Change or no change?

Mike at Techdirt asks the question of the day: Why Is There So Little Honesty In The Net Neutrality Debate?

Looking the array of interests assembled under the neutrality banner, it’s not surprising:

A. The end-to-end cargo cult, a group of people who understand virtually nothing about how the Internet is put together, but who nevertheless make their living explaining its implications. David Isenberg, David Weinberger, and friends.

B. Big Content companies who want free rein over facilities bought and paid for with other people’s money. Google, Yahoo, Skype, et. al.

C. Political bloggers desperate to win a significant victory in order to get their consulting rates up: Kos, Armstrong, Stoller, Kelly.

D. Bewildered PACs afraid changes in the network will reduce its effectiveness as the ATM for fringe causes: Christian Coalition, Moveon.org, and their ilk.

The only real potential source of understanding comes from Big Content because the others have no more than a pedestrian understanding of the issues (if that) and Big Content has no incentive to tell the truth.

The Telcos have been sticking much closer to the facts than the forces on the other side.

The bogus “net neutrality” debate comes down to one fundamental question: are we going to allow the Internet to adapt to the needs of the future or are we going to strangle it according to the design of the past?

The Internet is under stress by new delay-sensitive applications such as telephony and streaming video and new protocols such as BitTorrent that suck up all available bandwidth and cause delay for other applications. The old TCP-based congestion model doesn’t work any more, and people expect better QoS for their phone calls than they can reasonably hope to get on a network dominated by BitTorrent traffic. And the business model for the wholesale Internet has never been right.

The net neutrality sycophants with their hippie fixation on the past are the real barrier to technical progress for the Net.

It should be easy for policy-makers to simply dismiss any group that claims “new laws pending in Congress permit your ISP to censor web sites.” That charge is so far from reality it doesn’t warrant a serious response.

14 thoughts on “Change or no change?”

  1. “The Internet is under stress by new applications such as telephony and streaming video and new protocols such as BitTorrent that suck up all available bandwidth.”

    Please leave telephony out of that phrase. IP telephony (when used) eats up about 38 to 88 kbps of bandwidth for G.729 and G.711 CODECs and that includes packet overhead. G.729 is actually only 8 kbps payload so if they lowered it to 25 packets per second from 50 packets per second, the packet overhead gets halved but the packetization latency doubles to 40 ms from 20 ms. Multiplexing the VoIP streams or using packet header compression can alleviate that issue. But even worst case uncompressed G.711 isn’t really that much of a bandwidth hog compared to video streaming or BitTorrent which is the biggest hog.

  2. Telephony isn’t a significant source of load, it’s an example of an application that’s particularly sensitive to corruption by BitTorrent. I re-worded the post a bit to try and make it more clear.

  3. So true.
    My Dad uses a VOIP phone, and the first part of a conversation is usually “let me go pause BitTorrent”.

  4. Funny that you say that in response to a post on Techdirt that totally rips apart the telcos arguments as being totally bogus. It’s pointing out all the lies and dishonest statements in the Chicago Tribune editorial I see you wrote about elsewhere as if it was a *good* argument.

    I’m against net neutrality legislation, as you are, but I actually find the telcos are being a LOT more dishonest in this debate.

    You keep saying the internet is under stress but there’s little evidence that is true. As George points out, VoIP is a very low bandwidth intensive app. There are questions as to whether it could use some QoS, but to say that it’s a problem is silly. As far as I can tell, your argument is that BitTorrent is a problem. BitTorrent is only a problem if the telcos artficially limit bandwidth capacity. It’s a problem they’re creating on their own.

    Meanwhile your claims that Google et al want “free reign” are so laughable I wonder why it’s even worth responding. Would you like to pay Google’s bandwidth bill? You think Google would be buying up so much dark fiber when they get such a “free ride” from the telcos?

    Finally, while I agree that the idea that telcos would censor a website are silly, it’s not quite as easy to dismiss as you’ve stated. The CEOs of AT&T, Verizon and BellSouth have all very publicly stated that they would consider blocking certain websites if they aren’t willing to pay extra. So, you say it’s far from reality, but the telcos have all suggested it’s in their plans.

    I agree that it’s not likely to happen, mainly because the backlash would be extreme, but it’s not quite as ridiculous as you state. The bigger concern is degrading or blocking other services, such as VoIP or video in favor of their own. *That* seems very likely.

  5. You keep saying the internet is under stress but there’s little evidence that is true. As George points out, VoIP is a very low bandwidth intensive app. There are questions as to whether it could use some QoS, but to say that it’s a problem is silly. As far as I can tell, your argument is that BitTorrent is a problem. BitTorrent is only a problem if the telcos artficially limit bandwidth capacity. It’s a problem they’re creating on their own.

    This is wrong on so many levels…

    Bittorrent is a huge problem for telcos (more so for cable cos), because it consumes lots of Bandwidth at the Telco Access layer. (Customers swapping files with other customers). This is a pretty important part of the “Internet” from a customer facing perspective, and that bittorent traffic has impacts other customers. It’s a similar problem they had when HTTP first became popular, and ISPs installed web caches to “fix” the problem while the backhaul and transport networks were upgraded.

    So more precisely, Telco networks are congested and under stress. Obviously, to fix the problems you’ll need add more bandwidth (which they are doing) but you’ll also need to address the problem via ratelimiting or caching, as TCP based protocols like Bittorrent will aggressively consume as much bandwidth as possible.

    Obviously, caching is not a good idea as most bittorent traffic content is illegal, and caching would either make the ISP liable or make it easier for content providers to identify all the pirates. If you use bittorent, you probably don’t want your ISP to do this.

    So.. you’re only other option is ratelimiting, and it seems to be working fairly well as it contains the bittorrent problem.

    So in conclusion: You are correct in stating that the long haul / transport portion of the “Internet” don’t have the same bandwidth and capacity problems as the Last 10 miles /Metro portion of the “Internet”. But a system is only as strong as it’s weakest link, and thus Richard is correct in stating that this portion of the Internet is severly congested.

    Plus, the bittorent issue will need to be addressed again if Content providers decide that it’s a swell way to reduce their bandwidth costs and distribute content…. If they don’t want to pay for their portion of the network traffic, and the customers don’t want it to cause problems with their other applications, than charging for QoS is the only way to make everyone happy.

  6. Please leave telephony out of that phrase.

    While VOIP and telephony apps aren’t bandwidth pigs, they aren’t nearly as reliable as the PSTN in delivering voice.

    So in a sense, it is a stressor if you want your VOIP reliablity to be as good as that provided by the PSTN.

  7. Telephony is not a bandwidth hog in terms of the bandwidth needed to support a conversation. But it is different from things like email and web-browsing in that it is less bursty. People tend to talk pretty continuously for the length of a conversation. This could have an effect on oversubscription ratios (and bandwidth needs) if VoIP comes to be the predominant mode of telephony.

  8. I wonder why The Register links to such an ill-informed blog. While I don’t quite understand categories A, C and D, the second most important category is being left out: users.

    User already pay for their internet connection, and they pay in order to use Google, Yahoo and Skype. If Google doesn’t pay the telcos, does that mean I lose access to Google despite me already paying for that access? If Google does pay up, does that mean my connection will be free or cheaper? Or do the telcos charge twice for a single service?

    Ofcourse the suggestion that Google wants a free ride on facilities they didn’t pay for is completely laughable. As already mentioned, they pay more for their internet connection than you can probably imagine. But tell me this one thing: why should Google pay for something I want?

    If I’m using more bandwidth than I’m paying for, the telcos should charge me more, not Google.

  9. ser already pay for their internet connection, and they pay in order to use Google, Yahoo and Skype.

    Users pay for their best effort internet connection, which is unsuitable for PSTN quality VOIP.

    f Google doesn’t pay the telcos, does that mean I lose access to Google despite me already paying for that access? If Google does pay up, does that mean my connection will be free or cheaper? Or do the telcos charge twice for a single service?

    The only thing that is clear by your statement is that your a little Out of your depth. No one wants to block Google. It’s impractical at best, and business plan suicide at worst.

    Srvice providers would like to charge you extra for QoS, something that is not useful for Search Engines, Email, or conventional webpages. It is extremely useful for applications like VOIP, streaming video, or for traffic shaping bittorent, and other next generation applications that are delay or jitter sensitive, especially when those two types of applications (Conventional and Realtime) share the same wire.

    Regards,
    Max

  10. Nobody has ever said they would block or degrade google; in fact they have said the opposite all along.

    The biggest beneficiaries of the non-net-neutrality are the users, not the telcos. For example, if you are an at&t customer, you’ll be able to pay $15 for your 1.5 dsl, but you’ll be able to download a movie to watch on your tv through your set top box in a manner of minutes, instead of have to wait over night. That is because 3rd party providers will pay at&t to increase your bandwidth to them for the download. That benefits everyone- the content provider, at&t, and most of all, you.

  11. Adam V says:
    “So true. My Dad uses a VOIP phone, and the first part of a conversation is usually ‘let me go pause BitTorrent’.”

    You need a router that supports QoS and prioritization. They even have hacks for the Linksys WRT54 that will prioritize certain data stream types. Skype of course is a lot harder to identify. This way you don’t need to stop BitTorrent but its upstream will just get a lot slower. A simple solution is to just cap your BitTorrent’s upstream and downstream to leave some room but you still want QoS on your own router’s outbound stream.

    Max says:
    “Obviously, caching is not a good idea as most BitTorrent traffic content is illegal, and caching would either make the ISP liable or make it easier for content providers to identify all the pirates”

    BitTorrent supports encryption for the purpose of avoiding bandwidth caps and blockage. If the ISP would cache BitTorrent deep inside the cloud, it would certainly lessen the strain on the leaves of the Internet. Caching encrypted content for the purpose of optimizing Internet flow isn’t illegal since there is no way to tell what you’re caching anyways. All the ISP is doing is optimizing the pipe.

  12. Caching encrypted content for the purpose of optimizing Internet flow isn’t illegal since there is no way to tell what you’re caching anyways. All the ISP is doing is optimizing the pipe.

    I know where you’re coming from, but it is kind of a gray area, which explains why ISPs aren’t really eager to adopt the technology. The overwhelming percentage p2p traffic is arguably pirated material, unlike HTTP, and the caching systems haven’t been developed for the encrypted versions of bittorrent, much like SSL content is not cached today.

    Beyond that, placing such a device on network would allow for a single point of transactional logging for the p2p traffic, providing a single source of logging information that doesn’t exist right now. Such logs would be of great interest to the RIAA and MPAA, and I’m sure most ISPs already get enough subpoenas to keep them busy. You, as the ISP may not be liable, but it gives these organizations the ability to more easily go after your customers, which can be just as bad.

    What is really interesting is how Big Content is adopting the technology to lower their bandwidth bills, and how this tactic doesn’t really get mentioned in the context of the NN debate, but it really should.

    P2P has a lot of potential for large content providers, it reduces bandwidth costs by offloading the filetransfer to the end users. Disney already uses it, or is thinking about it, and I think a couple of the other networks are starting to use the system to distribute episodes of “Lost.”

    However, P2p has some drawbacks.. It’s slow, there are security problems, etc.

    In order to fix these problems, you’ll need to QoS or p2p caching systems. Obviously, content providers don’t want to pay for these, it defeats the entire point of switching to p2p delivery mechanisms in the first place… better to legally mandate that you get the service for free by passing garbage like the Markely and Snowe(job)-Dorgan amendments.

  13. No matter what anyone says, regulation is regulation. More laws do not enable the creation of new markets to serve the needs of customers. Laws designed by any group are created to protect that group’s interests. The best approach would be the removal of all barriers in this market. Let information providers, infrastructure providers and consumers sort out the winners. The only reason the telcos are in the middle of this thing is because of their legacy infrastructure. Let innovation determine the best option for completing the economic transaction of meeting my needs as a consumer. If Google wants to run fiber optic to my house and charge me $5/mo for unlimited usage, then you better stay out of my way.

    Check out this article on Somalia.
    http://news.bbc.co.uk/2/hi/africa/4020259.stm

    They don’t even have a functioning government yet have lower phone bills and in many cases better service than I have in the US with our seemingly endless layers of regulation designed to protect someone who is obviously not the consumer.

Comments are closed.