Leo Vegoda <leo.vegoda at icann.org> wrote
>>> "Revoked" must be clearly visible.
>> 
>> I disagree. I do not think the registry should publish a comment on why
>> a registration exists or does not exist and the word REVOKED is clearly
>> intended to imply that the registration was removed against the desires
>> of the registrant.
> 
> On those - increasing number of - occasions when an RIR discovers that
> information it was given in support of a registration, was untruthful or
> invalid then it seems to me entirely reasonable that the RIR should make
> it clear that what it had previously published, should not be relied upon.
> 
>> Publishing a registration (a positive act) but giving it a negative
>> status is likely to cause confusion, especially with automated
>> network-centric systems that ignore the status attribute value.
> 
> There will surely be technical solutions to that technical problem.

Why create a technical problem just so that you can create a technical solution? The actual problem is social and I suggest that that is where you should focus solutions.

>> I also think the example you give is unrealistic. If the ISP can see its
>> own object and a bunch of other objects then the problem is unlikely to
>> be that to be that the whois database is broken.
> 
> It's very realistic.  Nobody would be suggesting that the "whole database
> is broken".  What would be suggested is that some records are missing or
> the database has not been updated.  That would not necessarily affect the
> records of the ISP querying the database, as its own record would probably
> be significantly older.

Here and above you seem to be arguing two things: firstly, that the RIPE NCC's procedures or staff are considered so error prone that the removal of a registration from the database would automatically be assumed to be a mistake, rather than intentional. That seems unlikely to me but maybe I have missed the hollering and alarm at the frequent reports of errors. Secondly, you seem to believe that the RIPE NCC should not just maintain a registry of what is but should also annotate the registry with commentary about the organisations whose data are published there.

I recognise that there is a group of people who abuse the system. However, I think that commentary about that group should be confined to communication between them and the RIPE NCC and if necessary the courtroom. I do not believe it is appropriate for the RIPE NCC to publish reasons for its actions in the database. In fact, I believe that to do so would breach the current IPv4 policy:

   3.1 Confidentiality

   Internet Registries (IRs) have a duty of confidentiality to their registrants. 
   Information passed to an IR must be securely stored and should not be 
   distributed wider than necessary within the IR. When necessary, the 
   information may be passed to a higher-level IR under the same conditions of 
   confidentiality.

I should probably also point out that this is a requirement of section 10 of ICP-2, too:

http://www.icann.org/en/icp/icp-2.htm

Of course, policy can be changed but I am not convinced that I have seen a convincing case to change the current policy yet.

>> If some kind of mechanism is needed to allow network operators to check
>> that a prefix is not currently registered, then we should ask the RIPE
>> NCC to publish an easy to parse list of prefixes and the date on which
>> they were removed from the database. Presumably a prefix would remain on
>> the list until it had completed any quarantine period and is ready to be
>> re-issued.
> 
> We have been asking for exactly that for some years now, partly to allow
> reputational records to be reset as and when an allocation is recovered.

Has the RIPE NCC responded to these requests? If so, what has it said?

Regards,

Leo Vegoda