Hacker News new | comments | ask | show | jobs | submit login
Cable TV's plan to kill “over the top” internet streaming (medium.com)
164 points by steven on Dec 5, 2015 | hide | past | web | favorite | 73 comments

To me the obvious solution is to make it illegal for ISPs to get to get into a conflict of interest. So they should only be allowed to sell Internet access and not do anything else, especially not sell content. Given that they are running some of our countries most important pieces of infrastructure I think that's perfectly reasonable.

To an extent this happened in 2006 with the UKs physical phone and internet infrastructure.

It had been run by British Telecom, a privatised government department who were originally the only phone provider. They were required to spin off infrastructure into a sister company BT Openreach.

That company is required to give equal access to anyone who wants to provide phone services, including DSL, and its resulted in a fantastic variety of providers across the country. You can get everything including near free 30Mbps services with data caps and an awful router, unlimited 76Mbps service, right up to bonded DSL over multiple backbones. There's also cable if you want a faster connection.

It happened EU wide (actually EEA wide), as part of the EU telecoms deregulation.

While it is in general quite good, it suffers from limitations in how it is regulated that means that BT is milking Openreach for profits and upgrading the network very slowly. The whole thing could do with limits on how much Openreach is allowed to pay out in dividend based on how much they invest in infrastructure upgrades.

It's a matter of perspective though.

BT was ready to roll out fibre to the home back in the early 1990s for the delivery of TV. At the time BT had some of the most advanced fibre tech in the world.

The government of the day, however, decided to ban BT selling TV in favour of opening up the market to foreign owned cable companies, Telewest and NTL.

Without wide scpread Internet use that killed the investment case for fibre (and the U.K. fibre industry).

Today we're at a situation where none of the other players calling for BT's break up want to invest nationwide.

If they do break up BT it'll be great if you're in London, Manchester, Birmingham, Leeds but I don't expect much to improve anywhere else.

I think it's more likely that Vodaphone, Sky, Talk Talk will just try milk the existing infrastructure without further significant investment, as they currently do, and BT would be dis-incentivised from further investment.

Let's say you pass that law. Comcast splits off Xfinity as a separate company to sell just internet service, with low data caps. Comcast and Xfinity enter a deal to exempt Comcast's streaming service from the caps in exchange for some millions per year. Comcast gets some billions per year in subscriptions, because Xfinity customers can binge on all the Comcast "Stream TV" they want. New high-bandwidth services, like a young Netflix competitor, can't afford to pay off the ISPs, so they never get to compete with the incumbents. The situation is no better than if Comcast wasn't split up.

You've identified another target for regulation: ISPs should not be allowed to exempt traffic from data caps. This is part of a larger concept called net neutrality [1].

[1]: https://en.wikipedia.org/wiki/Net_neutrality

vertical integration!

Laws need to be created so that those who own the pipes can not unfairly compete directly or indirectly(split itself up and make deals with it's old self and or partners) with services offered on the global Internet and or interfere with the greater good the Internet provides society.

What Comcast is doing here is steamrolling the Internet for their greater good and I can't imagine this working out well for them. They are the most hated company.. this surely isn't going to help and I'm glad to see Susan Crawford publicize this.

Hopefully others like John Oliver will jump in too to further publicize this outlandish let's kill the Internet for our greater good anti-competitive move. They need their asses handed to them!!!

As a free market capitalist, I actually completely agree with this. Comcast, etc, actually impede the free market by forcing competitors to go through a monopolistic choke hold. Moving data around is a different thing that actually providing the data. If would be like the airlines actually owning the airspace. Or Ford Motor Company owning the roads and restricting Chevy's access to the roads. Under net neutrality, Chevy must be allowed on the road and be able to travel at the same speed, however under this analogy, Ford would be able to have a quota as to how many Chevys could use the road before paying a toll.

"Moving data around is a different thing that actually providing the data. If would be like the airlines actually owning the airspace. Or Ford Motor Company owning the roads and restricting Chevy's access to the roads."

Almost ... it's as if the airlines built the airspace and then owned it. It's as if Ford built the roads and then owned them.

Get into? Cable ISPs are already in a massive conflict of interest. They are selling everyone their data twice, just like 'voice, texts, and data' from wireless companies.

New FCC rule: Bandwidth may only be marketed as the true sustainable rate over a month. So a 300 GB/mo plan can only be described as a 1 Megabit/s connection.

DSL will start looking good.

That's not a practical metric, though. I'm on a campus with a 100Mbit/s connection, I frequently download large files (media and Linux ISOs), and still don't go above 200-300 GB/month. The 100Mbit/s figure is useful for informing me how fast a particular YouTube video will buffer, for example, but by your proposed metric, sustained 100Mbit/s would have to be advertised as 30TB/month. "20 Mbit/s with a 300 GB/month cap" is more informative.

I'd like to see the bandwidth measured from my computer to the wider Internet, not from the ISP's speedtest server hosted before the traffic sees the Internet.

At least on my ISP in New York, I can run a speedtest to the auto-selected server and see 330Mbps, but switch to a speedtest server farther away (requiring actual Internet transit), and see the rate drop to 10Mbps. So it's all very impressive that I am paying for a 300Mbps connection to the ISP's colo. But it would be better if I was paying for a 300Mbps connection to the Internet.

(Having a speedtest server hosted by your ISP is a great debugging tool, so I'm not asking them to stop doing that. It's great for figuring out if the bottleneck is my local link layer (WiFi, broken switch?), my Internet connection (broken router? squirrel chewing through a cable?), or simply the ISP's peering choices.)

Looking at these two comments, perhaps we may be reaching an agreement. There are only a relatively few numbers that really matter, although there's tons of complexity. Bandwidth should be a number giving sustained symmetrical throughput over a month + burst capability + download cap, all using the same units. All of those numbers should be 5+ hops away, and should be over a backbone to a server owned by somebody else besides the ISP. In addition, there would be some QoS to show that things like the burst number actually exists for the user 5%+ or more of the time (or some different metric). Heck, there could probably be a bunch more detail you'd add under there, but 3 numbers + access_to_wider net probably covers it.

So in the college example, they're getting (perhaps) .1/100/300

I think if we could just agree on a way to advertise it, we'd be ahead of the game. Net access is becoming like buying PCs: full of jargon and product differentiators meant more to confuse the buyer than assist them.

It's not that impractical. It is how traffic is quoted when renting space in a datacenter.

"Comcast 1 mbps with 100 mbps turbo"

"Comcast 1 mbps with 250 mbps turbo"

My wife and I are planning to move next year. It might sound nuts, but we've scratched off a lot of potential places because of a lack of unlimited internet connectivity.

Municipal fiber is very important to us and it's resulting in us focusing on places we didn't consider before.

Curious which places are at the top of your list?

Come join us in Myrtle Beach, SC (WhyNotTheBeach.com) where local telco HTC (htcinc.net) offers fiber to the home, without caps.

It is not nuts at all. Over 20 years ago I bought a house in the Berkshires. The town I am in has no broadband, and probably won't get it for years. If I were buying a house now I would NEVER consider this place, which I otherwise love. What's more, it might be impossible to sell.

Is your town participating in this project?


I agree it would be useful to have a list of such places!

In Russia a decade ago all the consumer internet was billed on usage. Like 40 kopecks per megabyte. Sometimes uplink was free and downlink was metered, sometimes more crazy schemes.

Guess what killed it? Competition. If you allow cables to be slung on top of rooftops / utility poles then soon you'll have a lot of cheap operators.

I was in Australia a few years ago and was flabbergasted that many ISPs were still operating on a complicated set of caps and on-peak / off-peak rules. I blamed the de-facto monopoly created when the copper got privatized and all sold to a single company, Telstra.

Telstra is the privatisation of the old government postmaster general (well after the split from mail post). The government department owned the copper, and telstra retained it as part of the privatisation.

As part of this agreement Telstra had to provide equal access to its network, and had to charge the same price for all users, so that the cost of calls was the same for people in regional areas (even though maintenance cost more).

Most of the issues relating to Austrlia's internet relate to the cost of overseas transit of information. There are only a few cables between Australia and the rest of the world, and that is the large driver of cost or at least it was in the early 00s

Yeah, Telstra was forced to lease copper to any ISP who wanted to use it -- but I strongly suspect anti-competitive business practices on that aspect (delaying hookups to non-Telstra customers, etc.)

Oh, the irony! Malcolm Turnbull paid Telstra for the copper networks so that the government could own them again. He did this so that nobody else could get fibre to the home.

I feel a bit sorry for him. He was forced into this by Tony "Shirt-front" Abbott, who knew nothing about technology (or how to run a country, for that matter). Now he's stuck with it - a millstone around our new Prime Minister's neck.

I guess the intense destabilization and white-anting now going on in the LNP will distract from this.

Is there any evidence at all that he was forced into this? Are there anecdotes of Malcolm arguing for an all-fibre NBN in the party room, only to be smacked down and forced into MTM?

There aren't. I concede that I could be projecting what I want reality to be, in this case.

I'm assuming this was rescinded shortly after because I'm in St. Petersburg right now and not only are competing ISPs plentiful, but the net is unmetered and fast as hell. I think we're paying about 650 rub a month for 40-100 mbps (they do the tiering weird where the more you pay the higher your minimum speed guarantee is). That works out to about $10 USD a month, or for perhaps a more reasonable comparison, about .9% of a monthly salary of 70,000 ruble (not unreasonable if you have some skills)

Usage-based-billing turned out to be not a complete disaster in Canada. ISPs were no longer incentivized to use deep-packet-inspection to interrupt or throttle connections. Regulations required incumbents to grant access to last-mile networks to smaller ISPs to encourage competition. Users had more choice -- plans differentiated on both speed and on bandwidth caps, so if you needed a small amount of data in a month but you needed it quickly, you could still save money by picking the right plan. ISPs were encouraged to build out their networks more, since increased speed led to increased usage led to more recurring income: if you wanted the fastest speeds now and could pay $200/month for it, you got it; if you didn't, you could get Internet faster than the average American "high speed" connection for something like $30/month.


thats what unlimited 250/20 cable costs in civilized world, goes to show how "Usage-based-billing turned out to be not a complete disaster"

You have to keep in mind that Canada has terrible population density compared to Europe/Asia. It's about the size of the US, but with 1/10th of the population.

oh, so you have good competition, low prices, high speeds and no data caps in big dense cities, right?

That's a really interesting perspective. I didn't know that Canadian internet providers operated that way. I wonder what the effect would be like in the US, given that we could muscle last-mile competition regulations through. That could change my mind about data caps.

Every time I find myself thinking maybe the Internet Service Providers (whether cable or mobile or DSL) are not evil monopolistic assholes doing everything in their power to provide worse service for more profit, I simply have to remember that the Internet was invented in the US, but we have among the worst consumer internet service among developed nations. The US is ranked 14th or worse (depending on how you measure and who you believe, but we're not in the top ten by any measure) in consumer broadband, in terms of speed and cost. The US is also much worse than leading nations on mobile service cost and quality. There is no scenario where the world leader in tech innovation should trail more than a dozen other nations on the most important piece of the tech puzzle, and yet we do.

This is the legacy of our telephone and cable monopolies. When monopolists "compete", they do it by using every crooked trick in the book. The telcos and cable companies have a significant number of politicians and regulators in their pockets. All to keep services cheap to provide and expensive to buy.

And, we're losing ground. Things aren't getting better for consumers, despite the new spectrum opening up; our government made huge amounts of low frequency spectrum available to telcos (at a price, of course), on the assumption that they'd use it to improve service and lower costs. No mobile provider offers actually unlimited data anymore, at any price. I used to have two unlimited mobile data plans at $45 and $55. Now, I have two plans that cost more and are capped at ridiculously small amounts of data. Land line providers have also now imposed caps, and have, with increasing regularity, tried to sneak through legislation allowing them to regulate the Internet to their own advantage.

Imagine a situation where a cable company can rate limit NetFlix to the point of unwatchability...or maybe just enough so you can't get a nice clear HD picture. Conveniently, cable company will be happy to sell you HBO, which includes HBO Go, and they'll generously provide unlimited access to that data. This isn't a paranoid theory, this is actually what the legislation they've been pushing every year or so is all about.

And, it's been going on for longer than most people have been aware of the Internet. My first company built web caching devices for independent ISPs, back before the telcos and cable companies used their monopoly to utterly destroy the independent ISP market in the US. They used illegal and anti-competitive practices on a massive scale, but when you destroy your competitors fast enough and thoroughly enough, they won't have the money to wage a legal battle to stop it.

Not that I'm bitter or anything.

This is exactly why Google is developping Google Fiber, Facebook has some crazy scheme involving building their own planes to provide bandwidth, and they are both spending fortunes in lobbying.

Worth pointing out that this is largely only an issue in countries without proper competition. To get away with a low cap in a competitive market you have to compete on price.

Unpopular opinion: I'd be happy to be capped at 10GB "fair usage" for a substantially reduced fee. Otherwise I'm just subsidising much heavier users.

I'd be ok with that if they would be required to charge a true fair cost... but then 10GB would probably be basically free and cable companies would not like that. They are only doing caps to increase profits, not because the network is overloaded.

Nobody should be required to charge a true fair cost. Competition should exist that makes that the workable choice.

That would be an even better solution (to stop forcing monopolies) but I'm not sure which of the two would be more likely to happen.

How are you subsidizing the heavy users when they don't cost more to the provider? Bandwidth is not a scarce resource.

My concern is that the incentives are not aligned with ever-increasing bandwidth. If my ISP can provide 100mbps and then milk that deployed infrastructure with data usage caps, they'll do it, unless competition appears.

The EPA mandates fuel economy standards for cars. Imagine if your ISP were held to a similar, ratcheting standard.

"Imagine if your regulators were not held powerless by those they supposedly regulate?"

Ok. It's hard for me to hold on to hope here.

The article strongly implies that widespread 300GB usage would not strain providers' capacity. As a consumer, I would certainly like it to be true. Is it?

Yes, for ground-based networks the capacity limitations are largely a question of expanding capacity at interchanges rather than any kind of actual network capacity.

If you are asking 'bout wireless....no, that isn't true.

For instance, TWC is quite profitable in my market with no cap and I've got a 100mbps in real-world usage most of the time.

Sort of, the capacity of fibers growth faster than the Internet usage. (But of course carriers need to upgrade their routers regularly.) Additionally most of the cost of ISP's is fixed cost, to build capacity, not variable with the total amount of data transferred. It makes sense to have a flat rate.

> Why did this happen on the mobile wireless side? Because no one stood up and complained early enough in the game. And there’s no real competition to prompt better behavior.

Its also a technology question...until the past couple of years, the caps on the mobile wireless side were needed because they simply didn't have the capacity.

> But it’s not that simple. The “cable” lane traffic (let’s call it Lane 1) includes not just traditional pay TV but also a host of data services — which today could be Stream and tomorrow could be telemedicine, business conference services, or anything else. And Lane 1 uses the internet protocol! This means that data packets are routed around independently, both upstream and downstream, and all that efficient internet magic is happening. From the user’s perspective, when he or she interacts with these things it will “feel” just like online, internet services are responding.

I think this definitely violates the spirit of net neutrality even in its current form. However, I don't see how they can move from streaming services which are sourced much further down the line than any internet traffic to dynamic traffic that needs total contact with the internet proper. From what I can tell, Comcast is arguing this service doesn't violate FCC regulations because it essentially works the same as on-demand cable. The media available for streaming is located close to the customer so it can be downloaded once from the internet proper and then streamed on demand without incurring any of the more valuable bandwidth further up the chain. I don't even think they need to prioritize the streaming traffic over regular traffic, since it is only eating into the much more plentiful local bandwidth.

If this distinction evaporates (as it would for telemedicine, business conference services, etc.), then they'd clearly violate the letter of FCC's net neutrality regulations. The fact that they used IP seems irrelevant; using anything but IP would require huge investment in custom hardware, so I don't see that as evidence of a future general internet services Master Plan (past the current streaming media Master Plan coming to fruition). To be honest, I'm somewhat on the fence about closing this loophole (though more towards the side of closing it) because it really is just like on-demand cable. It's clear that the data caps are being used in a predatory way, but I'm not sure that is the way to attack it.

I'm intrigued by the author's proposed solution, but it seems like something of a moonshot.

This is a HUGE problem. Addressing even a small part of it would net a fortune. Yes, others are already working on it. Yes, I know it will be difficult. That's what makes it exciting, right?

So, anyone want to kick around some ideas?

Barring anti-monopolist regulation or real competition, the only way to defeat these yoyos is to get serious about grassroots mesh networks that disintermediate them entirely.

Is there any serious indication a mesh-network can handle common network traffic and video?

No, not really.

The right way to solve the problem is how Chattanooga, TN and other such places have solved it: either install fiber optics and maintain and bill them as a utility, or open the existing fiber lines and maintain/bill them as a utility. [0]

This solution is so good that incumbent telcos have successfully lobbied to make it illegal in oh so many places.

A second, decent solution is how -IIRC- either the CR or Romaina did it... string up a communally-owned wired network from building to building, expanding the wired network until you get enough "customers" that you can pool the cash (and have the influence) to get a good link to the Internet.

[0] This was more-or-less what the telcos were supposed to do back in the 1990s with the tax breaks and other incentives they were given. As you can see, they failed to live up to their promises.

I think p2p streaming for chunks of a video would be interesting, and I'm sure people are working on it (and p2p in general, not just video). I saw some people talking about mesh networks in Bangkok, I'll have to investigate more to see how large it is.

Depends on the granularity and implementation. BitTorrent is a mesh network, at least logically.

Streaming video in the manner you describe is illegal unless you own the copyright or license it. ;)

But yes, you can do that if you are willing to spend $10k for a city block.

I forgot that there are countries that, unlike Australia, actually don't have quota caps on internet plans. Honestly it's not the end of the world so long as you have ISPs offering reasonable quota-exempt content (Netflix + iiNet <3).

That said I do see that huge, gaping conflict of interest between cable providers and content distribution.

Netflix streaming wouldn't exist for you to <3 if ISPs had prohibitive data caps when they were new. Do you want to freeze the internet as it is today, blocking out new entrants and innovations, so long as you get your Netflix?

The problem here is, what if you want to use a new competitor to Netflix, perhaps one with more content and a snappier UI? But you can't because they didn't pay-off your ISP? (In addition to them paying their own ISP, and you paying your ISP. And then your ISP not doing its job because no kick-backs.)

I'm actually OK with the idea of caps, as long as there is competition so they are reasonably high caps. Some overselling / bursty-ness of consumer ISPs is reasonable IMHO, and communicating that with caps is reasonably simple.

I live in Brisbane, and my current internet plan is 'unlimited' (TPG).

The biggest problem I have isn't the quota / data caps available, it's the fact I can't get any better than 2mbps upload speed, even if I switch from ADSL2+ to cable through Telstra.

So if I don't care about Netflix but want to watch e.g. a lot of twitch.tv I have to shop around for an ISP that has a deal with them? No thanks...

I don't know why you are getting down voted. I came here to make more or less the same comment: Australia has not slid of the face of the earth due to download quotas. Last month internode increased my quota from 250 to 1000 Gb at no additional cost.

Perhaps people are raging against usage caps because all hope of non-cable monopoly broadband in the US is lost.

Isn't this why people were mad about the Comcast/Netflix deal though?

Usage based paying is fine and in fact very very appropriate. It does not make sense to charge one person who uses 1000GB the same as someone who only uses 20 or 30 GB to check emails.

The real problem is that there needs to be more competition.

I agree that there needs to be more competition in the provider market, but I think the issue that the author and I have with usage-based billing is that it artificially stifles competition and innovation for a wide variety of industries.

As you say, the 20 or 30 GB user would pay less for using a reasonable amount of email-checking, occasional youtubing bandwidth. But that reasonable number is orders of magnitude higher than what the same user would have used 10 years ago. Innovation in improving user experiences has driven bandwidth usage way up, even for the unassuming customer. Artificially limiting that cap results in stifling innovation.

Why is that limit artificial? Unlike water or power, data transfer is not a reservable resource. Bandwidth is a shared resource, but unused data transfer capacity has zero value.

And as others have indicated in this thread, the largest providers are not on the losing end of the current service bargain. But they are losing their hold on unrelated services like broadcasting. Switching to a usage-based model from a rate-limited model in an industry where startup costs are enormous and providers make a healthy profit is strictly an anti-competitive strategy.

For you and I its fine, but I don't think most people have any concept of what uses their bandwidth cap. I regularly speak to people who are baffled that they're using all their mobile data allowance early in the month only to find out they do things like using YouTube videos as a music player when out.

Either data caps need to be way higher, or they need to go away. I find it bizarre that mobile operators can still get away with 0.5-1GB plans as the standard.

The battery usage tracking features built into Android have come a long way to make it easy to track down which apps are chewing through that limited resource - perhaps it is time to include the same functionality for bandwidth (technically data transfer)?

Actually, billing based on number of bytes transferred has absolutely no relation to the ISP's cost of providing the service. ISPs have a fixed amount of bandwidth to various upstream ISPs. If customers exceed that at certain times of the day (or always), it costs the ISP money to upgrade their connectivity. If the customers aren't exceeding that capacity, there's no additional incremental cost. So if all your customers check email at 9am, that could be more expensive than a bunch of people who are constantly using their entire 1Gbps to run Bittorrent.

Similarly, that's why Netflix and Youtube struggle to load their 5Mbps video streams, while Bittorrent works fine. Bittorrent will simply not get any data from peers whose routes take the packets through the congested link, but use all the others. If there's only one route to a service, that's more likely to hit congestion.

Anyway, if ISPs want to charge for bits transferred, they'd better be upgrading their peering links and charging you based on how you're increasing their costs. Otherwise they're just charging you Because They Can, which is the real reason behind data caps. (What are you going to do, switch to their competitor?)

>Usage based paying is fine

this is how you get people not getting medical treatment

Caps or pay per bit should be OK but only if the price per bit falls over time. Provider cost per bit surely is dropping. But lacking competition they may try to pocket the difference.

I once again ask - why not price internet like electricity. 5$ per TB. Pay what you consumed.

Because "internet" isn't a physical resource like electricity. Nothing is being consumed when you send 0s and 1s down the wire that's already been laid and paid for. When you load this webpage, it hasn't cost your ISP any additional money, so why do you want to pay them additional money?

ISPs' costs are a combination of fixed costs (hardware, labor, facilities, etc) and peak bandwidth capacity. They bill us in pricing that's a combination of a fixed base cost, and upgrade tiers that correspond to different peak bandwidth capacity. That's closer to "pay what you consume" in reality.

Brian Roberts https://en.m.wikipedia.org/wiki/Brian_L._Roberts

Look at him, smirking because he's got millions of people by the balls and there's nothing anybody can do about it.

He'd be one of the first people in the guillotine if there were ever a revolution. Which there won't be because he also owns a large news and media empire that keeps people pacified or angry about the wrong things.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact