Hi, On Tue, Aug 12, 2003 at 10:19:44AM +0100, Peter Galbavy wrote:

1. Everyone pussyfoots around the issue of RIPE =?= RIPE-NCC. As the funding for both are out of the same pockets, please STOP trying to make that distinction. If RIPE (as a natural monopoly) was classed like BT, then this practise would be seen as cross-subsidisation.

You're confused. RIPE isn't funded in any way. RIPE is "all of us". The funding goes to the RIPE NCC, which has offices, employees, and needs money to do the work that RIPE (we) ask them to do. [..]

2. RIPE, again as a natural monopoly, does NOT offer "members" the choice of opting out of the "fluffy stuff". RIPE should, IMHO, provide registry services ONLY and base its costs on that. The other hand waving, experimental, attempted standard setting stuff should be optional and extra. At the moment, those of us who just want IPes and ASes have to pay for others to play with their academic toys. Why ?

Because the majority of the members hasn't voted against it. [..]

Does anyone believe that RIPE is not a "natural monopoly" for IP registry services in Europe ? If it isn't, as I predicate it is, then I get choice and can take my business elsewhere. Going to an "ISP" is not the choice I can make, so don't try that one.

It is a natural monopoly in the way that you can't go elsewhere if you want RIPE member services. But then, how else do you want to do hierarchical distribution of a limited resource? On the other hand, if you just want IP addresses and AS numbers, you *can* go through an ISP (but it will reduce the number of options that you have). Gert Doering -- NetMaster -- Total number of prefixes smaller than registry allocations: 56535 (56318) SpaceNet AG Mail: netmaster@Space.Net Joseph-Dollinger-Bogen 14 Tel : +49-89-32356-0 80807 Muenchen Fax : +49-89-32356-299

I am not looking to break the natural monopoly, but rather I am looking to move to a situation where the "monopoly" stuff is walled off from the optional stuff that RIPE/RIPE-NCC management (and friends) use to pay for their own pet projects. I am happy to pay on a cost basis for the "monopoly" stuff, but I don't get a choice.

This was something i raised years ago prompted mainly by the test traffic white elephant which used registry monies to plod along and then eventually go on a subscription basis, and this something i had no intention of ever using but still the registry i was with paid its money and had no say over how it was used. Something that has been repeated over and over again is that the NCC should be a registry not a research dept for internet trends and toys. There should be 4 main functions 1.registry services (good and timely) 2. database repository (and related DB functions) 3. Training services (mainly for new LIR's to get them up to speed) 4. RIPE meeting organization All of which the NCC do well and should be commended, anything else is surplus to requirement there is one other thing the NCC has never done and never fully justified and that is a help desk you can talk to. Other RIR's do it. I think it helps cut away the sterile faceless image and lets you see what you are getting for your money - discuss Stephen Burley Internet Communications Consultant Africonnect

On Tue, 12 Aug 2003, Peter Galbavy wrote:

Stephen Burley wrote:

...

1.registry services (good and timely) 2. database repository (and related DB functions) 3. Training services (mainly for new LIR's to get them up to speed) 4. RIPE meeting organization

I agree with 1-3. I think 4 should be a "spare time" activity or one organised by an EU Internet "club". If 4 can be done part time by *one* person, then I am happy to pay for my part of the time.

i agree with 1+2+4. 3 should be covered on a cost basis by the attendees. if the cost of a training session is 5000euro (room, lunch, instructor time, etc) and 25 attend, then the cost should be 200euro per attendee, with no affect on the overall ncc budget.

Peter

Hank Nussbacher

Gert Doering wrote:

I don't understand that question. You were complaining that people make this distinction "just to confuse matters" - and I was trying to give a precise definition as for what is what.

Especially it's not "RIPE *and* the RIPE NCC are both funded in some convoluted ways". RIPE is NOT (and can't be, by definition).

I know it is not an English acronym, but perhaps someone could translate what RIPE stands for ? It is not "the European ISP club" is it ? It is the European IP coordination function (AFAICR). I dislike people pretending that RIPE-NCC membership fees do not fund RIPE activities. Else we would only see costs associated with running a registry in the *RIPE-NCC* annual report.

...

Because the system is weighted in such a way that getting a vote proposed, let alone voted on by any real number of people, is difficult to impossible.

I can't agree with that statement.

I have been quite successful in changing some of those pieces that annoyed me.

From reading the docs a few months ago, I need to find 5% of the membership to agree to any proposal I may make to be put before a RIPE annual meeting (for example to propose the RIPE/RIPE-NCC immediately cease all non-registry related activities and to begin a program of rationalisation that reflects

But that is / was your job ? If you have close to 100% of your working day dedicated to RIPE / regulatory affairs / whatever, then getting involved is easy. People in(/friends with people in) RIPE/RIPE-NCC are basically in control. Most members do not even have anyone who reads this stuff. Hence my comment about apathy. the needs of the industry). I have NO communication path to approach this notional 5% as I do not have access to a membership list, and I dislike spamming folks anyway. Not all members are on an open mailing list (AFAIK) - or rather not all those who are in a position to vote on behalf oif their company. Without knowing who are members, and how many are "5%" I cannot try to put forward a democratic proposal of the sort above. On the other hand, the executive board can put forward proposals unconditionally - or have I remembered wrong ?

In what way is "getting an AS number from the competition" something that's harmful for your business in the long run?

AS is OK - it is for the lifetime of the assignment, except the maintainer object is required for changes and last time I asked, new maintainer objects are only for members.

As far as I understand your situation, getting a PI address block through any other ISP, and announcing that via your AS (that you have already) should solve your needs without causing any competitive problems either.

PI is "wasteful" and not guarenteed (for some value of that word) routeable. PA is owned by the upstream, and also makes most multihoming impossible. I have pre-PI/PA space I can use for my own self, but some of this is not actually just about my specific case.

Isn't that exactly what the activity plan is about, which is agreed-upon on a very specific date that was announced *WELL* in advanced, and where every LIR can go and vote for or against? You can even bring proxy votes.

I oppose the whole concept of having a plan, not the plan. If the sole activity was registry services, there would be no need for a plan. Peter

On Tue, Aug 12, 2003 at 10:19:44AM +0100, Peter Galbavy wrote:

1. Everyone pussyfoots around the issue of RIPE =?= RIPE-NCC. As the funding for both are out of the same pockets, please STOP trying to make that distinction. If RIPE (as a natural monopoly) was classed like BT, then this practise would be seen as cross-subsidisation.

I agree. When I have to send off forms for allocations and such, I can't remember ever having said "RIPE NCC" instead of "RIPE". To most of us in the business - I think - "RIPE" is a general term for "those guys that will not give us more IP addresses unless we really, really need them." The distinction seems to be some bureaucratic necessity. I'd really rather the geeks ran this world :-)

2. RIPE, again as a natural monopoly, does NOT offer "members" the choice of opting out of the "fluffy stuff". RIPE should, IMHO, provide registry services ONLY and base its costs on that. The other hand waving, experimental, attempted standard setting stuff should be optional and extra. At the moment, those of us who just want IPes and ASes have to pay for others to play with their academic toys. Why ?

To some extend, I agree with your notion. It can be irritating, indeed, to have to pay up for something you don't really see a purpose to. However, it seems to me to be much like the whole "should our tax euros really be spent paying for research into things that do not have an obvious practical implementation (yet)?"-discussion. Whatever non-IRR-stuff RIPE is doing _may_ turn out to be the next Big Thing on the nets. It may also fade away into the distance. Somebody, however, have to pay for research into that which will pay my salary in ten years, and the companies that will probably prosper from it are a reasonable suggestion for where to send the bill.

3. The registry should be run efficiently, not just "quickly". From the reports that others have sent me off-list in the past, my suspicions are strong that there are basically too many staff at RIPE.

The great big problem with anonymous off-list mails is the fact that they are anonymous. (Not to you, I know, but to the list). If there's a problem, those who know of the problem should inform the lists.

Does anyone believe that RIPE is not a "natural monopoly" for IP registry services in Europe ?

Nope. It definitely is. However, such seems basically unavoidable given the fact that we _need_ a central office to distribute IP addresses and AS numbers. Note that RIPE is only a monopoly on IPs and ASs in the sense that Volvo is a monopoly on Volvo cars. If you want to drive a Peugeot or want to make your own ipv4-internet, seperate from the one you use now, neither Volvo nor RIPE has any say. Peter B. Juul, Uni·C (PBJ255-RIPE)

Peter B. Juul wrote:

To some extend, I agree with your notion. It can be irritating, indeed, to have to pay up for something you don't really see a purpose to. However, it seems to me to be much like the whole "should our tax euros really be spent paying for research into things that do not have an obvious practical implementation (yet)?"-discussion. Whatever non-IRR-stuff RIPE is doing _may_ turn out to be the next Big Thing on the nets. It may also fade away into the distance. Somebody, however, have to pay for research into that which will pay my salary in ten years, and the companies that will probably prosper from it are a reasonable suggestion for where to send the bill.

How is the IETF funded ? The IETF doesn't do addressing, but seems to exist without me paying for it. In this instance, IMHO, academics should be funded by academic models and maybe by corporate sponsorship, they should not leach off "us" by the back door. BTW RIPE is not a government department funded by taxpayers, else it would be subject to audits by higher bodies, not by the collective apathy of a membership.

If there's a problem, those who know of the problem should inform the lists.

I was asked in this instance to not repeat names as one of the people was not available to approve the release of information they collected. I am happy to support people who wish to speak up.

Note that RIPE is only a monopoly on IPs and ASs in the sense that Volvo is a monopoly on Volvo cars. If you want to drive a Peugeot or want to make your own ipv4-internet, seperate from the one you use now, neither Volvo nor RIPE has any say.

Using this metaphor, RIPE has a monopoly on traffic signs and road numbers. I can build a road, but I cannot join it to the rest of the public road network without their involvement. Peter

How is the IETF funded ?

My understanding is that the IETF is funded the same way as the *RIPE* is funded: - meeting participants pay a fee to cover the meeting costs, with some sponsored events added - the IETF chair Harald Alvestrand works for Cisco Systems so I assume Cisco pays for his time (just as the RIPE chair works for NIKHEF who pays for Robs time, just as I work for Tiscali who pays for my time (well if you ask my family they will probably claim I do this out of office hours so its my own time) When it comes to the RIPE NCC of the IETF (just waiting for the flames on that one...) the RFC-editor it is funded by a membership organisation (the Internet Society)

The IETF doesn't do addressing, but seems to exist without me paying for it.

You probably do indireclty; if you look at the list of Area directors and WG chairs, and even the working group members spedning time on improving the Internet you probably pay trough some product you use If you dont want to pay the RIPE NCC for the service I (or somebody else for that matter) could probably offer you Internet transit with IP address registration bundled as a service for your customers.

In this instance, IMHO, academics should be funded by academic models and maybe by corporate sponsorship, they should not leach off "us" by the back door. BTW RIPE is not a government department funded by taxpayers, else it would be subject to audits by higher bodies, not by the collective apathy of a membership.

I can agree on that principle - there will be an oportuity at the next RIPE meeting - both in the RIPE NCC services wg and in the general meeting to change the direction of the ship. -hph