Feedback from the WG chairs group on 2007-01
Hi APWG folks, we have received feedback from the Working Group Chairs Collective regarding the policy process on 2007-01: --- Dear APWG Chairs, The Working Group Chairs Collective has received your request to check consensus has been reached on policy proposal 2007-01. After reviewing the discussion, we could not reach consensus that significant changes between versions 1 and 2 of the document received sufficient attention within the Review and Last Call periods, and as such the proposal may not have received the appropriate levels of scrutiny. Due to lack of consensus within the group of Working Group Chairs, we ask that the proposal is reconsidered. --- This means that the proposal needs to re-enter review phase, and we'll ask you to closely look at it, and confirm that you're expressing support for the *latest revision* of the proposal. (Please don't start a big discussion now on the outcome - the policy process says that the group of all Working Group Chairs needs to agree that the process has been followed correctly. There were serious concerns, and thus there was no consensus on this - and the only way forward is to re-enter review phase, which we'll do shortly.) As I assume that we'll reach consensus again, work on the related proposals (2006-01 for IPv6 PI etc.) will continue in parallel. regards, Gert Doering -- APWG chair -- Total number of prefixes smaller than registry allocations: 110584 SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279
On 15 May 2008, at 10:02, Gert Doering wrote:
This means that the proposal needs to re-enter review phase, and we'll ask you to closely look at it, and confirm that you're expressing support for the *latest revision* of the proposal.
I have just now, as requested, looked closely at http://www.ripe.net/ripe/draft-documents/ripe-424-draft.html, which I understand to be the latest revision of the proposal, and have the following comments to make as part of the re-entered review phase. These comments include (a) a request for information from RIPE-NCC and (b) a proposal to remove certain text from the draft. 6.9 Anycasting TLD Nameservers Fine. 9.0 PA vs. PI Address Space 3 proposed changes: 1/3: PI concept, distinction from PA, contract options: fine; 2/3: Suggested warning on potential disadvantages of PI: fine; 3/3: Retroactive application of new policy to existing non-PA: problematic. It is certainly necessary to make explicit that the proposed new policy applies to existing PI assignments, and thus modifies the contractual conditions supporting any such assignment. I expect that existing contractual arrangements actually support fair and reasonable modifications to the terms and conditions involved. It would be useful to have confirmation from RIPE-NCC that this is indeed the case. I doubt that RIPE-NCC has formal competence to modify the terms and conditions (whatever they might be!) under which ERX assignments were made. It is therefore most likely futile for the RIPE community to request any such action of RIPE-NCC. I propose removing the text "including address space marked as Early Registration (ERX)". I see no other problem with this proposed change. By the way, and off-topic for the present discussion, I would see it as a useful exercise to encourage holders of ERX resources to "regularize" their position, perhaps along the lines of ARIN's "Legacy Registration Services Agreement". Best regards, Niall O'Reilly University College Dublin IT Services PGP key ID: AE995ED9 (see www.pgp.net) Fingerprint: 23DC C6DE 8874 2432 2BE0 3905 7987 E48D AE99 5ED9
Hi Niall, community, Niall O'Reilly wrote:
On 15 May 2008, at 10:02, Gert Doering wrote:
This means that the proposal needs to re-enter review phase, and we'll ask you to closely look at it, and confirm that you're expressing support for the *latest revision* of the proposal.
I am trying to do that once again and I am *not* able to express my support for the latest (Version 2) revision of the proposal and the draft doc. http://www.ripe.net/ripe/draft-documents/ripe-424-draft.html , for a couple of reasons: - the headline of this draft is "IPv4 Address Allocation and Assignment Policies for the RIPE NCC Service Region" which for me implies that it is supposed to apply to address space distributed as a service by the RIPE NCC to its Service Region. This seems to be confirmed (in Section 9.0) by the paragraph headline of "End Users requesting PI space should be given this or a similar warning:" Looking at the (1st block of) NEW TEXT talking about PI and routability seems pretty reasonable and consistent. However, looking at the next block of text labelled ADDITION TO DOCUMENT which starts to - out of context - talk about ERX (to me) seems grossly illogical and structurally misplaced. - trying to sneak in a retro-active requirement into a policy which would require the NCC to impose usage limitations, contractual work or new fees onto *legitimate* holders of legacy resources without, at the same time, offering some reasonable incentive for the resource holders to agree and comply, poses the risk of damaging the perception of the RIPE NCC as a professional, impartial and "reasonable" organisation trying to support the community. I thus propose to remove this additional text block (which seems to have been introduced without much review and consistency checking when producing Version 2). At the same time I'd like to explicitely voice support for the goal of - eventually - trying to bring the legacy resource holders (including AS numbers) under the umbrella of the RIPE NCC's regular resource management environment. The most reasonable point in time to do that seems to be the availability of the RPKI Service (in production quality) and maybe by way of a separate policy proposal - if a new policy is necessary at all at that point in time, rather than simply a new service offer. Regards, Wilfried.
Hi, this discussion is starting a bit premature... On Tue, May 20, 2008 at 08:00:56PM +0000, Wilfried Woeber, UniVie/ACOnet wrote:
Niall O'Reilly wrote:
On 15 May 2008, at 10:02, Gert Doering wrote:
This means that the proposal needs to re-enter review phase, and we'll ask you to closely look at it, and confirm that you're expressing support for the *latest revision* of the proposal.
Maybe my wording wasn't clear enough here. We are working on a new version of the proposal right now, and when it's done, "we'll ask you to closely look at it <and so on>". [..]
- the headline of this draft is
"IPv4 Address Allocation and Assignment Policies for the RIPE NCC Service Region"
which for me implies that it is supposed to apply to address space distributed as a service by the RIPE NCC to its Service Region.
This seems to be confirmed (in Section 9.0) by the paragraph headline of
"End Users requesting PI space should be given this or a similar warning:"
Looking at the (1st block of) NEW TEXT talking about PI and routability seems pretty reasonable and consistent.
However, looking at the next block of text labelled ADDITION TO DOCUMENT which starts to - out of context - talk about ERX (to me) seems grossly illogical and structurally misplaced.
This block has nothing to do whatsoever with the warning message - which is why it's in a *separate* text box, and not included in the "changed text for the warning message" box. Since this seems to be easy to misunderstand, we're currently reworking the structure of the change to make it more obvious that the "ADDITION TO DOCUMENT" block is *not* part of the "warning message to PI requestors", but of the general PI section. Independent of this, the ERX part will be moved to a separate policy proposal (see below).
- trying to sneak in a retro-active requirement into a policy which would require the NCC to impose usage limitations, contractual work or new fees onto *legitimate* holders of legacy resources without, at the same time, offering some reasonable incentive for the resource holders to agree and comply, poses the risk of damaging the perception of the RIPE NCC as a professional, impartial and "reasonable" organisation trying to support the community.
We've discussed this on RIPE55, and got *very* explicit support from the room that "yes, this needs to apply for existing assignments as well" (that was the part about the "old farts", remember?). Please read it up in the minutes. There is absolutely nothing "sneaky" here. The RIPE NCC management has voiced support for this as well, and not expressed any concerns about this part of the proposal. The "missing" incentive for the resource holders could be, among others, better quality of the data in the RIPE database, and thus less hijacking, less troubles with upstream routing, etc. - and in the long run, resource certificates, of course. I agree that we cannot do that for ERX space, because this is space that was never handed out by the NCC, and no ties to the NCC exist. So the plan is to put up a separate policy proposal for ERX ("if you want service from the NCC, like a resource certificate, the ERX space needs to be changed to normal PI space, with all accompanying obligations"). Gert Doering -- APWG chair -- Total number of prefixes smaller than registry allocations: 110584 SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (89) 32356-444 USt-IdNr.: DE813185279
Gert Doering wrote:
Hi,
this discussion is starting a bit premature...
Ah, sorry.... [...]
Independent of this, the ERX part will be moved to a separate policy proposal (see below).
Very good. [...]
The "missing" incentive for the resource holders could be, among others, better quality of the data in the RIPE database, and thus less hijacking, less troubles with upstream routing, etc.
Those are pretty weak ones as all of this is working already, and routing for ERX space in the general interest of the Internet, but
- and in the long run, resource certificates, of course.
this sh|could be The Real Thing[TM]. There may be a couple of other handles, like management of reverse delegation? This clearly is a service offered by the RIPE NCC, and with cost involved.
I agree that we cannot do that for ERX space, because this is space that was never handed out by the NCC, and no ties to the NCC exist. So the plan is to put up a separate policy proposal for ERX ("if you want service from the NCC, like a resource certificate, the ERX space needs to be changed to normal PI space, with all accompanying obligations").
Fully supported. Although, at the same time, I am disappointed that we are still waiting for the draft contract and the discussion of the type or the form of new membership, like fees and voting rights?! Which, by the way, also holds true for the "genuine" PI space part of the game ;-)
Gert Doering -- APWG chair
Wilfried.
Hello Wilfried,
Although, at the same time, I am disappointed that we are still waiting for the draft contract and the discussion of the type or the form of new membership, like fees and voting rights?!
This part is not for the address policy working group to decide. The idea is that there will be a new class of membership, but the details are for the NCC and its members to decide. - Sander
On 20 May 2008, at 21:59, Gert Doering wrote:
this discussion is starting a bit premature...
Sorry from me too. I was trying to avoid being late. 8-)
Maybe my wording wasn't clear enough here. We are working on a new version of the proposal right now, and when it's done, "we'll ask you to closely look at it <and so on>".
OK, fine! /Niall
participants (4)
-
Gert Doering -
Niall O'Reilly -
Sander Steffann -
Wilfried Woeber, UniVie/ACOnet