[members-discuss] [ncc-announce] [GM] Publication of Draft Charging Scheme Models 2024

On 18.04.23 16:37, Clement Cavadore wrote:
> On Tue, 2023-04-18 at 16:30 +0200, Gert Doering wrote:
>> So a reclaim mechanism is required - the real question is, of course,
>> what that mechanism should be.  Pay NCC staff to go out and ask LIRs
>> "is this ASN still in use? is the data accurate? can you return it?",
>> or create an incentive for them to come back voluntarily.
[...]
> But AFAIK, RIPE NCC already does this kind of mechanism for ASN which
> are not seen on the DFZ. They ask if it's in use, and why it's not seen
> on the DFZ (and there are legit use case, for example for IXP route-
> server ASN, or for out of internet routing domains).

Make it a formal requirement to *either* be publicly connected *or* 
actively maintain the corresponding aut-num object, send annual 
reminders to those who seem AWOL, and investigate (+ bill extra?) only 
those who still haven't edited the "confirmed as of ..." field two 
months later, then.

Speaking more generally, the impression I'm getting from this entire 
discussion is that we don't have an agreement on what the *basic 
principle* of RIPE's billing should be, much less on a specific charging 
scheme implementing any such principle:

-- As a not-for-profit org, RIPE has an interest to charge in direct
    relation to the effort/expenses it has to cover. The *strictest*
    possible form of doing that would, of course, be to have fees paid
    for every activity (trouble ticket?) they undertake, possibly not
    even fixed ones.
-- Since they *also* have an interest to generate a prima-facie-known
    income *before* expenses need to be paid and defaulters hunted down
    (who here would be willing to pay a fee for being *turned down* by
    RIPE because of due diligence?), they *actually* propose fees on
    resources held, as an *estimator*, however poor IMHO, of the
    activities that those will require/generate.
-- For the polar opposite, there are those who said that plentiful
    resources that were handed RIPE for free should still be free for
    further distribution. Which essentially is an "eternally fixed
    prices" model, with IPv4 IPs being the striking example of a
    resource that *ceased* to be plentiful later on (and is unlikely to
    return to "free" short of IPv4 getting globally disabled, no matter
    how and how hard RIRs might try to push for that).
-- Then there's the idea of assigning prices as a means of policing
    distribution/usage, in particular, to cause unused IPv4 to be handed
    back to RIPE. (Which means that RIPE would need to invoice *more*
    per IP than you can earn by renting/selling them on the free market
    - which will happily react by raising the price, etc. ad infinitum
    -, *or* somehow manage to disrupt said free market ... ho hum.)
-- Last not least (though I'm probably still forgetting some), those who
    would like RIPE to stay with identical fees for all members,
    independent of their size. Which ultimately translates into "pay per
    vote" ...

Two more thoughts:

If there is pressure to give "back" IPv4 space that you don't use 
yourself (and, again, the price for which you could instead sell them 
*already is* such pressure), it'll get more fragmented than ever, and 
thus eat more RAM etc. on every EBGP peer. Are we really *helping* IPv4 
there ... or handing it more rope to cut its misery ever shorter? :-3

Finally, I'd like to speak in favor of intentionally "overcharging" 
members with the charging scheme and "redistributing" the excess back to 
them (a year) later. Stripping RIPE down to the minimum viable budget 
may look good on everyone's invoice, but it also creates a 
representative of the community that is a defenseless pushover as soon 
as some commercial entity (with some money to spend, contacts to lobby, 
etc.) deigns to take over. Assuming that the "excess" stays roughly the 
same and *no* hostile takeover is attempted, it can be considered a 
deposit for the duration of one's membership, even though the figures on 
the bill *look* that much higher every year.

Kind regards,
-- 
Jochen Bern
Systemingenieur

Binect GmbH