ARIN Proposal

[Howard C. Berkowitz]

I'm responding to two positions here. The proposal drafters, as I
understand, have put out a fee structure. Michael Bathrick and others are
saying the fees are too large for them to pay.

To me, there is a problem with both positions.  I do not believe the
initial proposal should have suggested any actual numbers for fees, but
instead have discussed the functions to be performed and some estimate of
the workload  in terms of number and types of work requests (as opposed to
the resources needed to service them).

Consensus needs to form as to:
  1) The need for a registry (I think this is self-evident, but I'll put
     it down for completeness)
  2) The functions that registry will perform, including availability
     goals for public servers,
  3) An estimated workload of allocation requests, ASN requests, etc.

Once there is a consensus on these requirements, the means for carrying
them out can be examined.  There will be a set of fixed costs that must be
allocated over the expected number of direct customers (assuming that is
the only funding source), and there will be some set of variable costs that
varies with the workload.

If the services are accepted as needed, and it is agreed that it will cost
a certain amount to provide them, then those costs will have to be
distributed over the current and prospective customer base.  If those costs
are excessive for small ISPs, then either there needs to be a community
consensus that there is a non-economic reason to subsidize them (and there
may be), or, sadly, a certain amount of natural selection comes into play.
(I'd really like to play NFL quarterback, but I just don't have what it
takes...let's be honest...defensive back...gets to hit people).

But we now seemed locked in a battle over arbitrary numbers.  How can that
go anywhere?  Even if a new proposal changes the numbers, unless there is a
derivation of where the numbers came from, I see the flames continuing.

Howard Berkowitz

At 3:55 PM -0600 1/20/97, Brian Tackett wrote:
>On Mon, 20 Jan 1997, Michael D. Bathrick wrote:
>
>> I have to agree with David - the fees posted will do nothing but hurt the
>> small ISP, many of whom are running on the hairy edge of existance.  We
>> were never consulted when Network Solutions chose to impose a domain
>> registration fee - but at least that fee was within reason.  The fees
>> currently proposed are outrageous.
>
>*sighs deeply*
>
>OK....now it looks like I'm in the slightly odd position of defending this
>proposal. I'd like to just take a time out and aswer what seem to be the
>most common misperceptions of this plan:
>
>1) As explained by the proposal's primary drafter, the fee schedule is
>based on a yearly fee, charged on the cumulative amount of address space
>allocated in the previous year directly from the registry. If ISP X
>allocates 2 /16's, they pay $20,000 total. THe pricing is *not* $20,000
>for each /16, etc. At the end of the year your allocations are added up,
>and what you pay depends only on the range you fall into.
>
>2) The small ISP (non-multihomed) WILL NOT have to pay this price
>schedule, though they may have to pay a lesser amount based on upstream
>providers charging to recoup their investment. *ONLY* ISP's which obtain
>allocations *directly* from the registry will be paying the proposed fee
>schedule, and since it is currently not possible to obtain an address
>block smaller than a /19 (32 class C's) from InterNIC, there will really
>be no change in this aspect for small ISP's.
>
>3) The allocation guidelines will not change unless it is decided for
>other reasons to do so in the future. If you couldn't get a /19 now, you
>won't be able to get it then either, money irregardless.
>
>
>Now....can we PLEASE wait until we see the next revision of the proposal
>to start wasting time and bandwidth again? Kim has assured as that a good
>bit of work is going into making the next generation readable and clear,
>and that objections made previously on this list and others are being
>considered. All we're doing until then is rehashing the same arguments
>over, and over, and over, with no knowledge of whether or not said
>arguments have or have not been dealt with in the next revision of the
>ARIN proposal.