Just to be clear, this approach is awesome and completely correct. However, it's important to remember that this policy isn't stable. It's liable to change with the FCC chairman. Should the next administration choose to regulate ISPs more loosely, these gains will be reversed.
The only solution is for Congress to pass these regulations and enshrine net neutrality as a fundamental principle of internet regulation.
I agree that Congress sucks, but their website shows they enacted 79 whole laws in 2013. These include the "Small Airplane Revitalization Act of 2013" and the "Freedom to Fish Act".
Did you know that cell phone networks today are mesh networks? There is a huge amount of work going into mesh network development, though admittedly, there's still a long way to go.
All that said, even if we just had a functioning alternative (even if it were significantly slower), it would give us options when we need uncensored, unregulated communication. It's better than having absolutely no options when your government oversteps its bounds.
> Did you know that cell phone networks today are mesh networks?
I don't think this is true, at least not using Wikipedia definition of "mesh network".
Mesh Networks are where each node relays data to other nodes, with data "hopping" from node to node.
That does not happen on cell networks. Each phone relays data to a cell tower, but thats it. There's no "hopping" from phone to phone, or from tower to tower. (Your phone might move between towers, taking a phone call from one tower to another as you drive, but there's never more than one "hop". Your phone call never jumps through three cell towers before hitting backhaul, for instance.)
For the same reason that having a few Wifi hotspots backhauled in a building isn't a "mesh network", a cell phone network is also not a "mesh network".
Not quite true. If you are driving (say, along I-95 on a long trip) it's true that your phone will disassociate with one tower going out of range to a new tower coming into range, and the phone switches behind the towers will transfer "ownership" (more like association), but once you (if you aren't driving, or your passenger) accept a phone call, things get a bit different.
The cell phone is answered. So, it goes from switch O (the caller) to switch A (where your cell phone is associated with). Eventually, you'll move out of range of towers for switch A, and come into range of towers for switch B. But because the phone network is circuit switched, you can't just create a circuit from O to B. No, what happens is that A forms a connection to B, so now the call is going O -> A -> B. Talk long enough, and eventually, your call may end up going O -> A -> B -> C -> D -> E -> F.
So while it's true that your call doesn't go through multiple towers, it does go through multiple switches (as long as you are talking on the cell phone, and the cell phone is moving).
Are these links RF or Land Based? Are there any examples of fast (> 1 megabit) RF Mesh Networks with greater than 5 hops in the RF Mesh and, say, around 20 active talkers at once behind the last hops of the Mesh?
I'm happy to get 100 kbits/talker in that scenario with todays FHSS technology, 500 meters/hop, 1 watt power, 200 kHz channel width and 20 MHz of spectrum in the ISM band.
Of course, the advantage of Mesh Networks is they require little in the way of infrastructure, and scale up into the millions of nodes, with 50 million active node networks practical. But high data rates are not one of the properties of such networks.
Well, it's RF between the cell phone and the cell tower. Between the switches, it's most likely land based circuits, but really, they could be anything.
The relevant section is Sec. 7.3.2, "Description of the subsequent handover procedure ii): MSC-B to MSC-B'".
(They call your "O" by "A", your "A" by "B", and your "C" by "B'" (B prime). And an MSC is a Mobile Switching Center that usually handles about half a city's worth of towers.)
Is that while a call is in progress? That sounds like what happens when a cell phone that is not in use, transfers from switch to switch like you describe.
Or the instructor I had taught us outdated information.
AFAIK cell phones are "mesh" in a bit different meaning of that word. Remote BTS-es have high connectivity provided over radio, which creates a mesh. But it is still centrally administered. It is way closer to switches of your ISP having STP turned on than to the openlibernet.
Speaking of which, a nice incentive. I see FAQ expresses everything that should be told, in an honest way. Good read for someone who wants to know about mesh networks.
I would also love something like it to become commonplace. But as it is not a viable alternative to the Internet and does not add anything while the system has not went mad... I don't see it happening any soon.
Hey Irem, I'm doing a presentation on how to make a "mesh network" for a group of primarily web developers. I do a bunch of js robotics so I'm leveraging talking over a serialport, and am doing an xbee proof of concept project. I'd love to know more about existing hardware and systems that I could tap. Would you mind talking to me for a few? (contact info on my profile)
This sentence blows my mind, and I don't know enough to understand it. Could someone helpful point me to a relevant Wiki article or something to help me grok how this could be possible?
I don't think it's fair to say that as a generality. The stability and latency of the network very much depends on the routing protocol used as well as the strength of the connection between the nodes (assuming it's a wireless mesh). There are many routing protocols that are more robust. Some protocols are tolerant of devices moving, some are not. Some are self-healing, some are not. Some have to map out the full route to a device before sending data, some are only aware of the next hop.
There are a lot of options, each with their advantages in different situations. It's not so clear cut. It can be fast or it can be slow. It can be stable or not. It very much depends on the particular network protocol and setup.
Reading your comment I felt we agree on the facts. For personal opinions, I will reiterate: I don't see it happening in practice. Cool idea on so many levels, just the incentives aren't there to make it happen in current society.
Since I can't reply to your last two comments (too deeply nested I guess), I'll comment here.
I agree that right now it's not going to happen, no matter how many different efforts there are out there. For it to really happen I think we would have to come up with a plan that would actually work, in terms of creating a massive world-wide mesh network. There would have to be a lot of consideration for things like security, DDoS mitigation, speed, standards and upgradeability.
If developing a viable mesh network for the task weren't tough enough, there would also have to be a plan for migration which is really the hard part. Getting everybody in the world moved over to a new network would be nothing short of a miracle.
I wonder how this sort of thing will affect Google's fiber offerings. Say what you will about Google (I don't trust them, but I don't trust anyone...including YOU!) but they run one hell of a tight ship and they are aggressive when it comes to their business interests (note G+ integrations with all other GOOG services).
If Google Fiber comes to an area that's controlled by TimeCast (ComWarner?) will Google enter into an aggressive pricing war? I mean there were rumors several years ago that they intended to license the dead air waves to use as free internet. And I'd let Google inject ads into my browser if it meant getting that insane bandwidth - especially if it's for free.
Negatively. A heavy regulatory burden will raise Google's costs to enter into this business. In particular, any last-mile network that Google would build would be much less valuable to Google if the government requires Google to be "net neutral." Of course Google wants to discriminate in favor of its traffic--and why shouldn't it? It built the network!
Government enforcement of "net neutrality" is a great idea if we want to cement the role of the incumbent operators.
There's nothing in a net neutrality demand that would change Google's costs as such. One might argue that neutrality might impact Google's potential profits - but since Google already has such a large portion of the net's traffic, it's main business model has involved a neutral net and they have been a strong advocate. Since they profit from a large portion of Internet traffic, their aim has been to increase this traffic to thus increase their profits (IE, the ability to shake-down Netflix is not worth anywhere as much as the other ways Google makes money - see: "world's largest advertising company").
On the other hand, some of their motive for entering the ISP business has been make sure the current operators respect net neutrality and don't engaging in sleazy that exploit their monopoly (throttling, ad-substitution, etc). If net neutrality were guaranteed, Google might feel less urgency in entering this business. On the other hand, they aren't doing this at a loss so they may well continue.
Also, Google has a vast internal network that carries it's internal traffic. This network isn't part of the Internet at all and won't be affected ever by net neutrality issues.
I thought the ruling said they didn't have the authority to do this without classifying ISPs as common carriers. If they're not going to do that (per the end of the article) then what changed that makes it legal this time?
I took "The commission will not seek to immediately reclassify Internet service as a utility. Mr. Wheeler said that the commission will retain the right to do so, however, if its new rules are approved and appear not to be working adequately." to mean that they'll be using the utility-classification as a lingering threat if the ISPs don't follow along.
It's odd to see utility status classified as a punishment for industry. Shouldn't it refer to the technology (water delivery, electricity, possibly internet) and its role in the customers' lives?
It is in a way a punishment. What makes a company profitable? Creating things people want and need. What makes them really profitable? Creating things people really want and need. But if you build something people really need, you get classified as a utility, which limits your ability to profit from what you're building and also drives out most of the spark from the industry. Our water pipes are mostly 100+ years old, our electric grid is ancient, most of our power is provided by plants that are decades old, etc. Chicago decommissioned a 110 year old coal plant in 2012. Our utilities are not exactly places bustling with innovation, new ideas, new technologies. Regulation has effectively killed the once promising nuclear industry, for example.
I'm not sure this is a fair assessment. There's been tons of research into nuclear power, for example. Look at the generation IV reactor designs [1]. And there's been tons of interest and investment in cleaner electric power alternatives [2].
The examples you cite (coal plants staying open forever, water pipes, etc) are massive capital investments. That's the primary reason they've stayed open / haven't changed for so long, once you spend that much money you want to milk it for every penny it's worth.
Utility regulation is burdensome and costly, but I don't think it necessarily kills innovation in industry like you intimate.
> I'm not sure this is a fair assessment. There's been tons of research into nuclear power, for example. Look at the generation IV reactor designs [1].
Research isn't industry. The regulations don't prevent people from designing reactors; just from building nuclear power plants.
I don't think the statement "The Nuclear Regulatory Commission is the reason nuclear power hasn't taken off" is incompatible with the statement "Regulation has effectively killed the once promising nuclear industry."
Nuclear is a significantly more regulated industry than coal and natural gas. Consequently, it is cheaper to run coal and natural gas plants, just because there are large external costs associated with natural gas and coal (emissions, environmental damage) which are passed on to the taxpayer instead of the consumer.
It is quite possible that if natural gas and coal were forced to pay for their externalities (via regulation) nuclear would be more cost-competitive, and you'd see more advancement of nuclear technology in industry.
> Our water pipes are mostly 100+ years old, our electric grid is ancient, most of our power is provided by plants that are decades old, etc. Chicago decommissioned a 110 year old coal plant in 2012. Our utilities are not exactly places bustling with innovation, new ideas, new technologies. Regulation has effectively killed the once promising nuclear industry, for example.
In a privatized scheme, it is unlikely that infrastructure would be upgraded unless those upgrades paid for themselves in some way or were compelled by law. We are already seeing that telcos don't seek to improve the quality of infrastructure in many cases and instead seek a way to increase margins without having to make significantly expenditures on infrastructure.
But if you build something people really need, you get classified as a utility
This is misleading oversimpliication. The issue is the viability of competition, not really the "need" of the consumer. In the case of cable/telcos, competition is precluded by a variety of non-market barriers. These barriers create problems for functioning markets to capitalize competing entities, regardless of availability of capital, technology, or managemnet skill. So, in a sense the government is responsible for the "mess" of limited cable and telco competition. And they are (as a result) on the hook to limit the damage of their anti-competitive public policy. The logic of limiting competition is some variation on the theme of "natural monopolies"; this is the origin of "utility status". In other words, it is almost strictly speaking a supply side issue and a government policy issue. The levele of demand is really only a second-order part of the equation; although internect acesss is being blurred in a sense of becoming a <perceived> public good (like infrastructre). But in a certain sense, it is not really a public good. The latter are typical removed from the market altogether. ie, Roads are for the most part free. Telephone and cable service, have not really been treated this way. What is sort of different is that the Internet has been priced more like a road thank like cable service. The cable companies are of the view they are "entitled" to price it like a cable service because their strategy is to monopolize access and to extract rents from their "entitlements". The public interest, from a cultural and economic perspective, therefore of what goes over the pipes can be sepertated from the pipes themselves. The pipes should and could be treated like "servicies" for which people pay. But the cable companies could and most likely should be denied and access to the data and the economics that flow down the pipes. The government could legislate the cable companies asset value to zero, if they need to. They could pass legislation to destroy copyright's perpetuity value {etc}. There is plenty of leverage to bring the cable companies to bear to a public interest position. At least technically. In reality, we know who is already in the pocket of the cable companies though +D.
Probably the same number that dream of going to work at the intensely-hated digital utilities, aka cable companies.
Who in the world has ever said "My dream is to work at Time Warner Cable"? I wouldn't tell people I worked for them. My house might get egged.
Hell, I wouldn't be surprised if companies that design and operate water treatment plants aren't often more innovative than a company like TWC that has no competition on many of its markets.
Since 2002, the internet speed at my parents' house in Virginia has increased from about 256 kbps (SDSL) to 50 mbps (fiber). That increase of 100x isn't quite in keeping with Moore's law, but it's pretty damn impressive and talented people at those "intensely-hated digital utilities" are the ones behind it, backed up by the tens of billions of dollars those companies are throwing at the problem.
256 kbps was pretty slow even in 2002. Using that as your baseline is like bragging that you doubled the wealth of somebody who previously had one dollar to his name — big percentage improvements on small numbers are not huge accomplishments in the grand scheme of things.
My personal situation: My Internet is the worst it's been in years, more expensive than it's ever been, and I have no other option than Cox in my area. I know very few people who have seen any notable improvements in the past 10 years.
Lucky them. Since the late 1990's, the internet speed at my parent's house in Alabama has gone from 38.8kbps to... 38.8kbps. The house is in the sticks, in a bowl, served by ancient, poorly maintained telephone and cable lines.
The future has been here for decades (see South Korea, or much of Europe, or...), it's just very unevenly distributed.
One of the things that led to the current situation is municipalities granting favorable franchise agreements to get companies to serve places like that which the market wouldn't have served to begin with. I don't think we want more of that.
The future is here, in the U.S. too: http://www.akamai.com/dl/documents/akamai_soti_q213.pdf?WT.m... (see page 13). The U.S. ranks 8th in average connection speeds. It can't compete obviously with places like South Korea, where half the whole country lives in the Seoul metro area, but is better than much of Europe.
I agree that we don't want to provide favorable terms to companies who aren't interested in managing telecommunications infrastructure as a public utility.
I've heard the population density argument, before.
It's bunk. If it wasn't, we would expect to see reasonably priced 100mbit+ residential fiber service in NYC, SFO, SEA, and other densely populated metro areas. Instead, we see somewhat reasonably priced 1gbit residential service in Chattanooga, TN (1/10th SFO's population, and 1/20th NYC's), wherever Google deigns to pick a fight, and places like Sebastapol, CA (with a population just under 8,000 people).
I had a ~5mbps cable modem connection in southwestern Washington State miles from the nearest thing that could reasonably be called a city in 1999. 50mbps represents a 10x increase in 16 years.
I know of at least one in my circle of acquaintances, who is a highly intelligent and accomplished engineer. There's lots of fascinating stuff to do at a water company: industrial control systems (software and hardware), large structure engineering, resource management, etc.
Lots, it's an entire subdiscipline of civil engineering. So by that token, maybe we can also ask how many talented people dream of becoming legislative lawyers.
Yeah, there are many things that killed the nuclear industry, but I've never heard anyone claim it was too little regulation.
( I mean, yes, the regulations outlawing new nuclear plants obviously, but those regulations came about because most people judge the previous regulations to be inadequate. I don't agree with them, but I don't think deregulating the nuclear industry would help anyone. )
Too much regulation. The process made it inordinately expensive to bring new nuclear technology to market.
I'm not knee jerk opposed to regulation. But the best regulation identifies a market failure and addresses it in targeted way. Net neutrality regulation doesn't do that. It just implements some group's ideal of what the industry should look like.
Hang on a second, in what way is "last-mile connectivity providers cannot selectively extort content providers based on business considerations" NOT identifying and addressing a market failure?
I pay for the pipes on my end, I want the same data-rate regardless of whether the data originally came from a netflix server or a NBC-affiliated company's server.
If netflix needs to beef up their peering arrangements in order to get the data to comcast's network, that's one thing, but they absolutely should not have to pay an additional fee to comcast to protect themselves from being penalized.
That's not a market failure, but that's rather using a market failure as a hook impose a particular ideology about what the internet should look like. Hypothetical: say there was no last mile monopoly. The coax into your house was owned by the municipality, which allowed competing providers to plug in and provide the rest of the network. Would that eliminate the need for net neutrality regulation? No: companies could still pay ISP's to prioritize their packets all the way up to the municipally-owned coax into the customer's house. In other words, the regulation has nothing to do with the market failure ostensibly justifying it.
Being able to pay other companies to give you better service is an intrinsic advantage to being big, and is not the product of a market failure. For example, Amazon works out a special deal with the USPS for Sunday deliveries. That still gives a huge benefit to Amazon versus smaller competitors who can't afford to pay up. Being small comes with penalties, and the FCC shouldn't impose an ideological view of the internet that artificially removes these disadvantages.
I don't understand this argument at all. You're conceding that last-mile extortion is a market failure, then turning around and saying legislating against that last-mile extortion is "imposing ideology" rather than just addressing the market failure?
When I purchase internet connectivity, I want internet connectivity. I don't want varying grades of speed depending on who I'm talking to, I just want a dumb pipe. That's what I'm paying for. How is legislating the dumbness of the pipe 'imposing ideology'? The market is allegedly $/gb/mo, let providers compete on service, price and speed. If they can't deliver those GBs for those $s, they're free to charge more to consumers. The only ideology there is capitalism. What am I missing?
Note that I said in my original post, if Netflix doesn't have good enough peering arrangements on their end, then that's their problem and nobody thinks there should be legislation guaranteeing them cheap bandwidth. But most of us think they should be able to buy bandwidth to the consumer networks, via peering, at the market rates that everyone else pays, maybe with discounts for volume.
The last mile monopoly is the market failure, saying that internet connectivity should be just a "dumb pipe" is the ideology. Legislating wireline providers into being "dumb pipes" doesn't do anything to directly address the last mile issue, but instead uses the last mile issue as a hook for imposing an idealistic vision of what internet services should look like. There's no nexus between QoS or prioritization or the other sorts of things an ISP might want to do and the last mile monopoly.
A better legislative approach would be something like having municipal utilities own the last mile coax, and allowing different ISPs to plug into it. That would eliminate the market failure directly, and allow the industry to evolve into whatever structure makes sense. I think you wouldn't see them end up as dumb pipes in a competitive market. Some ISP would come along and make a deal with the NFL to have high quality streaming only on their network, etc. Just like how there's no monopoly in shipping but Amazon still has an advantage by being able to cut special deals.
Ok, that's our disagreement then. When I buy internet service, I and 99.999% of customers are trying to buy a dumb pipe. That's what it's marketed as, that's what the provider tells us we're buying.
I'd consider a patchwork of different providers with the same problem ("well, sprint has fast netflix but comcast has fast amazon") to be just as much of a market failure. That's bringing huge distortions to internet business, whether or not there's a monopoly on the last mile.
How about if we split it like so?
* Content-neutral QoS and prioritization isn't a market failure.
* Throttling netflix because they strategically threaten NBC (owned by comcast) is a market failure.
Would you agree with that sort of categorization?
How about the following?
* Throttling whichever service your customers are using the most so that you don't have to provide the bandwidth you sold to them under a contract, also a market failure.
Being out of compliance with what you think the internet should look like does not make for a "market failure." There has to be some sort of economic explanation: the existence of a negative externality, free-riding, etc. Even in competitive markets, companies still get together and cut deals and customers face choices like "Sprint with fast Netflix" versus "Comcast with fast Amazon." The mobile OS world is quite competitive, yet I still face choices like "Windows Phone with Office" versus "Android with Youtube." The restaurant industry is highly competitive, yet most places serve Coke or Pepsi but not both.
If you eliminated the last mile monopoly, which is an economically valid reason to regulate, with a targeted regulation, I still think you'd see a lot of the sort of activities that HN-ers dislike. There would still be a huge incentive for ISPs to vertically integrate and buy up content companies and cut special deals. What does Comcast care if they don't own the actual coax into my house, just everything leading up to it? Who can compete with them when they have a "killer app" like Hulu and their competitors don't?
So, if which ISP I happen to be a customer of completely dictates which streaming service I wind up using, then the actual merits of the product have become divorced from which product I buy. That's textbook market failure, right? Dollars aren't flowing to the best product or the cheapest, they're flowing to the product approved by someone in a completely different industry. The invisible hand isn't doing shit for me in that scenario.
Is the next phase of this discussion "the free market will create incentives to start new telephony companies that don't throttle your favorite website"?
For an analogy, should the electric company have a say in which brand CPU I use? Or should they just sell me watt/hours?
Is it a market failure that I can't use Nokia Maps on iOS, or Apple Maps on Android or Google Maps on Windows Phone? I'm no proponent of invoking the "magic market fairy" but to me situations like that fall far short of "market failure."
I guess it all comes down to you not seeing internet conectivity as a utility. Personally, I think it has a lot more in common with sewer/electricity, from the near impossibility of market entrance right down to the sewer/electric/phone lines being physically next to each other, than to putting a phone in a box and selling it to someone.
Heck, I'd even put mobile service as being closer to a normal consumer good than wired internet. Wireless they just need one tower every few miles.
There are some similarities, and a lot of differences. Most households have a fixed demand for power and water. Meanwhile, bandwidth demands have exploded in just 10 years. That's a blink of an eye in utility timescales. 100 year old water pipes are still usable. 10 year old network infrastructure can be obsolete. Public institutions simply aren't able to bring to bear the kind of capital investment necessary to keep up with changing internet technologies and growing demand. People complain about telecoms not investing enough in infrastructure upgrades, but our public infrastructure is has accumulated trillions of dollars in maintenance debt. Even as demand is stable, its a challenge to just maintain existing levels of service in our water, sewer, and power systems.
That's not a market failure, but that's rather using a market failure as a hook impose a particular ideology about what the internet should look like.
The ideology is what you think is the result of the market failure, whether it's being used as a "hook," or something else. Heck, whether it can even be defined as a market failure can been argued to be a value judgement these days.
You should probably back up your extraordinary claims with evidence, if you want to be believed (by anybody who isn't already anti-regulation/government).
I feel like I could sort of wrap my head around Net Neutrality in a world with just content providers and ISPs.
Once people start talking about CDNs and peering agreements, and all those massive middleman infrastructures most consumers don't even know exist, this issue gets orders of magnitude more complex. It seems unregulatable.
What is Netflix supposed to do if Verizon just passes data around its network to keep it "full" all the time and refuses to open up good datalines to Netflix CDNs? What is the network supposed to do if a major data provider ends up choking the lines just to move from one endpoint to the other, not even stopping at customers?
I doubt any rules would work here except for actual enforcement of anti-monopoly regulations on a locality by locality basis. Maybe Netflix needs to start rolling out fiber and competing directly with ISPs (most of them are already competing by providing content anyway).
Although legal action against ISP censoring and monopolization sounds beneficial, I'd prefer that government entities remove themselves from open Internet pursuits entirely.
If internet service were a luxury product, I'd agree. However, an internet connection is every bit as important as having a phone (land or cell) these days.
Assuming we completely de-regulate, you'd still have municipal governments striking these anti-competitive monopoly deals. In fact, there'd be no barrier to it at all, since we're de-regulated in this case.
Don't like Comcast? Better not move to City A, because they "own" it, even though some of the fiber was paid for by the residents.
If that's the problem, why not prohibit municipalities from creating local monopolies? Why have the FCC dictate limits on the design of network pricing and services everywhere, even where there are many options?
Limiting governments from making corrupt or short-sighted mistakes isn't usually considered 'regulation', but even if it is, this 'regulation' is also simultaneously local-franchise 'deregulation'.
And it's more sensible, since competition is a cure-all, than federal 'neutrality' rules about what services are allowable.
I agree, but I'm not sure how you can maintain these things without being owned by the people (govt) or a corp. I guess a regulatory body like the FCC then?
They can serve all they want, but the unfortunate fact remains that companies like Time Warner and Comcast are needed to actually provide the network service.
It's like having a gas station in the middle of a desert, with no roads within a 50 mile radius.
The point is not that they can avoid using Time Warner's infrastructure, it's that Time Warner will eventually stop offering content if nobody wants to view it. Then they will have no more incentive to throttle third party content.
I'm not suggesting that this will work, I'm disputing that the people don't have any power. People almost always have some influence over their situation. That doesn't mean they have a magic bullet, but there is almost always hope.
1) Internet isn't a luxury anymore, it's a necessity. One can't live in society today with much success without some form of Internet access.
2) The free market is not at work when it comes to the proverbial "pipe" that's been laid - the Internet backbone. Getting a fiber cable to your house, to your neighborhood, and to your city was a task that cost a lot of money, and was paid for in part by the US government and state governments. Despite this fact, companies like Time Warner do not share "their" infrastructure to 3rd party ISPs. This, in turn creates an artificial monopoly where Time Warner unfairly benefits from publicly funded infrastructure.
People can't avoid Time Warner, because in many places, they have literally no other option. There is no hope for these folks.
How many people have no access to Verizon Wireless, AT&T wireless, Tmobile, Sprint, or telephone DSL?
There is competition in the residential Internet business.
If you say that "all those alternatives are too slow," I say: too slow for what? Too slow or too throttled for Internet video: perhaps. But the argument you are making is that "one can't live without some form of Internet access." Most people have Internet access available, in multiple forms.
I agree that Internet is not a luxury, but 10 or 50 megabit Internet is. True, you need Internet to apply for jobs and even in some cases to interact with your government. Do you need 50 megabits for that? No.
People can avoid Time Warner. They might not get their Netflix with their slower Internet, but the government should not be getting involved to preserve something as trivial as streaming video.
> How many people have no access to Verizon Wireless, AT&T wireless, Tmobile, Sprint, or telephone DSL?
Different product, same problem, possibly even the same company. Switch out "Time Warner" with "Verizon" (a reality for some), and the problem exists anew.
There is not competition in the residential Internet business.
I would guess that most people are more interested in watching something than they are in fighting Time Warner Comcast. As such, if the cable ISPs throttle Netflix (or de facto throttle it by choosing suboptimal peering arrangements), even if consumers would have preferred Netflix, they will watch cable because they don't have any other option.
That only works in places where a free market exists. In the case of ISPs there are generally very limited options and no room for new competitors. That being the case, we don't have much of a free market or opportunity for proper capitalism to govern the entities involved. In situations such as that, unfortunately, we need a governing entity to set some ground rules to protect the consumers.
There's always room for new competitors, if that were the goal. About 20 years ago, almost no place had 'broadband' - only dial-up. Since then, most places have gained 2 or more wired broadband options, and 2 or more wireless options, that are all many multiples faster than dial-up.
The wireless technology keeps improving. New wired infrastructure can be run – as with Google Fiber or other initiatives. The sensors and robots for such installs also keep improving. And for sticky 'last mile' wire bottlenecks, sharing of that conduit could be required, or a 'minimal universal neutral' service option be required where there are no other options.
Instead, the FCC seeks to limit how all broadband services, everywhere, must work, to respond to an abuse threat that remains almost entirely theoretical. That itself is an anti-competition, anti-differentiation move. It also adds another lever of power by which the FCC's other regulatory tendencies – such as censoring indecency or enabling wiretapping – could be applied to the internet. (In the UK, Australia, and South Korea it is agencies like the FCC which maintain compulsory national internet black-lists.)
In my experience all recent Netflix slowness is a product of congestion at peering points between Netflix's ISPs (Cogent, Level 3) and the consumer broadband ISPs (Comcast, Verizon). There's a good article in the WSJ today on the dispute:
For super high bandwidth things (video today) it makes sense to serve as close to the end user as possible. This probably means putting the server in your ISP's network. This probably violates what most people think of as net neutrality. But it makes sense from a technical standpoint. I think net neutrality needs to update it's thinking to accommodate this model. Maybe what should be neutral is the pricing for putting a server inside the ISP's network? How do you treat the ISP's own video service? Force it to be a different company that pays the same as netflix to store a server there?
There are a few interesting choices in your comment that make it appear to be an Astroturf:
1. You state that it is "your experience" that "all" recent Netflix slowness is a result of peering contract disputes. Unless you are a network operations engineer at one of these companies how could you have this experience?
2. You link to the WSJ, which is not a credible source when discussing big media companies.
That said, you may well be a real person, in which case I would recommend changing your wording on commonly astroturfed topics.
OK, if you want to get into the nitty-gritty: My company uses Cogent as an ISP, which means we were very interested in figuring out why our VPN performance became junk starting around last October. Some traceroutes and extended pings pointed the finger at the Comcast/Cogent peering point in NYC. A little Googling confirmed that we were obviously collateral damage in Netflix's ISPs ongoing disputes with the home broadband ISPs.
Since then, it's been easy to correlate poor performance on my Netflix streaming with a a 60ms+ drop across a Cogent/Comcast peering point when I ping one of my company's servers from home.
There have been several articles on tech news sites discussing this dispute[1][2]. I linked to the WSJ article because it was fresh this morning and provided a good summary of the state of play.
On the other hand, accusations of actual throttling of Netflix traffic on a broadband provider's own network have been frequent but I haven't seen definitive evidence that it's happening on Comcast or Verizon.
Personally, I'm pretty pissed that Comcast is taking my $70/month and failing to provide within a factor of 50 of advertised rates to one of the internet's most popular services. They seem to be playing chicken with Netflix and their own customers. It would rebound badly on them if they had decent competition, but franchise agreements are what they are.
I don't count "internet points" very often, but it's pretty amusing for someone with a grand total of 137 HN karma to accuse me of being a shill.
Sean isn't a shill. He is a tech lead at one of the major media distribution companies who has a zest for getting at hard data instead of speculating.
I said it once, I will say it again it is Netflix' fault for not shaming bad ISPs. Netflix controls the source code for their media player and they control their servers. With 99% accuracy they should be able to detect which ISPs are screwing them and release this in a report to consumers.
I wasn't accusing Sean of being a shill. I was pointing out the absurdity of his accusation that I might not be a real person, since by the (silly, of course) way that HN keeps track of these things, I've contributed much more here than he has. His comment was condescending and dumb, and he deserved to be called out on it. Think I'm wrong? Argue with me, don't start calling names.
Netflix does publish ISP numbers, and the trend over the past 6 months for Comcast and Verizon has been awful, as is shown in the "not credible" WSJ article I linked.
I think you're reading too much into it. On HN, it is extremely common for people to make very authoritative sounding posts, especially counter intuitive ones. This is the case whether the topic is net neutrality, Godel's Incompleteness Theorem or the existence of black holes or whatever. You just take it with bags of salt and move on. Also, for many, the Wall Street Journal is not only a credible source but a battle cry. It's just very common on HN, don't think it is automatically astroturfing or pure propaganda (sometimes it is).
Some dirty ISPs are refusing to use the CDNs that Netflix makes available to them instead of straightforward packet discrimination. I blame Netflix for not shaming Verizon and such with CDN usage stats.
Basically it's a play on "grass roots" movements. Powerful groups attempt to mimic grass roots, populist movements by hiring a bunch of employees to pose and regular people and write comments on forums, etc. And as you know, astroturf is fake grass.
There may be a technical differences between bad network performance due to some active throttling configuration in the network hardware, and bad network performance because a company doesn't want to make changes to a network interconnect. To the end user sees both cases as "throttling" in terms of poor performance.
Net Neutrality should be about how we come to terms with companies that are pure ISPs trying to deliver data their customers want vs companies that stand to increase profits and hurt competitors through active throttling or, in the peering case, simple inaction.
First, most Netflix content isn't served from AWS, but from their CDNs running on a variety of ISPs (lately, a lot of Cogent).
Second, the VPN trick doesn't really tell you anything except that the end point of the VPN has a better path to Netflix's CDNs than the route your ISP is using. Your ISP doesn't have to act overloaded routes to the content, which they want Netflix to pay to upgrade.
1) I don't think he meant that it was being served off of AWS, just that was his VPN example. Any in the US should do.
2) VPNs add latency and are still limited by the bandwidth of your connection. Given that Netflix has a decent CDN, in theory your VPN-box shouldn't have all that much better a path to a Netflix server than you do. So, assuming that the paths to netflix are mostly equal, a fast VPN-Netflix experience vs a crappy local connection one would imply the ISP is to blame (for whatever reason malicious or not).
The whole point is that Netflix's CDNs aren't "decent".
That is, they're probably "decent" (quite good, I'd even imagine), but the load is so high so as to cause traffic problems anyway. The question was "how can I tell if it's Netflix or if it's my ISP?" and the answer is, "You can't."
If they're actually overloaded, as opposed to there being peering failures to the consumer ISPs, why would there be a difference between VPN and not-VPN?
To answer both you and your sibling comment - there would be a difference in performance, depending on which CDN is delivered to you. I don't know of a way to request a specific CDN from Netflix, and if the one that's the closest to you is having perf issues, that doesn't mean the one closest to your VPN will have the same perf issues.
I was discussing paths to the CDN, not the performance of the CDN's servers themselves. If the servers are the bottleneck, and not the network, shouldn't performance be roughly the same (or worse over) the VPN verses the local connection?
On that note, can anyone recommend a home router that can easily route all traffic through a proxy server?
This appears to be the only reliable way to stream video during peak hours in my neighborhood, and I'm done with my ISP's suggestion that the problem would go away if only I'd reboot my modem again.
Not to get too in-depth with home network troubleshooting, but are you on a wireless network? Interference may be a cause.
I know it's fun to blame ISPs for all the problems under the sun, but it's actually rarer than one might think that the ISP itself is the cause of a home networking issue.
Wired. Also have tried going direct through the cable modem. ISP's techs can't resolve this after many calls and site visits. (Some techs say this is the expected behavior of the connection.)
While streaming in Firefox, reported speed rate jumps to expected (i.e. non-peak rate) when running through a proxy server outside my ISP's network. Speed drops instantly when the proxy setting is disabled. Flipping the proxy switch on/off literally changes my observed speed by ~10+ Mbps. This experiment has has been confirmed over weeks, different devices, etc.
So...I'm not blaming the ISP, but using a proxy server clearly works for me. So I'm going to do that so I quit wasting time trying to figure out why it doesn't work. I just need to be able to do it at the router level and not the computer level so it works for e.g. the Apple TV.
But, even your "diagnostic technique" is flawed, and doesn't necessarily point at the ISP. If we merely look at what you've said, all we know is that your connection to Netflix's local CDN provider is lower than required for optimal performance.
We can't say the ISP is in any way at fault here.
Try installing DD-WRT onto your router. It has features for things like this.
> We can't say the ISP is in any way at fault here.
FWIW though -- I'm not blaming my ISP, specifically because I don't have enough data and don't care enough about assigning blame to get it. It wouldn't matter anyway, I'm basically stuck with this ISP.
I worry that the rules will forbid ISPs from having explicit whitelists/blacklists, but will do nothing to prevent them from having undersized and saturated transit links then demand payment for peering with content providers like Netflix.
What would happen if consumer ISPs were required, for each customer, to choose a single point (at the metro level, probably) where they are required to peer openly? That allow all Netflix-like companies access to customers, but since it's a point local to the customer, would not impose any long-distance bandwidth requirement on the ISP.
They need to regulate ISPs as common carriers. Anything short of that will result in ISPs constantly lobbing to get more say, until they get what they want. We can NOT rely on the free market to sort the ISPs out because there is 0 competition in that market.
While I agree that right now there is zero competition in that market, the reason is because we didn't let the free market decide. We gave government approved monopolies to them and now we're seeing the consequences.
What does a market with competition looks like? 20 cables from 20 different ISPs hanging outside my window? ISP is a utility, like water, electricity, and gas. Maybe once we have good wireless technology it would be different.
> Maybe once we have good wireless technology it would be different.
Or a bunch of satellites in low orbits, which is something Google wants to do. A number of companies have tried this idea, so far without economic success, but as time passes it's becoming more practical, and eventually it will probably become feasible. Once it does, many companies will do it, and you will be able to choose which company you want to get satellite Internet service from. Finally there will be competition.
The problem with low orbit satellites is the backbone connection speed. The way I understand it, the individual connections to each subscriber isn't really a problem. But say you have a satellite serving 10k clients at one time. If you have a service with any scale, that's realistic. To provide a good quality service to those users you need to get at least a 200 gigabit connection to an Internet backbone. Getting that over fiber isn't difficult, but over air, not sure if the technology is there yet.
I am so looking foreword to satellite Internet, mainly because my new home is off the grid, and I have no Internet at all there. The best option I found so far gives me 25 gigs per month for $70. I would gladly pay $140 for 50+ gb but that's not even an option. Unlimited, fast, and ubiquitous Internet would be amazing, but in my research I have yet to see any company that's even close to offering that.
If any one can recommend a satellite Internet company with high or now bandwidth caps, I am all ears. My budget is something under $200/m for unlimited or 200+gb/m.
> The problem with low orbit satellites is the backbone connection speed.
Yes, and system congestion, which you also mention. This is why so many different ideas are being considered -- balloons and solar-powered high-altitude drones to name two. But satellites, in spite of their drawbacks, at least in principle would have longer lives and wold present a lower per/subscriber cost.
Because of the propagation delay you mention, satellites work very poorly with VOIP, in fact existing synchronous-orbit satellite methods won't work with VOIP at all. Future low-orbit satellites might be more acceptable, but propagation delay will always pose a problem for anything but terrestrial fiber.
If the FCC classifies a carrier as a common carrier, does that apply just to the service, or also to the infrastructure itself? If cable companies were made into a common carrier would cable television also have the same regulations applied to it?
The only solution is for Congress to pass these regulations and enshrine net neutrality as a fundamental principle of internet regulation.