The Looming Cable Monopoly
January 4, 2011 10:41 AM Subscribe
Law professor Susan Crawford takes a moment to explain to all of us why we should be wary of Verizon's decision to suspend FiOS rollout across the country and the resulting likely domination of the high-speed internet access biz by the cable companies in a short (for a legal journal) paper in the Yale Law and Policy Review.
If you're intimidated by any combination of the words "Yale", "Law", and "Policy" in the same URL, Ars Technica has the rundown in a moderately easier to access blog post.
If you're intimidated by any combination of the words "Yale", "Law", and "Policy" in the same URL, Ars Technica has the rundown in a moderately easier to access blog post.
I doubt Verizon was gambling on doing away with Net Neutrality to implement FiOS networks to begin with. This smells like a gambit to me. They can justify stalling the implementation by calling it cost cutting to the shareholders, they can blame the FCC for making it financially infeasible although it is certainly not infeasible.
Now our Galtian Overlords can blame Obama, write new internet access rules, and our internet bills will skyrocket. The good news is that you won't be able to get to Wikileaks on the New Internet anymore.
posted by Xoebe at 10:59 AM on January 4, 2011 [2 favorites]
Now our Galtian Overlords can blame Obama, write new internet access rules, and our internet bills will skyrocket. The good news is that you won't be able to get to Wikileaks on the New Internet anymore.
posted by Xoebe at 10:59 AM on January 4, 2011 [2 favorites]
...though it (Verizon) will continue to build-out where it had previously announced service (Washington, D.C., New York City and Philadelphia, namely).
Phew. I might actually have an alternative to Comcast? In the fullness of time.
posted by fixedgear at 11:02 AM on January 4, 2011
Phew. I might actually have an alternative to Comcast? In the fullness of time.
posted by fixedgear at 11:02 AM on January 4, 2011
Cable vs RBOC is just a fascinating example of the prisoners dilemma played out in real life. With marginal cost of service approaching zero, and very low incremental capital costs any competition is going to be disasterous, on the flip side a lack of competition virtually guarantees consumers overpaying relative to the capital costs. Only real solution is to force cable and RBOCs to be regulated like utilities. In 2010 broadband isn't a luxury.
posted by JPD at 11:05 AM on January 4, 2011 [15 favorites]
posted by JPD at 11:05 AM on January 4, 2011 [15 favorites]
If Verizon actually gives FIOS to Manhattan and Brooklyn then I will eat my shoe.
posted by Threeway Handshake at 11:06 AM on January 4, 2011 [2 favorites]
posted by Threeway Handshake at 11:06 AM on January 4, 2011 [2 favorites]
I am once again ever so happy that we still have Cincinnati Bell in SW Ohio. They continue to roll out there FiOS across the tristate area, although very slowly.
posted by Mick at 11:09 AM on January 4, 2011 [1 favorite]
posted by Mick at 11:09 AM on January 4, 2011 [1 favorite]
Hey, I've got a great idea: Let's quit making stuff and turn our society into an "information" society. And then we can hand over the information gate-keeping keys to private corporations. What's not to like?
posted by Benny Andajetz at 11:13 AM on January 4, 2011 [13 favorites]
posted by Benny Andajetz at 11:13 AM on January 4, 2011 [13 favorites]
Hey, remember when the phone company was a giant monopoly and the feds thought it was a good idea to break it up to re-introduce competition in the telecom market?
Neither do I.
posted by tommasz at 11:14 AM on January 4, 2011 [2 favorites]
Neither do I.
posted by tommasz at 11:14 AM on January 4, 2011 [2 favorites]
Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal.Right?
posted by swift at 11:19 AM on January 4, 2011 [1 favorite]
Yeah, meanwhile I live in Canada and have two options for Internet: The phone company, and the cable company. The bandwidth caps are Scroogian, and the overage charges are ridiculous.
posted by blue_beetle at 11:21 AM on January 4, 2011 [2 favorites]
posted by blue_beetle at 11:21 AM on January 4, 2011 [2 favorites]
The DC FiOS rollout is expected to be complete between 2018 and 2022. I really wish that I were making that up.
Considering that all of the regulators live here, you'd think that Verizon and Comcast would be trying to impress them. Instead the two companies seem fairly confident in their ability to manipulate the government at their will, and seem to have been fairly successful in doing so.
(That said, despite my many misgivings about Verizon as a company, FiOS was an absolutely phenomenal service, back when I lived in the service area. I frequently pulled down more than my rated bandwidth, and never encountered any downtime or throttling. These days, I'm lucky if my Comcast connection works more than 20 hours a day, and I get 25% of my rated bandwidth at any given moment.)
Many of my friends have switched to Clear's ~1.5mbps WiMAX service, because Comcast is so unreliable and slow (mull that over for a second if you will).
Honestly, I think that Verizon stopped the rollout because it was simply too expensive. I have far more faith about Verizon than I do about the cable companies. I've never encountered a cable operator that wasn't complete scum.
posted by schmod at 11:21 AM on January 4, 2011 [2 favorites]
Considering that all of the regulators live here, you'd think that Verizon and Comcast would be trying to impress them. Instead the two companies seem fairly confident in their ability to manipulate the government at their will, and seem to have been fairly successful in doing so.
(That said, despite my many misgivings about Verizon as a company, FiOS was an absolutely phenomenal service, back when I lived in the service area. I frequently pulled down more than my rated bandwidth, and never encountered any downtime or throttling. These days, I'm lucky if my Comcast connection works more than 20 hours a day, and I get 25% of my rated bandwidth at any given moment.)
Many of my friends have switched to Clear's ~1.5mbps WiMAX service, because Comcast is so unreliable and slow (mull that over for a second if you will).
Honestly, I think that Verizon stopped the rollout because it was simply too expensive. I have far more faith about Verizon than I do about the cable companies. I've never encountered a cable operator that wasn't complete scum.
posted by schmod at 11:21 AM on January 4, 2011 [2 favorites]
I think that Verizon stopped the rollout because it was simply too expensive. Clearly this is the case. Not that it is expensive relative to current prices, but that you'll never recover the original capital outlay because you are competing against cable companies who already have wires in the ground, so the first thing they'll do is undercut you on price for as long as price > direct cash costs of supplying you with broadband (which is pretty close to zero actually)
posted by JPD at 11:24 AM on January 4, 2011
posted by JPD at 11:24 AM on January 4, 2011
In a rare display of common sense, my city's government got Verizon to agree to run FIOS to every single house in town if they wanted to connect any. If it had been up to Verizon, they would have just run fiber to the rich neighborhoods and left the rest to Comcast. Unfortunately, they've got three more years to finish the rollout and so far my side of my street hasn't been hit yet.
posted by octothorpe at 11:25 AM on January 4, 2011 [1 favorite]
posted by octothorpe at 11:25 AM on January 4, 2011 [1 favorite]
I don't really have faith in Verizon (I live in Northern Virginia and they are not rolling out FIOS to my area (apartment dwellers, racially and economically mixed)), but I do want Comcast to have a viable competitor. I'm certainly considering FIOS availability in my house search however as I have coworkers who have it and their reports are glowing.
posted by longdaysjourney at 11:27 AM on January 4, 2011
posted by longdaysjourney at 11:27 AM on January 4, 2011
schmod - I'm also in DC and also waiting for FiOS on my block. In the meantime I have the Clear 1.5 mbps service and.... it's not. Not even close. In the evenings (when presumably most people are home using it) it's close to .75 mbps if I'm lucky. It's infuriatingly slow and I just hope I can hold out on it long enough for Verizon to get FiOS to my door so I don't have to go back to that soul sucking hell hole that is Comcast.
posted by Inkoate at 11:29 AM on January 4, 2011
posted by Inkoate at 11:29 AM on January 4, 2011
It probably didn't help that every town in Massachusetts treated Verizon's FiOS rollout as a gravy train and tried to extort payments from Verizon in return for permission to build out the town's damned infrastructure. Because that's a good idea.
This is what Boston got in return for its massively corrupt, completely irrational planning permission system. And somehow no one learned anything. If we had a more competent government in place then they could tell Verizon to screw and build out the networks themselves, but nope: better to scuttle private industry efforts and then replace those efforts with nothing whatsoever.
posted by 1adam12 at 11:30 AM on January 4, 2011 [2 favorites]
This is what Boston got in return for its massively corrupt, completely irrational planning permission system. And somehow no one learned anything. If we had a more competent government in place then they could tell Verizon to screw and build out the networks themselves, but nope: better to scuttle private industry efforts and then replace those efforts with nothing whatsoever.
posted by 1adam12 at 11:30 AM on January 4, 2011 [2 favorites]
About 10% of users in the U.S. are still on dialup. So who is on dialup here?
posted by VikingSword at 11:52 AM on January 4, 2011
posted by VikingSword at 11:52 AM on January 4, 2011
Give 'em a minute to finish loading the thread.
posted by cairnish at 11:56 AM on January 4, 2011 [11 favorites]
posted by cairnish at 11:56 AM on January 4, 2011 [11 favorites]
Hey, I've got a great idea: Let's quit making stuff and turn our society into an "information" society. And then we can hand over the information gate-keeping keys to private corporations. What's not to like?Whaaat? Sounds like you're implying that private corporations didn't hold all the keys to the manufacturing society. Which is crazy so obviously I'm misunderstanding.
posted by Skorgu at 11:57 AM on January 4, 2011
If Verizon actually gives FIOS to Manhattan and Brooklyn then I will eat my shoe.
Uh.... I live in Manhattan and have FIOS. Much as I hate Verizon (and I do -- I remember what Verizon is) I have to say it's a great service, albeit at an inflated price.
posted by The Bellman at 12:03 PM on January 4, 2011
Uh.... I live in Manhattan and have FIOS. Much as I hate Verizon (and I do -- I remember what Verizon is) I have to say it's a great service, albeit at an inflated price.
posted by The Bellman at 12:03 PM on January 4, 2011
FWIW, one rule of thumb I've often heard is that it costs $50 - 70k / mile to lay buried cable.
That's a long payback period for residential service in a competitive environment.
posted by ZenMasterThis at 12:09 PM on January 4, 2011 [1 favorite]
That's a long payback period for residential service in a competitive environment.
posted by ZenMasterThis at 12:09 PM on January 4, 2011 [1 favorite]
My Verizon FiOS bill is like $55/month but it's rock solid and provides dependable speeds. I think it went down once in 2010.
I'm in Northern NJ, about 10 miles east of Harlem. They announced service in mid 2005, I got it in January 2006. It was happy day when I told Cablevision that I was dropping them.
posted by exhilaration at 12:23 PM on January 4, 2011
I'm in Northern NJ, about 10 miles east of Harlem. They announced service in mid 2005, I got it in January 2006. It was happy day when I told Cablevision that I was dropping them.
posted by exhilaration at 12:23 PM on January 4, 2011
despite my many misgivings about Verizon as a company, FiOS was an absolutely phenomenal service, back when I lived in the service area. I frequently pulled down more than my rated bandwidth, and never encountered any downtime or throttling.
And the tv doesn't recompress HD signals down to the level of complete shitburger, which is also a bonus. With TimeWarner it had gotten to the point that I'd sometimes tune to the SD version of some channels because they had better picture quality than their horribly macroblocked "HD" versions.
posted by ROU_Xenophobe at 12:24 PM on January 4, 2011 [2 favorites]
And the tv doesn't recompress HD signals down to the level of complete shitburger, which is also a bonus. With TimeWarner it had gotten to the point that I'd sometimes tune to the SD version of some channels because they had better picture quality than their horribly macroblocked "HD" versions.
posted by ROU_Xenophobe at 12:24 PM on January 4, 2011 [2 favorites]
And remember when you get big enough the providers pass packets to each other "for free" at the meet points as part of peering agreements *AND* most of the incumbents didn't have to pay for the land to get service to your location.
So any newcomers have quite the barrier to entry.
posted by rough ashlar at 12:25 PM on January 4, 2011 [3 favorites]
So any newcomers have quite the barrier to entry.
posted by rough ashlar at 12:25 PM on January 4, 2011 [3 favorites]
Whaaat? Sounds like you're implying that private corporations didn't hold all the keys to the manufacturing society. Which is crazy so obviously I'm misunderstanding.
When we had a larger manufacturing base ( and unions, and government regulators with balls) we had a larger spectrum of opportunity - people with less education and smaller skill sets could still get by. Sure, corporations still had the keys but power was much more spread out.
Incredibly, very few people saw the problem to moving to an information and services economy: a large group of people who could participate before would have a much harder row to hoe. That's, literally, a drag on everything and everybody. Before, if you worked with your back, you didn't need information. Now it's essential.
To allow a few essentially monopolistic providers to dictate how this information is disseminated is insane. Our economy (par for the course nowadays, unfortunately) takes a back seat to their right to make a profit. And their size prevents real competition.
Seems like we recently had a discussion about the detrimental effects of monopoly communication providers, doesn't it? Didn't we decide that the situation was dire enough to break them up? Must have been a dream.
posted by Benny Andajetz at 12:54 PM on January 4, 2011 [3 favorites]
When we had a larger manufacturing base ( and unions, and government regulators with balls) we had a larger spectrum of opportunity - people with less education and smaller skill sets could still get by. Sure, corporations still had the keys but power was much more spread out.
Incredibly, very few people saw the problem to moving to an information and services economy: a large group of people who could participate before would have a much harder row to hoe. That's, literally, a drag on everything and everybody. Before, if you worked with your back, you didn't need information. Now it's essential.
To allow a few essentially monopolistic providers to dictate how this information is disseminated is insane. Our economy (par for the course nowadays, unfortunately) takes a back seat to their right to make a profit. And their size prevents real competition.
Seems like we recently had a discussion about the detrimental effects of monopoly communication providers, doesn't it? Didn't we decide that the situation was dire enough to break them up? Must have been a dream.
posted by Benny Andajetz at 12:54 PM on January 4, 2011 [3 favorites]
I was very excited to find that my new has had Fios service as an option. Ordering service was one of the first things I did after I moved. Later, I was astounded to learn that most people on my street haven't bothered to connect despite much higher speeds and similar prices to everything else. For a while, according to the Verizon guy, my house was the only one connected to our entire "node" (I probably have the jargon wrong, the box that sits off the main line designed to feed a neighborhood).
Some neighbors are on Comcast, some are on DSL. The one's I've talked to just don't want to bother to switch. Comcast is VERY aggressively trying to keep people from switching, making multiple visits per year to every house and offering temporary sweetheart deals.
Just saying that even if Verizon pays to lay the lines, many people just stick with their previous provider for all kinds of reasons.
posted by Patapsco Mike at 1:01 PM on January 4, 2011
Some neighbors are on Comcast, some are on DSL. The one's I've talked to just don't want to bother to switch. Comcast is VERY aggressively trying to keep people from switching, making multiple visits per year to every house and offering temporary sweetheart deals.
Just saying that even if Verizon pays to lay the lines, many people just stick with their previous provider for all kinds of reasons.
posted by Patapsco Mike at 1:01 PM on January 4, 2011
I don't think people care about speed so much as price, and price discovery can take a while. Even then within a certain pricing umbrella a lot of people don't perceive the hassle of changing being worth the effort.
Benny - I don't get what you want? The government to own the telecoms infrastructure? You can't have competitive providers. The math just does not work. All you can do is regulate them. You can't even split the wires form the service, because the service itself is basically selling nothing. All you've got is the wires.
posted by JPD at 1:10 PM on January 4, 2011
Benny - I don't get what you want? The government to own the telecoms infrastructure? You can't have competitive providers. The math just does not work. All you can do is regulate them. You can't even split the wires form the service, because the service itself is basically selling nothing. All you've got is the wires.
posted by JPD at 1:10 PM on January 4, 2011
All you can do is regulate them.
Bingo.
But we can never forget they are utilities. They are different than other for-profit enterprises. Letting them arbitrarily set prices or determine services or choose where, or if, they deploy is not regulating them. Breaking them up and then allowing them to reconsolidate is not regulating them.
posted by Benny Andajetz at 1:16 PM on January 4, 2011 [3 favorites]
Bingo.
But we can never forget they are utilities. They are different than other for-profit enterprises. Letting them arbitrarily set prices or determine services or choose where, or if, they deploy is not regulating them. Breaking them up and then allowing them to reconsolidate is not regulating them.
posted by Benny Andajetz at 1:16 PM on January 4, 2011 [3 favorites]
Verizon realizes that FIOS is going to compete with 4G cellular.
Verizon just rolled out LTE. LTE gives download speeds of 5-12 Mbps, and LTE's theoretical maximum is way higher. Most FIOS customers get 10-15 Mbps.
Now think about this thing called "the future." If in 5 years people are going to be walking around with LTE smartphones, i.e. if every member of a household has their own personal 5-12 Mbps connection to the internet, why the hell would you pay another $50/month to get a "landline" internet connection?
Verizon sees the writing on the wall. The future is always on, high bandwidth wireless internet. FIOS will rollout to markets that will be late to LTE (i.e. "late majority" users.)
posted by Pastabagel at 1:30 PM on January 4, 2011 [7 favorites]
Verizon just rolled out LTE. LTE gives download speeds of 5-12 Mbps, and LTE's theoretical maximum is way higher. Most FIOS customers get 10-15 Mbps.
Now think about this thing called "the future." If in 5 years people are going to be walking around with LTE smartphones, i.e. if every member of a household has their own personal 5-12 Mbps connection to the internet, why the hell would you pay another $50/month to get a "landline" internet connection?
Verizon sees the writing on the wall. The future is always on, high bandwidth wireless internet. FIOS will rollout to markets that will be late to LTE (i.e. "late majority" users.)
posted by Pastabagel at 1:30 PM on January 4, 2011 [7 favorites]
if every member of a household has their own personal 5-12 Mbps connection to the internet, why the hell would you pay another $50/month to get a "landline" internet connection?
Um, maybe because the famed Google / Verizon net neutrality compact was written in such a way which preserves (sort of) neutrality across landlines, but doesn't for wireless?
posted by hippybear at 1:34 PM on January 4, 2011 [6 favorites]
Um, maybe because the famed Google / Verizon net neutrality compact was written in such a way which preserves (sort of) neutrality across landlines, but doesn't for wireless?
posted by hippybear at 1:34 PM on January 4, 2011 [6 favorites]
I'm not sure thats the case - especially given VZ doesn't own all of Verizon wireless. The RBOC's aren't as rational about the substitution of mobile data as you think they are (or should be). They've spent so much trying to defend their local loop.
posted by JPD at 1:35 PM on January 4, 2011
posted by JPD at 1:35 PM on January 4, 2011
Oh, it sure is the case, according to the FCC.
And then there's wireless. "Verizon and Google issued a proposal for open Internet regulation that would exclude wireless, except for the proposed transparency requirements," says the FCC. So it wants more information on whether this is a good idea.
posted by hippybear at 1:40 PM on January 4, 2011
And then there's wireless. "Verizon and Google issued a proposal for open Internet regulation that would exclude wireless, except for the proposed transparency requirements," says the FCC. So it wants more information on whether this is a good idea.
posted by hippybear at 1:40 PM on January 4, 2011
Mod note: Comment removed. Do not pull shit out of people's profile and throw it in thread. Thank you.
posted by cortex (staff) at 1:43 PM on January 4, 2011
posted by cortex (staff) at 1:43 PM on January 4, 2011
Yeah, meanwhile I live in Canada and have two options for Internet: The phone company, and the cable company. The bandwidth caps are Scroogian, and the overage charges are ridiculous.
Same here with the added bonus that the phone company owns the cable company.
posted by ODiV at 2:17 PM on January 4, 2011
Same here with the added bonus that the phone company owns the cable company.
posted by ODiV at 2:17 PM on January 4, 2011
I've got no problem regulating the phone/cable/fiber/whatever companies like utilities, i.e. harshly. Just so we're clear.
I don't buy that the manufacturing industry was any less centralized than the tech industry is but I don't think it's provable in either direction. Also:
posted by Skorgu at 2:46 PM on January 4, 2011
I don't buy that the manufacturing industry was any less centralized than the tech industry is but I don't think it's provable in either direction. Also:
Incredibly, very few people saw the problem to moving to an information and services economy: a large group of people who could participate before would have a much harder row to hoe. That's, literally, a drag on everything and everybody.Everyone saw it. The meme describing it (roadkill on the information superhighway) rose to prominence and descended into ridicule on the basis of that observation and it all happened a decade ago. The Digital Divide is doing the same thing now. You're right, the fact that our functional literacy rate is about 70% is A Problem but the solution isn't to bring back manual labor so the uneducated have something to do.
Before, if you worked with your back, you didn't need information. Now it's essential.Bullshit. If you didn't have information you accepted a lower wage, saved less, got screwed on everything from pipes to politics and probably didn't even know it. Information has always been power; it's just a little bit more visible now.
posted by Skorgu at 2:46 PM on January 4, 2011
I had fios @ my old house for about 8 years, we were one of the test areas. I just bought a different house about 10 miles away, and when I called Verizon, they sent a crew to lay the fiber to my house...which is in a seriously rural area. After having 25mbps speed, I can't imagine going back to dsl or cable.
As an aside, the tech they sent to do the install saved my dogs life...and Verizon is now my fave faceless corporation.
posted by dejah420 at 2:52 PM on January 4, 2011
As an aside, the tech they sent to do the install saved my dogs life...and Verizon is now my fave faceless corporation.
posted by dejah420 at 2:52 PM on January 4, 2011
Skorgu:
I was, as usual, being (just a bit) hyperbolic. Even though I still think we do need more manufacturing and manual labor opportunities, I think we're on the same page.
posted by Benny Andajetz at 3:00 PM on January 4, 2011
I was, as usual, being (just a bit) hyperbolic. Even though I still think we do need more manufacturing and manual labor opportunities, I think we're on the same page.
posted by Benny Andajetz at 3:00 PM on January 4, 2011
Great article, thanks. My favorite line:
We are about to confront a well-coordinated cabal of local monopoly cable providers.
Yup.
posted by bearwife at 3:13 PM on January 4, 2011
We are about to confront a well-coordinated cabal of local monopoly cable providers.
Yup.
posted by bearwife at 3:13 PM on January 4, 2011
MetaFilter: the tech they sent to do the install saved my dogs life
posted by Splunge at 4:29 PM on January 4, 2011 [2 favorites]
posted by Splunge at 4:29 PM on January 4, 2011 [2 favorites]
And remember when you get big enough the providers pass packets to each other "for free"
This is sort of true, but I think you're being misleading. Getting a zero-cost peering agreement isn't a function of size as much as it is of traffic direction. What you pay for, at the higher tiers of the network, is imbalance.
If you 'source' many more packets than you 'sink,' or vice versa, chances are you'll have to pay for bandwidth. Residential customers and ISPs pay their providers because they have way more downstream traffic than they do upstream traffic (and what upstream traffic they do have is generally not being 'pulled' from somewhere else, it's stuff like Bittorrent which creates upstream traffic more or less incidentally as part of a download). Big colos generally have to pay, too, because they tend to be pushing traffic unequally in the opposite direction. Essentially it's a supply and demand problem; if you want access to what the ISP has more than they want access to what you have, then you're going to pay.
Where you get the "gentlemen's agreement" type of peering arrangement are between networks that have basically equal amounts of bidrectional traffic. They could charge each other, but since they each equally desire access to the others' network, it would basically cancel out, so they just write off whatever small difference remains. You used to see stuff like this between small networks (universities especially) in the early days of the Internet. Now it occurs mostly between the big backbones, but that's a result of a whole different process than the consolidation of residential/smb last-mile ISPs.
The problem is that there's basically no way that average users or even small networks are ever going to come up with a win-win versus someone like Comcast, or even a smaller local ISP. Access to Comcast and the rest of the Internet is worth much more to the individual subscriber than access to Joe Schmoe's LAN is to the ISP. As long as the Internet remains predominantly client/server, that's probably not going to change.
But the existence of $0 peering arrangement isn't, in itself, indicative of anything bad going on. It's actually a system that works quite well ... if anything, we should be very nervous about extortionate schemes that attempt to end-run it (residential ISPs charging content providers for 'access', in addition to what those content providers are already paying to their own backbone providers).
The actual backbone operators, insofar as I've had much interaction with them or the people who work for them, are not the villains in this play.
posted by Kadin2048 at 5:33 PM on January 4, 2011 [2 favorites]
This is sort of true, but I think you're being misleading. Getting a zero-cost peering agreement isn't a function of size as much as it is of traffic direction. What you pay for, at the higher tiers of the network, is imbalance.
If you 'source' many more packets than you 'sink,' or vice versa, chances are you'll have to pay for bandwidth. Residential customers and ISPs pay their providers because they have way more downstream traffic than they do upstream traffic (and what upstream traffic they do have is generally not being 'pulled' from somewhere else, it's stuff like Bittorrent which creates upstream traffic more or less incidentally as part of a download). Big colos generally have to pay, too, because they tend to be pushing traffic unequally in the opposite direction. Essentially it's a supply and demand problem; if you want access to what the ISP has more than they want access to what you have, then you're going to pay.
Where you get the "gentlemen's agreement" type of peering arrangement are between networks that have basically equal amounts of bidrectional traffic. They could charge each other, but since they each equally desire access to the others' network, it would basically cancel out, so they just write off whatever small difference remains. You used to see stuff like this between small networks (universities especially) in the early days of the Internet. Now it occurs mostly between the big backbones, but that's a result of a whole different process than the consolidation of residential/smb last-mile ISPs.
The problem is that there's basically no way that average users or even small networks are ever going to come up with a win-win versus someone like Comcast, or even a smaller local ISP. Access to Comcast and the rest of the Internet is worth much more to the individual subscriber than access to Joe Schmoe's LAN is to the ISP. As long as the Internet remains predominantly client/server, that's probably not going to change.
But the existence of $0 peering arrangement isn't, in itself, indicative of anything bad going on. It's actually a system that works quite well ... if anything, we should be very nervous about extortionate schemes that attempt to end-run it (residential ISPs charging content providers for 'access', in addition to what those content providers are already paying to their own backbone providers).
The actual backbone operators, insofar as I've had much interaction with them or the people who work for them, are not the villains in this play.
posted by Kadin2048 at 5:33 PM on January 4, 2011 [2 favorites]
As an aside, the tech they sent to do the install saved my dogs life
?
That sounds far more interesting than another rehashing of the myriad reasons to hate telecom providers. Please, tell us the story about your dog.
posted by ryanrs at 5:50 PM on January 4, 2011
?
That sounds far more interesting than another rehashing of the myriad reasons to hate telecom providers. Please, tell us the story about your dog.
posted by ryanrs at 5:50 PM on January 4, 2011
Disclaimer - I own and run an ISP in Utah.
We have enough servers to make peering with a giant worthwhile. We're also still providing residential service which uses just as much bandwidth in the opposite direction. For a short time, while it was worthwhile, I had a connection to Palo Alto where I had the potential to peer with many large networks. The ones who had network managers who realized the base value in peering peered with us. The ones who had execs who saw peering as somehow "stealing bandwidth" made geographic requirements that the smaller ISPs could never meet. In my mind as long as the network on the other side had enough capacity to peer with us, it was always a win-win. The packets got to their destination faster and had more backup in case of disconnect of other peering.
Like peering, residential access is quickly consolidating to monopolies and duopolies. In our region, it is Qwest and Comcast. Qwest made a decision in 2008 not to allow 3rd party ISPs to connect their customers over their newer infrastructure, fiber-to-the-node, where fiber is run to a neighborhood and then ADSL2 is run from that node. It provides connections up to 40Mbit download. Comcast has never allowed anyone else to use their network. Our customers are going to either of these competitors, because they literally have no choice over 10Mbit. I don't blame them for the choice, I would do the same thing in their shoes.
However, there is another network in Utah. The municipal UTOPIA fiber-optic network. Any data provider that wants to can provide service on it, and we were one of the first. We can provide a symmetric 50Mbit connection on it for about $80/month. Since cities bonded together in 2004 to build this network, it has been beset by baseless lawsuits by the incumbents and legislation from conservatives who scream that it is government competition against the free market. Ironically, we have more competitors on UTOPIA than any other area, yet it is the only part of our residential business that continues to grow.
My summary on network neutrality is simple - municipal networks like UTOPIA. Like roads & airports, the short term cost is high, but the long term benefit is greater. If you're on a municipal network and you don't like the way your ISP is providing service, or throttling your bits, switch to a competitor.
I fear the US will fall far behind on our data infrastructure, because we are still trusting the telcos and cable companies to make the right choice for all of us. As the linked essay examines, that isn't in their best interest.
posted by pashdown at 6:00 PM on January 4, 2011 [13 favorites]
We have enough servers to make peering with a giant worthwhile. We're also still providing residential service which uses just as much bandwidth in the opposite direction. For a short time, while it was worthwhile, I had a connection to Palo Alto where I had the potential to peer with many large networks. The ones who had network managers who realized the base value in peering peered with us. The ones who had execs who saw peering as somehow "stealing bandwidth" made geographic requirements that the smaller ISPs could never meet. In my mind as long as the network on the other side had enough capacity to peer with us, it was always a win-win. The packets got to their destination faster and had more backup in case of disconnect of other peering.
Like peering, residential access is quickly consolidating to monopolies and duopolies. In our region, it is Qwest and Comcast. Qwest made a decision in 2008 not to allow 3rd party ISPs to connect their customers over their newer infrastructure, fiber-to-the-node, where fiber is run to a neighborhood and then ADSL2 is run from that node. It provides connections up to 40Mbit download. Comcast has never allowed anyone else to use their network. Our customers are going to either of these competitors, because they literally have no choice over 10Mbit. I don't blame them for the choice, I would do the same thing in their shoes.
However, there is another network in Utah. The municipal UTOPIA fiber-optic network. Any data provider that wants to can provide service on it, and we were one of the first. We can provide a symmetric 50Mbit connection on it for about $80/month. Since cities bonded together in 2004 to build this network, it has been beset by baseless lawsuits by the incumbents and legislation from conservatives who scream that it is government competition against the free market. Ironically, we have more competitors on UTOPIA than any other area, yet it is the only part of our residential business that continues to grow.
My summary on network neutrality is simple - municipal networks like UTOPIA. Like roads & airports, the short term cost is high, but the long term benefit is greater. If you're on a municipal network and you don't like the way your ISP is providing service, or throttling your bits, switch to a competitor.
I fear the US will fall far behind on our data infrastructure, because we are still trusting the telcos and cable companies to make the right choice for all of us. As the linked essay examines, that isn't in their best interest.
posted by pashdown at 6:00 PM on January 4, 2011 [13 favorites]
If in 5 years people are going to be walking around with LTE smartphones, i.e. if every member of a household has their own personal 5-12 Mbps connection to the internet, why the hell would you pay another $50/month to get a "landline" internet connection?
You may be right and it's an excellent point in the short to mid-term, but consider this long-term: The bandwidth limit of fiber is nearly unlimited, the only limit is what equipment is on the ends. A fiber optic may be only 10Mb today, but that same fiber thread could do 10Gigs with the right equipment. So as wireless gear gets cheaper/faster, there's no reason landline fiber gear can't do the same. And wireless (radio) will reach a limit of speed physics long before fiber (light) does.
posted by stbalbach at 7:54 PM on January 4, 2011
You may be right and it's an excellent point in the short to mid-term, but consider this long-term: The bandwidth limit of fiber is nearly unlimited, the only limit is what equipment is on the ends. A fiber optic may be only 10Mb today, but that same fiber thread could do 10Gigs with the right equipment. So as wireless gear gets cheaper/faster, there's no reason landline fiber gear can't do the same. And wireless (radio) will reach a limit of speed physics long before fiber (light) does.
posted by stbalbach at 7:54 PM on January 4, 2011
I use the Minneapolis city-regulated 6 mbps wifi for $25 a month. It's awesome, but of course getting beyond 6 mbps with commodity wireless gear is probably not feasible. Sadly, last I heard the ISP is losing money and can't raise prices due to its contract with the city, so it's unlikely to be replicated elsewhere.
The proper solution needs to separate the last-mile physical connection from services like voice, video, and internet access. I envision the government picking a single company that would then sell fiber lines to individual homeowners, at a cost of $2-3k from what I've heard. These lines would go to neighborhood cabinets, which would be serviced by competing ISPs. Some ISPs would offer to cover the cost of a fiber line in exchange for a long-term service contract, much like FiOS does already. But of course this requires rolling over current vested interests and is thus unlikely to happen.
posted by miyabo at 9:10 PM on January 4, 2011
The proper solution needs to separate the last-mile physical connection from services like voice, video, and internet access. I envision the government picking a single company that would then sell fiber lines to individual homeowners, at a cost of $2-3k from what I've heard. These lines would go to neighborhood cabinets, which would be serviced by competing ISPs. Some ISPs would offer to cover the cost of a fiber line in exchange for a long-term service contract, much like FiOS does already. But of course this requires rolling over current vested interests and is thus unlikely to happen.
posted by miyabo at 9:10 PM on January 4, 2011
And wireless (radio) will reach a limit of speed physics long before fiber (light) does.
Radio waves and light travel at the same speed. What do you mean by that?
posted by Threeway Handshake at 11:08 PM on January 4, 2011 [1 favorite]
Radio waves and light travel at the same speed. What do you mean by that?
posted by Threeway Handshake at 11:08 PM on January 4, 2011 [1 favorite]
Means visible light has higher information density than radio. Bit-speed instead of crow-flies-speed.
posted by breath at 11:51 PM on January 4, 2011
posted by breath at 11:51 PM on January 4, 2011
> Radio waves and light travel at the same speed. What do you mean by that?The executive summary of the issues at hand:
a) There's a reason why everyone uses switches instead of hubs on their Ethernet networks. (Putting everyone in the same collision domain KILLS network performance.)
b) You can push more elephant, faster through the Panama Canal than you can a soda straw. (A fiber connection is an extremely fat pipe. A link used by a cellular modem? Not so much.)
posted by simoncion at 12:03 AM on January 5, 2011
Japan has had FTTH (up to 1 Gb/sec., depending on where you live- better coverage in newer buildings and major metro areas) for a number of years now. All-you-can-eat services for around $50-70 USD/mo.
In the wireless realm, there's everything from 3G, HSDPA, and WiMax with ever greater coverage and bandwidth. Lots of competition. All-you-can-eat services for less than $40-50 USD/mo.
Having enjoyed very fast Internet for a number of years in Tokyo, it's hard for me to imagine single-or-even-doule-digit Mb/sec. Internet service in the US.
Successful broadband models are myriad (Japan and S. Korea are just two) I don't see why the FTC and many White House administrations can't fix this- it's a huge issue for the future of the competitiveness of the US.
posted by gen at 12:27 AM on January 5, 2011
In the wireless realm, there's everything from 3G, HSDPA, and WiMax with ever greater coverage and bandwidth. Lots of competition. All-you-can-eat services for less than $40-50 USD/mo.
Having enjoyed very fast Internet for a number of years in Tokyo, it's hard for me to imagine single-or-even-doule-digit Mb/sec. Internet service in the US.
Successful broadband models are myriad (Japan and S. Korea are just two) I don't see why the FTC and many White House administrations can't fix this- it's a huge issue for the future of the competitiveness of the US.
posted by gen at 12:27 AM on January 5, 2011
b) You can push more elephant, faster through the Panama Canal than you can a soda straw. (A fiber connection is an extremely fat pipe. A link used by a cellular modem? Not so much.)
Did you really just use the Series of Tubes analogy? It doesn't quite work here. The RF band actually has tons of bandwidth. The real problem is that everyone has to share it. Want more bandwidth from fiber? Simply lay another fiber, or kick off some of your users. The cost of the fibers themselves is close to nothing, and they're almost always installed in gigantic bundles for this reason.
posted by schmod at 6:45 AM on January 5, 2011
Did you really just use the Series of Tubes analogy? It doesn't quite work here. The RF band actually has tons of bandwidth. The real problem is that everyone has to share it. Want more bandwidth from fiber? Simply lay another fiber, or kick off some of your users. The cost of the fibers themselves is close to nothing, and they're almost always installed in gigantic bundles for this reason.
posted by schmod at 6:45 AM on January 5, 2011
The RF band actually has tons of bandwidth.
Well, I guess "tons" is subjective, but relative to optical fiber it certainly doesn't. There's more bandwidth available just in the 'C band' of optical fiber (1530-1565nm) than in the entire usable RF spectrum. (8.565×10^15 hertz according to Wolfram; that's 8,500 THz.)
Of course, we have gotten very good at utilizing the available bandwidth of RF, and do so far more efficiently than we use optical fiber ... channel spacing on a radio system might be measured in kilohertz (if you can even talk about 'channel spacing' at all; with some of the spread-spectrum stuff it's harder to define), while most optical wavelength multiplexers have 100 or 50 GHz spacing, and that's considered "high density". But there's a lot of bandwidth there, and I don't think there's any reason why we won't be able to get much narrower in time; WDM systems were apparently only made practical in the late 70s, while similar techniques for electrical telecommunications had a century head start, and ~90 for radio.
Most optical systems you see today are the light-based equivalent of old spark-gap radio transmitters, just blasting out a signal over a huge portion of the spectrum. This is acceptable because the symbol rate is so high, you can use time-division schemes to fake up multiple channels ... but there's a whole different direction for us to spread out in.
The real question to me is whether enough demand will exist to justify the high installation costs of fiber, rather than using radio or the existing TWP/coax stuff that's in the ground. It seems that if we could justify hauling coax out to just about every house in the country just for extra TV channels that deploying fiber in order to get everything that a truly broadband Internet would bring would be a bargain. But I think we're looking market failure squarely in the face; there's no way to build the demand without the infrastructure, and there's no way to justify the infrastructure buildout without demand.
posted by Kadin2048 at 9:44 AM on January 5, 2011
Well, I guess "tons" is subjective, but relative to optical fiber it certainly doesn't. There's more bandwidth available just in the 'C band' of optical fiber (1530-1565nm) than in the entire usable RF spectrum. (8.565×10^15 hertz according to Wolfram; that's 8,500 THz.)
Of course, we have gotten very good at utilizing the available bandwidth of RF, and do so far more efficiently than we use optical fiber ... channel spacing on a radio system might be measured in kilohertz (if you can even talk about 'channel spacing' at all; with some of the spread-spectrum stuff it's harder to define), while most optical wavelength multiplexers have 100 or 50 GHz spacing, and that's considered "high density". But there's a lot of bandwidth there, and I don't think there's any reason why we won't be able to get much narrower in time; WDM systems were apparently only made practical in the late 70s, while similar techniques for electrical telecommunications had a century head start, and ~90 for radio.
Most optical systems you see today are the light-based equivalent of old spark-gap radio transmitters, just blasting out a signal over a huge portion of the spectrum. This is acceptable because the symbol rate is so high, you can use time-division schemes to fake up multiple channels ... but there's a whole different direction for us to spread out in.
The real question to me is whether enough demand will exist to justify the high installation costs of fiber, rather than using radio or the existing TWP/coax stuff that's in the ground. It seems that if we could justify hauling coax out to just about every house in the country just for extra TV channels that deploying fiber in order to get everything that a truly broadband Internet would bring would be a bargain. But I think we're looking market failure squarely in the face; there's no way to build the demand without the infrastructure, and there's no way to justify the infrastructure buildout without demand.
posted by Kadin2048 at 9:44 AM on January 5, 2011
Means visible light has higher information density than radio. Bit-speed instead of crow-flies-speed.
But the speed limitation is not the result of the media carrying that information. The lack of speed of radio is hamstrung by the protocol pushing the data, not because it is radio waves.
a) There's a reason why everyone uses switches instead of hubs on their Ethernet networks. (Putting everyone in the same collision domain KILLS network performance.)
There is an equivelant of switches for wireless radio.
Again, this is also hamstrung by our protocols. If you started from scratch right now and eliminated the current reservations of the wireless spectrum, and wanted to make "broadband" over radio, we could all be easily streaming 1080p video to our phones.
When I hit the post comment button, it'll be sent over HTTP via my wireless network at home, which is able to push 300mbps using just a tiny sliver of the radio spectrum. The limitations aren't because of the media, it is because of our implementations of them.
posted by Threeway Handshake at 11:11 AM on January 7, 2011
But the speed limitation is not the result of the media carrying that information. The lack of speed of radio is hamstrung by the protocol pushing the data, not because it is radio waves.
a) There's a reason why everyone uses switches instead of hubs on their Ethernet networks. (Putting everyone in the same collision domain KILLS network performance.)
There is an equivelant of switches for wireless radio.
Again, this is also hamstrung by our protocols. If you started from scratch right now and eliminated the current reservations of the wireless spectrum, and wanted to make "broadband" over radio, we could all be easily streaming 1080p video to our phones.
When I hit the post comment button, it'll be sent over HTTP via my wireless network at home, which is able to push 300mbps using just a tiny sliver of the radio spectrum. The limitations aren't because of the media, it is because of our implementations of them.
posted by Threeway Handshake at 11:11 AM on January 7, 2011
Spread spectrum isn't a panacea. There's a fundamental limit to how much information you can convey over a particular amount of spectrum (channel), given various transmission characteristics. (This is defined by the Shannon-Hartley theorem, and the channel capacity of a particular noisy channel is generally called the Shannon limit.)
Digital spread spectrum products work by taking a signal and (unsurprisingly) spreading it out across a wide bit of spectrum, so that it's more resistant to interference. But it doesn't suddenly manufacture new bandwidth; it's not magic. It lets you operate multiple transmitters with a lot less coordination than you'd need with traditional narrow-band transmission methods.
E.g., traditionally if you had 10 channels each with 10kHz bandwidth, you might space them out at 10kHz intervals (or, realistically, maybe more so they have guard bands in between them). Ideally you could put them all in 100kHz. This requires a fair bit of coordination; you don't want one accidentally transmitting on the other's allocation. But with DSS you could take each signal, smear it out across 100kHz, and then let them each go nuts. They each have more bandwidth available, but there's more noise (due to the other transmitters). But the net result is — or ought to be, anyway, again in an ideal case — that it's a tradeoff between bandwidth and noise.
But the important part is that it's a tradeoff, not a free ride. It's still a shared medium, and this is where a switched fabric is totally different from a shared medium like radio. DSS is a very clever way of allowing interoperation, but it's not doing what a switch does: breaking up the collision / interference domain so that different stations aren't interfering with each other. If I have 100Mb switched Ethernet and a good 10 port duplex switch, I should be able to get the full 100Mb between any combination of 5 pairs through the switch, simultaneously. (A lot of cheap switches don't do this very well, of course. But the topology should allow it.)
Radio is more like old coax-based Ethernet, where there's just a single collision/interference domain. You can get very clever about making stations play nicely with each other, but in the end everyone is connected to everyone else and sharing the same bandwidth.
posted by Kadin2048 at 8:52 PM on January 7, 2011
Digital spread spectrum products work by taking a signal and (unsurprisingly) spreading it out across a wide bit of spectrum, so that it's more resistant to interference. But it doesn't suddenly manufacture new bandwidth; it's not magic. It lets you operate multiple transmitters with a lot less coordination than you'd need with traditional narrow-band transmission methods.
E.g., traditionally if you had 10 channels each with 10kHz bandwidth, you might space them out at 10kHz intervals (or, realistically, maybe more so they have guard bands in between them). Ideally you could put them all in 100kHz. This requires a fair bit of coordination; you don't want one accidentally transmitting on the other's allocation. But with DSS you could take each signal, smear it out across 100kHz, and then let them each go nuts. They each have more bandwidth available, but there's more noise (due to the other transmitters). But the net result is — or ought to be, anyway, again in an ideal case — that it's a tradeoff between bandwidth and noise.
But the important part is that it's a tradeoff, not a free ride. It's still a shared medium, and this is where a switched fabric is totally different from a shared medium like radio. DSS is a very clever way of allowing interoperation, but it's not doing what a switch does: breaking up the collision / interference domain so that different stations aren't interfering with each other. If I have 100Mb switched Ethernet and a good 10 port duplex switch, I should be able to get the full 100Mb between any combination of 5 pairs through the switch, simultaneously. (A lot of cheap switches don't do this very well, of course. But the topology should allow it.)
Radio is more like old coax-based Ethernet, where there's just a single collision/interference domain. You can get very clever about making stations play nicely with each other, but in the end everyone is connected to everyone else and sharing the same bandwidth.
posted by Kadin2048 at 8:52 PM on January 7, 2011
We have FiOS, which really is lovely, but we don't receive cable TV. It's been like this for more than 5 years. Comcast is mostly over it, maybe just a bit wistful at times, but Verizon seems a bit desperate in their pitching of woo lately. Really, it's us, not you Verizon.
And by "us" I mean comfortably streaming three 3 HD Netflix programs at once, with plenty of bandwidth left over for web browsing.
posted by NortonDC at 8:59 PM on January 7, 2011
And by "us" I mean comfortably streaming three 3 HD Netflix programs at once, with plenty of bandwidth left over for web browsing.
posted by NortonDC at 8:59 PM on January 7, 2011
Threeway Handshake, 300 Mb/s sounds nice, but understand that the demonstrated capacity of fiber stands at 69.1 Tb/s, or 69,100,000 Mb/s, over roughly 150 miles. That's a pretty big gap.
posted by NortonDC at 9:10 PM on January 7, 2011
posted by NortonDC at 9:10 PM on January 7, 2011
« Older It’s totally unknown to me what you just asked me... | 10 Year Anniversary Crying Edition Newer »
This thread has been archived and is closed to new comments
posted by Dr. Twist at 10:58 AM on January 4, 2011