Hi David, Thanks for your useful information and valuable comments. I have no objection to people using free identifiers, but I don’t see a strong reason stopping people pay RIRs for using IPv6. I wish companies from IoT industry become APNIC members and apply IPv6 only for their services. This is the area APNIC will continue to grow after IPv4 exhaustion. With supercomputing in new datacentres and AI support, similar to assigning single IPv6 address to each electronic smart device, mapping single IPv6 address with each non-electronic item for hosting unique information of that item is possible. It is a real Internet for everything. There are solutions to address security and privacy concerns. I will consider all feedback from the community and modify the proposal. BTW, I will be on travel in next few weeks and might not answer questions timely. I will see you all at APNIC 58 in New Zealand. Kind regards, Guangliang (Benny) ================
________________________________ From: David Conrad Sent: Wednesday, August 14, 2024 2:43 AM To: Guangliang PAN Cc: Wesley; [email protected] Subject: Re: [sig-policy] prop-161-v001: Using IPv6 for Internet of Things (IoT) -- correct version Hi Guangliang, On Aug 12, 2024, at 11:02 PM, Guangliang PAN <[email protected]> wrote: Use the same example I mentioned, IPv6 has a key benefit on anti-fake.[…] The use cases you’ve described are remarkably similar to those used for https://www.gs1.org/standards. In practice, there are many reasons why systems like these were not universally accepted, not least were privacy-related and the fact that many in industry didn’t like the idea of their products being identified in a global registry. Technically speaking, the IPv6 address is assigned to a node hosts a page (URL/FTP/whatever) containing the information about a non-electrical object. The IPv6 address is always connected (online) and respond to any query from the Internet. From this point of view, IPv6 address is not actually assigned to non-electrical item but to the interface hosting the information on the Internet about that item. Since there has to be a gateway to represent/proxy/translate for non-connected objects, why not use a (free) PEN or a (free) self-generated UUID instead of (e.g.) having to pay yearly for an IPv6 prefix that will never be directly routed? RFC1881 - “IPv6 Address Allocation Management” created in 1995 defines how IPv6 will be allocated […] There is NO mention of IPv6 could not be used for non-electrical items. RFC4291 - “IP Version 6 Addressing Architecture” created in 2006 provides guidance on how to use IPv6 in the networks. […] There is NO mention of how to use IPv6 for non-electrical items. There is also no mention of using IPv6 addresses as confetti, but that doesn’t suggest it would be a good idea. There is an argument that can be made that IoT devices can/should be numbered with IPv6 as it seems likely that (eventually) the cost of hardware/software/power/etc. for implementing a full TCP/IP stack will fall sufficiently for it to make sense to use it instead of different/simplified protocols, although those devices should obtain their addresses via SLAAC/DHCPv6 to ensure routability. This obviously doesn’t apply to non-connectable objects. Regards, -drc
_______________________________________________ SIG-policy - https://mailman.apnic.net/[email protected]/ To unsubscribe send an email to [email protected]
