BitTorrent throttling and cable bandwidth lawsuits

Class action lawyers have sued Comcast for throttling users of the bandwidth-intensive P2P application BitTorrent, and the Federal Communications Commission by a 3-2 vote has declared the cable provider’s practice unlawful. (UPI, Aug. 3; Janko Roettgers, “The FCC Rules Against Comcast. Now What?”,, Aug. 1). But Insight Communications CEO Michael Willner defends the general need for some practice of this sort (Jul. 28; via Class Action Blawg):

[A reader/commenter who has filed a class action suit against Comcast suggests] building whatever capacity needed to give consumers all they use. I’d love to do that but it’s a self defeating process for any ISP with relatively high upload speeds to do so.

Here’s why. My company is accountable to the nearly half million broadband customers on our network. But when we provide relatively high upload speeds (1 meg and better), Internet users all over the world are directed by their P2P software to come to us before they go to slower providers. Within a few days, we simply are unable to handle the load leaving unmanaged consequences to take over, slowing everyone on our network no matter what they are doing. We could add more and more capacity, but the cycle simply starts all over again, bringing even more people to our network for uploads. We never get to the point where we would be able to build enough upload capacity to accommodate everyone from New Zealand to New Brunswick.

So we really only have two choices: We can limit all of our customers’ upload speeds making our network far less attractive to the downloader in New Zealand. That is the net effect of what DSL does. Or we can allocate a disproportionately large amount of upload capacity to our heavy upload users, but limit it fairly.

On some possible technical fixes, see Iljitsch van Beijnum, “IETF: find more peer-to-peer bandwidth, but use it sparingly”, Ars Technica, Aug. 3.


  • If throttling back speeds is required, Comcast shouldn’t go around telling customers they have download/upload speeds of X when it is actually limited to Y (where Y<X)

  • agreed this is a very simple case on failing to uphold their end of a contract ( now while i will say all cable speeds are somewhat variable, there is a set limit it shouldnt fall below)

  • The above posters are right on. If you don’t have the network to back it up, don’t advertise high upload speeds. You want to have upload speeds for casual users but penalize heavy uploaders, advertise usage limits. Tell your users how you intend to limit thier service clearly, and before they sign up. Why is that too much to ask?

  • Have to agree with the other commenters – looks like Comcast entered into contracts it could not fulfill.

    Now, they were probably very BAD contracts that SHOULDN’T be fulfilled… which means that Comcast should not have entered into them, and must now pay the penalty for not fulfilling them.

    Which is exactly what SHOULD happen, right? Am I missing something?

  • “So we really only have two choices”

    Um, actually, there’s a third choice which makes a lot more sense: metered bandwidth. There’s a reason that gas and electric companies don’t have this overselling capacity problem, and it’s that everyone has to pay for the amount of power or gas they use. For some reason, this model is never implemented for home internet users, so of course everyone saturates the line. It’s a pretty classic tragedy-of-the-commons situation, so I don’t understand why they don’t switch to billing by the megabit or whatever. Seems like it would be a lot easier than selling “unlimited bandwidth” and then putting in these stupid Rube Goldberg methods of foiling people who actually try to take advantage of it… it’s as if you opened an all-you-can-eat restaurant and then started kicking out all the fat people.

  • In my opinion, this isn’t actually a network neutrality issue at all. This is an oversubscription and false advertising case.

    Comcast advertises high bandwidth, right? That’s the entire selling point of their service. However, Comcast has found that the vast majority of their users don’t actually use more than 1% of the advertised bandwidth and merely delight in the quicker loading of images and youtube videos. These users tend to use high amounts of bandwidth in tiny spurts and together put only the smallest burden on the network. Knowing this, Comcast has declined to upgrade their infrastructure to accommodate rising levels of subscription. They rake in the bucks because almost no one is policing the contract, so to speak.

    Everything would be going fine with this except for the five to ten percent of their user base that actually does attempt to use the bandwidth that Comcast advertises as available. Comcast’s lines are so oversubscribed that even 5 to 10 percent of their users making partial use of their individual bandwidth can completely overwhelm the network. Naturally, the response to this is to choke off these users. In many ways, it’s similar to how insurance companies behave when historically healthy patients suddenly catch something like cancer- suddenly they become a problem instead of a profit center and the insurance company wants to chase them off.

    In fairness to Comcast, they are going into otherwise unserved markets and offering a service that is 80% as good as broadband for probably 90% of their users. Presumably as these markets become more mature, other providers will appear to challenge Comcast’s weak service levels, horrible customer service and incompetent installation people.

    Comcast can’t switch to metered bandwidth because it would either drive off an enormous portion of their users or end up undercharging them compared to current prices. A price low enough to retain most of their higher end users (especially the early adopters) would almost not charge the casual users, from whom they currently receive a generous 40-50 dollars a month. A metered price high enough to keep charging 40-50 dollars a month to the casual user would chase the serious users off to leased T1 lines (that have ironclad guarantees of both bandwidth and uptime). Such an outcome would create the very serious possibility that these hardcore users would then set up miniature ISPs of their own at a rate far below 40 dollars a month. This is becoming a greater threat the cheaper that wifi and similar become available.

  • Gas an electric companies (and restaurants) are actually selling a deletable resource though. Data isn’t “used up” when it’s transmitted. The problem is far more akin to overbooked airline flights. Comcast offers access to a data pipe that can transport a certain amount of data, but it sells that access based on the idea that not everyone is actually going to use it constantly. Look at what he’s saying: he’s couching it in blaming other sources, but the end result is that if too many Comcast customers are uploading at the upload speed Comcast offers them, other Comcast users suffer. Metered bandwidth might create an incentive for people to use less bandwidth, but it wouldn’t change the fact that Comcast is overselling its network, just give it an excuse to do it more…not alleviating the problem that if everyone got online at the same time, they wouldn’t have the network capacity that they sold to them.

  • It is not clear to me how many “casual users” would be chased of by metered usage. After all, the current model subsidizes the heavy users. It could well be that a much lower base fee with usage fees would produce a much lower price for those casual users while deterring the heaviest users from jamming the system. Fee schedules could even take into account off-peak pricing so that large transfers would be more economical during periods of lower volume.

    As for losing those “Hard core” users, aren’t they the ones taxing the system? Losing them would mean pushing back the point where your infrastructure would need upgrades and sounds like a winning move.

    I looked over my agreement. I was promised speeds up to a point but no promise was made that I would always get that level of service. In fact they mention some but not all items that might lower the performance.

    So I don;t see how they violated the contract.

  • There’s a reason that gas and electric companies don’t have this overselling capacity problem, and it’s that everyone has to pay for the amount of power or gas they use.

    The problem with the gas and power analogy is that the user has control over how much power or gas they use. The end user can set their thermostat to 72 degrees in the winter and will save money compared to another homeowner with an identical house that sets his theromostat to 73 degrees. Usage is governed by the end user.

    That is not the case with internet access. The destination – the site – determines amount used. A site that is graphic intense and flooded with multimedia will require much more bandwidth than a simple site with text and a few simple graphics.

  • I have no sympathy for Comcast on this one (I am a Comcast subscriber). The CEO lays it out fairly clearly:

    “So we really only have two choices: We can limit all of our customers’ upload speeds making our network far less attractive to the downloader in New Zealand. That is the net effect of what DSL does. Or we can allocate a disproportionately large amount of upload capacity to our heavy upload users, but limit it fairly.”

    His 2 choices are:
    1) Tell the truth about our speed and appear to be less competitive.
    2) Lie about our speed in order to beat the competition.

    Mr. Willner obviously prefers option 2. This is really a variation of a joke told by Flip Wilson and others about a lemonade stand with a sign reading “All the lemonade you can drink 5 cents”. When a customer buys a glass, drinks it and asks for another he is informed it will cost another 5 cents. Buyer brings up the text of the sign, but the seller explains that one glass IS all you can drink for 5 cents.

    Mr Willner mistook that joke for a business plan.

  • Actually, most people are missing the point Michael made: that when Comcast increases available bandwidth, people who are NOT his customers flock to his network, because of the way P2P networks work. It’s more like adding highway lanes and noticing that most of the traffic increase is NOT coming from the local areas, but from far away, thus inconveniencing the locals. To preserve access to the highway, the police restrict access to the onramps. Not a perfect solution by any means.

  • gitarcarver, it’s not the sites that are causing the problem, it’s the P2P networks driving traffic to Comcast’s network. (and that’s not to say that P2Ps are doing anything illegal.) The overriding issue is size of transfered files, and frequency of requests at peak times. But even if it were true, the end user determines where he goes and therefore how much bandwidth he consumes.

    Please someone explain to me how we can have unlimited use of a limited resource. Especially when a very large percentage of users are not contributing significantly to the problem, but are being penalized by the few.

  • The bandwidth is coming from his network, its coming from his users who are providing uploads. The user of the network is who ever payed for the bandwidth. If I upload an image to a website, its not the website using my upload bandwidth…

    To use the above analogy, I’ve payed for the right to use both the onramp and the offramp of the highway (note that comcast advertises speeds for both). Now the highway is crowded, so the police wont let me up the unramp…

  • As I pointed out above (apparently it was too long a post for anyone to bother reading) metered use won’t scare off casual users, it will UNDERCHARGE THEM compared to current rates. The average home user is probably consuming about a dollar of service in return for about 40-50 dollars in monthly fees.

    Any metered rate high enough to charge the average user 40-50 dollars a month will scare other users off to services that guarantee upstream and downstream speeds for a slightly higher price. I am personally considering getting a small business DSL because of the endless bullshit that I put up with from Comcast in terms of downtime and two years of forged TCP RST packets that have kept me out of some of the better torrent communities.

  • I don’t see why a metered usage must be really inexpensive for the casual user. Just set a connection fee (at or slightly below the current fee) plus a usage fee with a base allocation set several times what the casual user uses. Then send an automatic email monthly to the user showing their usage and how they really aren’t anywhere near the base amount. Plus lots of advertising of other products.

  • The hardcore users are the only ones that need the high bandwidth and they are the early adopters that allow these companies to begin market penetration in any given area.

    Comcast is in a bind because of two overlapping phenomena:
    -competing technologies that guarantee bandwidth levels like SDSL and T1 are more expensive than cable, but not by so much that jacking up rates for home users won’t cause them to begin switching. When this happens, expect prices for these competing services to begin dropping as they go from a specialized business service to a more widely adopted service. As I tried to point out above, things like Wifi make it much easier for a power user to lease a T1 line for about 360 dollar a month and set up his own Wifi based ISP for all the casual users in his neighborhood. He then pockets the profits that used to go to the cable companies and gets his own T1 to play with most of the time.
    -more and more “casual users” are developing a taste for high bandwidth and are getting frustrated by stuff like forged TCP RST packet, which is actually a disruption more than a slowing down of bandwidth. Comcast is essentially creating a whole population of disgruntled users who are itching for an opportunity to leave them for a competitor. If Comcast doesn’t start upgrading its infrastructure soon and providing what the customers expect, it will start losing users to increasingly cheap SDSL.

  • john,

    I know what Comcast is saying. The proposed solution was “metered bandwidth.” I was addressing the proposed solution as opposed to commenting on the alleged “problem.”

    (And while the user does determine where he goes, he does not determine, nor may he even know, how much bandwidth a site uses before he visits that site.)

    Please someone explain to me how we can have unlimited use of a limited resource.

    Please explain to me how with the increasing speed of computers, systems, routers, servers and transfer rates you see this as a “limited resource.” In other words, is the “limiting” factor that of companies willing to invest in new equipment and infrastructure?

    That is an artificial limitation.

  • Bill, the problem is that any pricing structure that restricts the power users enough to hide Comcast’s lack of infrastructure will inevitably chase them off entirely to ISPs that serve businesses.

    Comcast is already putting out feelers in this direction by:
    -choking off everyone’s upload bandwidth at around 40k total. This is what they mean by “up to 500k upload speed, but we guarantee only best ‘effort'”
    -forging TCP RST packets anytime it detects anyone seeding a torrent. This is essentially the same as picking up the phone and in your voice, telling the person you are talking to “sorry, gotta go” and then hanging up. This is in addition to the very tight bandwidth restrictions outlined above.
    -banning users who go over 1TB a month in total bandwidth in either direction. This is to get rid of the newsgroup users who saturate their downstream bandwidth for weeks at a time.

    Make sense? I’m telling you that a large portion of Comcast’s customers and certainly all of the power users are sick and tired of their behavior and just looking for any competitor who is even 10 percent better or 10 percent cheaper than them. We are talking a really low bar here.

  • The throttling is not the problem, per se. If Comcast had been honest about the throttling instead of surreptitiously injecting forged TCP reset packets into their customers’ connections, there would be less of a problem.

  • I hate to sound like a marxist, but I think that the market has failed in this instance. We have cable companies and phone companies who own all the broadband infrastructure in this country and who essentially refuse to compete with one another on price or features. They bother oversubscribe, they both refuse to spend any money on upgrading infrastructure and they both give generally bad experiences to their customers.

    So what is the best free-market solution to this problem? I’m not sure there is a purely free-market solution, as much as it pains me to admit it. The closest I can come is either
    a) bet on some new technology that doesn’t require laying a mountain of wire or fiber everywhere and hope this forces the cable and dsl people to start competing again. Even this approach I think ultimately fails because eventually there would be a stable equilibrium with a 3 way oligopoly instead of the 2 way one we currently have.
    b) let state governments take the initiative and start laying fiber on their own. Many US states are larger than the high-broadband countries like south korea and could easily afford to do universal broadband if they were willing to put other priorities further back. The benefit of this approach is that it solves the problem of figuring out which policies are the best. If universal broadband doesn’t actually bring value to the economy in proportion to the cost, other states won’t follow suit. If it does produce benefits, letting states try their own approaches lets us find problems with specific policies.

  • And at the very least, FCC needs to stop Comcast from doing the TCP RST thing. It’s not even a legitimate form of traffic shaping. They’re interfering with communications. They’re blocking. That shouldn’t be allowed.

  • Did anybody read the original article? The majority of bandwidth isn’t being used by Comcast subscribers, it is being used by people from outside Comcast to access the files being shared by Comcast’s users. These P2P systems make ever computer they are on into a virtual server, hosting the shared files. In my opinion Comcast would be with in it’s rights to either ban P2P or to charge more for it.

  • Jim Collins, we read it but rejected it offhand as a completely asinine argument. People from outside the network only use the Comcast network so far as necessary to connect to Comcast users. There isn’t some additional usage the outsiders would have beyond allowing the Comcast uploaders to send data. Comcast isn’t some sort of backbone provider to the world’s ISPs.

    Or perhaps you fail to realize that Comcast has long claimed that users will get speeds of hundreds of KB per second upload speed and the people are complaining that Comcast is not only choking them off to less than 50Kbps but also using packet forging to break connections so the applications using the bandwidth don’t work properly.

  • Yes, Jim, we understand that you 1) have upload issues with Comcast, 2) think the solutions is more infrastructure (paid for by people who aren’t using the existing BW, let alone increased BW) 3) think that real world problems have virtual world solutions. We get it.

    While it is very true that Comcast has acted at least liek jerks, if not out-right illegally, the solution isn’t to demand that a technologically inferior company suddenly take on the attributes of a superior one, but bypass them.

    As far as setting up mini-WiFi ISPs goes, check your regulatory structure and get back to me. I won’t wait up: it’s around 3000 pages of dense legalese. You’ll want a good lawyer.

    Comcast serves many people moderately well, and a smaller number exceedingly badly. And the fact that they didn’t see it coming speaks only to their amateurism.

    And if you don’t think BW is a scarce commodity, I suggest that, if it weren’t we wouldn’t be having this conversation.

    And none of this excuses sending RST packets, OK?

  • gitarcarver, the problem isn’t the first time a user loads the high needs website, but repeated visits, when he DOES know what’s going to happen.

    And still, 90% of the problem is with 10% of the customers.

  • The destination – the site – determines amount used. A site that is graphic intense and flooded with multimedia will require much more bandwidth than a simple site with text and a few simple graphics.

    This isn’t entirely true. You don’t have to accept all of the stuff that the site tries to send you. A well-behaved site will only send a modest amount of data at first and will let you decide to what extent you want the rest of the images, audio, video, etc. Sites that immediately download a ton of Flash are obnoxious. As soon as I hit one, I break the the connection. A good browser will give you a lot of control over how much traffic you generate.

  • The central flaw in what Comcast is doing is that it is discriminating against a particular type of usage, not setting limits on total usage or rate of usage or metering on the basis of usage. I occasionally use BitTorrent, exclusively for legal software downloads. If I want the current Debian Linux ISO, I can download it the usual way or via BitTorrent. The bandwidth that I will consume is identical in the two cases. The same approximately 700MB have to get from wherever the ISO is to me. The difference between the two approaches is that if I use BitTorrent I will likely get different pieces of the download from different sources, thereby more evenly distributing the overall load on the network. This is a GOOD THING. If my ISP wanted to go to an arrangement with a rate cap or total bandwidth cap or metering, that would be perfectly fine with me (assuming, of course, that the rates are reasonable.) What I do not want them doing is discriminating against BitTorrent traffic in favor of other traffic. It isn’t any of their business how I use my bandwidth anymore than it is the electric company’s business what channels I chose to watch on my TV.

  • John:

    Doesn’t matter if it is legal or not, most tech people have only the vaguest notion that FCC rules apply to them. There are certainly no practical obstacles to charging your neighbors for bandwidth sharing.

    Regardless… the solution in the short term (besides dropping the stupid TCP RST shit) is either:
    a) for Comcast to stop advertising their service as broadband, especially the part where they advertise “up to x kbps.” It’s a deliberate deception because they already know EXACTLY how much bandwidth the traffic shaping equipment gives the customer and it is way below what they advertise.
    b) for them to build infrastructure so they can deliver what they advertise. I’d be happy with 100k up. I’d be ecstatic at 200k up let alone the nearly 500k they promise in their sales pitch. I just want to be able to seed a torrent and use the internet for anything else at the same time.

    I recognize that the market doesn’t rotate around me, but I think you need to recognize that Comcast, Cox, Earthlink, AT&T etc all have much higher margins and much crappier service than they would if they had to actually compete for my business. Unfortunately, we’ve moved away from intra-platform competition (thanks again W!) towards inter-platform pseudo-competition in which the consumer gets to choose between two precisely identical packages, one of which is offered with basic cable, the other offered with basic phone service. This is not competition. At the very least we should bring back the notion of letting anyone offer service over the existing lines, with everyone paying rent to the line owner. Everyone still gets their cut, but the provider with the best actual service, best backbone lines, best customer service and best price gets all the business. This is just a matter of who owns the wires going to your house.

  • John,
    the problem isn’t the first time a user loads the high needs website, but repeated visits, when he DOES know what’s going to happen.

    Yet the site still determines the amount of data it will try to send.

    And still, 90% of the problem is with 10% of the customers.

    It would also be the fault of Comcast for not maintaining and upgrading their network.

    This isn’t entirely true.

    There are some controls in a browser that will help control bandwidth issues. That doesn’t mean that the site is not reponsibile for what it wants to load, or the amount of bandwidth it will use. Secondly, any good commercial webdesigner will always load the graphics and the ads that generate revenue for the site first, and then the content that the user wants after that. So while it is great that you kill the site before it loads completely or don’t visit the site again, the fact remains that with metered bandwidth, not onlyhave you used some bandwidth, you have used bandwidth that by your own admission is worthless to you. You have essentially paid for nothing.

  • I currently live in an area where metered access is the rule. I pay $10/gb transferred in or out. The problem I see with the debian iso example from above is that ftp will use much less total bw than bt will. FTP needs to transfer the file data and some amount of control information. BT will desire to send out just as much file data as you receive, with attendant control data, so using BT will in effect double the amount of bw used if fair allocation across the network is assumed.

    Before I moved to my current location I had looked into T1 service, at no time did I find T1 without metering, which would seem to kill that avenue for all but the most hardcore p2p users.

  • Jim W,

    If you’re cool with the legalities, who am I to interfere? Let me know when you’re up and running in my neighborhood…

    And I agree completely with the notion that the companies would have better service etc if they had to compete; but I really don’t want my streets torn up every month (instead of every OTHER month). And I’d love to get the DOWN speeds I’m promised, let alone the up speeds. But we can’t really blame W for this: the problem is in the monopoly at the neighborhood level.

    And I still say it’s a limited commodity, regardless of how unlimited you think it could be. gitarcarver is busily making that point. But let’s do a thought experiement:

    Imagine that Comcast increases it’s capacity 100x. Should be enough for everyone? Now, that costs big $: probably in the 100’s of millions to add capacity to the servers, routers, fiber, switches and people to manage it. how long before the network saturates?

    Given how long it took for the last upgrade, and multiplying it by 10, you would have unfettered bandwidth for approximately:

    6 months. Maybe less. I might have dropped a significant figure, in which case it would be 5 years, but given that I do this for a living, I don’t think so. Think it through: if they roll out such a big upgrade, people will flock to it. Especially people who like to upload, because they’re the ones having the toughest time right now. Second are the rest of the P2P universe. Suddenly, there’s no limit to the number of torrents you can start! Fabulous, right? Until there’s a limit. then, we’re right back where we started, but now the non-P2Pers are paying for the increase in available bandwidth, and not using it.

    My suggestion is that P2P packets be treated differently (as has beens suggested by others) and alotted a percentage of total bandwidth. When that percentage saturates, so be it. Maybe if people scheduled their torrents for non-peak times that would help, but it’s at best a temporary thing.

  • This reminds me of the late 1970’s, when I was in high school . . . we used a DEC-10 (Digital Equipment Corp.) terminals, but almost everyone else in the school district (schools, departments, principals, board members, etc.) had to timeshare [remember that term?]. As an example, if student A was playing a football game on his terminal while board member B was sending a long proposal to employee C, the system slowed down for everyone–not too much for A, but noticeably for B and C. If too many people got on the system and had large programs or complicated ones running at the same time, the entire system slowed WAY down. Sure, you might get X baud width advertised by DEC, but reality says you got maybe 33%-50% of speed X.
    That is what Comcast is in right now, seems to me–almost everyone wants high-speed, NOW, at the same time, without realizing that such would slow down any network.

  • As THE system/network admin for a quite small ISP, I understand and agree with their attitude of trying to provide the best service, but throttling the people who abuse the privileged.

    Here’s the difference: In our terms of service on the contract, it’s spelled out in over sized letters that excessive bandwidth usage will be warned and then throttled to preserve the availability of network resources for the majority of our customers.

    The difference between us and them?

    When I throttle someone, I make sure they know I’m doing it and why. If they’re not willing to control their habit so that it doesn’t impact prime time (after 10P and before 7A), then I throttle. Commcast lied to everyone and denied they were doing it. If they had come out from the beginning and said “These people are putting a drain on our network resources beyond what we feel is an acceptable limit, so we’re throttling them” then I don’t think there would be a big issue.

    Commcast should say the same thing I do. If you don’t like our policy, go buy internet from someone who’s policy you do like. It is, after all, a buyers market.

    JFYI: We only have 9M of aggregate bandwidth with 6M to the internet. Considering all that, we have over 150 DSL customers (1.5M/384K) and 2000 dial up customers who seem to manage not to use up all our oversold resources at the same time.

    Just my $.02

  • Imagine that Comcast increases it’s capacity 100x. Should be enough for everyone? Now, that costs big $: probably in the 100’s of millions to add capacity to the servers, routers, fiber, switches and people to manage it. how long before the network saturates?

    John, with all due respect, your “thought experiment’ fails on many levels. First, you make the assumption that the “upgrade” would happen to all places at once. The upgrade in service would be incremental and “leapfrog” based on profitibility. In other words, if you assume that “D” is the fastest speed that is currently available, you are assuming that Comcast would go from “A” to “D” all at once. They do not. They would go from “A” to maybe “C” in some areas, and then from “B” to “D” in others. Secondly, your assumption that more people would be required may be false as well. What once took 100’s of people to run can now be run by 10 people. Third, you make the assumption that as equipment wears out, it will be replaced by similar equipment, rather than upgrading. Fourth, I believe that you are assuming that internet usage will increase when the last study I read said that use is actually declining.

    I submit to you that part of the problem here is Comcast’s own making. They are the ones that are advertising VoIP services. They are the ones that advertise downloading music. They are the ones that are advertising downloading videos or watching video online. They are the ones screaming “use more bandwidth!”

    The thought experiment now changes. Assume that you have 10 customers that are running P2P and dl’ing 4 gigs a day. Using all the features that Comcast wants to provide and has advertised for you, 90 customers are using half a gig of bandwidth a day.

    We can say that the system saturation point is 50 gigs a day (hypothetically) and given that, who is the real problem? Who has created the real problem?

    The scale of usage that the so called “10%” are using may not be even remotely close to what the 90% are using. Yet it is easier to blame the 10% because in the minds of the public, those who use P2P are evil and breaking the law. It is easier to blame them rather than say “the services we advertised and offer have overloaded our system.”

  • It seems that many of the commentators are attributing the above quote to Comcast. If you read even the byline, it is the Insight CEO who is saying it. I realize that the lawsuit and activities being discussed involve Comcast, but that is no reason to bitch about Comcast based on something Insight leadership said.

  • John, you can blame W and his FCC because they revoked the rules that make the local loop neutral and allowed competition. The phone and cable people lobbied to have monopoly control over their lines and they got it. As a result, every area in the country gets a choice between only two services.

  • So let’s see where we are:
    gitarcarver thinks that increasing a business’s infrastructure won’t take any additional help. That’s just dumb. We’re talking about a massive infrastructure change: it will cost a bloody fortune in personel costs, even if some of the drones can be bullied into working more (good luck with that, BTW: ever worked with cable people?) And of course it happens over time, last I heard there was no way to do everything at once. the point was what happens over a block of time. that’s why it’s a THOUGHT experiment, not real world. And if internet use is declining (I have seen NO evidence for this, the opposite in fact: although INDIVIDUAL use is leveling off, overall use is up sharply as a result of increased accessablility over 3G networks) And the problem we’re talking about isn’t downloading x Gb/DAY, but at what time and at what priority.

    AND OF COURSE the problem is of Comcast’s making. Who said anything else? But they can’t work their way out of it by building an unmonitored network, is all I’m saying.

    BTW, there is a major difference in network usage between VOIP and movie downloading and P2P: They all have completely different impacts, so you can’t luimp them like you did. VOIP is essentially free compared to movies and P2P, and comparing P2P/movies requires more analysis because torrents etc. CAN get out of control (1 person can start many multiple torrents, most movie sites would limit the number of simultaneous downloads, which is what I suggest Comcast do for P2Ps)

    Anyway, if you want to believe that Comcast or Verizon or anyone can build their way out of the problem, go right ahead.

    Jim W, could you check your dates? I believe that you’ll find that the relevant rules were developed under Clinton. I despise both 41 and 43, but not every problem is of their making.

  • Regardless of who is to blame, I think the old local loop neutrality rules were good for consumers and only bad for the providers to the extent that it denied them monopoly rents for their service. We should bring them back in my opinion.

    I was thinking about this a lot last night and on the commute this morning and it occurred to me that congress could tweak the property rights surrounding the local infrastructure that runs to the end-user, maybe give them some sort of equitable right in the disposition of their lines. This would let the original company that laid the lines retain ownership (disrupting ownership patterns could have other impacts down the line) but the consumer gets to choose service providers and backbones.

  • […] In case you missed it, yesterday’s post on the disputes over bandwidth, cable speeds and BitTorrent has prompted an unusually rich discussion with contributions from many knowledgeable readers — I know I’ve learned a lot. Check it out here. […]

  • any good commercial webdesigner will always load the graphics and the ads that generate revenue for the site first, and then the content that the user wants after that.

    My browser’s adblocker, which I have enabled for most sites, stops the ads.

  • gitarcarver thinks that increasing a business’s infrastructure won’t take any additional help. That’s just dumb.

    Yet history within the industry has shown just that.

    AND OF COURSE the problem is of Comcast’s making. Who said anything else?

    You and Comcast for two.

    Anyway, if you want to believe that Comcast or Verizon or anyone can build their way out of the problem, go right ahead.

    Which is it? If Comcast can’t build itself out of the issue for which they are partly responsible, then why are they trying to do just that? If you don’t think that increasing the infrastructure will help, go right ahead.

  • I think John is a troll. Half his comments conflict with stuff he has said previously, and his style is way too abrasive to be anything other than an attempt to provoke.

  • Ook. Ook. That’s me, the ol’ troll. Feel free to have ted or walter verify my email address.

    I really HAVE to learn not to reply to people who have a demonstrated inability to RFC.

  • Getting rid of monopoly rents on the local loop is great for the guy who lives in a big city, but it sucks for those of us who live in the boondocks. There are still a significant percentage of people who have access to no broadband Internet access at all. Forcing whoever wires us up to charge low fees and give other companies access at artificially low prices just ensures that we never get wired up.

    It has in fact had precisely this consequence for me. Broadband was planned for my area in mid 2000, and the plan was killed when the FCC changed the rules.

    I want the first company to be able to charge monopoly rents. Because right now I can’t get broadband for any price. Seriously, I’d have to pay about $750/mo for 1.5Mb/s each way.

  • It seems to me that Comcast has a bad business plan in that they don’t offer different subscriptions for different users. Here in Switzerland, the ISP’s offer packages based on the user needs. For example, Swisscom offers DSL starting at 9 Sfr/m (1 USD = 1.05 Sfr). For this you get 300Kbit download and 100Kbit upload and pay 2.40 Sfr/hr. This is perfect for people who use the net mostly for email and specific information lookups. The next package is 49 SFr/m and gives download speeds between 300 and 3500 Kbits and upload of 300 Kbit and no charge for connection time. The next package is 59 SFr/m and gives download speeds between 3500 and 5000 Kbits and upload of 500 Kbit and no charge for connection time. The final package is 69 Sfr/m and gives download speeds between 8000 and 20000 Kbits and upload of 1000 Kbit and no charge for connection time. The offers very clearly states the range of speeds guaranteed and indicate that actual speed depends on a number of factors. The cable companies offering internet services also have similar package deals.

  • I think we’d all be happy if we could get 8000 and 20000 Kbits and upload of 1000 Kbit and no charge for connection time for $69 a month. Right now with Verizon I’m paying $60 a month, and for the last 2 day my connection has dropped a total of 79 times and when I can get a download test to run I’ve had a massive 75k down and 10 up. I’ve disconnected the router, and connected my computer directly to their modem, their tech support insists that it’s my modems fault that my connection keeps dropping. Yay Monopolies.

  • Wow, I didn’t know the S word was kosher here. Yay?