Who’s responsible for the internet’s capacity?

AT&T thinks that Netflix is trying to pass off the cost of network connections to end customers. There have been a few different displays of the architecture of the internet. Netflix operations at a different layer than AT&T does – Netflix is an application, so it runs on a layer above the network layer, which AT&T operates. Netflix doesn’t really care who actually sends their bits to the end user – they just care that they get there in a fashion that enables high def video. To this end, they purchase bandwidth from a company, mostly Cogent, and I pay Comcast (others pay AT&T) for me to receive those bits from the bandwidth provider of Netflix. I pay Netflix for access to their content.

Based on this payment model, if there’s not enough bandwidth for Netflix and I’m paying AT&T or another ISP for accessing Netflix, it’s up to them to make sure I have that connection. Content is King, so for me, it’s most important that I can access what I want when I want. That’s why I have an ISP so they can let me see what I want.

I think that the best analogy for content trumping gate keepers are the examples of higher premiums from popular channels. In some cases Timewarner cable pushed for lower rates to show a specific channel to their subscribers. In this example Forbes points out that ESPN costs $5.54 per viewer, they wanted to lower that price and pulled the channel out of rotation. This made a lot of people unhappy and in some cases people left Time warner over the issue.

Essentially, this is the same thing that is happening with Netflix. The ISPs don’t want to pay to upgrade their infrastructure to ensure that the consumers of media online (many of these people paying for higher download speeds and higher data caps). Netflix is providing a service that these people are willing to pay for but cannot control how the ISPs interact with their intermediaries so is in a tough spot. It’s a target because of it’s popularity and has no control of how anything gets to a specific user. That’s why it’s looking at the peer2peer model (which is how Skype keeps their rates low) so it won’t need to go through Cogent and will likely burden other parts of the network very differently.

I believe that if an ISP cannot meet it’s advertised speeds 90% of the time, then they need to update their infrastructure to meet my needs. If they throttle a popular service I’m watching and thus make it unwatchable, they need to upgrade their infrastructure. Most ISPs have an extremely high profit margin, which means that it’s coming out of their infrastructure investment and are not actually adding value.

There are many companies that are responsible for the capacity of the infrastructure and all of them can negatively impact our ability to use the internet. However, from an end user perspective, my ISP is on the hook first, then everyone else.

What’s the difference between Ma Bell and Comcast?

If you were born in the 80’s or before you know that Ma Bell was the only phone company in town. Born any later than that you were born into a world without a single monopoly for telecommunication. That’s right, we’ve had a point in our collective history where there was only a single phone company. There are rules in place that prevent something similar from happening with Comcast, but we’ve been there before. However, I believe there are critical differences. AT&T knew they were a monopoly and they were a state sanctioned monopoly. They did everything in their power to keep prices down to prevent being broken up. AT&T actually had a broader monopoly than what Comcast could ever hope to have. They made the phones that worked on the line, they made all the telecom technology that made it work, and they designed the services that made it work. This is something called a natural monopoly, which I’ve written about before. A former founder of Comcast has declared Comcast a natural monopoly.

The biggest difference between Comcast and AT&T, back in the day, was that they did everything they could to keep the government happy. Was it perfect, no clearly not, there were shady business practices, but we as a society benefited greatly from Bell Labs. To this stay is still one of the greatest research facilities that ever existed. If it wasn’t for Bell Labs our current way of life would be very different. I highly suggest checking out the book on it.

Comcast claims to be pushing innovation with their X1 Xfinity platform, but that’s not really true, it’s simply a new operating system pushing content. Voice activation isn’t innovation and if that’s your main selling point then you’re in serious trouble. As I mentioned yesterday, the Netflix deal is a major concern, the Verge is saying the Internet is fucked and that we need to be contacting the FCC daily to un-fuck it.

I’m not entirely sure that the FCC can fix it. Congress has greatly hamstrung the FCC in dealing with internet companies, furthermore, their solution of calling the internet a Utility won’t work. If you aren’t aware we’ve had big pushes to deregulate the utility industry which unfortunately hasn’t really made rates better in many cases or in the long run. I think that it’s fair to say that in the telecom industry this is true as well. The impact of the AT&T break up has been this long term collection of conglomerates that continually increase price as well as “Fees” which similar to baggage fees are hidden from the “price” of the service. So, treating the internet like a utility isn’t going to work. What we need to do is treat it like a road.

Everyone that uses a car on the road is taxed based on use (Gasoline taxes) everyone pays for a portion of the maintenance based on other local taxes too. No, these aren’t perfect and are going to be under pressure based on hybrid and electric cars – and new models are being proposed. Of course one way to do this is through toll roads (which really never work) or through some sort of black box in the car to measure mileage (which no one wants).

Essentially it’s a pay for bandwidth consumed, so if you’re a high consumer of bandwidth you’d pay more, but the rates need to be realistic and the goal would be to cover expenses and continually improve service while making it cheaper. Which brings me back to AT&T – the president of Bell Labs had one mantra anything could be tested but only if it could lead to a “Better, cheaper or both” network. A public internet similar to a road that was paid to continually get cheaper, better, more secure, and faster is the only way to truly un-fuck the internet. It’s not likely to happen because it’s not a capitalist response. However, the internet these days is similar to public transpiration – it’s goal isn’t to make money, it’s goal is to enable economic activity. If think of it that way, then we can see the long term benefit of the whole economy rather than singular actors.

Is Net Neutrality regulation commie nonsense?

Network Economy

Regulation’s a bad thing, right? Personally, I think there are instances where regulation is an amazingly good thing that drives innovation. We also need to be cautious about who is saying regulation is good or bad. Back in the 90’s we’d hear that regulating in anyway to prevent acid rain would cripple business and kill our economy. This clearly didn’t happen, we have acid free rain for the most part, we have more productive manufacturing than ever. We also hear that regulating CEO pay by median rather than average is significantly more complicated to the point that a place stacked full of MBA’s can’t figure it out. Then there are regulations that pick winners like Solyndra and turns out to be a disaster. These cause higher taxes and are actual drains on the economy (personally I’m on the fence about experimenting with new technologies and having the government support them, but that’s me).

What about the FCC “regulating” net neutrality? I think that it’s important to look at how this all started. First, I’ll start with a bit of a history with the telecoms, then move to how the internet was developed, and move to comparisons between other monopolies.

AT&T has been described as a natural monopoly. This was partially helped by the US government because the government wanted coast to coast telephony and selected AT&T as the standard for that activity. This gave AT&T incredible market strength, but was also extremely fragile as it was continually under threat of being broken up for being a monopoly (which is was). To do everything they could to avoid this, the geniuses at Bell Labs continually designed ways to keep their costs down, improve quality, and make very thing better. They also had some government deals that helped them a lot (military contracts for telecom stuff, like the first satellite). The value of AT&T’s network grew every time a person joined the network.

The fact that one person joined Network A over Network B could further impact the growth of that network. Let’s say Person A is friends with 5 people and is already on Network A, it’s likely, if they are really good friends and A is known for making good decisions, that those five people will join A on Network A. The value increases by more than simply 5, because all five of those people can talk to each other as well as every other person they know on Network A. Now if Person A has more friends, but not as good of friends and they actually are better friends with Person A’s friends they will also likely join Network A. This sort of cascade effect will continue to happen. This is also known as Metcalfe’s law.

When AT&T was force to break up, all of that interoperability remained. Instead of one big monopoly there were regional ones instead. As we’ve seen over time, these same regional operators have slowly re-joined back into 2 Bells versus the non-Bells. AT&T being split is a type of regulation for sure, but it did spur some interesting competition for a time.

How the Internet was designed:

The internet was originally designed to operate in many different application layers. Essentially the bottom of the stack was Internet Protocol which was agnostic to the type of information being sent across it. At the time, the most efficient method was over Ethernet so there was not any requirement to be concerned over the application medium. Over time there would be some concern, but that was really addressed by the protocol.

What would happen is that the applications that required information to be sent on either end would translate the information to be used by the layer below it to send out, such as a web browser to the OS, to the network driver to IP, across the internet to the network driver to the OS to the web server application. Across this entire process the actual data being sent was unknown to any of the nodes in between the application layers. (If you’re interested in this check out Internet Architecture and Innovation).

Of course the companies providing the bandwidth for that did not want to find itself in a similar role as they had after the break up of AT&T where they were forced to become “dumb pipes” for whatever people wanted to send across their network. To prevent this they created capabilities like deep package inspection and other tools to identify what content was being shipped across their lines. This also was the beginning of violating “True” net neutrality.

Why were they dumb pipes? Because they were defined as a common carrier to increase competition across the land line providers and ISPs the telephone companies had no choice. This lead to the explosion of ISPs like AOL, Century Link, and so on. What has happened since? The broadband lines have been ruled that they are not “Common Carriers“. Meaning that the data across the line can be treated however the companies that own the lines want.

Why is this bad in a network economy?

In a network economy, being able to fully control anything and everything can be very bad for the consumer if there is no other option. Now, you could argue that there are options, but in most cases because of other monopoly rules there are few options for allowing a new ISP.

A perfect example where a network monopoly isn’t a big deal is in Smart Phones. The iOS App Store is a natural monopoly in a network. The more people using the iPhone the more valuable it became and more app developers developed apps. It never became a problem that Apple regulates the entire experience BECAUSE there were other networks you could shift to, such as Blackberry, webOS, Windows (whatever mobile version you want to include), and, of course, Android. All of these ecosystems offer very different options for devs. Additionally, within Android there are competing App stores which further benefits the consumer. If there were no other competitors to iOS and it’s App Store the constraints that Apple puts on their product would likely be viewed as very anti-competitive and a type of “foreclosure.”

Market foreclosure is using one monopoly to enable another monopoly. Now, regardless of if you think that this should have happened or not, it did. Microsoft was hit for using it’s Window’s OS to foreclose on the internet browser market and was looking to do the same with their music player. What resulted was that MS was required to offer other browsers when a new Windows OS was launched and helped to reduce the market share of IE.

How does this apply here? Comcast is already trying to do the same with Netflix in the streaming video business. Comcast owns the content (Universal, NBC, etc), the connection (Comcast Cable ISP), the rules (data caps), and if they want to charge to access their network or not. Eliminating the rules of net neutrality tilt the table in the direction of Comcast to a degree that Netflix may never recover. If Netflix, at one point 2/3 of all internet traffic, had to pay for every bit they streamed to allow for an enjoyable streaming experience they would be bankrupt in very short order.

I get that Comcast’s of the world don’t want to be dumb pipes, they own the content and that’s king. However, not every ISP owns content (Verizon/AT&T) so they aren’t at such an advantage to companies like Netflix. However that’s where AT&T’s data plan comes in. Which would essentially level the table compared to Comcast. We, as end users, wouldn’t see any benefit out of this. It’s not that our subscription fees would lower or we’ll magically get faster internet. This is simply rent seeking behavior and bad for the economy overall. Only true new competition can lead to that. Changing these rules have zero impact on that competition.

What it does do though is negatively impact the creation of new businesses that want to stream video or provide a novel product that requires high bandwidth and equal rights to streaming. Removing the protections on net neutrality dramatically increases the cost of streaming that otherwise could go into building that startup’s infrastructure. Think of the problems at Twitch.TV with their growth. My subscription fees pay for the growth of the network that I subscribe to regardless if it’s something like Twitch or Comcast. Anything else will go to shareholders and CEOs.

Could we develop other options like a Mesh network? It’s possible, but for that to work the option would have to be a public/private venture. Most citizens aren’t going to help create that and likely don’t have the technology savvy to do so. To further complicate this issue many ISPs are actually pushing to make it illegal for cities to create their own ISP.

In many cases regulation is bad for business. However, in cases like net neutrality it’s returning the net to it’s roots and enabling much stronger competition based on the merits of the company providing the service, not the arbitrary whim of network owner.

Sponsored data and YOU!

This could be your lucky day, your cellular provider is going to start offering packages where certain content doesn’t cost you anything in your data cap. This is awesome. You can start streaming more and more video/music/whatever it is that you stream from your favorite services. However, not all of your favorite services will be free of data charge! So make sure that you tell your favorite service that YOU want THEM to sign up and make their content data cap free to you! All those service providers have to do is pay your cellular provider money to stop the data caps! No, seriously, AT&T wants to do this.

Is this a problem? I think it depends on who you are. For a consumer in some cases this is pretty awesome. Let’s say you love to watch video games being streamed on Twitch.tv by your buddies over at KBMOD and Twitch decides to pay money to prevent your data from being charged against your data cap. But you’re also a huge fan of MLG and MLG just decided to start their own Twitch competitor but they can’t afford to pay those same fees. Well, guess you’ll be only watching MLG from your PC or on wifi. Too bad your favorite shows are on while you’re not able to use Wifi though! O well, Twitch is there for you though!

This is a niche market obviously. Not everyone cares about watching someone play streaming video games or even streaming video games to your phone so you can keep playing a game you were playing from home. A lot of people care about TV and movies though. We can look at this as something that’s really analogous to what Comcast was trying to do to Netflix close to two years ago. In April of 2012 Comcast announced that its Xfinity streaming service would not be charged against your Comcast data caps while Netflix streaming service would be. Netflix’s CEO argued that this violated Net Neutrality because it provided preferential treatment to one source of data over another.

What is Net Neutrality? Well, there are two different arguments, which I discuss in a blog here, where one is saying everything must be treated equally, while the other one argues that there are nuances and we can treat data differently because we need to “Groom” our networks. Internet and network purists believe that you shouldn’t even be able to determine what the data is or what the source of that data is if you’re a point along the network, just where it most recently was and where it needs to go next. The only application that can read the data in the package is the application that requested it.

AT&T’s plan, similar to Comcast’s, is in violation of Net Neutrality and the FCC will step in to regulate this type of “service” because it’s, in the end, bad for the consumer. Unfortunately, there are limitations to what the FCC can do and even potentially what AT&T can do.

There has been much more of a push for encryption and it’s likely that these pushes may actually enable more of a return to the true meaning of Net Neutrality. If all of our data is fully encrypted, deep packet inspection tools (which tell if the data you’re getting is video, music, or whatever), won’t work very well as that information will be encrypted. Furthermore, if your application’s data is all encrypted and AT&T won’t be able to tell if your data is your data then there’s no value in paying for “privileged” data status from AT&T.

It’s one of the reasons why I’ll likely support company’s like Wickr, an encrypted Snapchat competitor, which told the FBI to screw itself when they were asked to put a backdoor into their encryption. It’s important that we work to protect our data and support companies that do so in terms of Net Neutrality and encryption.

What companies do you support that encrypt and fight for net neutrality?

Data caps can we get rid of them?

Changes are afoot in the mobile industry. We’ve been seeing a sea change with how T-mobile looks at their customers. They’ve been changing the contracts and potentially even eliminating them. They’ve separated the phone payment plan from the monthly access fees which is an imortant first step in completely separating the purchase of a phone from the carrier that provides the services. This will eventually reduce the amount of junk files installed because  the carriers can’t force the providers to install them. I think that the reason why the Nexus 7 isn’t on Verizon Wireless is that they are refusing to install the bloatware that the Verizon is trying to force upon Google. I think this is a good thing. Google has also filed a suit against VZW to force them to allow the Nexus 7 on their network. This is going to break the carriers power over the user. Carriers don’t want to be dump pipes for the internet. The wired and cable companies are actively fighting against this by puttng data caps on downloading content. At the same time they’ve experimented with this, the wireless providers have been much harsher. However, these caps weren’t a problem in the past. According to BGR a website that focuses on mobile issues has noted that the amount of data that wireless users has begun to regularly break the data caps. T-mobile has also offered all of their plans without datacaps. This will likely push many data users on verizon and AT&T to switch over to these networks. This will likely put more pressure on the carriers than they have experienced in over a decade. I look forward to this increase in competition especially since I refuse to go over my data caps and greatly limit the amount of data that I use over the wireless networks. I believe we’ll see some interesting changes in the future.

I think that these changes will also preclude the merger of T-mobile and sprint as the US market needs additional competition not less. I think that there will be changes in all of the companies marketing and product offerings, but likely T-mobile will be the largest winner with Sprint in second.

NSA Bulk Metadata ruled likely unconstitutional

Today was a pretty big day for privacy fans. The NSA’s bulk collection of metadata has been ruled likely unconstitutional. Why is this a big deal? It’s “Just” metadata. Well, as the CBS 60 Minutes report showed the NSA is able to convert that information into a network. Networks show everyone that you talked with and despite assurances otherwise that phone numbers weren’t used, it’s fairly easy to unmask a person in a network based on the network characteristics. I wrote a blog post about this a while back that talked about a paper showing the power of metadata. I think it’s important to reiterate here what that is.

In the article, titled Using Metadata to find Paul Revere, the author explains by using who talks who it is possible to construct a large network and that it was likely to determine the major players of the US revolutionary war. Just using club membership, it wasn’t even what they talked about, just what groups they were members of and how they were all associated. Based on the metadata Paul Revere is a pretty central figure and knows a lot of the other leaders of the revolution.

The NSA would take this view and say, “See it could have caught those terrorists back before the revolution!” However, the judge in this case says that the government did not do a sufficient job showing that this actually worked. It is, in fact, likely that the British had some of this membership information but wasn’t able to put it to good use. In this case, the judge ruled that the collection of Bulk metadata is a violation of the 4th amendment.

What can we expect next? Well, the ACLU has a very similar case that is being heard. If the judge rules differently the Supreme Court may need to weigh in to deal with the problem once and for all. Which depending on how these cases are dealt with could be a good or bad thing.

It is unclear at this point how this will change the conversation in DC, it will likely just lead to more denials from the NSA and White House. They will argue it’s still legal and that they will appeal to the highest court that they can. If they lose this case, it will likely lead to a lot of other questions being asked and possibly calls for impeachment and resignations. I would not be surprised if some of the more extreme on the right call for Obama’s arrest as well.

The other piece that is of interest to me is the question about the companies that have been complicit with sharing of our metadata. Are they going to be in the clear or not? In the case of AT&T there was a law that protected them retroactively. I am interested to know if that will also be ruled unconstitutional as it enabled the government to break the law farther than it could have before.

In general this is something really good, but I believe it opens many more questions than it answers about the long term repercussions of this program. I will continue to blog about this topic!

Book Review: Idea Factory, the history of Bell Labs

Yea, I know I’ve just been doing book reviews.

This book was amazing. I had no idea of all the different things that Bell Labs produced from the mid 1920’s until the 1970’s and later. The book focused on the high point of Bell Labs innovation run. It followed the career of several, at the time, famous and prominent scientists that were employed at Bell Labs. Please such as Mervin Kelley (vastly improve the vacuum tube and was a long running director, VP and President of the Labs), William Shockley (inventor of the transistor) Brattian (inventor of a different kind of transistor), Claude Shannon (inventor of the field of Information Science), John Pierce (inventor of passive and active satellite). These there were many others, however, they each had significant impacts on how our modern society works.

The book does an excellent job in explaining some of the basics of how the research was conducted, what work needed to be done to make it work on an experimental level, the method of transferring the invention into innovation or a full product and the goal of each of these inventions. Mervin Kelley was famous for saying that to implement a change in AT&T’s network the new technology must be “better or cheaper or both.” This prevented a great deal of frivolous technologies from being implemented into the telephone network. Additionally, this was required to ensure that AT&T was always able to work towards reducing rates for subscribers as they were a “natural” monopoly.

This was a time when research was done to ensure that the network would be operational for 30 years without malfunction. This required huge investments in quality control and required that additional costs were built into the network for redundancies and protection. In fact Statistical Process Control was invented at Bell Labs to ensure proper quality.

How did all of this work? Well, there were two factors going on here. First, Bell Labs was able to hire the best and brightest to work on interesting problems. Second, the scientists had a continually evolving project that always needed more innovation. These two combined with a freedom to explore allowed the scientists to delve into basic and applied research. In some cases they did not know how or why something would work, but felt that it would improve the quality of the telephone network.

One of the goals of AT&T was to create a coast to coast network with universal service. This required the company to figure out how to address signal decay due to distances over several miles. To address this the company developed the vacuum tube repeater, which significantly increased the distance a voice call could travel. The manufacturing of a tube was extremely difficult and expensive. Bell Labs felt that there had to be a different way to create a repeater. Over the next 20 years they investigated off and on (with a break for WWII) how to make semiconductors work as a repeater. Bell Labs was capable of making this sort of investment because it had a guaranteed revenue stream and a mandate to continually improve the network. These two together allowed the Labs to do work that they otherwise would not have been able to investigate.

This is a very different model for innovation than we currently have in any organization. Universities come close, but they fall short in the fact that the professors are continually required to apply for more money and seek permission from someone to pursue their work. Bell Labs was much more relaxed about this.

This innovation method is also very different than some of the historic events in the US, such as the Manhattan Project or the Moon Landing. Those were single goals which allowed the focus of a great group of minds.There was never any intention of keeping those minds together for the next big project. Bell Labs had the ability to do this.

There are some organizations that should be able to do something like this. The National Labs are one, but there’s no direct business need so even this doesn’t exactly work. An organization like TNO in the Netherlands, which focuses more on practical matters could increase the amount of basic research they conduct in various different areas. TNO is structured differently than the National Labs in the US, because they are expected to work closely with both industry and universities. This gives each of the groups a strong business focus and could serve as a pipeline from basic research into business activities for the companies that work with TNO. However, at this point TNO does not perform these activities.

I give this book a 4.5/5. It was extremely well written, well organized and dealt with some amazing subject matters.