On 31 Jul 2017, at 8:10 PM, Randy Bush <randy@xxxxxxx> wrote: > ... > in our meager experience, pure ipv6 with no hack to get to the ipv4 > internet is not viable from the user perspective, period. it won't be > for many years. yes, this is a tragic failure; and we know where the > root causes lie. > > nat64/dns64 is useful in constrained circumstances. the ietf meeting > network is not one of them; as there is far too wide a variety of users, > some with laptops they can't even control due to IT departments, many > use vpns, applications are still incompatible, ... this is yet another > tragedy. > > the biggest obstacle to improvement on both these fronts is the denial > that there are serious problems, and hence far too little serious work > being done to ameliorate them. > > but, no matter how hard we work, there is not going to be a magic day > when it all just works. ipv6 is on-the-wire incompatible with ipv4; > so it is gonna be hacks; and hacks are tecnhnical debt. Randy - Agreed on all of the above – it doesn’t appear that forced usage of IPv6+nat64 is really compatible with an IETF network "where people can get their work done with absolutely minimal speedbumps” However, the draft proposing same notes that more IPv6/nat64 usage at IETF may very well help to identify applications, services, or use cases that are not entirely compatible with that setup, and that seems worthwhile - particularly if the lessons learned are published in a I-D afterwards. To that end, would it be possible for those behind the "IPv6-only Wi-Fi for IETF” draft to partner with the NOC at the upcoming meeting to help debug issues that are reported on “ietf-nat64* and publish the results? The IETF does have a wide variety of users, which is really helpful in finding edge cases with protocols that might have issues and otherwise go undiscovered. If it were may clear (e.g. during the opening) that there is a goal to put the “ietf-nat64” network through its paces and that there would be a post-mortem I-D as a result, I suspect quite a few folks would at least give that setup a try for a while, with the goal of finding and reporting breakage. /John