This archive is retained to ensure existing URLs remain functional. It will not contain any emails sent to this mailing list after July 1, 2024. For all messages, including those sent before and after this date, please visit the new location of the archive at https://mailman.ripe.net/archives/list/[email protected]/
[address-policy-wg] 2014-02 New Policy Proposal (Allow IPv4 PI transfer)
- Previous message (by thread): [address-policy-wg] 2014-02 New Policy Proposal (Allow IPv4 PI transfer)
- Next message (by thread): [address-policy-wg] 2014-02 New Policy Proposal (Allow IPv4 PI transfer)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Jan Ingvoldstad
frettled at gmail.com
Mon Apr 21 21:14:36 CEST 2014
On Mon, Apr 21, 2014 at 11:34 AM, Tore Anderson <tore at fud.no> wrote: > > These requirements are in my opinion out of touch with operational > reality; networks evolve and change over time, and given enough time, > pretty much all assignments made will end up being used in a different > way than what the original criteria was. The PI assignment request form > asked for an very detailed criteria, with a breakdown of each individual > subnet in the network, and a listing of all the equipment that would be > used, including the manufacturer name and model numbers. So if you in > the 1990s received an assignment that you had said was for a dozen brand > new Sun UltraServers, you better not have replaced those with modern x86 > hardware, or you have invalidated your assignment! :-O > > Some NCC staffers have told me that the way they've logically "solved" > this impossible requirement was to consider that whenever the > criteria/purpose changed, the original assignment was returned, and a > new one consisting of the same block was issued for the new purpose. > Then they could just "optimise out" the middle steps where the > assignment was removed and re-added. That approached worked (up until > Sep 2012 anyway), but I think it would be much better if the policy > didn't worry so much about the "original criteria", but rather focus on > whether or not the assignment conforms to the address policy in effect > at any given time. If it does, there is no reason to call it invalid. > > Well said. When specifying a purpose for a use, it is quite difficult to provide something that is generic enough to last for the lifetime of operations, rather than the lifetime of something else entirely. -- Jan -------------- next part -------------- An HTML attachment was scrubbed... URL: </ripe/mail/archives/address-policy-wg/attachments/20140421/5be0935e/attachment.html>
- Previous message (by thread): [address-policy-wg] 2014-02 New Policy Proposal (Allow IPv4 PI transfer)
- Next message (by thread): [address-policy-wg] 2014-02 New Policy Proposal (Allow IPv4 PI transfer)
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]