Why is that bad? I am not a Netflix customer. Why should I subsidize my neighbors who all want to watch Game of Thrones at the same time?
Why is that bad? I am not a Netflix customer. Why should I subsidize my neighbors who all want to watch Game of Thrones at the same time?
John Levine wrote: :>Netflix is already paying various ISPs to ensure that its customers :>get a decent streaming experience, including Verizon, Time Warner :>Cable, Comcast and AT&T, but the company has voiced its concerns more :>than once regarding these payments, urging the FCC to look into these :>agreements.
:Why is that bad? I am not a Netflix customer. Why should I subsidize :my neighbors who all want to watch Game of Thrones at the same time?
You already do. Verizon, Comcast, et al. have the internal network capacity to deliver the content to their customers. What they lack is the interconnect with the networks of content provider. They simply want to be paid twice for doing one job. They want their customers to pay them for being able to use the internet, and then they want content providers to pay them for being able to deliver that content to ISP's customers. Nice work, if you can get it, but not exactly honest.
While Verizon FiOS has rather high capacity, that is the exception that proves the rule, not the general case. Internet services were not built to replace Cable TV. High-quality video takes orders of magnitude more capacity than most other Internet applications. So expanding the network to support them can be a big deal, not just a bigger cord at the interconnection point.
This is the "but the network is there, my Blue Box call didn't cost them anything" argument. The short-run incremental cost of any one unit of demand tends to be zero, unless it hits the next incremental point and becomes very high. Hence costing and pricing tend to smooth out the curves and use averages.
Netflix is attempting to game the system. If you or I called up Verizon and demanded a 10 Gbps free service to your house because you're the coolest kid on the block and everybody else wants to connect to you, they'd laugh you out. You want service, you pay. Netflix doesn't want to pay for their connections. It's that simple. They are abusing the Internet and bleating about "neutrality" to try to destroy it in order to make it into their private Cable TV domain, regulated for their benefit.
IMNSHO, the general case is that home users have high-speed connections that can handle the load of a single video stream. If Verizon expected its customers to use only a tiny fraction of that capacity during a time when new services, ideas, and methods were popping up every hour, then I think the company deserves to suffer: Verizon is, after all, in the business of predicting consumer and business demands for its "shared use" products, and has over a century of experience in finding ways to meet those demands. Remember that Bell Labs scientists *invented* many of the statistical models which are used to predict the behavior of emerging markets!
Leaving aside the definition of "high quality video", I'll point to youtube, to facepage, and to all the other online diary sites which welcome images, plus iTunes, Amazon, etc. How come *that* demand (which includes video transfers) can be met, but not NetFlix's load? Is it because Apple paid up? Youtube? Facepage? Amazon? Somehow, I doubt it: Netflix is small enough to pick on, and therefore we're witnessing a corporate-sponsored shadow play.
I disagree with your premise: Blue Box calls not only didn't cost "Ma Bell" anything, they *made* money for her! After all, the method used was to dial an 800 number and then trick the network into rerouting the call - but when the diverted connection was completed, the company that had the 800 number *WAS* *CHARGED* *FOR* *THE* *CALL*!
In other words, Blue Boxes always caused harm to innocent third parties - the 800 number owners - and their use was, for that reason, an inherently criminal act. The "short-run" cost was borne by corporate entities, at the current tariffed rates for "800" calls, so it was *NEVER* zero.
That might seem a small matter, unless we consider that the underlying method did *not* create an incentive for Ma Bell to fight Blue Box fraud: she was, after all, getting paid! In that respect, the pheepers were acting as employees of the phone companies, since they generated revenue for those firms.
In like manner, there is a division of interest between Verizon's customers and the corporation vis-a-vis "High speed Internet" services, since Verizon seems to resent actually being asked to serve a "Sunday afternoon load" when it priced its services at fixed rates to attract consumer-grade customers. I think the capacity *is* there.
Verizon's claims boil down to "We didn't see this coming", and to wishful thinking about the good-ol-days when every regulatory agency rubber-stamped every rate increase. Those days are gone, and Verizon's corporate culture can't adapt to a world where there are other providers and other competitors. It's not the "blue box" argument - it's the "We're the /phone/ /company/!" claim.
Sorry, and no offense, but I feel you're being overly simplistic. Netflix is relying on its customers to pay for the transport layer, it's true - but *every* Internet user pays for the transport layer, and Verizon is trying to have their cake and eat it too, by crying "No fair!" when one service doesn't cut them a check for the privilege of being successful. Verizon is already getting paid, AFAIK, and just because the corporation doesn't want to adapt doesn't mean I'm willing to foot the bill.
Verizon wants to live in an "as if" world: a world where every data transfer required a separate pair of wires, and those with more data to move *had* to pay more. Verizon is acting as if it's entitled to deny the changes the Internet has brought to the telecommunications business, and in addition, to act as if it is entitled to change tolls for a bridge that was designed at taxpayer expense, built with private funds, and is being lawfully operated (with occasional backups, I admit). Verizon isn't willing to add any extra lanes, only to demand that motorists pay them extra because they /used/ to run the only ferry.
Fred Goldstein wrote: :On 12/29/2014 12:26 PM, David Scheidt wrote: :> John Levine wrote: :> :>Netflix is already paying various ISPs to ensure that its customers :> :>get a decent streaming experience, including Verizon, Time Warner :> :>Cable, Comcast and AT&T, but the company has voiced its concerns more :> :>than once regarding these payments, urging the FCC to look into these :> :>agreements. :> :> :Why is that bad? I am not a Netflix customer. Why should I subsidize :> :my neighbors who all want to watch Game of Thrones at the same time? :> :> You already do. Verizon, Comcast, et al. have the internal network :> capacity to deliver the content to their customers. What they lack is :> the interconnect with the networks of content provider. They simply :> want to be paid twice for doing one job. They want their customers to :> pay them for being able to use the internet, and then they want :> content providers to pay them for being able to deliver that content :> to ISP's customers. Nice work, if you can get it, but not exactly :> honest. :>
:While Verizon FiOS has rather high capacity, that is the exception that :proves the rule, not the general case. Internet services were not built :to replace Cable TV. High-quality video takes orders of magnitude more :capacity than most other Internet applications. So expanding the :network to support them can be a big deal, not just a bigger cord at the :interconnection point.
Verizon, comcast, et al are being paid to provide internet service to their customers, by their customers. If the service they claim to be providing isn't what they're providing, they should either change their claim, or change what they provide. The facts are that all these networks have serious capacity in their cores, and aren't overloaded because of streaming video, and most of them have plenty of capacity to their customers. (Comcast, in particular, has so much excess capacity that they're trying to turn their customers' [homes] into wifi access points [which] Comcast [uses to] collect money from [other] users of the wifi networks.)
:This is the "but the network is there, my Blue Box call didn't cost them :anything" argument. The short-run incremental cost of any one unit of :demand tends to be zero, unless it hits the next incremental point and :becomes very high. Hence costing and pricing tend to smooth out the :curves and use averages.
Yes? What's your point? Comcast, Verizon, et al are selling a service. They're happy to provide that service as long as you use the services thye think you should, or at least not those they think you shouldn't. Their pricing should include the per customer capital costs of providing that service.
:Netflix is attempting to game the system. If you or I called up Verizon :and demanded a 10 Gbps free service to your house because you're the :coolest kid on the block and everybody else wants to connect to you, :they'd laugh you out. You want service, you pay. Netflix doesn't want :to pay for their connections. It's that simple. They are abusing the :Internet and bleating about "neutrality" to try to destroy it in order :to make it into their private Cable TV domain, regulated for their benefit.
Nonsense. Netflix is prefectly willing to pay for connections. What they don't want to do is pay for something that the end customer is*ALREADY PAYING FOR*, which is the delivery of bits to their home.
Verizon, Comcast aren't running networks in the usual sense of internet networks, where the traffic into the network is roughly that of traffic that leaves the network. They're net traffic sinks, because their customers are consumers, not content providers. They want to be paid twice to provide internet service to their customers: once by their customers, and once by netflix. Their customers pay for access to the internet, verizon, comcast, et al should provide that access. Their customers are not paying for access to those sites and services which comcast, verizon et al find most profitable for comcast, verizon et al.
I'm paying Comcast about the same amount now for 25Mbps as I was paying Continental Cable when I first got cable Internet about 15 years ago. I know that much of that growth was almost free, due to Moore's Law and other technological improvements. But there's enormous capital costs to replace infrastructure with more efficient versions, and Moore's Law doesn't apply to human resources (except insofar as employees are replaced by computers and robots).
Whether they "saw it coming" is not very relevant if their revenue sources don't permit accomodating the growth in demand. The reason we're even having this debate is because consumers are dead-set against a "pay as you go" pricing scheme for Internet services. They've become accustomed to flat-rate pricing, and that results in a "tragedy of the commons" -- consumers will use as much bandwidth as they can, and providers will have to increase capacity, with no corresponding increase in revenue to pay for it.
It only makes sense to look to the other end of the connection for that revenue.
A general case is one that encompasses almost everyone, not the single largest case. I am thus looking at the broader ISP market, including the wireless guys. I'm currently looking at the options for bringing broadband access to the remote places in Massachusetts that still don't have it. Stringing fiber to the forest would cost a lot. Wireless might be cheaper, but its capacity is much much lower. A TV White Space channel can carry 12 Mbps. As a shared medium serving multiple users, how much TV viewing can it handle? Now, do we say "sorry, either it's Netflix or you send letters", or do we allow a general case where ISPs just might not be able to become de facto cable distributors?
Verizon's FiOS has plenty of bandwidth to the subscriber. Of course it bottlenecks upstream. Internet demand does not work like PSTN demand. A (wireline) call's a call, 64 kbps period. Hence the statistical models work.
Internet apps keep expanding to fill available capacity. TCP in fact is designed that way, to be predatory, seeking out the bottleneck capacity. So you can't prevent loss; expanding supply necessarily expands demand. Nor do the statistical models work, because it's a truly cockeyed (that's a euphemism based on its first two syllables' Indo-European homonym) feedback loop. So they might try to stay ahead but then Netflix jumps in and pushes the curve an order of magnitude or so and demands that others pay for it.
Yes, they do mostly pay. They do it via the existing peering mechanism, which is a mutual exchange of value. Those other sites tend to create much higher value *per bit*, because their "eyeballs" use fewer bits/minute. Even YouTube, which is mostly small Flash videos viewed on a computer, not filling a big screen TV like Netflix. Still images are much smaller in load.
In that case the owner of the 800 number was being ripped off. So it did cost someone something. Not Ma Bell necessarily, if nobody caught the error, but it wasn't literally costless.
And that again is the problem with the Netflix model. The Internet is not the PSTN! On the PSTN, carriers pay or collect a rate per minute based upon the (hideously complex, the CABS mess) classification assigned to that call and their role in carrying part of it. Nobody argues over the value of the call, just over whether the price is "just and reasonable", which comes up when the FCC rejiggers the tariffs. On the Internet, there is no legal distinction between an ISP and a customer. (What's a university?) So it's all a matter of perceived value. If I as an ISP see value in more connectivity to you, I offer to peer. If I as an ISP see more cost than value in connectivity, I offer to sell you service (money equalizes the value). Netflix is trying to get out of paying, by demanding that the Internet become regulated like the PSTN and that they be given a special favored role. It is the very worst type of regulatory gaming, not just trying to get ahead but doing it by blowing up a working, critical, but very fragile, system.
You can think what you want, and there may be capacity there, but ISP prices aren't regulated, so if you impose more costs on the industry, prices are bound to rise.
The ISP side has never been regulated; it's Netflix who's crying for rate regulation. Verizon is not my favorite company, to put it mildly, but like an attorney defending an ax murderer because everyone is entitled to due process, I'll defend them for the sake of everyone else.
Internet users pay for peak-rate transfer. Cable modems are more common than FTTH, and cable modems have shared capacity, not dedicated. So what your neighbors do impacts your service. The only reason residential rates per BPS are a tiny fraction of commercial rates (price out a 100 Mbps business line lately?) is that they are assumed to have a very low *average* use, compared to business lines that are assumed to have many users already stat-muxed onto it, and thus maintain a higher average. Netflix blows that model.
Verizon's *wire* rates should be regulated, as they are monopoly owners of local loop plant. The present vertically-integrated model is broken and cannot be fixed, though fixing it is unthinkable in captured Washington. Nonetheless, I am distinguishing between the lower-layer telecom portion of Verizon's business and the Internet itself. Netflix is trying to blow up the unregulated higher-layer Internet for its advantage. You are invoking fear and loathing of Verizon based on everything else they do, which is understandable. But I'm defending the ax murderer here because he had an alibi in this armed robbery, and just being a usual suspect is not the same as being guilty.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.