(from article) Is there any hope for the likes of Tizen, Firefox OS, and Windows on phones and tablets?
How did Windows Phone get in that group. That's the 3rd largest ecosystem and growing rapidly with multiple billions behind it. It has shipped and is shipping. Unitwise it is over 1/3rd of of iOS sales. Definitely 3rd place but not marginal, or non-existant.
So you would have a bunch of VPN providers sending huge amounts of dummy traffic through Verizon's network to a fake or real address on the Level3 side (say 40 exabytes / day or whatever it would take to balance out). How is that going to be cheaper than just paying the penalty? Verizon is going to be charging those VPN companies a fortune.
2 stop off vs. one. That's twice as much of a hit to the internet. And would stop them from paying Verizon only in so far as they pick up larger VPN fees.
There exist today plenty of phones on the market that do have keyboards. Motorola droid, the Black Q10 are both examples of quite good phones with keyboards. We have sales data. Customers vote with their dollars.
. The same is true for small smart phones.
There are tons of small smartphones: Galaxy S4 Mini, HTC One mini, HTC Sense 6, Moto E, Sony Xperia Z1 Compact...
If you haven't bought a smartphone in 3.5 years you are too picky or too cheap to be worth servicing too. The last 8 years have seen massive smartphone improvements. This is what computing was like in the 1980s and early 1990s.
But they're essentially in a monopoly position in many markets and what recourse does Netflix have if Verizon decides to amp up the charges or their cloud service start sucking donkey balls?
Verizon is directly in thousands of data centers and connects to hundreds of other cloud services. Just host in any and handoff directly to Verizon with no peering at all.
They do that for cellular. They mostly are trying to avoid that for lad since except for very heavy users extra land traffic isn't a problem. For those extra heavy users because of Netflix its the same thing.
Case 1: Level3 pays Verizon for peering, level3 charges Netflix, Netflix charges customers. Money goes customers -> Verizon Case 2: Verizon charges the customers a surcharge for traffic.
Work with Google Fiber & municipalities to build hi-speed & wireless that bypasses the monopolies
If the municipalities were willing to fund internet for home / small business we could have a much more rational system in the United States where internet becomes a heavily subsidized if not free utility and so essentially 100% of the population uses it. The costs are bundled in taxes and upgrades to networks can be paid for with tax free municipal bonds. Yes that would be much better. But Netflix has nowhere near the pull to get those kinds of changes. The current group of crazies in power aren't funding infrastructure maintenance to things like dams or the national health system.
What you are suggesting with VPN wouldn't help Netflix. Level3 would just have peering problems with more carriers and moreover Netflix's traffic would have much longer latencies and would be using a lot more resources as it bounced around. The global VPN providers probably couldn't handle all of Netflix's traffic today even if they wanted to. Far cheaper for everyone would be to just host a copy of Netflix's infrastructure on Verizon's cloud and eliminate the handoff entirely for Verizon's customers, the entire transaction happens inside Verizon's network. But evidently just paying the peering is even cheaper than that.
Doubt Verizon is going to tell you the REAL cost of LTE. Doubt you have any access to real data about that either. Fess up! You pulled those numbers out of your ass.
I have access to some of their real costs. I know some of the people who've laid down the fiber they are using and I know what that cost. I know some of the places they are hosting from. I can assume what Verizon spends on their towers is commensurate with what Sprint spends (and I do know Sprint's costs). AT&T is winning business from Verizon right now, especially in embedded because their LTE costs are lower. While it might still have a probability above 0 that Verizon has covered up their real costs and is aggressively losing business to AT&T it ain't much above 0.
The more likely scenario is Verizon is telling the truth and you are just upset because everything in the world isn't free.
Yes I know how much bandwidth costs. It costs a ton to build and a ton to maintain. It is people on slashdot who think bandwidth is free because they've never had to deal with the tremendous costs in anyway. They want it to be cheap so they like to pretend it is cheap. The same as people who don't like the complexities of housing policy so want to pretend houses can be built for $1000.
Try doing even small parts of what a Verizon does and then be critical of what tremendous service they are giving you for $50 / month.
There are limits to density depending on signal strength. You want signal strength for penetration. Verizon, AT&T... are hitting those problems in cities.
There is an easier way to solve your rural problem. Have the rural government purchase the internet and make it a tax. In just about every suburb or x-burb something like $1000 / house would be plenty to get them fully wired up for 100mbit+ data connections. Have the local government pay for the wiring, or subsidize the wiring and the problem is solved. In terms of rural there may not be any economically viable way to connect them. Cellular data might provide an answer but its unclear if it does, because a lot of rural customers are also poor and mostly unwilling to spend for data.
Which BTW since you like the analogy of electricity, having the governments massively subsidize the cost (effectively buying it for the population) is precisely how the country ended up getting wired. Starting with local governments, then states then federal as the areas got less and less dense.
the most data you can buy is 100 GB per month
Nonsense. You can buy 100t blocks from Verizon to allocate among phones. The most you can buy on a consumer grade plan is 100g to allocate between phones that's very very different. Once you start using that kind of large scale data they want you buying through an agent.
It's really not that hard -- it's been done in the past for other types of service.
It is that hard. Lots of services aren't economical to provide anywhere or only in very concentrated areas. There are all sorts of services one can buy in New York that aren't even available in other large cities. There are all sorts of services one can buy in cities that aren't available in suburbs... There is nothing unusual about internet. The size of the customer base / density matters, it matters a lot.
. But it's way too much to be paying $20 or $30 or more for that kind of download. Especially when, most times, the *content itself* is also very expensive.
Then mail DVDs back and forth. Netflix still offers that service. That's the way people did it up until a few years ago. That's cheaper bandwidth.
As loudly as they will cry crocodile tears at the prospect, we simply have to regulate them, and yes, cut down their profit-making potential a bit, in order to usher in an enormous boost to the rest of the economy. Content creators, advertisers, online goods retailers like Amazon and Valve, etc. are chomping at the bit to sell reasonably-priced goods to consumers that are chomping at the bit to purchase them. But the carriers are acting as a bottleneck, preventing this potential from being realized,
That's true of all sorts of corollary services for goods. If everyone had a 5000 sq foot home: gardening services, furniture companies, carpet cleaners, car dealers... would make a fortune. Its only the fact that there is a price difference between what it costs to sell 5000sq of home vs 800 sq feet that prevents all that economic gain. If there really were that much economic gain to be had then internet would be provided via. a system where the people selling the goods would be the ones paying Verizon to get access to customers and problem solved.
____
As for the comments about 5g. You are missing the point about the caps. You are also missing the actual financials. A gig of data was often around several hundred dollars to buy. The caps were essentially unlimited based on the phones not really offering services that made large data usage appealing. Even the customers on dataplans were mostly doing stuff like compressed email or compressed text webpages. The usage has entirely shifted. Today customers on a 5gig cap are likely to be not be using.1%-1% of that cap but more like 25-90% of that cap i.e. 100x as much data for the same price.
Certainly Verizon could meter data but retail customers hate metering. So what you get are metered plans with the metering obscured partially. You can't really talk about these plans as if the obscured metering was the real price.
So Verizon made a bet that customers wouldn't use the unlimited data that they sold them and they lost. Tough!
Not tough. They never sold them performance guarantees or dedicated bandwidth. The agreement allows for throttling. So tought on both sides.
It looks like Verizon should start offering plans that reflect the actual cost to supply. Those that use the most pay the most.
They do offer such plans. They don't offer it at residential levels since they don't want to be explaining it to 30m customers. Rather they expect the Netflixs of the world to understand and handle it on their side.
I would like to see the last mile become a common good but other changes would be needed to ensure fairness.
You aren't asking for a free market, quite the opposite you are asking for last mile to be a highly regulated utility and/or socialized.
Colin Nederkoom used a VPN link to show that he could access Netflix at almost 10x the paltry 375 kbps he was getting on his Verizon 75 Mbps symmetric service.
That's fine. He isn't going to have any affect on the peering. Once tens of millions of people are doing that regularly it does affect peering. What works for an individual doesn't work collectively.
The underlying issue is that Verizon is muddying the waters by being both a residential ISP AND a Tier 1 provider.
Not really. Assume there were 3 companies:
V1 = a residential provider only (replacing Verizon). V2 = Tier1 provider for V1 Level3 = provider for Netflix
The Netflix usage by V1's customers would still create peering problems between V2 and V1.
I don't know how it would EVER be possible to have nearly symmetric traffic with them when their small-customer base is in the millions as end-user traffic is HIGHLY asymmetric.
Not all. For example dropbox goes the other way. But mostly I =agree. Anyone providing traffic that is mostly designed for end user base shouldn't be on peering. Level3 is the first company to be hitting this because Netflix is too big to ignore but mostly it should be paid peering, Level3 should be passing that cost on to Netflix and Netflix passing that cost on to their paid customer base.
There's a reason why i mentioned VyprVPN apart from them being the service used by Nederkoom to bypass Verizon's throttling - they have an online storage service as well so if customer's take advantage of this, it might help balance the traffic if Netflix acquires them.
Netflix doesn't have to do that sneaky stuff. Verizon sells a cloud solution that is excellent. Netflix could just host out of the Verizon cloud to serve Verizon customers . Then they never have a peering problem at all since the entire transaction takes place within Verizon. Or they could host at a colo along one of Verizon's main pipes. That's much easier.
But there's a much simpler solution that Netflix could implement - a bytecounter & traffic gen plugin that would send random data back when the user is connected to the streaming service and if they implement it unencrypted, they could have Level3 discard it silently as soon as it hits their network from the Verizon side.
That would solve the peering problem but it would double Netflix's data usage. It would be far cheaper to pay for peering.
The customers have not paid nearly what their usage is worth. If the customers were paying commercial ethernet prices for their 20 mbit - 100+mbit plans Verizon would be thrilled with the Netflex consuming as much bandwidth as possible. The consumer pricing is based on most consumers using only a fraction of their possible bandwidth because they use it sporadically and most servers can't fill the allotments. Multiple hours of video changes that equation.
Verizon could equally well charge customers a premium for using Netflix on their network, i.e. the opposite of net neutrality.
The subsidy isn't nearly that high, more like $17-22 month: $350-450 over 20-24 months. And they aren't paying for what they are using. Verizons price for pay as you go is $15 / gig. The $5 / gig price assumes you go under. That allows retail to average about $8 / gig for gigs used.
Those people on unlimited plans should be switching. Verizon probably should make switching to a plan commensurate with their current cost structure a no brainer.
Interesting question is if Verizon believes that using more than 4.7GB per month is unreasonably high, then why does Verizon offer individual data plan buckets of 6GB and even 8GB?
They offer more than that. You can get 20g+ plans. You just pay $5 a gig (contract) or $15 / gig (pay as you go) for those plans. Those customers are paying for their usage.
Yes I agree that would be the most reasonable thing to do. Or they could put them on some sort of prepay monthly plan and just eat the cost of the first month. Verizon should be moving those customers off those plans. For the vast majority of customers on their unlimited plans they probably have dumb phones or feature phones that only use EVDO data and their "unlimited" data is around 2m / month. They aren't the problem. The people streaming movies are. They probably should be forced to switch.
That's like saying stealing oranges from the grocery store isn't really theft because the tree, the farm, the delivery trucks, the warehouse and the rent on the store were already paid for. The stuff sold by the store, the oranges, are what pay for the tree, the farm, the warehouse... That's where the money comes from.
Pretend for a moment that Verizon retail and Verizon wholesale were two separate companies. Verizon retail acts like an MVNO and buys up chunks of minutes (prime, weekend, nights), SMS, data....
Case 1: They then mark those minutes, SMS, and data up and sell them to you having you pay a fixed fee for each minute of phone, each SMS and each k of data. OK that's a fair system for everyone.
Case 2: They notice that most of their customers enjoy thinking of their minutes, SMS... as "free" and like to relax on the network. So instead of of charging you for the minutes and SMS, k of data they just sell you a large block that's capped at below their cost that's more than you would use. You buy a bundle. If you actually use up the bundle they lose money on you as a customer that month. This is a bit more complex.
Case 3: For the vast majority of customers the bundle limits are so much higher than their usage so the best way is to sell them unlimited on the cheap stuff: minutes and SMS while selling them bundles for the expensive stuff (k of data). This is basically Verizon's current plans.
Case 4: There are some customers who managed to find a loophole where they have unlimited for the expensive stuff and pay for the cheap stuff. On those customers Verizon retail always loses money. That money comes from their other retail customers.
_____
That's pretty close though oversimplified to the reality of Verizon's internal ERP
No it isn't. If I ship 1000 packages a day using UPS and they handoff to the postoffice for local delivery on 500 of those packages that's a nice deal. If I then go up to 1100 packages and pay the postoffice directly for the extra 100 that doesn't change the situation on the first 500.
Deal 1: Netflix pays Verizon to carry a small percentage of their traffic directly Deal 2: Level3 pays Verizon extra to carry the bulk of the traffic.
It is a free market. The home delivery part is of course a regulated utility but a free market for your town. But mostly if you are consuming tremendous amounts of Netflix you probably are a customer that's a net loss for Verizon. I'm sure if they were tracking fully loaded cost they would be happy if you cancelled.
It's well within Level3's capability to spread the load among all the peers
Of course it is. But how does that help? Most of the smaller players would have to buy more bandwidth themselves to carry any substantial portion of Netflix's traffic. They don't have the capacity to accept most of the traffic at the handoff in Oregon. So Level3 would end up having to shift traffic all over the USA themselves, handoff all over the place and then have those smaller players handoff to Verizon. The performance would be both unpredictable and quite bad for Netflix customers. And in the end, its still going to have to get handed back to Verizon for last mile, so all these small providers are going to be asymmetric.
Do owners of popular venues pay to upgrade the interstate 50 miles out of town?
Our road system is maintained by the government to serve the common good. The internet is much more of a free market. That being said, in areas where there is very little justification for a road and a business that wants one, they often do pay to get their access to the interstate upgraded.
By the way, Netflix HAS ALREADY PAID both Comcast & Verizon - http://www.theverge.com/2014/4... [theverge.com] That was months ago. So the issue now is between Verizon & Level3.
AFAIK that's two different agreements. Netflix is having Verizon carry some of their traffic directly. Then there isn't peering obviously and Verizon doesn't care if the traffic is asymmetric. In theory if Netflix bought enough Verizon to cover delivery to all Verizon customers that would solve the Level3 / Verizon problem as there wouldn't be any handoff. But they aren't doing it they are only doing it for some of the traffic. Probably places where the Verizon / Level3 handoffs are flooded to get those customers good service. The issue still remains because Level3 still carries most of Netflix's traffic.
Is Netflix going to have to pay extortion money to EVERY major provider in the world if their traffic causes asymmetric bandwidth between peers?
I wouldn't call it extortion, I'd call it paying for services used. And the answer is likely yes. Though the more reasonable thing is that it get wrapped in the Level3 bill. If Netflix wants to generate huge traffic to require expensive traffic upgrades their customers should be the ones to pay for those upgrades.
If it were up to me, I would tell Comcast & Verizon to go fuck themselves and start buying up every multihomed / distributed VPN service I could find and sell that as a premium service with some goodies to sweeten the pot for my customers.
How would that help? If a packet is going to get from Level3 Oregon to say a Verizon home in Philadelphia it at some point is going to have to hit a peering location between Level3 and Verizon. Unless you add a bunch of other 3rd parties in there. In which case it is still going to have to happen indirectly. What is a VPN going to do to fix that?
Slashdot Mirror
How did Windows Phone get in that group. That's the 3rd largest ecosystem and growing rapidly with multiple billions behind it. It has shipped and is shipping. Unitwise it is over 1/3rd of of iOS sales. Definitely 3rd place but not marginal, or non-existant.
So you would have a bunch of VPN providers sending huge amounts of dummy traffic through Verizon's network to a fake or real address on the Level3 side (say 40 exabytes / day or whatever it would take to balance out). How is that going to be cheaper than just paying the penalty? Verizon is going to be charging those VPN companies a fortune.
Netflix -> X, X-> Verizon -> Customer vs.
Netflix -> Verizon -> Customer
2 stop off vs. one. That's twice as much of a hit to the internet. And would stop them from paying Verizon only in so far as they pick up larger VPN fees.
Blackberry is designed for with messaging in mind.
There exist today plenty of phones on the market that do have keyboards. Motorola droid, the Black Q10 are both examples of quite good phones with keyboards. We have sales data. Customers vote with their dollars.
There are tons of small smartphones: Galaxy S4 Mini, HTC One mini, HTC Sense 6, Moto E, Sony Xperia Z1 Compact...
If you haven't bought a smartphone in 3.5 years you are too picky or too cheap to be worth servicing too. The last 8 years have seen massive smartphone improvements. This is what computing was like in the 1980s and early 1990s.
Verizon is directly in thousands of data centers and connects to hundreds of other cloud services. Just host in any and handoff directly to Verizon with no peering at all.
They do that for cellular. They mostly are trying to avoid that for lad since except for very heavy users extra land traffic isn't a problem. For those extra heavy users because of Netflix its the same thing.
Case 1: Level3 pays Verizon for peering, level3 charges Netflix, Netflix charges customers. Money goes customers -> Verizon
Case 2: Verizon charges the customers a surcharge for traffic.
If the municipalities were willing to fund internet for home / small business we could have a much more rational system in the United States where internet becomes a heavily subsidized if not free utility and so essentially 100% of the population uses it. The costs are bundled in taxes and upgrades to networks can be paid for with tax free municipal bonds. Yes that would be much better. But Netflix has nowhere near the pull to get those kinds of changes. The current group of crazies in power aren't funding infrastructure maintenance to things like dams or the national health system.
What you are suggesting with VPN wouldn't help Netflix. Level3 would just have peering problems with more carriers and moreover Netflix's traffic would have much longer latencies and would be using a lot more resources as it bounced around. The global VPN providers probably couldn't handle all of Netflix's traffic today even if they wanted to. Far cheaper for everyone would be to just host a copy of Netflix's infrastructure on Verizon's cloud and eliminate the handoff entirely for Verizon's customers, the entire transaction happens inside Verizon's network. But evidently just paying the peering is even cheaper than that.
I have access to some of their real costs. I know some of the people who've laid down the fiber they are using and I know what that cost. I know some of the places they are hosting from. I can assume what Verizon spends on their towers is commensurate with what Sprint spends (and I do know Sprint's costs). AT&T is winning business from Verizon right now, especially in embedded because their LTE costs are lower. While it might still have a probability above 0 that Verizon has covered up their real costs and is aggressively losing business to AT&T it ain't much above 0.
The more likely scenario is Verizon is telling the truth and you are just upset because everything in the world isn't free.
Yes I know how much bandwidth costs. It costs a ton to build and a ton to maintain. It is people on slashdot who think bandwidth is free because they've never had to deal with the tremendous costs in anyway. They want it to be cheap so they like to pretend it is cheap. The same as people who don't like the complexities of housing policy so want to pretend houses can be built for $1000.
Try doing even small parts of what a Verizon does and then be critical of what tremendous service they are giving you for $50 / month.
The contract has you reverting to month to month. They can terminate.
Fine prioritizing traffic paying lots for traffic paying very little.
There are limits to density depending on signal strength. You want signal strength for penetration. Verizon, AT&T... are hitting those problems in cities.
There is an easier way to solve your rural problem. Have the rural government purchase the internet and make it a tax. In just about every suburb or x-burb something like $1000 / house would be plenty to get them fully wired up for 100mbit+ data connections. Have the local government pay for the wiring, or subsidize the wiring and the problem is solved. In terms of rural there may not be any economically viable way to connect them. Cellular data might provide an answer but its unclear if it does, because a lot of rural customers are also poor and mostly unwilling to spend for data.
Which BTW since you like the analogy of electricity, having the governments massively subsidize the cost (effectively buying it for the population) is precisely how the country ended up getting wired. Starting with local governments, then states then federal as the areas got less and less dense.
Nonsense. You can buy 100t blocks from Verizon to allocate among phones. The most you can buy on a consumer grade plan is 100g to allocate between phones that's very very different. Once you start using that kind of large scale data they want you buying through an agent.
It is that hard. Lots of services aren't economical to provide anywhere or only in very concentrated areas. There are all sorts of services one can buy in New York that aren't even available in other large cities. There are all sorts of services one can buy in cities that aren't available in suburbs... There is nothing unusual about internet. The size of the customer base / density matters, it matters a lot.
Then mail DVDs back and forth. Netflix still offers that service. That's the way people did it up until a few years ago. That's cheaper bandwidth.
That's true of all sorts of corollary services for goods. If everyone had a 5000 sq foot home: gardening services, furniture companies, carpet cleaners, car dealers... would make a fortune. Its only the fact that there is a price difference between what it costs to sell 5000sq of home vs 800 sq feet that prevents all that economic gain. If there really were that much economic gain to be had then internet would be provided via. a system where the people selling the goods would be the ones paying Verizon to get access to customers and problem solved.
____
As for the comments about 5g. You are missing the point about the caps. You are also missing the actual financials. A gig of data was often around several hundred dollars to buy. The caps were essentially unlimited based on the phones not really offering services that made large data usage appealing. Even the customers on dataplans were mostly doing stuff like compressed email or compressed text webpages. The usage has entirely shifted. Today customers on a 5gig cap are likely to be not be using .1%-1% of that cap but more like 25-90% of that cap i.e. 100x as much data for the same price.
Certainly Verizon could meter data but retail customers hate metering. So what you get are metered plans with the metering obscured partially. You can't really talk about these plans as if the obscured metering was the real price.
Not tough. They never sold them performance guarantees or dedicated bandwidth. The agreement allows for throttling. So tought on both sides.
They do offer such plans. They don't offer it at residential levels since they don't want to be explaining it to 30m customers. Rather they expect the Netflixs of the world to understand and handle it on their side.
You aren't asking for a free market, quite the opposite you are asking for last mile to be a highly regulated utility and/or socialized.
That's fine. He isn't going to have any affect on the peering. Once tens of millions of people are doing that regularly it does affect peering. What works for an individual doesn't work collectively.
Not really. Assume there were 3 companies:
V1 = a residential provider only (replacing Verizon).
V2 = Tier1 provider for V1
Level3 = provider for Netflix
The Netflix usage by V1's customers would still create peering problems between V2 and V1.
Not all. For example dropbox goes the other way. But mostly I =agree. Anyone providing traffic that is mostly designed for end user base shouldn't be on peering. Level3 is the first company to be hitting this because Netflix is too big to ignore but mostly it should be paid peering, Level3 should be passing that cost on to Netflix and Netflix passing that cost on to their paid customer base.
Netflix doesn't have to do that sneaky stuff. Verizon sells a cloud solution that is excellent. Netflix could just host out of the Verizon cloud to serve Verizon customers . Then they never have a peering problem at all since the entire transaction takes place within Verizon. Or they could host at a colo along one of Verizon's main pipes. That's much easier.
That would solve the peering problem but it would double Netflix's data usage. It would be far cheaper to pay for peering.
That's opex it doesn't count capex.
The customers have not paid nearly what their usage is worth. If the customers were paying commercial ethernet prices for their 20 mbit - 100+mbit plans Verizon would be thrilled with the Netflex consuming as much bandwidth as possible. The consumer pricing is based on most consumers using only a fraction of their possible bandwidth because they use it sporadically and most servers can't fill the allotments. Multiple hours of video changes that equation.
Verizon could equally well charge customers a premium for using Netflix on their network, i.e. the opposite of net neutrality.
The subsidy isn't nearly that high, more like $17-22 month: $350-450 over 20-24 months. And they aren't paying for what they are using. Verizons price for pay as you go is $15 / gig. The $5 / gig price assumes you go under. That allows retail to average about $8 / gig for gigs used.
Those people on unlimited plans should be switching. Verizon probably should make switching to a plan commensurate with their current cost structure a no brainer.
They offer more than that. You can get 20g+ plans. You just pay $5 a gig (contract) or $15 / gig (pay as you go) for those plans. Those customers are paying for their usage.
Yes I agree that would be the most reasonable thing to do. Or they could put them on some sort of prepay monthly plan and just eat the cost of the first month. Verizon should be moving those customers off those plans. For the vast majority of customers on their unlimited plans they probably have dumb phones or feature phones that only use EVDO data and their "unlimited" data is around 2m / month. They aren't the problem. The people streaming movies are. They probably should be forced to switch.
That's like saying stealing oranges from the grocery store isn't really theft because the tree, the farm, the delivery trucks, the warehouse and the rent on the store were already paid for. The stuff sold by the store, the oranges, are what pay for the tree, the farm, the warehouse ... That's where the money comes from.
Pretend for a moment that Verizon retail and Verizon wholesale were two separate companies. Verizon retail acts like an MVNO and buys up chunks of minutes (prime, weekend, nights), SMS, data....
Case 1: They then mark those minutes, SMS, and data up and sell them to you having you pay a fixed fee for each minute of phone, each SMS and each k of data. OK that's a fair system for everyone.
Case 2: They notice that most of their customers enjoy thinking of their minutes, SMS... as "free" and like to relax on the network. So instead of of charging you for the minutes and SMS, k of data they just sell you a large block that's capped at below their cost that's more than you would use. You buy a bundle. If you actually use up the bundle they lose money on you as a customer that month. This is a bit more complex.
Case 3: For the vast majority of customers the bundle limits are so much higher than their usage so the best way is to sell them unlimited on the cheap stuff: minutes and SMS while selling them bundles for the expensive stuff (k of data). This is basically Verizon's current plans.
Case 4: There are some customers who managed to find a loophole where they have unlimited for the expensive stuff and pay for the cheap stuff. On those customers Verizon retail always loses money. That money comes from their other retail customers.
_____
That's pretty close though oversimplified to the reality of Verizon's internal ERP
No it isn't. If I ship 1000 packages a day using UPS and they handoff to the postoffice for local delivery on 500 of those packages that's a nice deal. If I then go up to 1100 packages and pay the postoffice directly for the extra 100 that doesn't change the situation on the first 500.
Deal 1: Netflix pays Verizon to carry a small percentage of their traffic directly
Deal 2: Level3 pays Verizon extra to carry the bulk of the traffic.
It is a free market. The home delivery part is of course a regulated utility but a free market for your town. But mostly if you are consuming tremendous amounts of Netflix you probably are a customer that's a net loss for Verizon. I'm sure if they were tracking fully loaded cost they would be happy if you cancelled.
Of course it is. But how does that help? Most of the smaller players would have to buy more bandwidth themselves to carry any substantial portion of Netflix's traffic. They don't have the capacity to accept most of the traffic at the handoff in Oregon. So Level3 would end up having to shift traffic all over the USA themselves, handoff all over the place and then have those smaller players handoff to Verizon. The performance would be both unpredictable and quite bad for Netflix customers. And in the end, its still going to have to get handed back to Verizon for last mile, so all these small providers are going to be asymmetric.
Our road system is maintained by the government to serve the common good. The internet is much more of a free market. That being said, in areas where there is very little justification for a road and a business that wants one, they often do pay to get their access to the interstate upgraded.
AFAIK that's two different agreements. Netflix is having Verizon carry some of their traffic directly. Then there isn't peering obviously and Verizon doesn't care if the traffic is asymmetric. In theory if Netflix bought enough Verizon to cover delivery to all Verizon customers that would solve the Level3 / Verizon problem as there wouldn't be any handoff. But they aren't doing it they are only doing it for some of the traffic. Probably places where the Verizon / Level3 handoffs are flooded to get those customers good service. The issue still remains because Level3 still carries most of Netflix's traffic.
I wouldn't call it extortion, I'd call it paying for services used. And the answer is likely yes. Though the more reasonable thing is that it get wrapped in the Level3 bill. If Netflix wants to generate huge traffic to require expensive traffic upgrades their customers should be the ones to pay for those upgrades.
How would that help? If a packet is going to get from Level3 Oregon to say a Verizon home in Philadelphia it at some point is going to have to hit a peering location between Level3 and Verizon. Unless you add a bunch of other 3rd parties in there. In which case it is still going to have to happen indirectly. What is a VPN going to do to fix that?