William Allen Simpson wrote:
The NSI registry whois seems to be falling down and cannot get up. We need a good system of distributed whois servers, just as we have a robust system of distributed DNS servers.
<start of rant> I think you misunderstand and inadvertently misrepresent the issue. NSI's position (and indeed the major driver of it's financial and market position) has been that the data provided to NSI during the process of registering a domain is proprietary, confidential, and "copyrightable" by them. Looked at one way, they're right. There is no reason that they should make available to *anyone*, not the least their competitors, who their customers are, and how to contact them. In the days of the total monopoly (the old days) this was arguable. Now it isn't. That is *not* to say that the roots of whois are not noble, and appropriately in the public domain. Or that the data should not be available today. I am merely pointing out that absent some kind of larger "in the public interest" issue, they should not have to make available their customer data. On the other hand, I believe that there *is* an overwhelming public interest issue as far as typical whois type data is concerned. The ability to reach responsible parties during network events like DOS attacks, bogus announcements, domain failures, spam runs, etc. are necessary. Of course, the urgent need for a sales droid to inform one that the/she can get better service elsewhere is *not* one of these events :-) I believe that the value of whois data is on a downward spiral, accelerated by the registrars who themselves create contact data that is of no value (have you noticed the increasing presence of technical and admin contacts in domain registrations with "no-valid-email@" as the address portion)? There was a time when SOA data served a purpose. Unfortunately that data is generally outdated, and useless as well (I use it to contact enablers of spam, so I know how bad it is). So, a fundamental decision has to be made: Is it important for parties that control DNS data (domain names and address space) that are connected to "the network" to be identifiable to the community at large, and reachable? If so, a replacement rfc needs to be developed. While it can be argued that due to the fact that there are so few networks that are default free, it should only be necessary to identify the provider at the top of the tree, and that contacting them should start a cascading process that ultimately gets the problem solved, frequent events experienced by us all prove this wrong. I believe that without the traditional good sense championed during Jon Postel's time (the enlightened days) the entity that today could, should, and won't solve this is ICANN. They have neither the balls, nor apparently the clue. Shame on them. And on us for letting it happen. <end of rant> I believe this discussion is operational, and valid for this list. Failure to solve this issue in some way will cause further network problems. -- Rodney Joffe CenterGate Research Group, LLC. http://www.centergate.com "Technology so advanced, even we don't understand it!"