<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style></style></head><body lang=EN-US link="#0563C1" vlink="#954F72"><div class=WordSection1><div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>>I see pros and cons for both accepting this proposal and rejecting it.<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>>One thing I'm curious about.. ARIN has run out of IPv4 space.. <o:p></o:p></p></div><div><p class=MsoNormal>>Has this stopped any new startup from doing business or what?<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>--<o:p></o:p></p></div><div><p class=MsoNormal>>George<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Hi George,<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>It seems like a common belief among the proponents of the various soft-landing policies that new entrants will find the lack of registry-provided IPv4 blocks to be a significant barrier to entry.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>We sell IPv4 blocks to new entrants quite frequently, and as has been posted on this thread, even a /22 can be extended to cover quite a few customers. Doing the math, the cost per Ipv4 address per customer served (using non-Belgium CGN) turns out to be less significant than most other costs that companies have to incorporate into their plans.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>If an address costs $15 and it serves 30 customers, the new entrant is spending $.5 per customer for a network component that is intrinsic to the ability to provide service. <o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>My guess is that this cost is far less than hardware or even advertising costs per-customer.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>And ARIN never has had to worry about fixing their final /8 policy to counter attempts to game the system via serial new LIR applications, as in RIPE.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>ARIN does not have to impose a moratorium on transfers of a certain class of addresses, as in APNIC.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>In retrospect, I think the ARIN policy was far-sighted and quite effective. I have heard no voices bemoaning the situation here from my perspective as a broker.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Finally I wanted to mention that should RIPE implement this policy, address prices would not double, the relation between this policy and pricing is very tenuous. It’s certainly not like the IPv4 market price has a mathematical relationship to the RIPE entrance fee divided by the allocation size for new entries. Now it’s 2000 Euros and you get 1024 addresses. There is no relationship between those numbers that yields a current price ($15) and the current price has risen over the last two years while the RIPE fees and allocation size has not changed.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>I am against the policy as written, I think the better route would be to reach complete exhaust everywhere and a enable a global transfer market without “classes” of addresses.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Regards,<o:p></o:p></p><p class=MsoNormal>Mike<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p></div></div></div></body></html>