<br><br>On Tuesday, September 24, 2013, Randy Bush wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">> IPv4 will, undoubtedly, eventually run out.<br> <br> ipv4 will be around after everyone on this list today is retired.</blockquote><div><br></div><div>Very good point. </div><div>I can hate it - but I would not bet much on you beeing wrong - even if I am going to retire a few years after you. </div> <div>(...)</div><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <br> we would like it if the registries acted as registries, i.e. recorded<br> accurately who controls/owns what address space. we thought that was<br> their primary responsibility</blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"></blockquote><div><br></div><div>Fully agree. </div><div><br></div><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> the registry system has a choice. it can make it difficult for users<br> to register, transfer, sauté, whatever address space, in which case<br> they have some chance at a reasonable level of accuracy in their<br> registries, or they can make it difficult. in the latter case, they<br> will become vestigial organs whose function we will soon not even be<br> able to remember. and we sure as heck won't be sending money to them.<br> <br> the choice is ours. today, dealing with the registries' petty rules<br> and bureaucracy is a major pita. this policy proposal tries to make<br> it easier.</blockquote><div><br></div>Since the start of my first encounter with the RIR Ripe 20 years ago we have gradualy<div>reduced the documentation to demonstrate the need for addressess (have had to send receipts for equipment and such a couple times - we do not do that anymore) - thus there is more trust in the LIRs self assesment of need than it used to be. </div> <div><br></div><div>I think the important shift now is that this proposal actualy transfers the responsibility for conservation fully to the LIR - much in the same way as the responsibility for aggregation has been with the LIR for years. </div> <div><br></div><div>(in the end of 80s the central registry even tried to look after aggregation by delivering a then class B - now /16 instead of 4 class C /24)</div><div><br></div><div>I would argue that we have a strong tradition to move in the direction of deregulation ad more trust in the LIR in this region </div> <div><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br> it really makes small difference what we choose. the internet will<br> route around whatever impediments we place. it's a big river now.</blockquote><div><br></div><div>Right. </div><div>It is probably fairly easy to transfer IP Addressess from company A to B in most regions trough other means :</div> <div><br></div><div>company -A forms subcidiary A-registry ltd. A-registry Ltd is sold to B and then merged with B. </div><div><br></div><div>Just is slightly more expensive and keeps the lawyers busy. Actualy - in my country I think I could to the legal work myself - its not that complicated. </div> <div><br></div><div>(and i have a real life example for a /8 that has made its way from Norway in the RIPE region to USA in arin Arin region this way - in a slightly more complicated manner with a few more steps). </div> <div><br></div><div>A side effect of allowing addresses to be more or less freely tranfered is that it could actualy bring more addresses to the market - when early adopers figure out that they can run their entire corporate/governmet network on private addresses between a NAT keeping only their servers offering public services on public addresses - the same way as most companies has had to do for years. </div> <div><br></div><div>As soon as the market price is higher than the cost to redesign and renumber this will start to happen. </div><div><br></div><div>We should probably take this further and split the policy in 3:</div><div> <br></div><div>- a clear registration policy that applies to everybody and requires addresses to be registered</div><div>- a transfer policy - focusing on proper identification and authorisation of transfers</div><div>- an assignment policy for the last /8 from the RIR and whatever address space was returned </div> <div><br></div><div>I belive that if we want to have _need_ included in the policies it should be in the last two - and we need to come up with a needs criteria and documentation of such that can be audited by an independent party - if we want theese to have some real world legitimacy. </div> <div><br></div><div>I think. <br></div><div></div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <br> randy</blockquote><div><br></div><div>- hph</div><div><br></div><div>PS: At the last ICANN meeting the US GAC representative read a long prepared statement on freedom and as little regulation as possible. </div></div><br> <br>-- <br><div dir="ltr"><div><div style="color:rgb(0,0,0);font-family:AvenirNext-Regular;font-size:22px;text-align:left">Hans Petter Holen<br></div><div style="color:rgb(0,0,0);font-family:AvenirNext-Regular;font-size:22px;text-align:left"> Mobile +47 45 06 60 54 | <a href="mailto:hph@oslo.net" target="_blank">hph@oslo.net</a> | <a href="http://hph.oslo.net" target="_blank">http://hph.oslo.net</a></div></div><div><br></div></div><br>