Views on Internet Technology Adoption

For this week’s article, I wanted to let Eliot Lear talk about technology adoption.

Jari Arkko, IETF Chair

Views On Internet Technology Adoption

The IETF’s relevance in the marketplace was the subject of a workshop held by the IAB in December in Cambridge UK on Internet Technology Adoption and Transition (ITAT). This wide ranging workshop took stock of how things have gone since the publication of RFC 5218, and considered some modern day protocol successes, challenges, as well as new approaches to understanding how we can evaluate likelihood of success of a protocol. The workshop featured both “greatest hits” to re-examine, such as DNSSEC deployment, as well as new found successes such as Bitcoin. We delved into incentive modeling, or why various stakeholders would implement and deploy our work, evolutionary economics and how layers might come and go, as well as systems biology, where our classic hourglass is analagous to their bow tie.

So what did we learn?

  • Success in the case of DNSSEC in particular has come to the .SE ccTLD domain thanks to lots of hard work, free software to enable easy key management, close coordination between the registry, registrars, and users, and a model where those who sign their domains actually get a small bonus.
  • The IAB should take care to consider the notions found in RFC 5218 of failure, success, and wild success, when considering birds of a feather sessions. And of course as we may do just that, it makes sense for those who are going to propose BoFs to consider that document as well.
  • We also considered the notion of bundling of technologies and how bundling of protocols might help or hurt their individual success. This is relevant with protocols such as DANE that require DNSSEC, as well as in the httpbis working group, many of the participants of which would prefer to implement HTTP2 only on top of TLS. What are the costs and externalities that might lead to success in these cases? While an early model was presented to the workshop, more work is needed. What might come out of it is a better understanding of the notion of Mandatory to Implement, something that is of course quite relevant to the folks in RTCWEB.

What we take away from our discussions is the possibility of one or more research groups and the potential for collaboration between researchers in this field and our engineers and leadership, in the hopes of gaining better understanding of how to have our work succeed.

For more information about the workshop or the papers presented, please have a look at, and stay tuned for an upcoming workshop report with a few additional comments at the IAB plenary.

Eliot Lear, member of the IAB