Re: [arch-d] ipv4 and ipv6 Coexistence.

[Vittorio Bertola <[e-mail]>]

> Il 26/02/2020 11:45 Stephen Farrell <[e-mail]> ha scritto:
> 
> If you do want to rehash the cryptowars debate again,
> maybe at least change the subject line? If doing so,
> and I hope you decide not to, because there's no gain,
> please recognise that there are many (me included) who
> do not share the views you quoted for sound technical
> reasons (e.g. [1] is perhaps the most-cited recent
> paper on that) that have garnered rough consensus in
> the IETF every time we've done that rehashing.

I think that no one here wants to reopen that discussion - we all agree that, from a technical standpoint, putting backdoors in cryptography is a bad idea. However, I wholeheartedly share the rant that Toerless posted; the problem is that, outside of the IETF, there are real public order issues that the IETF refuses to address, just continuing to deny their relevance or to suggest that they are technically impossible to solve. 

This behaviour will not make the issues disappear, nor the push by non-technical stakeholders to get them addressed, so either the IETF can come up with a workable technical reply to "what should we do then to enforce the law on a fully encrypted Internet?", or the push will simply break the Internet as we know it, and then you will have more countries building great firewalls, or mandating really bad technical solutions by law, because no one built a good one for them.

There could be solutions to the above question that impact the architecture of the Internet without breaking the communication protocols - for example, related to where the endpoints are located and who runs/gets access to them. Can this be done securely, with a reasonable compromise between privacy and access, etc.? I don't know, but AFAIK the IETF until now has refused to even discuss the issue - and even if it is mostly a policy issue, sound technical know-how and possibly bits of specific protocol development are likely to be part of any good solution. (Think of the RDAP protocol designed to handle global tiered access to domain registrant information, though it's not really working yet because the policy part has not been agreed - still, without building the appropriate technical tool there would be no way to address the policy issue.)

I understand that this is hard because it is not something that technologists can do alone; it requires interdisciplinary cooperation among many open minded people (open minded in the sense that they have to listen to and acknowledge very different views on what the Internet should be and do). But I see this as the big architectural challenge in the near future.

-- 
 
Vittorio Bertola | Head of Policy & Innovation, Open-Xchange
[e-mail] 
Office @ Via Treviso 12, 10144 Torino, Italy