TheNetStudio.com

Internet Music Collaboration

Think.Code.Repeat.

The Architect's Blog
Feb 21st, 2010

Uncomplicating the issue of Net Neutrality.

Why ANOTHER Net Neutrality article?

There is a tremendous amount of double-speak and obfuscation about Net Neutrality. In large part, I think that is because the stakes are high, and the outcome is far-reaching. We are entering an era where Internet technologies are likely to be the primary platforms for economic growth, influence, and power. To control the rules of Internet traffic will be to control the foundations of an enormous economic powerhouse. For that reason, I believe it's important for netizens to understand Net Neutrality for what it is.

Let me be up front about this: I am FOR Net Neutrality. Without the existing neutral Internet and the open technologies that is has spawned, TheNetStudio.com could never have been created. That means that I have a vested interest in keeping the Internet neutral and open, and my analogies will skew toward the positive nature of Net Neutrality.

Having said that, I still welcome both corrections and competing views.

A Gedankenexperiment...

Let's start out with a gedankenexperiment...

Let's say I have a big old pipe. I'll call it a "backbone". It can continuously send and receive up to 1 million small balls per second. I'll call that "traffic". Inside each one of those balls I can put a little piece of information, and I can send that information in both directions. I'll call those balls "packets", and myself I will call a "provider". For the sake of simplicity, let's say that although there has been a lot of experimentation, I have found that it is best to keep all the balls the same size, shape and weight.

Soon I discover that I want to send larger pieces of information than can fit into a single ball. In order to send those larger and larger types of information I figure out that I can simply break them into small pieces, put those pieces inside several balls, and send them as a collection. At the other end they simply unpack the balls and re-assemble the information. For all intents and purposes all the balls are exactly the same, except for the information that is inside each one of them.

Now let's say I connect my backbone to some customers. I sell them connector pipes in increments of 1,000 balls per second. So my customers can buy a pipe that can send/receive 1000, 2000, 5000, or even 10,000 balls per second. I'll call that "bandwidth" and the bigger the connector pipe they want, the more I charge them. Since I am a tech company I need an acronym for this, so I call it BPS (Balls Per Second). And we all agree that this is fair.

Pretty soon everyone figures out that it is much more efficient to color code the balls to describe the information that is inside of them before they put them into the pipe. It makes the whole process easier and faster. They don't actually HAVE to do that, but everyone agrees that it just makes everything better. My customers all agree on color standards: Yellow for email, Blue for video, Red for music, etc... Everything else stays the exact same, and the coloring process does not change the balls in any measurable way. I'll call that process "ports and protocols".

Life is good. The whole system works great. I'm making money, and my customers love this great new way to send and receive information.

Gedankenexperiment Part II: Monitoring

I like to watch my backbone, because I'm curious about how my customers are using it - Plus, I think the colors are kinda pretty. I start to notice that when I first sold those 2000BPS pipes they were fairly underutilized - mostly with just yellow balls being sent back and forth. But now, I'm noticing that along with the yellow balls being sent and received, there are now also a TON of blue and red ones. And it looks like it is growing.

I realize that even though I sold all those customers 2000BPS pipes, I didn't anticipate that they would fully utilize them. Worse yet, I have a bunch more customers coming online, and I realize that I may not have a big enough backbone to support all the pipes I have sold, if everyone keeps sending all those red balls. Rather than turn away customers because my backbone might not big enough, I just keep signing them up. I realize that I could build out my backbone to 2 or 3 Million BPS, but that will cost me a lot of money.

Maybe it would just be better to charge more for the red balls. I'll call this "QoS". I will tell my customers that if they want to send and receive the red balls, I need to charge them more to make sure their red balls get a higher priority to ensure they get to to where they are going at the speed that they are already paying for.

Now I have to figure out how I am going to explain to everyone that this would be good for them...

I actually know that as long as my backbone is big enough (I'll call that "over-provisioned"), QoS won't even matter. I know that QoS is only applicable when my pipe is so congested that I have to decide which balls I will send first, and which I will delay or drop. But I don't think I should tell anyone that. So I just tell them that I cannot sustain the pipe that I already sold them, and it is their fault because... Well... Because they are using it too much. I argue that, if I don't start charging extra, I will have to build out my backbone to support all my customers (who are already paying me for the bandwidth I have sold them), and that this would force me to raise prices.

I also argue that I really need QoS to support new kinds of critical time-sensitive applications. But I don't mention the fact that time-critical communications are already being accomplished using separate dedicated pipes (I'll call those "dedicated circuits") that I also lease. I cite the need for QoS to support real-time applications such as tele-medicine, even though the whole system (backbone, connector pipes and the devices attached to them) use technologies that are still too immature, insecure, and unreliable to trust anyone's life to them.

Why Support QoS?

So then: In our thought experiment, why would anyone support this QoS thing?

Well... in our thought experiment, our provider might think this would be a great way to extract more money from wealthy companies that would gladly pay more to provide a "reliable" service to their customers. And conversely allow them to effectively lock out small entrepreneurs, by making it too expensive for them to compete based on reliability.

For our theoretical "provider", this could be a very lucrative idea since it essentially rewards them for overselling, under-building, and letting the backbone get congested. Unfortunately it would also destroy the level playing field that currently exists, by favoring companies that can afford to pay more to ensure their traffic is faster and/or more reliable than their small competitors. This would likely destroy the small competitor, even if they have a superior product, but shallow pockets.

But more interesting... For our thought experiment - Let's imagine that there is a group of people that have figured out that high bandwidth content on this backbone thingy is disruptive to their existing business models. Let's imagine that they have realized that - By encouraging, supporting, and championing our theoretical "provider" to make it prohibitively expensive for entrepreneurs to create new and innovative network applications, they can effectively slow down the rate of innovation/competition. Maybe even prevent some small bootstrap entrepreneurs altogether. This might give them the time they need to contain their markets. Then they could attempt to ensure that only big businesses and deep pockets can ever use the power of the "backbone" for reliable, high speed content delivery. I'm not saying anyone has thought of this, just that it could be an interesting strategy that a non-neutral Internet would encourage.

So, without the protections of Net Neutrality - Where would that leave the bootstrap entrepreneur with a cool idea and just a few thousand bucks?

It could leave them priced right OUT OF THE COMPETITION!

Unable to compete just on the merits of their product.

But Don't Take My word for it.

Final Thoughts

Today the Internet is largely neutral. It was designed that way on purpose, and Net Neutrality is without question the primary reason that the Internet has risen to such prominence in such a short time.

If you were involved with computers and networks before the Internet, you know what it used to be like: Proprietary, unwieldy, complex, vertical, prohibitively expensive (in the order of millions of dollars), and with very little capability.

Pre-Internet computer and network vendor lock-in was pervasive and downright oppressive. Only the wealthiest companies could afford the luxury of computers and networks. Conversely, the Open Internet has enabled exponentially more capability for exponentially more people at exponentially cheaper prices.

Without Net Neutrality, it would be far too easy to return to that same environment.

A Net Neutrality law will simply codify the Open Internet that exists today, and ensure that it remains so.

Ask yourself this question: Who would be against that? And why?