Net Neutrality Day Of Action
Today, July 12th, is the Net Neutrality Day Of Action.
More than 70,000 websites, online services, and Internet users are participating including Twitter, Amazon, Netflix, Kickstarter, Etsy, Reddit, OK Cupid, Airbnb, Facebook, Google, Spotify, Soundcloud, Mozilla and AVC.
The grassroots power of the Internet is how we won the strong net neutrality rules that are now in place and are threatened by the new leadership at the FCC. The big telcos have their people in power now. But we can keep fighting with our grassroots efforts. They have worked in the past and I hope they will continue to work to keep the Internet an open and level playing field for everyone.
If you want to participate with your website, blog, or social media profile, go here and join this online protest.
Comments (Archived):
Counterpoint: http://blog.erratasec.com/2…
The article isn’t quite right about this: “Another example in Oliver’s show is whether search engines like Google or the (hypothetical) Bing can pay to get faster access to customers. They already do that. The average distance a packet travels on the web is less than 100-miles. That’s because the biggest companies (Google, Facebook, Netflix, etc.) pay to put servers in your city close to you. Smaller companies, such as search engine DuckDuckGo.com, also pay third-party companies like Akamai or Amazon Web Services to get closer to you. The smallest companies, however, get poor performance, being a thousand miles away.”Netflix is a big company that pays AWS.
netflix runs their own CDNs and leases from others also. they already have paid interconnect deals with big ISPs. http://blog.streamingmedia….
A-ha, thanks for extra info.
As I understand it, these type of interconnect deals break net neutrality.
paid interconnect is legal, paid prioritization is not. here is a good overview of the difference: https://www.linkedin.com/pu…whether the difference is substantial in the long run is a subject of debate.
Very good link, thanks.It is interesting how interconnects and content caching/hosting influence the outcomes. I believe, or perhaps just wish, that the edges of the network will get smarter eventually. We will have interconnected routers with huge caches working together at the neighborhood level.The last mile will be smarter.
Monopoly bottlenecks can and do occur everywhere in the informational stack. At the core or edge, at the top or bottom. It depends on the application, technology and context. Basically supply and demand parameters. Therefore there should be no distinction.
Exactly.News at 11:00: “Big companies with money have an advantage over your small shitty company with less resources”.
Well, WhatsApp started really small … 50-ish engineers servicing 900 million users.
Errata conflates many things; in part because net neutrality itself both conflates and treats separately things that shouldn’t be.The concept of net neutrality is a contrivance by those who are not willing to recognize the packet data world’s scaling on the back of the competitive voice markets of the 1980s-90s. So they are willing to throw network theory to the wind or reform it in ways that suit them (ie crypto). This approach in the end is either unsustainable or results in the opposite of what they originally intended because of immutable network laws. The internet is a perfect example, with silos and near monopolies everywhere and standards that take 20 years to get to 20% penetration.Regulatory oversight requires 2 simple things for generative and sustainable, digital, internetworked ecosystems: 1) mandate and foster interconnection as far out to the edge and down the stack as possible; and 2) foster market driven settlements that provide the necessary incentives and disincentives to all actors.This perspective runs counter to both sides positions. The edge access providers have (near) monopolies that stem from regulated licenses to public rights of way (physical and spectrum). Yet they have little control or influence over demand for the product they sell. They have been players pulled along in a game controlled by over the top players. But at the same time, those over the top (core) service and content providers have little direct influence on edge investment. If they would both recognize the weakness of their respective positions and figure out a way to work together the world would be a better place.
how would things be if BMW had built the autobahn in Bavaria?
Your point being that in order to use the autobahn you’d have to own a BMW and not a Porsche? Once again entertainment and pleasure non essential.
in a way yes. Porsche would have built the autobahn in Baden-Wurttemberg. then cartels may have formed through bilateral agreements.
Just as I called the Telecom Act of 1996 a well-intentioned farce that revealed the worst of regulatory policy, so too is net neutrality farcical and a contrivance.
A VPN like Pure VPN is a way to get around net neutrality changes
At what point in America did we get away from the thought of intense competition? We are regulating cell phones and the internet using a 1930s bill that was ill contrived in the first place. This isn’t natural law.Take the shackles off everyone and make the big boys and girls compete.Net Neut or No Net Neut isn’t the point. Both stink. I would prefer #Compete with lots of individual choice. Put power in the hands of a single person to make up their mind how to spend their money and put an individualized package together.My point is broader than just the internet. We don’t compete in a lot of industries today. Competition is good. Not everyone should get a trophy.
I will listen to this strategy (which in concept I agree with) when your statement addresses all the lower income people who will suffer as it sorts itself out.Without that this is textbook abstraction, which while conceptually correct, is unapplicable in my view.
Competition will drive prices down and service up allowing poor people more access than they have today. Look at rural areas of the country. In the town I spend a bit of time in they can’t get internet service to a lot of the county. Kids sit in cars to do their homework outside of establishments that have wireless. It’s not that they can’t afford wireless in their home, but lack of competition means no one brings it to them. The rich will always have more and better. You cannot adjust for that except at the tip of a spear.
I don’t disagree with the current state nor with the validity of your idea for the future.I simply don’t believe that jumping to free markets without safeguarding the transition for the poor is acceptable.What’s your plan for that?
Given the pace of technology, competition will take care of the poor. It works in other marketplaces, it will work with internet and cell. Right now, they aren’t being serviced and Net Neut isn’t going to help them. It’s only going to further stratify and eliminate competition among the oligarchy.
.Of course, they’re being served. They can use the local library. In the ATX, they have hundreds of computers which can access porn. Bit of a scandal. Also, at Austin Community College.JLMwww.themusingsofthebigredca…
Local Library. Exactly. This is a bedrock principle. We don’t give everybody (access to or ownership of) free automobiles either even though they are essential in modern day life.
Telecom (and information) networks are much different from automobiles. With the latter there is a variety of choices and classes of service and alternatives. Everyone’s context is different. But from a network effects perspective, It’s all 1 way, store and forward.Information (communication) networks are essentially 2-way and have enormous network effects where the value grows geometrically and requires everyone to connect to participate in that model. Therefore it should be possible to develop mechanisms to share the wealth more to get to greater network effects.Pushing universal service and making information networks freely accessible is in large part why the US rose to dominance in the 21st century. Don’t you want to keep the US great?
The cry goes out, we just will NOT force a person who is not able to pay to go without needed medical care. We won’t let them starve, either. And we won’t force them to sleep on the streets. We won’t force a wife to stay with an abusive husband or force children to remain with abusive parents.Why? One is, “There but for the grace of God go I.”Another reason is, being poor can happen to nearly anyone, and we don’t want people to live terrified of being poor.Another reason is, we don’t want to see poor, sick, starving people being victims of crime and freezing to death on the streets.Another reason is, in Western Civilization, e.g., read Trump’s recent speech in Poland, we have some view on, call it, human decency, or we want to live in a decent society.Another reason is, part of Western Civilization is to alleviate human suffering. Heck, we don’t even like to see cats, dogs, kittens, and puppies suffering.
hundreds of computers which can access pornA quick check of this lists shows that the initiative is supported by Pornhub, Pornmd, Redtube and Youporn. The first is actually a premier above the fold sponsor (note how the list alphabetizes after the initial supporters). Other than that appears to be a nice number of high bandwidth video and audio apps logoed.https://www.battleforthenet…
dunno…not as certain as you are.so easy for the rich and educated like us to think market strategies fix shit for people hard pressed.usually not true.
I don’t see the connection either.
See my points above about network effects and relationship of value and cost. The bottom 30% are losing out on a relative and absolute basis at the rate of moore’s law. Price per bit could be 90% below where it is and in many instances “free” or highly subsidized. None of the current regulatory approaches will get us there.
.Hard to see how the “poor” are exposed to anything other than better prices through competition. How do you see it otherwise?I remember the continuing decline of cell phone prices from $1.50/minute and Cellular One and Hugh McCaw.The access to the Internet provided by municipal libraries is at an all time high. The price there is FREE.What’s the case for this impacting the poor in a negative way at all?JLMwww.themusingsofthebigredca…
I remember yelling at my ex wife when she was talking to me “$1 a minute $1 a minute”.Also remember working for an attorney in college who was contemplating getting a trunked ‘car phone’ (this was pre-cell service) and he was salivating at the idea of being able to charge clients $150 an hour (whatever it was I have no idea) even if he was paying $5 per minute at the time. It was like a gift from god, being able to charge for time in the car commuting or traveling.https://en.wikipedia.org/wi…
It’s a winner takes all model. Value accrues to the core while costs are more or less uniformly borne at the edge. And that’s regressive when it comes to less dense (rural) or lower volume (poor). That’s why we had a regime of USF for 50 years in return for a government granted monopoly.The income and bandwidth disparity is both real and growing as a result of the enormous digital network effects over the past 30 years. The tide is certainly lifting all, but the gap is growing. And the latter is a reverse network effect that is not sustainable.We need to reintroduce settlements that serve to equilibrate value with cost (it ain’t crypto and ICOs, btw) and provide inter-network and inter-actor price signals that provide incentives and disincentives to clear supply and demand at the margin. See my comments elsewhere in this thread that provide additional context.
.Competition anoints winners. Winners are rewarded for winning.Density — customer density — is always a critical factor in how consumers respond.Austin, TX is always going to have better Internet service than Steamboat Springs, Colorado because of density and buying power of that density.It will also have better service because of the nature of the individual consumer.JLMwww.themusingsofthebigredca…
You are looking only at Pareto Principal and not fully appreciating Moore’s and Metcalfe’s laws.Monopoly tendencies exist everywhere in communication networks; brought about by many factors; density being only one. As communications is important for the reasonable functioning of any economy and society it can generally be viewed as a public good. Where do you want to draw the line to exclude certain people?Moreover, I can prove to you that the costs of connecting those whom you would exclude are much less than the value derived from the greater network effects resulting from their inclusion.Our regulators have not really modeled these new digital ecosystems well and the systems in place to address universal service are terrible. I’m hoping we implement market driven settlements.
all the lower income people who will sufferExamples of that suffering? In what way? Internet access is free in many places (as others have pointed out) and in bandwidth to be able to do anything that I would consider essential. Access to entertainment and those benefits of the internet are not essential. Access to sites that stream where you can learn something and better yourself can be done out of the house (for free).
non-net neutrality actually makes internet access cheaper, because it allows for content providers to subsidize delivery costs, like facebook is doing with free basics and t-mobile is doing with various partnerships with youtube, google music, spotify, mlb, etc.
1934 was all about universal service and divvying up the spoils of these brand new information monopolies that developed in the latter part of 19th and early 20th centuries. They were monopolies (telephone, radio, TV) because they hadn’t figured out a way to share the physical and spectrum rights of way. For the most part, they still haven’t; or at least they haven’t regulated that sharing in; ie mandated interconnect. (aka Title II, but not really)Nearly 70% of all internet access occurs through shared spectrum that is 1/20th that of licensed spectrum in total bandwidth. Just think about that. Also, the only “real” regulation that mandates interconnection is the power controls of the devices which effectively subdivides the spectrum into nano-cellular or edge-densified architectures. From that sprung the 802.11 wifi ecosystem that was never intentioned or expected when the rules were passed (between the 1930s-1990s).Elsewhere see my points on NN and flawed business models so I don’t repeat myself. Following a fairly light touch prescript of mandated interconnection as far out to the edge and bottom of the informational stack AND fostering and monitoring market driven inter-actor settlements we can have competition and universal service. And it would be enormously stimulative to incumbent and new entrant alike; as we saw in the 1980-90s when such rules were in place in a variety of communication sectors.
I don’t follow you: We regulate electric power, telephone, cable, water, sewer, etc. to the home because we want just one such connection per house and not several. So, the owner of that connection has a natural monopoly which we like to regulate.In particular, how do you propose to handle the last mile of copper, coax, or fiber in the ground?Sure, maybe the solution is a lot of wireless, e.g., with a lot separation in frequency and, then, space via high end beam forming.
Internet is not water and sewer. It might be beams. With less regs, innovation happens
> In particular, how do you propose to handle the last mile of copper, coax, or fiber in the ground?Sure, leave the last mile wide open, unregulated, for anything other than copper, coax, or fiber in the ground. So, leave the last mile wide open to wireless, via beam forming, satellite dishes, etc.And for anyone who wants to put in new, say, fiber in the ground, where there already is copper, coax, and/or fiber in the ground, let them do it and own it with relatively little regulation.E.g., if can get more bandwidth from cooper via tricky ADSL, etc. then go for it.
.Much of what is wrong with any “attack” on Net Neutrality is that it is purely speculative. It is not happening at scale right now. The ills are “might happen like this” stuff. Reality is different.What IS happening — in places like Austin By God Texas — is massive and effective competition. I get 1gig service from any one of seven different suppliers at less than what I was paying for 60mbps.I get a bundle of 1gig Internet, phone, and cable for a declining price plus I get some kind of monthly “loyalty” rebate because I’m — face it — SPECIAL. My cable service is so comprehensive I can get live feeds from the NSA (OK, that’s not true, they’re a day old.)In almost every instance, competition is the answer. What is not the answer is to hang more Christmas ornaments on an ancient bit of telecom legislation which has been growing dusty since the advent of party lines.I would also opine that this is a perfect example of why there are too many regulations. We are arguing about regulations passed by a formerly Dem controlled FCC which has now shifted to a Rep controlled FCC.I hate that the issue is political though I freely admit I don’t really get the politics of it all.If I were a competitor in the Internet/tech business, I would love to compete v the cable companies. Love it.JLMwww.themusingsofthebigredca…
The conservative, free-market Republicans have been supporting the socialist monopolies for 20+ years. The social Democrats have been supporting the competitors for 25 years. BTW, the competitors lost 12 years ago; but the real loss began with the TA96 21 years ago.As for competition in the last mile, there really isn’t much and you are in the minority. If there were real competition costs would be 90% lower. If you don’t believe me go back to my digital wireless work in 1996. The same holds true for gig access today; which you (or 99% of others) don’t really need for quite, quite some time.BTW, I am not pro/con NN as you see in my other comments; more ambivalent to it. The business model of networks is fundamentally flawed in a world of rapid supply/technology obsolescence and rapidly growing and bifurcating demand. Economists just don’t know how to regulate or approach this business.
.It is not accurate to suggest that there is such a thing as a “conservative, free-market Republican.” They may self-describe themselves that way, but they really don’t exist.Once a Republican gets to DC, he becomes whatever is the most expedient ideology to ensure re-election and becomes malleable to whatever lobbyist can deliver the requisite funding to get re-elected.The lobbyists exist to help the government pick winners and losers, not to foster merit based competition.This is not unique to Republicans, but with their ill-perceived support of business, they are a little more obvious.I wish it were otherwise.There are not many House or Senate campaigns which will turn on the issue of net neutrality.As to the nature of competition, it is like a virus. If it can grow somewhere, it can grow anywhere.I disagree with the notion of technological obsolescence on two fronts: fiber has legs and the cost of becoming obsolete is baked into the normal cycle of depreciation. Equipment of all types has a finite life.Hell, even the accountants and GAAP and the Tax Code recognize that. It is a cost of doing business.When one errs, it always seems to be better for the consumer when one errs on the side of more rather than less competition.JLMwww.themusingsofthebigredca…
I am working on 60ghz MIMO technology that is 1/10th-1/20th the cost of fiber in the last 1000 feet. It will happen in the next 1-2 years.But ironically, it will likely increase, not decrease the amount of fiber out there.The economists and accountants have no idea what the marginal supply/demand clearing (cost/price) equilibrium is. The business models are broken. Just ask Google who couldn’t make a go of it.We can and should be repurposing copper as in some cases it is cheaper in numerous instances and complementary in most others to both fiber and wired.
.You are conflating the accounting treatment of plant and equipment — which accountants have no problems with — and the financial performance of the enterprise. Two entirely different things.The business types may not have accurate projections — broken models as you note — as to how the market will respond, but the accountants have no problem understanding how the physical plant is depreciated.JLMwww.themusingsofthebigredca…
The intersection is called ROI. If I price my services based on a fiber plant with an expected amortization of 40 years, I am signalling an expected RoR on my capital (technology). But if a new tech comes that does the same for 1/10th the cost where does that leave my pricing; and my stranded investment?Recently I’ve been analyzing a lot of IoT and sensor networks. Everyone’s so excited about sensors and devices with 10 year batteries for a variety of reasons. The only problem is the market is changing so quickly that within 12-24 months an entirely new performance envelope (and business model) might exist from essentially the same device and you have to go out and replace that sensor before you recouped the original investment.That’s what’s been happening in 1 to 4G of digital wireless over the past 20 years, only no one has really noticed because penetration went from less than 8% to over 100% (effectively). And the service providers have been able to maintain artificially high prices even as they sold a service for $1 while only providing 30 cents of actual service (called wifi offload). They were able to do this principally because of regulations (frequency auctions) which inhibited new entrants and not being required to provide interconnection or choice at the edge of their networks.That said, Steve Jobs forced interconnection, or equal access (none of this NN nonsense) on them back in 2007 and the resulting application ecosystem explosion is what pulled 4G supply through. Otherwise we’d all still be mostly on 3G and texting across silos. The service providers were unwitting beneficiaries of an arcane regulation (Part 15) that took what was considered junk spectrum and made it shared. But it’s also what is disrupting them. Have to love the law of unintended consequences.Now, because there is no more subscriber growth and their value proposition has withered in the face of OTT providers, the carriers have to hype 5G when there is no real technology and no real business model. Same for the aforementioned IoT.Bottom-line: biz models have to change to reflect the dramatic technology obsolescence. They have to figure out how to fill the pipes quickly and radiate consumption from high volume low margin users to low volume high margin users as quickly as possible. That’s what will address the digital divide; not govt subsidies.That means knowing your marginal cost ex ante and pricing to that cost. WRT informational networks it implies horizontal scale and implementing settlement systems east-west between actors and the regulator insuring there are sufficient market mechanisms for efficient pricing north-south between layers.It means embracing sharing and it means forming internetworked settlement systems that equilibrate the value captured at the core (or cores) of networks with the costs borne at the edge. The latter is exactly what the NN camp opposes. Does that make them fools or hypocrites?
.To state the obvious — ROI is a term of art in the financial analysis realm. We were talking about accounting, not financial analysis, no?JLMwww.themusingsofthebigredca…
“I disagree with the notion of technological obsolescence on two fronts: fiber has legs and the cost of becoming obsolete is baked into the normal cycle of depreciation.” If your assumptions are wrong and you under-depreciate, there will not only be a write-off in the future, but your pricing will likely have been wrong.
.In the collision between GAAP and tax books, every depreciation schedule in the history of the world has been wrong. It’s a crap shoot.That’s why they have the “impairment” rules to make such adjustments. Businesses, understandably, are often fans of accelerated depreciation. Lots of telco/cable companies got in trouble for expensing things that should have been capitalized.Pricing is a dynamic process. It is not simply a matter of figuring out one’s cost and marking it up 10%. During the life of an asset, if its “cost” — meaning its all in depreciation cost — changes (because it’s now fully depreciated) then the margins change.This is just simple business accounting. Nothing more.JLMwww.themusingsofthebigredca…
Isn’t industry consolidation, where long term growth opps for most of these companies truly lies, antithetical to the notion of competitive markets? That’s the direction the industry is moving to, particularly w/ the growth of OTT distribution on entertainment related content. We’re looking at an inevitable land grab w/ M&A that will lead to monopolies similar to what exists w/ GOOG & FB. Competition in theory makes sense, but it reality the market is and will continue to move towards consolidation that will only inhibit competition, safeguards be damned.
.One of the obvious outgrowths of competition is winners and losers.Consolidation is simply the winners picking over the bones of the losers or buying market share. Either end up with the same result.Google Fiber was announced and then everybody else got into the game to blunt their perceived advantage.The first 1gig fiber in Austin, TX was Grande Communication. Everybody else (Google, ATT, TimeWarner) announced it first, but Grande did it first.Competition picks winners and losers and then the winners do their victory dance.JLMwww.themusingsofthebigredca…
> Economists just don’t know how to regulate or approach this business.Special case of: Economists just don’t know how to regulate or approach business.There is a famous economist K. Arrow. He won a Nobel prize in economics long ago. There is another economist Hurwicz who won a Nobel a few years ago.There is a famous paper in mathematical economics by Arrow, Hurwicz, and Uzawa. The paper is heavily about non-linear optimization from the Kuhn-Tucker conditions. Kuhn and Tucker were long at Princeton.Well, when I was a grad student, I took an unsolved problem in the Kuhn-Tucker conditions and solved it. My work also solved a related problem stated but not solved in the Arrow et al. paper.From that work, I can 100% absolutely, positively guarantee you that my work and the work in the Arrow et al. paper has absolutely nothing at all to do with anything real or even relevant to anything at all practical in any real economy. Nothing. Nichts, nil, zip, zilch, zero. Some of what I proved in my work, that I later published, is surprising, curious, etc. and crucial for what I did with the Kuhn-Tucker conditions but otherwise just useless.To me, academic economics has essentially nothing at all to do with anything real in any real economy and one of the worst aspects of academics.
“Agreed,” says this disconsolate and disaffected former disciple of the Austrian School. You might appreciate this article: http://bit.ly/2uqxcsjThe more I look at network effects and theory the more I believe a lot of answers to everything lie in understanding the math behind networks. It is extremely complex and has to do with ex ante clearing an infinite range of supply that is constantly evolving at the pace of technological change and ever growing and dynamic demand. (So no finite sets).In the world of nearly 3 billion smartphones how many homescreens are exactly identical? (answer: not many, if any, unless factory settings were retained)Outliers and exogenous variables rule the day; hence the need for open systems to be generative and inter-networked. It is the latter math that is even more complex. Darwinism was shaped by an industrial society calling for survival of the fittest. But hasn’t it been always the freaks that survive, adapt and make the new reality? Ecosystems colliding with ecosystems. Today’s economists cannot in their wildest imagination solve for this.Further, want really good voice recognition? Have the platforms work together. Want AI to work, have the different approaches work together. Most everyone thinks in silos and try to capture value at the core. Everyone and every system is an island. Sorry, that’s not natural.Everything needs to be understood at the margin in network math. There are no averages. I don’t think winner takes all math (which is what most of our socio-economic institutions predispose) is the natural mathematical order. Look no further than ICOs.Rather an intricate form of equilibration of value and cost is the natural form of networks (and inter-networks) in the cosmos. I wake up every day thankful that no single one of my 80 networks or systems in my body is dominant (aka the winner). Somehow they mostly all work in harmony.I get the simple part of network math, but the more complex stuff is beyond my pay-grade; maybe you can solve? How do you apply chapters 9-12 from Notes on Hilbert Space to create a Universal Service Settlement System? That would be a start.
Just read the article by John Rapley you referenced. He is nearly as bitter as I am!Apparently that article was part of a book. I didn’t look up Rapley to see what his background was.The article was weak in that he (A) gave broad opinions but (B) gave nearly no references to details. He mentioned Leontief a lot but even there gave no references to full statements of just what Leontief said. So, really, his article failed even common high school term paper writing standards.In places in his article, maybe I can provide some support for his claims from what I already know. Still I would have wanted a lot more detail to take his article seriously.Without the details, I can’t take him very seriously, but that doesn’t mean he is wrong — he may still be correct. His broad description that academic economics goes through periods of quasi-religious orthodoxy may be correct.It may be that big, solid examples of quasi-religious orthodoxy exist. If so, then likely it would be important to detect any such examples, especially important for someone caught in the middle of such potentially dangerous nonsense. So, economics may be caught in some quasi-religious orthodoxy; Rapley didn’t give a very solid argument that that was true; but still he may be correct; and given the importance of such a danger we should “look out”.The axioms for a Hilbert space — a complete inner product space — are short and simple. Some of the consequences are amazing. John von Neumann gave the axioms and named the definition after David Hilbert — okay.But to use von Neumann’s definition, need something that satisfies the definition (axioms), and that’s the first challenge. That is, the Hilbert space definition and consequences are just some phantom in the sky without a specific example.Or for something simpler, can give a definition of a continuous function. Do such exist? Sure: Darned near anything write down using high school algebra will be continuous. Can also give a definition of a compact set. Examples? Sure, e.g., the closed interval [0,1]. Any such closed interval. Any set both closed and bounded. Then what? Well, any continuous function on a compact set is bounded and achieves its bounds (doesn’t just come close but actually gets there exactly). Moreover, the usual criteria for the calculus integral existing are that the function is continuous on a compact set. When that that argument was worked up around 1900, it was cute and did make calculus more solid.Some of what was good about the definition of a Hilbert space is the range of important examples. So, IMHO what von Neumann did was look at some decades of math, see a lot of examples where essentially the same arguments were being used over and over, boiled down the crucial assumptions for those arguments, presto, bingo, came up with his definition of Hilbert space, then applied the old, frequently applied, arguments to give the basic consequences of satisfying the definition. So, with the definition of a Hilbert space, make the definition and show the consequences once and apply them many times.One example: Take the set of real valued random variables X such that the expectation E[X^2] is finite. Yup, that’s a Hilbert space. But to know this, have to show that that set satisfies the definition. The tough part is completeness. Yup, completeness holds. If think about random variables for a minute, that completeness could hold is amazing. Actually the proof is short.To apply the Hilbert space definition and consequences to networks, economics, etc., need to be clear on just what have that is a Hilbert space and need to prove it satisfies the axioms of a Hilbert space. Just typing this, I see no obvious examples of a Hilbert space in networks or economics.The concept of a Hilbert space is commonly misapplied. The biggie example is in quantum mechanics. There we are supposed to consider wave functions; each particle is supposed to be a wave function. Over time the wave function is to supposed to be a solution of a version of the very old, quite general wave equation.So commonly in physics, it is claimed that the wave functions form a Hilbert space. Well, okay, maybe. But quantum mechanics also emphasizes that each wave function is both continuous and differentiable. Tilt! Wrong! The wave functions can be points in a Hilbert space, but they cannot form or be a Hilbert space because, of the assumed inner product definition and the resulting norm, a sequence of the continuous, differentiable wave functions can converge to something that is not differentiable or continuous and, thus, not a wave function. So, the set of all wave functions fail the completeness part of the definition.So, commonly people are sloppy about such things. Not good.The Kuhn-Tucker conditions are still nice. Mostly they give necessary conditions for optimality; sufficient conditions from the Kuhn-Tucker conditions need a lot from concepts of convexity so are not very general. But for the necessary conditions, intuitively, are in a cave and want to find the lowest point. The walls of the cave correspond to mathematical constraints. The height of the floor of the cave is from the cost function are trying to minimize. Well, in simple terms, if put down a marble and it starts to roll, then are not at the lowest point. Or, to be at the bottom, any direction downhill is blocked by the constraints, that is, the walls of the cave. Can write out some nice, simple linear math for how the blocking works, but that math holds only when the constraints are nicely enough behaved, that is, not pathological. To have the constraints nicely behaved, there are some constraint qualifications. There are lots of those, so can look through the full collection until maybe find some qualifications that apply to the constraints you have at which time can know that the linear expressions hold. Some of the constraint qualifications (CQs) are easy to satisfy but don’t hold very often. Other CQs are harder to satisfy but hold more generally. So, there is some interest in knowing which CQs imply which others. Two of the more tricky CQs are from Zangwill and Kuhn-Tucker. What I showed was that these two CQs are independent. One direction was easy. The other direction needed a bizarre counterexample, and that is what I showed had to exist. For that I needed a theorem; I stated and proved it; it was a surprising, curious theorem.I’ve seen some somewhat practical applications of the Kuhn-Tucker conditions; they are good applied math; but that the Kuhn-Tucker conditions have anything important to do with a real economy is from smoking really strong funny stuff.Why? Because that a non-linear optimization problem has anything to do with economics is from that same funny stuff. The optimization problem is likely from considering utility functions which for anything real in economics are also from the same funny stuff. Utility functions are imaginary things out in the ozone somewhere: Can argue from some axioms, etc., that utility functions exist, but about need some quasi-religion to take the axioms very seriously. Even if believe in utility functions, for anything very real will never see one; yes, in some cases can try to make an estimate of a utility function, but the results will always be crude and nothing like anything precise enough to apply the Kuhn-Tucker conditions to a real case.Or, the whole idea that in a real economy people make buying decisions based on solving a really complicated non-linear optimization problem using the Kuhn-Tucker conditions from some generally very complicated utility functions they have no chance of having is fantasy beyond the ozone. But, apparently can get a Nobel prize doing that!
Separate issue is that access to super fast internet is a red herring anyway. Nobody needs super fast internet (such as you are describing). That is of value primarily for being able to stream entertainment. Entertainment is not essential. Further almost everyone can get news over the air and by broadcast tv. Most websites are viable to use with ordinary low powered run of the mill broadband. And as you pointed out there are places to get free internet the library being only one example.What this boils down to: Companies that do video streaming or want to do anything requiring a large amount of bandwidth not wanting to pay for the extra bandwidth they are using. Or companies that want to profit off of entertainment (an example I read was NBC potentially limiting access to the Olympics by an ISP they work with) not being able to get that content because of exclusive deals. Not essential.
But could happen if telcos are set free to manage and discriminate the packets flowing through their network segments as they wish. In my view, legislation should only assure that every packet is treated equally. Nothing political or speculative about it.Guarantee that and everything else falls into place.The only complexity I can see is lawyers writing the rules à la JOBS Act.
The way you convince people that are ridiculous anti regulation is to tell them that if you allow this that their liberal tech elite ISP is going to slow down Drudge and Breitbart to make it unusable.
Here is the problem. The ISP’s have in essence been granted a government right of way to build their services.Like or not using either wires or wireless to transmit your signal means you have been granted a government right of way.I cannot cut the wires going past my house.I cannot jam the airways going through my house.This is not a theoretical argument.Ed Whitacre CEO of AT&T and SBC said it best in 2005: http://www.washingtonpost.c…Yes, having done a ton of work for telco’s it burns them that companies like Google and others make money over the web and they don’t get a cut.Remember what phones were like before the iPhone? You paid a huge tithe to get anything on the phone.It’s great you have so much competition in Austin. But the biggest sin as you said you can make is not knowing what you don’t know.Want to know how many broadband companies I have access to? One. One. The cable company. Verizon doesn’t want to string FiOS, and you can say a wireless competitor can come in, but guess what? You need to do the backhaul on Atlantic Broadband who specifically prohibits that and has crappy speed anyway.Sure they’d love to throttle my Netflix account. I actually suspect they do.But they only are in business because they are allowed to run a wire past my house, and on the poles in the countryside. Sure it costs them money and that is why I pay them $2k a year, but I do not want somebody throttling me.No other utility does. The water company does not charge different for me watering plants or filling pools than using that water to drink, bathe, and clean which are critical.The electric company does not charge me different for charging my laptop or phone which are critical to my career versus some little nice to have light my wife has in the basement.The viewpoint that just let people compete and it will all be great does not work in certain areas where the commons are critical to providing service.Federal Reserve. Environment. Public Right of Ways.All of these are critical for certain companies to compete. If you use or can abuse these, the better economic decision without regulation is to abuse these. That is the role of government.
.The presence of utility easements at the edge of property is a reality. The gov’t controls those easements.I hear what you are saying and get it that ATX is a very good market for service.On Earth as it is in Austin By God Texas!JLMwww.themusingsofthebigredca…
I agree with both. But if you get to use the easement you subject yourself to regulation. In my mind that is a very simple trade.My point is this. For people that oppose any regulation they end up coming off as a horses ass, and realistic reduction of regulation (which I am so for) gets ignored.We have to have regulations. If we had no regulation why would I pay for trash service? I’d just throw it out of the back of the truck on the way to work.I am sure you are for reasonable OSHA regulations. Do they take it too far? Yes. But there has to be a baseline of safety in the workplace.This is a very baseline regulation.
.I am all for “reasonable regulation” in the same way there has to be rules for a basketball game before the competition begins.I am also in favor of prophylactic regulation such as the US SEC is supposed to enact and enforce.I am not in favor of regulation which anoints and entrenches winners at the expense of competition.I am not in favor of reckless competition which masks illegality — such as the unfettered creation of ICOs pouring out of Zug, as an example.It is easy for me to say these things, but it is very hard to actually draw the lines and make them work.@willykaram:disqus astutely pointed out the lower cost of foreign Internet services and the numbers do not lie. We should be trending in that direction. That may require a little gov’t heavy handedness to get there. We shall see.JLMwww.themusingsofthebigredca…
.One further point about easements — they get murky when the water utility, the sewer utility, the gas utility, the electric utility are all city owned.When that happens, the enforcement is also the user of the easement.JLMwww.themusingsofthebigredca…
I do not disagree. I do not like selling off the assets to private companies, but I can agree to selling off the right to run it for a fixed bid for a fixed period of time.The problem gets worse the bigger you are, because then layers of bureaucracy get entrenched.My small town runs very frugally. But the college town next to me runs with gold plated equipment. You can tell by just the trucks.
My understanding is that all that is happening now is the network neutrality issue has been moved from the FCC to the FTC which, by history, is a more appropriate place.
There are certain topics where I have to choose to believe certain people as being credible since I haven’t taken the time to become the experts many of you who have commented are. And in this case I believe in Fred. So if you are trying to order a burrito or bowl from boloco today, you’ll run into this first. As we say “even a burrito aspires” to make a difference 😉 https://uploads.disquscdn.c…
How do you know that your interests are aligned with Fred’s interests who is an investor?
Such an important issue! I am glad to see so many big sites participating in this protest.
Fake news man. Google and FB are trying to keep the analytics to themselves. They paid handsomely for that adtech monopoly. Pull your head out dude.
second round this year I’m doing this *sigh*
Last I heard, nothing has really changed on net neutrality. Instead, what has changed is that the net neutrality issue has been moved from the FCC to the FTC where, maybe, tradition says it should be anyway.
We gotta stop this – I just got letter with Cox Cable now capping internet traffic and charging more for use.
.Why not?It’s where our tax dollars flow for exactly this reason — public access to knowledge and learning. Seems a perfect use case.I am anxious to learn why it is not.I cannot tell you how much pleasure I enjoyed as a child going to the library.JLMwww.themusingsofthebigredca…
.Here is Grande Communications 1gig page. Look around for the other service types.http://mygrande.com/interne…No more homework assignments.JLMwww.themusingsofthebigredca…
Will, do you even know the regulatory environment before during and after divestiture? How good is your knowledge of the 1913 Kingsbury Commitment or TA34. Read my comments elsewhere in this thread. Net neutrality is a contrived notion because we didn’t apply equal access (interconnection) symmetrically under TA96; which created all sorts of market inefficiencies.Furthermore, we’ve not replaced the inefficient mechanism(s) for universal service with market driven settlements. The internet is settlement free and net neutrality argues against two-sided settlements which are essential for generative, networked, digital ecosystems. And that has resulted in the current mess of near monopoly siloes at the core and edge. The digital divide continues to widen, not decline. The business models of the providers are broken and we are all trading our privacy for access to information.Be happy to continue the discussion offline.
In another world it’s called equal access (aka interconnection) and it should be applied symmetrically across the informational stack along with terminating settlements. The latter is anathema to the settlement free internet which lacks broader “inter-network effects” and is increasingly being dominated by a few monopolies. Why has it taken IPv6 20 years to get to 20% penetration? Why do 2 companies control 80% of one of the biggest pieces of the revenue pie?
Sorry meant to post my deleted comment to the above thread.
The innovation can happen at the top or bottom, edge or core. I am not conflating anything. Look at my framework on my website which is pretty rigorous. As for arguments of 90s and 00s; not sure what you mean. I’m on record saying they didn’t get TA96 right; and I was an extreme minority back then. The inability for IPv6 or new protocols to gain traction in a networked ecosystem without settlements is as germane to the discussion as interconnection.
Equal access and interconnection should be applied symmetrically throughout the informational stack as required or necessary depending on the context. NN does not do that. It only deals with one side and is perfectly happy to maintain edge access monopolies and there control over interconnection at the WAN/MAN edge. Also NN argues against 2-sided settlements and the ability for the centralized party to subsidize or make access free. The settlements will also serve as price signals providing necessary incentives and disincentives to clear supply and demand in a complex ecosystem. That’s why I used the IPv6 example. Anyway, NN has deep flaws.
.Your point is well made. A lot of the world did not have a legacy system to abandon or to modify. Still, your point is important.In addition, the American telcos were and are powerful.JLMwww.themusingsofthebigredca…
.The facts speak for themselves.I spent a year in Korea and they, essentially, had no phone system when I was there in the 1970s.Hardest working people in the world and the second best thieves.When I was there, they had no steel. Not even reinforcing bars. Within 5 years, they were the world’s foremost ship builders.JLMwww.themusingsofthebigredca…
Been anti-telco for 27 years and always focused more on competition. I’ve seen all the bad legislation; but there has been some good stuff. Most of the positive stuff that has resulted, though, has been unintended and not entirely planned. My two favorites are part 15 leading to 802.11 wifi and ILECs move to flat rate pricing in the 1980s which paved the way for the internet to scale in the US.But I’ve also seen all the bad business models and I know what the true costs could be in a perfect world; basically 90%+ below where they are now. Right now I’m focused on the settlement systems that will make the IoT business case scale and bring gig access to the masses. A lot of my ideas revolve around empowering the end-user and the edge. That’s what most network folks fear; but they shouldn’t.
Regulation, applied symmetrically to cable, telco, wireless, should mandate interconnection as far out to the edge and down in the stack as possible. The quid pro quo to that is for the regulator to introduce, foster and monitor 2-sided settlement systems. They should not be involved in setting prices. Once we have those two features in place the market will drive universal broadband.
.I get what you are saying, but there is an absurdity to calling for intrusive regulation as a prelude to letting the market drive things. It feels like chemotherapy.”I am in favor of free markets, but only after we’ve had a really big dose of a regulatory enema.”Just another exercise in the gov’t picking the winners and losers with help from the lobbyists.JLMwww.themusingsofthebigredca…
The govt starts the monopoly with RoW restrictions and frequency allocations. (Basically as a solution to the tragedy of the commons.) And even in Part 15 shared spectrum their power control regulation is what created the wifi market (unintentionally). So they have a big say or impact from the very beginning. I am not in favor of them setting standards and prices. But they can mandate interconnection to the allocated spectrum or RoW (aka layers 1-2) and guide and monitor inter-network/actor pricing (across all layers and between all actors).The only thing missing then is for the market to formulate a “Market Driven Settlement System For Universal Service”.Note that Shapiro and Varian have it all wrong in this 1997 WP on network effects and universal service subsidies. http://bit.ly/2umzY1I They equate or perceive network value with value or cost at the edge. They argue against subsidies in almost all instances. That’s just not right. They are actually, in the process, retarding network effects and value.Value is captured at the core of the network and is geometric. Cost is more or less uniform at the edge (there are marginal differences as you pointed out, as well as marginal differences in value as they point out) and linear. They are missing this key aspect of network theory and therefore do not achieve sustainable business models. We need Universal Service Settlement Systems that equilibrate better the value captured at the core with the costs at the edge.For instance a rural household might hold tremendous value to a business in a city or an urban friend or relative. Likewise urbanites who travel to and through rural places want great 4G coverage. In their model broadband does not reach the rural markets based on the costs/values perceived in those local markets. There are no effective mechanisms in today’s business models to equilbrate the value from one area to the cost in another area; even thought it is quite obvious.BTW, with respect to ICO’s they believe all the value is distributed evenly. And that somehow each unit value equates to the total network value. Both are wrong and absurd notions. “Network Effects Gone Wild.”
.OK, Mike, you’re a solid nominee for the Best Boy award. Any thoughts on Trump’s visit to Paris for Bastille Day?Dial it back?JLMwww.themusingsofthebigredca…
We need something like Godwin’s Law for Trump in these conversations.
.Trump bad.I was watching the Trump-Macron arrival at Les Invalides in Paris. I got distracted. Sorry.JLMwww.themusingsofthebigredca…
One more homework assignment. It says upto 1gig. Here is the arbiter of speed: http://beta.speedtest.net/Also PCMag had an article. http://www.pcmag.com/articl…I am really impressed if you truly are getting 1 gig.
.I check it frequently using Ookla and I get 5ms ping, 887 up, 775 down consistently. I am Ethernet hardwired to the router which has an interior block (meaning fiber to the block next to the router rather than at the exterior).I can get the full 1gig but I have to be running a “clean” laptop with an Ethernet connection and nothing on the computer.I am using a 4 x i7 HP laptop with a three monitor docking station (Dell 34″ curved monitor and Dell 25″ side monitors) running Windows 10 Professional and Chrome.The speed is there.JLMwww.themusingsofthebigredca…
I am impressed. That is close enough to be 1gig.
Impressive, in particular the fact that it is symmetric. You could perfectly run a small site with that. Do they put any restrictions on how you use your upstream bandwidth?
Amazing times we are living in:Let’s see: My biggest Web page sends for about 400,000 bits (plus the ads which I don’t send because they go to the user directly from the ad networks). So, with your upload data rate of 775 Mbps, that would be a peak rate of775 * 10^6 / 400,000 = 1,937.5Web pages per second. With five ads per page and decent ad targeting, that would be about $2 per 1000 ads displayed for2 * 5 * 1,937.5 * 3600 * 24 * 30 / 1000 = 50,220,000dollars a month peak revenue or about half of that on average 24 x 7 for ballpark $25 million a month.Nice bandwidth!
> You could perfectly run a small site with that”Small”? See my post for $25 million a month.