Response to Infrastructure Questions
Response to questions about my infrastructure comments.
22-Jun-2006Version 2: 2023-05-28 17:09:26

I’m responding to Gerry Faulhaber’s questions about my infrastructure comments. While I presume calling me “advanced” is like calling me “special” his skepticism is typical and it’s worth responding to his questions including where’s the movement that I’m supposedly leading.

I’ve already done the “movement” thing. I created infrastructure when I made home networking happen while at Microsoft. That time I was able to take advantage of the fact that the home wires, including phone wires, were owned by the user. I didn’t write code (other than for testing out my approach) – just rewrote memos until I got the networking and OS groups to buy into the idea but even then it didn’t catch on until Microsoft had to figure out what to do with the ADSL hose that was going to come into the homes. Not only did I deal with the code in Windows, I needed to make sure that we had a “no new wires” option for “just works” networking. I used the existing (owned) phone wires as the physical transport – wireless wasn’t yet an option.

Today’s situation is actually quite similar but on a larger scale. While the US Congress continues to add epicycles to the Regulatorium there is growing awareness of a different way of thinking about tele/communications and that the current model is not sustainable. I go into more (both too much and too little) on this in

The big challenge is that this is not an incremental shift from telecom but something entirely different – what David Reed has called a phase change. I’d say a paradigm shift if that weren’t such a cliché.

In the short term “Network Neutrality” serves the same role as “shared web access” did in trying to explain why you would want a network in your home. The idea that all your devices, be they printers or light bulbs, should be connected hasn’t quite caught on yet.

So, for now it’s no surprise that people translate “Internet” (dropping the The) into “Broadband” in the same way that they so MP3®. Instead of “digitized music”. One of the important lessons you learn in business is that you don’t give people what they ask for – you have to translate it into what they’ll buy.

Remember that the carriers’ so-called “wireless Internet” was really WAP or the “billable Internet”.

This was because the carriers and the Regulatorium don’t have the concept of infrastructure. We accept their pronouncements because we tend to confuse mastery of the Regulatorium’s arcania with wisdom. The failure of WAP didn’t mean we didn’t want mobile connectivity – just the opposite.

If people don’t understand the concept of connectivity as infrastructure they will ask for muni-WiFi so they can have some Web along with their latté. You see an inverse example of this in today’s Boston Globe which has an article about the state trying to force the Turnpike Authority to provide cell phone service in the new tunnels.

In both cases we have a failure to recognize basic infrastructure and thus we have to pay for a special transport to support a particular service.

It’s difficult to have an incremental path from the service model to the infrastructure model – it’s like those optical illusions, you may see one face or the other but not a mixture. Perhaps one can subsidize a transport but it’s not the same dynamic.

We (Gerry and I) seem to agree that community owned Cable TV is a bad model – it’s just the carrier model in the miniature. We have the same problem with for subscription-based or sponsor-paid Wi-Fi. Neither approach is surprising given that people are used to broadband as an extension of television and they confuse the web with the Internet.

What we don't have yet are applications that use connectivity as infrastructure. Yet we can create our own opportunities even now – we can take traffic data from web sites and mix it with map information using the current location. These kind of mash-ups are becoming common on PCs and would work just as well in ones car – provided one had a platform rather than just closed devices.

The same kind of infrastructure arguments also applies to platforms. Unfortunately the carriers locked down the platforms too. Why else do we pay more for single purpose navigation devices than for what the equivalent platforms plus software cost?

If you frame the problem in terms of the Regulatorium every applications is special and difficult to implement and you can’t justify building a transport for each new service.

Once you understand the concept of infrastructure you find that it would cost far less to treat the current transports as common infrastructure.

I often find it’s the technical people who have the most objections. If they learned about networking from the current Internet they might not realize that it was built as a prototype and many of the accidental properties – such as using the IP address for naming and routing – were just temporary kludges. One of the biggest is the whole idea of the backbone as a single cloud.

I would write about the technical issues but don’t want to get lost in the details. The main point of the end-to-end argument is that is that we can composite solutions. This includes deriving the Internet from cooperation between the various local systems rather than having a single “Internet Inc” that enforces the rules.

For now I’m not concerned with whether there is a “Movement” as much as whether I can help people understand the concept of infrastructure. I am joining those who are advocating NN as a tactical step but I think that the stage is set for recognizing the value of infrastructure.

The press picks up on the controversies – if you don’t have a good funding model for connectivity it becomes a political issue. If you take a far more low key approach and experiment with deploying infrastructure there’s no story. If the deployment doesn’t go as planned you debug the process – it only fails if you have a rigid test and the stakes are high.

If you look you’ll find a lot of Wi-Fi deployments in various shapes and forms. The FCC notes an anomaly in that so called “rural broadband” deployments are ahead of schedule because of Wi-Fi. It’s treated as a footnote rather than a harbinger of things to come.

One reason for keeping the deployments low key is that the Regulatorium is actively trying to ban them. They are right to view any infrastructure as being “unfair” because it threatens their rationalizations. Today we have the complication of the FCC having an additional role as guardians of our morality and our security.

It’s useful to advocate having full access to the rights of way (AKA, the poles) but there is a lot of regulatory baggage – they are often owned by other private entities such as power companies.

Yet the market forces are inexorable.

The dial up Internet trumped Interactive TV and as we see with HDTV/TV it’s already happening again – more value in downloading video than in building a special multizillion dollar HDTV infrastructure (or is DTV – we can’t expect Congress to know or care).

What would happen when people learned that they could do far better than E911 and USF if they took advantage of connectivity but as long as we must petition Congress we are up against an all-pork diet. Fear and pigging out do slow thing down.

If the concept NN does nothing else but reduces the effort to stifle local infrastructure deployment it will make a big difference in accelerated local deployment and mooting the pork deliver system.

Imagine what would happen once we can talk about “infrastructure” and “connectivity” in their own right without having to translate it all into the language of the Regulatorium.