yes its true, all of it - the internet doesn't really exist, so it must be.

Monday, November 25, 2013

The Role of Religion in Revolutionary Network Architectures

I'm looking forward to the forthcoming IAB workshop on Internet Technology Adoption and Transition for lots of reasons (catch up with many people, navel gaze about important topic, maybe even find out what works and what doesn't)!)

However, I'm looking through all the papers that will be presented and am worried that we are missing a very big factor in technology's success or failure, and that is faith.

The papers to be presented break into 4 rough groups
1. Economics -
     e.g. how do markets and commons interact.?...
     how do various tricks bundling, regulation play out?

2. Process -
     what do patents do to things?
     how does the ietf capitalize (or not) on research?

3. Ecology
     does the hourglass emerge always?
     how is diversity helpful (or  hindering)?

4. Technology
   what makes a protocol tick well?
   what pieces of the current experimental world (ICN, bitcoin) will make it to      prime time?

All good, but all roads that have been trod several times before in the communities - in general, economics has not had a great track record in prediction, and bio-inspired stuff is fun, but again doesn't match the details.
It is always worth studying the process and use cases are well worth documenting of course, but what bugs me is that there are so many potential failures we havn't looked at, and what do they have in common?

For me, it is the lack of a fervour, and what is more, persistence in the face of strong adversarial reaction - when we started deploying IP (I am not talking about the mega-ARPA projects, I am talking about the "going into schools at weekends" and "laying out community nets in small towns" movements in the mid 1980s), we were conducting a missionary movement - I recall also giving courses on TCP/IP to hoards of commercial folks despite their seniors in their companies still buying all sorts of CCITT (now ITU) and ISO (now nowhere to be seen) products being pushed by big companies and government agencies (GOSIP - Government OSI Procurement, was the official religion).

We persisted on this for 20 years - we still do....but we are now the official religion.

so now what happens if you want to introduce new tech, you need to make it the underdog - IPv6, no good at all - DNSSec secure BGP? hopeless basket cases. You need something that
a) is really barking mad, but might just work
b) has the feel of overthrowing an older dogma
c) inspires faith, even when the evidence is thin....

but you also need to think long term - 20 years is too short - its generational.
And most of us in the game are from the previous generation, and we need to get out the way.....but of what?

1 comment:

Roch Guerin said...


While I totally endorse your arguments regarding the importance of a), b), and c) in the rise of the Internet, I would argue that the Internet emergence is actually the product of what has often been called a paradigm shift (god, I really hate that term, but it does capture what I intend to mean) in (network) design principles brought about by a technology disruption.

The general concept is not mine, but eloquently articulated in a very nice little book by Brian Arthur from Stanford and the Santa Fe Institute entitled “The Nature of Technology: What it is and how it evolves”. However, I think it readily applies to the Internet and its eventual emergence as the new dominant network technology.

Specifically, the technology disruption I am referring to is the semi-conductor revolution and the corresponding rise of computing. I will argue that the Internet and more broadly packet networks correspond to a paradigm shift in how networking/communication was realized. Until then and even if we did incorporate new capabilities afforded by semi-conductors, it was done within the confine of existing design principles. If you look at a 5ESS switch, it is basically similar in principles to the design behind human phone operators connecting wires through a patch panel to setup a voice call. We have replaced the humans and the physical wires by transistors, but the core principles are unchanged.
The Internet upended all this by essentially putting computing in the network and in the process introducing totally different designs that were better capable of taking advantages of the new capabilities made available by ever faster processors.

The main reason for this lengthy preamble in responding to your post is that many of the (new) technology adoption challenges we are facing today are of a very different nature than the type of transition that was behind the emergence of the Internet. Most are of a more incremental nature, but they can nevertheless be tremendously important and play a huge role in sustaining or improving an existing technology.

This means that making them happen is likely to call for different approaches, and I would argue that even if it is less exciting, in many cases their success or failure will boil down to basic economics.

Blog Archive

About Me

My Photo
misery me, there is a floccipaucinihilipilification (*) of chronsynclastic infundibuli in these parts and I must therefore refer you to frank zappa instead, and go home