On 19/08/2010, at 11:51 AM, Terence Zhang YH wrote: > Hi Terry, > > Thanks for your comments. > > The statistics you show is correct, there are not many CI assignments these two years. > CI requirement is not large, but it's important. Since ICANN recently launched > New gTLD program, IDN and IDN ccTLD fast Track, > we can expect modest increase of new IDN TLDs and gTLDs in the next few years, > and that will coincide with our entering into final /8. > Right, modest. and even with the updated stats from Sanjaya, 36.7% utilised over 7 years still says to me that there is head room there. >> The question in my mind relates to if 203.119/16 is 100% set aside for critical infrastructure assignments or not, given that section 11.3 doesn't actually say. If so then I struggle to see what real live problem prop-085 is going to solve. My belief is that the final /8 will have its assignment policy set and will be all consumed well before enough new critical infrastructure organisations can form and apply to use up the remaining space in 203.119/16 which would imply a restraint in the wrong direction. >> > > What I understand is, even if that block is reserved and available, assignments/allocations > from that block still have to be justified according to some policy criteria. yes. as would be expected for a very constrained resource. > Currently final /8 policy ONLY allow allocations, so even if there are enough space in 203.119/16 when we enter final /8, > critical infrastructure users still have no way to justify their needs using '11.3 Critical Infrastructure Policy', > they have to justify their needs under allocation policy : > 9.3 Criteria for initial allocation > 9.4 Criteria for subsequent allocations > Which they might have difficulty to justify, ie. they may not be able to show the need of /22. > o.k. If you say so. Although originally that wasn't my interpretation of the effect of prop-62-v002. But it seems as written to be the case. > >> If 203.119/16 isn't set aside for just CI applications and other member applications can encroach on it, then I think you might want to consider that to be the low hanging fruit instead of heading toward the last /8 policy space. >> > > According to the final /8 policy '9.10 Distribution of the final /8', > the final /8 doesn't mean a single stand alone /8 block, it means > 'When the total remaining space in the unallocated APNIC address pool reaches a threshold of a total of one /8' > right. > So, > If the 203.119/16 still have available space when we enter final /8, it's a component of the final /8 space, > it's reasonable to continue make CI assignments from it. yes. > > if the 203.119/16 is used up when we enter final /8, that shows the need is steady, it's reasonable > to open another block for it. sorry "if".. that is stretching it a bit.. And that is the problem I have. With still over 60% of the 203.119/16 still remaining and a consumption rate at such a low level, and even with the new IDNs (which generally go to the existing gTLD/ccTLD) and TLDs I would rather see an pragmatic analysis of the proposed CI demand backed by facts which coincide by the last /8 mark to say that the CI /16 would be consumed by then and there would be a very very high likely-hood of another block from the /8 required for CI.. if we aren't all on IPv6 by then. My other point is at the final APNIC /8 point AND the 203.119/16 is consumed I see very little difference between a new CI organisation and a LIR. Why should we bless one category in particular?