[ This message states my personal opinion and *not* the opinion of the RIPE NCC. I realise that NCC staff commenting on controversial"political" issues in a personal capacity can be considered inappropriate. However I feel I have to speak up just *because* I personally invested a lot of energy into the RIR processes and also into the gestation of ICANN, the ASO and its processes. In other words: I think I earned the right and *have the obligation* to speak up.] Hans Petter, I just read your messages analysing the legal prose. While this is good work and highlights quite some symptoms, it does not analyse the disease. A much more clear symptom of the desease is in the latest amendment to the ICANN-USG agreement http://www.ntia.doc.gov/ntiahome/domainname/agreements/amend5_09192002.htm "... ICANN agrees to perform the following activities and provide the following resources ... Work collaboratively on a global and local level to pursue formal legal agreements with the RIRs, and to achieve stable relationships that allow them to continue their technical work, while incorporating their policy-making activities into the ICANN process. ..." This clearly shows that ICANN does not want to work the way the RIRs are successfully working. ICANN has no respect for, and maybe still does not understand our long established, well working, bottom-up, transparent way of making policies. There are more symptoms of this in all areas of ICANN activity. A good example is the current "AXFR conflict" with some ccTLDs. They script for these problems is: - ICANN staff drafts or changes policies for reasons only apparent to ICANN staff and, possibly, the US Government in some cases. - ICANN board ratifies changes without discussion. THIS MEANS THAT THERE IS NO PROCESS OTHER THAN ON PAPER. - ICANN staff applies policy - 'victims' are surprised since they did not hear anything via what they thought of as their part of the "ICANN Process". - ICANN staff says: "This is policy ratified by the board, please use the 'ICANN Process' to change it again. Meanwhile just comply." THIS MEANS THAT THOSE WHO DEPEND ON ICANN FOR SOMETHING ARE SCREWED. In the particular 'AXFR conflictt' I refer to above, ICANN has deliberately jeopradised the stability of the DNS of some pretty large ccTLDs over an extended period of time. I have lost all confidence in ICANN as an organisation, the ICANN process and many of the people involved, both board and staff. I have the strong suspicion that I am not the only one. *This* is the disease: Loss of confidence. It cannot be addressed with adding paper "Core Values" and even more paper process while maintaining the level of substance at a minimum. I personally think the RIRs should *now* walk away from ICANN for better or worse. This means we have to do our home work well: - Create a process for global policy coordination. We have thes substance of it: the ASO. We just need a legal shell outside of ICANN and -maybe- some minor adaptions to the substance. - Create a process for 'process appeal', i.e. where someone can go if the global policy coordination does not follow its own rules. Maybe professional arbitrators can be used. Certainly not another politically loaded body. - Make sure that those who look at ICANN for influcene into the process find their needs sufficiently addressed by the current regional processes plus the "new" global process. I suspect that this is largely a matter of educating people. Hans Petter, imagine you had invested the time it took to analyse the ICANN generated legal prose into drafting this. I think it would have been a much better investment. We should have a very open discussion in this working group on how to proceed with address policy coordination on the global level. If the RIPE NCC, and the other RIRs, should decide to walk away from ICANN they need our full support. The RIPE NCC also needs help with doing the work outlined above together with the other RIRs. Respectfully Daniel
Daniel, On Mon, Oct 07, 2002 at 11:22:21AM +0200, Daniel Karrenberg wrote:
I personally think the RIRs should *now* walk away from ICANN for better or worse.
I agree. We should spend our time on building a new framework. A good start could indeed be by building upon the existing ASO. However, it is also important to consider the relations between the RIRs and IETF/IAB etc.. David K. ---
to paraphrase from a private conversation: note that the iana function is the only formal link between the ietf and the registries, and we should be careful of what we break. the ietf does not want to start writing rir (and N other fiefdoms) consideration sections in rfcs. there are a number of different roles of the iana function, what different parties need from the iana function, and their/our respective relationships to and through the iana. the rir position seems to be to break away from the iana. the ietf position, such as it is, seems more to coordinate the non-dnso iana functions in the iana in a way well detached from icann dnso politics. randy
Major point: In thepast I have been a strong advocate of keeping all the functions of the historical IANA together in the new structure that eventually became ICANN. From the start I, and others, have been concerned that names would poison this mix in all sorts of ways. This has happened and we seem to agree on that point. Minor point: There are not that many RFCs that would need RIR considerations. Ignored: Reference to "fiefdoms". I do not know what the RIR position is, I only speak for myself. My opinion is that it would be best that the IETF and RIR functions of IANA be taken out of ICANN because I have lost all confidence in ICANN. If this set of functions could be called "the IANA" that would be very nice; just continue to call the names stuff "ICANN" and most people's model of the world will not change an iota. Once the proper parts of the name space are delegated to "IANA" the connection points between "IANA" and "ICANN" will be quite minimal, almost non-existing. Tell me why this would not work. Daniel
--On 8. oktober 2002 11:06 +0900 Randy Bush <randy@psg.com> wrote:
to paraphrase from a private conversation:
note that the iana function is the only formal link between the ietf and the registries, and we should be careful of what we break. the ietf does not want to start writing rir (and N other fiefdoms) consideration sections in rfcs.
So a design criteria for whatever scenario we are looking at should be to maintain this link. (ie co-organisation of the IETF-IANA function and the RIR-IANA function.
there are a number of different roles of the iana function, what different parties need from the iana function, and their/our respective relationships to and through the iana. the rir position seems to be to break away from the iana. the ietf position, such as it is, seems more to coordinate the non-dnso iana functions in the iana in a way well detached from icann dnso politics.
I think the root of the concern is the strong link between the percived de facto policy makers (ICANN staff) and the IANA staff. I belive this concern is only true if it is in fact so that ICANN staff makes policy desicions (or suggests them to the board who ratifies them without public process) The cause for this concern is that while some of us see something as a policy desicion others don't. The very fundamental question is perhaps: - Are there any desicions at all to be taken by ICANN which does not require an open transparent process ? Looking carefully at that question I realise that it can be generalized: - Are there any desicions at all to be taken by X which does not require an open transparent process ? where X = {ICANN, RIPE NCC, ARIN, APNIC, IETF, ASO, ...} and this it is where it gets interesting: - the concern is exactly the same at both sides of the ICANN vs RIR discussion... -hph
I think the root of the concern is the strong link between the percived de facto policy makers (ICANN staff) and the IANA staff.
indeed. and it is not at all clear any link is needed.
The very fundamental question is perhaps: - Are there any desicions at all to be taken by ICANN which does not require an open transparent process ?
yes. details of dealing with technical matters and disputes where private information is exchanged. e.g. occasionally confidential data are disclosed for protocol number allocations. jon postel preferred not to air the internal squabbles of country X over fools who thought cctld admin was a power, as opposed to responsibility. randy
Disagree. The RIR"s seem to think there is a world that is a) includes ICANN b) doesn't include ICANN. What IMHO, believe is that should the RIR's depart, which I'm against, this would open the oppertunity for other organizations to take control. ITU, being one of them. This would be a particularly BAD IDEA Its also interesting to note that at least one RIR hasn't really put this policy issue to its members as a propsed policy. Really good for bottom up processes :| There should be a "formal policy proposal" that goes to the membership at large on the topic of "Should the RIR's bail on IANA / ICANN". I haven't seen one. John Brown speaking personally
-----Original Message----- From: owner-lir-wg@ripe.net [mailto:owner-lir-wg@ripe.net] On Behalf Of David Kessens Sent: Monday, October 07, 2002 7:50 PM To: Daniel Karrenberg Cc: Hans Petter Holen; RIPE Local IR Working Grouo Subject: Re: [lir-wg] ICANN Reform
Daniel,
On Mon, Oct 07, 2002 at 11:22:21AM +0200, Daniel Karrenberg wrote:
I personally think the RIRs should *now* walk away from ICANN for better or worse.
I agree.
We should spend our time on building a new framework.
A good start could indeed be by building upon the existing ASO. However, it is also important to consider the relations between the RIRs and IETF/IAB etc..
David K. ---
Its also interesting to note that at least one RIR hasn't really put this policy issue to its members as a propsed policy. Really good for bottom up processes :|
some of the RIRs seem have a tradition of making policy in the back room and voting on engineering in group face-to-face meetings with very random proposals which almost all get voted 'yes'. needless to say, but i'll say it anyway, for an ietfer this seems ill-considered. randy
say it isn't so..... we lost bottom up somewhere along the lines......
-----Original Message----- From: owner-lir-wg@ripe.net [mailto:owner-lir-wg@ripe.net] On Behalf Of Randy Bush Sent: Monday, October 07, 2002 8:35 PM To: John M. Brown Cc: RIPE Local IR Working Grouop Subject: RE: [lir-wg] ICANN Reform
Its also interesting to note that at least one RIR hasn't really put this policy issue to its members as a propsed policy. Really good for bottom up processes :|
some of the RIRs seem have a tradition of making policy in the back room and voting on engineering in group face-to-face meetings with very random proposals which almost all get voted 'yes'.
needless to say, but i'll say it anyway, for an ietfer this seems ill-considered.
randy
At 04:35 AM 10/8/2002, Randy Bush wrote:
Its also interesting to note that at least one RIR hasn't really put this policy issue to its members as a propsed policy. Really good for bottom up processes :|
some of the RIRs seem have a tradition of making policy in the back room and voting on engineering in group face-to-face meetings with very random proposals which almost all get voted 'yes'.
needless to say, but i'll say it anyway, for an ietfer this seems ill-considered.
John, Randy, Please name horse/rider and cite chapter/verse, *then* we can discuss more meaningfully. For the RIPE NCC my impression is, that the policy process is suffering from *too much* confidence of the community in the RIR rather than too little. This tends to make all parties complacent and mistakes can happen. However this needs to be put into perspective: too much confience is a problem that is *orders of magnitude* less grave than no confidence, which is ICANN's problem. Daniel
some of the RIRs seem have a tradition of making policy in the back room and voting on engineering in group face-to-face meetings with very random proposals which almost all get voted 'yes'.
needless to say, but i'll say it anyway, for an ietfer this seems ill-considered.
Please name horse/rider and cite chapter/verse, *then* we can discuss more meaningfully.
at kokura and rodos: o political changes with icann were not discussed anywhere except in plenary, and were done as presentations by the 'powers', not formative open discussion. o things such as golden v6 allocations for anything 'important' were discussed in a wg, voted (with almost no representation by folk who run routers, i.e. will pay the costs), and done with long laundry lists of ideas for who might deserve golden prefixes, and all were approved. it was a childish land grab.
For the RIPE NCC my impression is, that the policy process is suffering from *too much* confidence of the community in the RIR rather than too little. This tends to make all parties complacent and mistakes can happen.
the RIRs don't have a tradition of open fora for the political issues. heck, we did not used to have these political issues, so no surprise. but this means we need to create these processes. similarly, the world used to be O(100) engineers. technical and addressing policy was made in global consensus, often coordinated by jon postel and the iana. we've grown. we now try to distribute the process. but, in doing so, we have lost the technical core (which did things such as cidr etc) and have a large influx of well-meaning folk who, unfortunately, do not have the experience or the scars. so, no blame. but we are slipping sideways in non-good ways. randy
At 10:42 AM 10/8/2002, Randy Bush wrote:
o political changes with icann were not discussed anywhere except in plenary, and were done as presentations by the 'powers', not formative open discussion.
I can only speak for Rhodos: As you may recall I tried to open the discussion by asking some quite inconvenient questions to Rob. But if the people there do not want to discuss what can one do. Maybe having a more BOF like setting would be good. Any bright ideas?
o things such as golden v6 allocations for anything 'important' were discussed in a wg, voted (with almost no representation by folk who run routers, i.e. will pay the costs), and done with long laundry lists of ideas for who might deserve golden prefixes, and all were approved. it was a childish land grab.
Sorry I am not aware of the issue. What worries me is that I cannot find a reference to what you talk about in the minutes of either the IPv6 DB or this WG. This is bad. Can you point me there? Maybe the lack of stable discussion here is a result of the perceived lack of currnet operational importance of IPv6. In other words: Noone really cares (yet). This creates room for all sorts of ideas and little checks and balances. [Myh personal opinion about golden prefixes: When I still took part in address policy discussions I have made it quite clear that all the RIRs should do is provide a *registry* of prefixes declared special by *anyone* asking. The most a policy about this should do is to create categories of self-declared specialness, e.g. "dns-root-server", "dns-cctld-server", "address-whois-server", "www-search-engine", "isp-home-page", .... . Those rinning routers could then decide to use this registry and which categories or entries they would treat specially.] Daniel
On Tue, 8 Oct 2002, Daniel Karrenberg wrote:
o things such as golden v6 allocations for anything 'important' were discussed in a wg, voted (with almost no representation by folk who run routers, i.e. will pay the costs), and done with long laundry lists of ideas for who might deserve golden prefixes, and all were approved. it was a childish land grab.
Sorry I am not aware of the issue. What worries me is that I cannot find a reference to what you talk about in the minutes of either the IPv6 DB or this WG. This is bad. Can you point me there?
I don't think Randy is referring to the RIPE region. In that region, I do know, however, that the discussion, which I seem to remember took place in the IPv6 wg and the LIR wg mailing lists, is difficult to follow because of the way the indexes are threaded. Someone had some comments about not being able to follow the discussion if they just clicked in the URL mentioned in the minutes. This is however only a technical glitch and the information is most certainly there and accesible to all. It may have even been corrected by now. So give it another try.
Maybe the lack of stable discussion here is a result of the perceived lack of currnet operational importance of IPv6. In other words: Noone really cares (yet). This creates room for all sorts of ideas and little checks and balances.
Possibly, though I have the feeling interest in IPv6 is growing rapidly. Joao
On Tue, Oct 08, 2002 at 11:19:03AM +0200, Daniel Karrenberg wrote:
randy wrote:
o things such as golden v6 allocations for anything 'important' were discussed in a wg, voted (with almost no representation by folk who run routers, i.e. will pay the costs), and done with long laundry lists of ideas for who might deserve golden prefixes, and all were approved. it was a childish land grab.
Sorry I am not aware of the issue. What worries me is that I cannot find a reference to what you talk about in the minutes of either the IPv6 DB or this WG. This is bad. Can you point me there?
Just for the record: no such things happened in the ipv6 wg at the last RIPE meeting. It's the lir wg who sets policy. David K. ---
--On 8. oktober 2002 12:07 -0700 David Kessens <david@IPRG.nokia.com> wrote:
On Tue, Oct 08, 2002 at 11:19:03AM +0200, Daniel Karrenberg wrote:
randy wrote:
o things such as golden v6 allocations for anything 'important' were discussed in a wg, voted (with almost no representation by folk who run routers, i.e. will pay the costs), and done with long laundry lists of ideas for who might deserve golden prefixes, and all were approved. it was a childish land grab.
Sorry I am not aware of the issue. What worries me is that I cannot find a reference to what you talk about in the minutes of either the IPv6 DB or this WG. This is bad. Can you point me there?
Just for the record: no such things happened in the ipv6 wg at the last RIPE meeting. It's the lir wg who sets policy.
The action from the lir-wg on the NCC was: http://www.ripe.net/ripe/meetings/archive/ripe-43/presentations/ripe43-plen ary-lir/sld008.html 43.4 NCC Continue the process to move the 6bone under the framework of the RIRs Maybe this was not clear enough but this is in my opinion (and this is perhaps a place were my opinion counts :-) NOT a policy decision , but an action on the NCC to continue the process already started. I would expect a complete policy proposal to be presented to go to the final policy approval process once all concerns were settled. I'll make a note of labeling policy decision different from work in progress actions in the future. Regards, Hans Petter Holen lir-wg Chair
Daniel, On Tue, Oct 08, 2002 at 11:19:03AM +0200, Daniel Karrenberg wrote:
At 10:42 AM 10/8/2002, Randy Bush wrote:
o political changes with icann were not discussed anywhere except in plenary, and were done as presentations by the 'powers', not formative open discussion.
I can only speak for Rhodos: As you may recall I tried to open the discussion by asking some quite inconvenient questions to Rob. But if the people there do not want to discuss what can one do. Maybe having a more BOF like setting would be good. Any bright ideas?
It seems that our good old mailing lists didn't loose any of it's capability to allow people to discuss this very topic in an extremly public place. David K. ---
Daniel, On Tue, Oct 08, 2002 at 10:20:28AM +0200, ext Daniel Karrenberg wrote:
For the RIPE NCC my impression is, that the policy process is suffering from *too much* confidence of the community in the RIR rather than too little. This tends to make all parties complacent and mistakes can happen.
The problem seems to be more related to that fact that most members of the RIRs have dayjobs and they increasingly care about running their own businesses. They have very little spare time to participate in the policy making process so they have very little choice but to accept the policy as proposed by the RIR because they don't have the resources to come up with their own proposals. David K. ---
For the RIPE NCC my impression is, that the policy process is suffering from *too much* confidence of the community in the RIR rather than too little. This tends to make all parties complacent and mistakes can happen.
My diagnosis is that there is to little participation from the community. This is not only the RIPE NCCs fault. This could be a community failiure, process failiure, or even failiure of the chair of this wg or something else. -hph
At 09:45 PM 10/8/2002, Hans Petter Holen wrote:
My diagnosis is that there is to little participation from the community.
This is not only the RIPE NCCs fault. This could be a community failiure, process failiure, or even failiure of the chair of this wg or something else.
Hans Petter, this diagnosis is a difficult one. I have been worrying about this 'problem' ever since RIPE meetings have grown beyond 50 attenders. And I have been very concerned, to the point of being distressed, about the lack of a vivid discussion many times. Rob is my witness. After a while he developed quite some routine to calm me down on these occasions. ;-) I am now convinced that one cannot do more than provide an open forum and bring the important issues to the table together with proposals how to deal wit them. If there is no big discussion, implement things as proposed and do not waste too much energy worrying whether there was enough discussion. Take the lack of discussion as a vote of confidence. I understand that anyone who is committed to open discussion, such as yourself, will sometimes have the feeling that there is too much confidence expressed by lack of discussion. That is true. But one cannot do more than provide the forum and offer every opportunity for discussion. If the community has too much confidence there is little one can do about it. Either it is justified or it is not, but worrying about it is wasting energy. Daniel
agreed siegfried On 23 Oct 2002 at 9:04, Daniel Karrenberg wrote:
At 09:45 PM 10/8/2002, Hans Petter Holen wrote:
My diagnosis is that there is to little participation from the community.
This is not only the RIPE NCCs fault. This could be a community failiure, process failiure, or even failiure of the chair of this wg or something else.
Hans Petter,
this diagnosis is a difficult one.
I have been worrying about this 'problem' ever since RIPE meetings have grown beyond 50 attenders. And I have been very concerned, to the point of being distressed, about the lack of a vivid discussion many times. Rob is my witness. After a while he developed quite some routine to calm me down on these occasions. ;-)
I am now convinced that one cannot do more than provide an open forum and bring the important issues to the table together with proposals how to deal wit them. If there is no big discussion, implement things as proposed and do not waste too much energy worrying whether there was enough discussion. Take the lack of discussion as a vote of confidence.
I understand that anyone who is committed to open discussion, such as yourself, will sometimes have the feeling that there is too much confidence expressed by lack of discussion. That is true. But one cannot do more than provide the forum and offer every opportunity for discussion. If the community has too much confidence there is little one can do about it. Either it is justified or it is not, but worrying about it is wasting energy.
Daniel
Hi, On Wed, Oct 23, 2002 at 09:04:00AM +0200, Daniel Karrenberg wrote:
I am now convinced that one cannot do more than provide an open forum and bring the important issues to the table together with proposals how to deal wit them. If there is no big discussion, implement things as proposed and do not waste too much energy worrying whether there was enough discussion. Take the lack of discussion as a vote of confidence.
I just want to state publically that I trust Daniel, Hans-Petter, and the other RIPE people involved in this matter to do the right thing. Not being a legal expert, and not being a native english speaker either, it's very hard for me to directly comment on the ICANN proposals. On the other hand, I have come to trust you guys over the last years, and see no reason to distrust you now. Of course, if you change the existing open processes into a "the RIPE NCC rules the world!" dictatorship, we'll come and get you... Gert Doering -- NetMaster -- Total number of prefixes smaller than registry allocations: 48282 (47686) SpaceNet AG Mail: netmaster@Space.Net Joseph-Dollinger-Bogen 14 Tel : +49-89-32356-0 80807 Muenchen Fax : +49-89-32356-299
I am now convinced that one cannot do more than provide an open forum and bring the important issues to the table together with proposals how to deal wit them. If there is no big discussion, implement things as proposed and do not waste too much energy worrying whether there was enough discussion.
and when there are six proposals, often with overlap, all are then accepted. i have seen this. a wonderfully degenerate form of design by committee. not very good engineering. randy
At 02:51 PM 10/23/2002, Randy Bush wrote:
and when there are six proposals, often with overlap, all are then accepted. i have seen this. a wonderfully degenerate form of design by committee. not very good engineering.
Bad things can always happen. I do not know what you are referring to, but has that been fixed since then. Otherwise I will be happy to try starting a lively discussion once more ;-). Daniel
Daniel, Thanks for you comments, | I am now convinced that one cannot do more than provide an open forum | and bring the important issues to the table together with proposals | how to deal wit them. If there is no big discussion, implement things | as proposed and do not waste too much energy worrying whether | there was enough discussion. Take the lack of discussion as | a vote of confidence. To a certain extent I agree, but to act responsibly I have to ask myself constantly wether I am doing the ringth things or whether wthings could have been done differently to archive a better result. I think of this as a constant process of improvement. | I understand that anyone who is committed to open discussion, such | as yourself, will sometimes have the feeling that there is too | much confidence expressed by lack of discussion. That is true. | But one cannot do more than provide the forum and offer every | opportunity for discussion. If the community has too much confidence | there is little one can do about it. Either it is justified or it is | not, but worrying about it is wasting energy. Maybe you are right, but maybe there are things I could do to facilitate better discussions. Maybe the procedure on http://www.ripe.net/ripe/wg/lir/howto_develop.html should be reviewed to see if we can increase the likelyhood of qualitative discussions. -hph
At 04:29 AM 10/8/2002, John M. Brown wrote:
What IMHO, believe is that should the RIR's depart, which I'm against, this would open the oppertunity for other organizations to take control.
ITU, being one of them. This would be a particularly BAD IDEA
Not unless the RIRs screw up enough to loose the confidence of their communities. Daniel
Or gov's get concerned enough that the ITU directs its treaty bound orgs to "take over". If gov types get concerned enough they can and will pass legislation that can, and will (with the power of law) take any and all of this away. Don't believe this can happen, look at various digital rights laws. DMCA in the US for example. Bad law can and does happen, it takes years to fix and lots of money. By dumping ICANN/IANA and walking, that could be viewed as "screwed up enough" to those on Capital Hill, Brussels, etc.
-----Original Message----- From: owner-lir-wg@ripe.net [mailto:owner-lir-wg@ripe.net] On Behalf Of Daniel Karrenberg Sent: Tuesday, October 08, 2002 2:14 AM To: john@chagres.net Cc: 'David Kessens'; 'Hans Petter Holen'; 'RIPE Local IR Working Grouo' Subject: RE: [lir-wg] ICANN Reform
At 04:29 AM 10/8/2002, John M. Brown wrote:
What IMHO, believe is that should the RIR's depart, which I'm against, this would open the oppertunity for other organizations to take control.
ITU, being one of them. This would be a particularly BAD IDEA
Not unless the RIRs screw up enough to loose the confidence of their communities.
Daniel
At 02:22 PM 10/8/2002, John M. Brown wrote:
By dumping ICANN/IANA and walking, that could be viewed as "screwed up enough" to those on Capital Hill, Brussels, etc.
Not convinced at all. Daniel
participants (9)
-
Daniel Karrenberg -
Daniel Karrenberg -
David Kessens -
Gert Doering -
Hans Petter Holen -
Joao Damas -
John M. Brown -
Randy Bush -
Siegfried Langenbach