[exelius]

I don't know that this is the solution (or even that there's a problem). I know my opinion isn't very popular on HN on this issue; but I continue to share it because I feel it's important that people understand the view from the other side of the fence. I expect to get downvoted because people disagree with me, but then magical internet points never really mattered much to me.

I think all this talk of the "slow lane" is a bit tinfoil-hat. Companies like Comcast have no interest in slowing down web site traffic; in fact they do a lot of QoS to make web browsing faster and more responsive. This type of traffic (DNS, HTTP requests, online gaming, etc.) tends to get put in a high-priority QoS class: the data transmitted is often small and it greatly improves the user's experience. ISPs have a huge incentive to make this type of traffic as responsive as possible; and given the low bandwidth requirements, this should definitely be possible. It makes their service "feel" faster to the customer and it's the right thing to do for the customer.

Video streaming services are another story. They don't need to be responsive because they pre-cache a lot of data; in fact the right thing to do from a technical perspective is QoS them into the basement. Video can handle this; it's high bandwidth and low latency. The thing is, streaming video accounts for about 80% of peak Internet traffic. A small percentage of users (~30%) are starting to overload the ISP's last-mile networks with video traffic.

The types of high-bandwidth scenarios that the ISPs will be pushing the "fast lane" on are going to be almost exclusively video streaming services. Video streamers have had to pay CDNs for years anyway if they wanted their videos to stream quickly. The idea is that because these services have such a disproportionate effect on bandwidth usage, they need to contribute economically to avoid a tragedy of the commons [1] situation. Your average website or app that's not pulling 1.5+ mbit/s over an extended period of time is likely going to be fast regardless because it's in the ISPs best interests to make it that way.

[1] http://en.wikipedia.org/wiki/Tragedy_of_the_commons

[devcpp]

>Companies like Comcast have no interest in slowing down web site traffic

Yeah? How about money?

As a comms engineer, I too think about the technical implications of this.

But you have to remember that the people who push for the abolition of net neutrality are mainly the finances guys, i.e the ones responsible for bringing as much money as possible to the company. And when you put yourself in their shoes, all of a sudden you get dreams of Comcast turning the Internet into the same kind of market it has in cable television. And you very quickly forget about QoS, bandwidth, latency and the job of "delivering the bits" and just think in terms of profits. You wouldn't know what most of those terms mean anyway...

[exelius]

Well, in short, any web site with any money to pay the ISPs is likely already running on a CDN that's been paying the ISPs for a decade. The portion of that market that's not yet monetized is likely very small. Besides, slowing down general-purpose web traffic makes it look like your service just totally sucks ass.

But more generally, slow-loading websites make the ISP's service look shitty. There's almost no cost for them to just QoS the low-bandwidth stuff up and it makes people think their service is better.

[orthecreedence]

> But more generally, slow-loading websites make the ISP's service look shitty

Exactly, so people will just leave that ISP and sign up with one that doesn't throttle traffic. Except most people have only one ISP in their area. Yay free market!

[exelius]

Even in a truly free market, most areas would only have one ISP. See how Verizon and AT&T have dramatically scaled back or stopped their FiOS/U-Verse rollout plans. If a company with easements and existing infrastructure (and thus lower costs) has decided it can't justify the investment, how would anyone else?

[orthecreedence]

I was saying "Yay free market" as in "a truly free market is a bad idea as seen by local ISP monopolies," not as in "our free market has failed us and needs to be freer."

In a truly free market, there would be one company that owns everything, including infrastructure. Not a world I'd like to live in.

[randac]

You only need look outside the US to see that's garbage.

[a_c_s]

As it is Netflix pays for bandwidth and the end user pays for bandwidth.

Why should Netflix pay disproportionately more for its bandwidth than a non-video streaming service?

[tempestn]

Exactly! Netflix is already paying for the upload bandwidth they're using, and Comcast customers are already paying for their download bandwidth. That should be the end of it. The problem is, because of services like Netflix, more people are actually using the bandwidth they're paying for. Comcast takes issue with this because they can no longer massively oversell their infrastructure.

[redthrowaway]

>Companies like Comcast have no interest in slowing down web site traffic

You can't think of a few sites and services that TV cable providers would like to slow down? The issue isn't that the general web will get slower, but rather that cable companies will become the arbiters of which services survive and prosper.

[pdq]

The other part that virtually no one talks about is how this would be regulated. Are we going to have "internet inspectors" from the FCC doing periodic surprise checks of every ISP? Will ISPs be required to have a neutrality license with full historical audit logs?

This reminds me of when the TSA took over airport screening after 9/11, over the belief that only a government agency could screen passengers securely. The net effect was longer lines, more expensive travel, less rights for flyers (ie either you get a body scan or a pat-down, and no more liquids on flights), and in general less happy travelers.

[devcpp]

Same way the NSA would be regulated against spying on the general population: remove money from the equation.

Think about it, the only reason Comcast is doing all this is so they can charge premium fees on certain services. If you don't allow them to do so (which is very easy to enforce, as customers can report such pricing policies), they have no incentive to be against net neutrality.

This is why the current status quo has worked for so long: ISPs have no way to legally make profits out of "premium" traffic, so they (generally) don't apply outrageous QoS measures. Money is the only incentive, and removing that incentive solves the problem without the need for active policing.

[zo1]

"This is why the current status quo has worked for so long: ISPs have no way to legally make profits out of "premium" traffic, so they (generally) don't apply outrageous QoS measures. Money is the only incentive, and removing that incentive solves the problem without the need for active policing."

And because they weren't allowed to charge extra for "premium" traffic, they essentially switched over to the over-subscription model in order to make profit. The tighter you squeeze, the stronger it oozes out the sides. And never in the place or way you wish it would!!

[wmf]

QoS mostly decides which packets get dropped, but ISPs shouldn't be dropping any significant fraction of packets.

(I was tempted to downvote you for your persecution complex, but that would only validate it so I'm torn.)

[exelius]

Packets get dropped when links are saturated as they often are during peak hours. I know a network admin would think "well the links should never be saturated" but the economics of that just don't work in the consumer ISP space. You aim for a target level of service that means certain portions of the network will be saturated, and use QoS to ensure things like DNS and HTTP remain responsive.

Regardless, the point is that QoS exists to keep non-video traffic from getting trampled out by video traffic. Video traffic is such a disproportionately large amount of total Internet traffic that virtually the only services that would be significantly impacted by being in the "slow lane" are video services. The links wouldn't even be saturated in the first place without video traffic.

EDIT: Also wanted to add that congestion is almost guaranteed with adaptive bit rate streaming. Netflix will use as much bandwidth as is available up to like 9 mbit/s.

[wmf]

the economics of that just don't work in the consumer ISP space

A citation is needed here. Somehow consumer ISPs got along for 15 years without "permanent congestion".

[exelius]

Because they didn't have 50% of their users trying to view 6 mbit/s video streams.

EDIT: Just wanted to clarify that 50% is a number I pulled out of my ass. The general point is that the percentage of users who stream video online on a regular basis is increasing more quickly than the economics/logistics allow the ISPs to perform network upgrades.

[AnthonyMouse]

> The general point is that the percentage of users who stream video online on a regular basis is increasing more quickly than the economics/logistics allow the ISPs to perform network upgrades.

This is another citation needed point. Bandwidth consumption has always increased over time, as has the performance per dollar of routing equipment.

[mhandley]

Well, the goal of a properly designed transport protocol is to transfer X bytes in 0 seconds. TCP doesn't quite achieve that, but that is its goal. This pretty much dictates †hat for any reasonable sized download, somewhere on the path is congested. If that congestion isn't at the server (it usually isn't), and isn't in the backbone (it usually isn't), then it's either your consumer ISP or your computer (it usually isn't). ISP's tend to oversubscribe backhaul to make the economics work, so if you're not getting your full linkspeed, then it's likely the backhaul that's congested, and everyone else is seeing congestion too.

There are exceptions of course. Older OSes have ridiculously small TCP receive windows, and older TCP congestion control algorithms have trouble filling the pipe. But these shouldn't really be the main problem these days. It doesn't apply so much for streaming either, because modern streaming protocols such as MPEG-DASH will select a lower bitrate stream if they sense congestion.