[ipv6-wg] Have we failed as IPv6 Working Group?

*knock knock* is this mic on ?

Once upon a time during the FOSDEM post-conference gathering with organizers I challenged the notion that the dual stack was the state of the art. The organizers, being themselves in various tech roles, agreed. It was interesting to get to those days bleeding edge. Of course with NAT64. 

Thus FOSDEM became the very first large-ish (above 5K clients) WiFi setups in the world to go with IPv6-only+NAT64 in the default SSID. The announcement was made during the opening session, which everyone attended.

We maybe had 20% clients remaining on the default “bleeding edge”. Which is a testament to the immense power of human laziness aka the law of the default: it was several hundred times more compared to the previous year when they SSID was “on the side”. 

Mind you, this was the time when Android still wasn’t able to connect to IPv6 only on WiFi, and when iOS was flapping WiFi every minute, if connected to v6-only access. That time is long gone - every subsequent year the %% participation on the default IPv6-only access SSID nearly doubled. 

Were there complaints ? In single digits, sure. (And about complaints in general: no temperature in the room can help avoid having complaints about it being too cold or too hot. Some of those complaints came from dual stack SSID and general confusion that happened first year) - if there is anyone from rest of the FOSDEM network crew of those years reading this, please correct me if I am missing anything.

But the overall feedback from the participants was overwhelmingly positive. And there was a lot of bug reports, questions, etc. directed at makers and vendors that didn’t support IPv6. 

Since then it’s business as usual and no one is even asking.

So, IPv6-only access with NAT64 was a good thing to do for a technology-focused gathering of people.

For an unrelated business-focused gathering that would not have been a useful idea.

RIPE and IETF (and the vendor conferences for that matter) - being the mix of the two, are tricky to decide.

To me it all depends on an answer to a simple question: “do we want to lead the trends or do we want to follow them, and which ones ?” NB: both are valid business strategies.

What you care about with an IPv6 only access network at a short event is not necessarily just testing the apps. You care about the lasting memory and perception of that event that will be projected outward. It stopped being purely technical several years ago.

I hear “but there are non-technical people who just wanna get their job done”. Ironically from my experience they are the least ones to have problems, because they don’t use fancy outdated VPNs. And even if they do - they are empathetic enough to understand the modern 20-year technology still has kinks to iron, and smart enough to switch over to “-fallback” SSID or similar. Would you not be in their position ? If no, I am sorry for you. If yes - why would you think of them as lesser capable humans?

I hear “but if people experience problems with IPv6, it gives it a bad rap”. To which I reply - the opposite of love is not hate, its irrelevance and  indifference. Make it the best possible experience and let the glitches drive the improvement. This is the same way everywhere - business, relationships, knowledge.

I hear “but dualstack works fine”. Sure, but for a 4-day long event so does, from a layman’s practical standpoint, pure IPv4-only. Those in dire need of 128-bit address can use their corp VPN or - if they don’t have it - Cloudflare’s Warp+.  The latter works beautifully over any legacy - and accelerates the web experience too! [shoutout to Mahtin. Thank you so much!]

When I look at today’s application architectures and latest trends, I see a lot of parallels between today’s IPv4 Internet and PSTN from 30 years ago. It’s fascinating to imagine how it all looks in 30 years from now, in another perspective, when it inevitably changes again. 

Provided that whole climate thing still allows us to hang around :-)

Thanks for listening if you made it till here.

Sometimes I wish I could make a sequel to my now probably biggest career contribution (“NATs are good” video), but ironically that startup is no more, they got bought, and it’s all now a serious business, no kidding.

—a


> On 11 Oct 2019, at 09:30, Philip Homburg <pch-ripeml at u-1.phicoh.com> wrote:
> 
> In your letter dated Thu, 10 Oct 2019 23:56:12 +0200 you wrote:
>> * Gert Doering (gert at space.net) [191009 19:57]:
>>> Troopers runs their main conference wifi with NAT64.  If I'm not
>>> mistaken, so does FOSDEM.
>> 
>> True. 
>> FOSDEM was Dualstack till 2013 and then switched to IPv6-only in 2014. 
> 
> FOSDEM is similar to the RIPE meeting in that they have both a dual stack
> SSID and a NAT64 SSID.
> 
> The difference is that FOSDEM promotes the NAT64 SSID as the main one and
> the dual stack SSID as the fallback.
> 
> 
>