Geoff Huston wrote:

It would also be helpful if the details of the allocation were removed from the RIPE NCC stats file (ftp://ftp.ripe.net/pub/stats/ripencc/delegated-ripencc-latest/)

"ripencc|ES|ipv6|2001:7f9::|32|20040719|allocated"

as well as from the RIPE WHOIS database - after all its no longer allocated space we are talking about here. [..]

...

inet6num: 2001:07F9::/32 org: ORG-CA59-RIPE netname: CONSULINTEL-20040713 descr: Experimental Prefix ======================== Duration of assignment: ======================== Start date: 2004-07-19 End date: 2006-07-19

YEAH You can get free IPv6 address space from RIPE apparently. As the above has now become: 8<-------------------------------------------------------------- inet6num: 2001:07F9::/32 org: ORG-CA59-RIPE netname: CONSULINTEL-20040713 descr: Experimental Prefix ======================== Duration of assignment: ======================== Start date: 2004-07-19 End date: 2007-08-01 ======================== Details of Experiment: http://www.consulintel.euro6ix.org/prefix.htm ======================== country: ES admin-c: JPM128-RIPE --------------------------------------------------------------->8 For anybody wanting to dodge the LIR fees, clearly just request an experimental allocation under the premises that you do some experimental stuff, then you can use it for your commercial activities. Fortunately it is now sort of documented (see the new url which actually contains information). After 3+ years of not being so. It is of course funny to read "Some of those activities may behave disruptively at some moment". Funnily all the experiments mentioned there (without actual results which is something that the experimental request requires one to publish publically, that is also mail onto the mailinglist) have all been carried out already by several other entities who have been paying all those years for the prefix they used. To sum a few things up which are clearly skewed: They are testing with 'multihoming', but the doc states "the prefix is announced in a stable and aggregated way to the routing table", clearly that is not the thing that SHIM6 is supposed to do as SHIM6 is not touching the routing tables. Also multihoming implies multiple prefixes, not a single /32. Teredo is also out of scope for that document as Teredo has it's own IANA prefix. "Support to phase-out of the IPv6 experimental network - 6Bone - IPv6 Day http://www.ipv6day.org" Since when is webhosting an experiment? And why does that require a /32? "Research and development of Tutorial's contents and distributed demonstrations, providing tunnels for the participants when required - Number Hands-on Tutorials for RIRs: - ARIN XVII - RIPE 52 - AFRINIC" Two of these RIR's are outside the RIPE region, next to the fact that these RIR's usually have their own IPv6 connectivity, thus one doesn't need to use your own connectivity. As for documentation purposes there is 2001:db8::/32 I still wonders WHY there is such an experimental prefix, seeing that these 'projects' are clearly funded and very commercial and thus can also follow the normal guidelines, the same as every other company is doing. A single /48, received from their upstream provider is all they need. Clearly from the routing tables there is a single, US based, not even close to Spain, uplink, who are very capable of providing a free and gratuit /32. Thus I wonder why RIPE NCC is being bothered with this work. Apparently RIPE NCC doesn't mind doing it either, otherwise they would not have extended it. It's also a bit odd to see that the participants of this 'experiment' don't read the IPv6 mailinglist, they don't seem to have responded why they didn't stop announcing the prefix. Of course now they seemed to have received another free year of usage of the prefix. Greets, Jeroen