Democrats unveil legislation forcing the FCC to ban Internet fast lanes | Hacker News Reader

Democrats unveil legislation forcing the FCC to ban Internet fast lanes

washingtonpost.com

161 pointsspenvo104 comments

Instead of limiting my ability to enter contracts and chose the Internet/streaming service I want, I would much rather they devise a methodology for measuring and enforcing connection performance.

For example, "broadband internet" means at least 50Mbit. "50 Mbit" means that 90% percent of the time, there is a sustained throughput of at least 5MB/s for a TCP connection to a server in any data-center across the continental US (assuming the data-center link itself is not congested). "Unmetered" means that per month, the user can transfer a volume of data of at least 10% of the volume corresponding to a saturation of the connection 24/7 (1300 GB for a 50 Mbit connection as defined above).

Under such a rule, Comcast's connections could only be sold as "Dial-up+". And if broadband is not available in an area it's fair game for the municipal or state authorities to provision it with public funds as a basic need for a modern citizen; since the incumbent private provider obviously can't get a return on it's investment (what OTHER reason for not developing your infrastructure could there be, right ?)

I had to create this account just to correct this. "Broadband" means the data is modulated on to a wide spectrum signal, on the same copper as the POTS network. Anything from good old 512kbit/s upwards can be "broadband". I say "can be", because if it's not using a wide spectrum signal, e.g. fibre-optics, then that's not broadband.

It wasn't a statement of fact, it was a wish list that started with "For example"

Ok, thank you for pointing that out. However I still don't think it's right to "re-purpose" words. Remember what happened when people tried to re-purpose kilo for 1024 etc.

"Broadband" is a well-defined technical term originating from physics. But "broadband internet" (what the OP was referring to) is a vaguely defined marketing term that is often used by ISPs to deceive consumers as to the speed of their connections. The OP is suggesting that that term be repurposed as a clearly (legally) defined one that improves internet access in America without introducing onerous regulations.

The FCC does have a definition of broadband: currently 4 Mbps, possibly to be redefined to 10 or 25 Mbps. http://www.washingtonpost.com/blogs/the-switch/wp/2014/05/30...

The trouble with all this is that it is reactive rather than proactive.

There is a simple solution that would put an end to all this BS, and this is to separate the last mile from the middle mile the same way that the old AT&T was broken up to split local from long distance service. (This is how they do it in the UK and UK people are very happy with their internet service.)

In this case you get a choice of different middle miles and different television services from different providers. If work got around that "the Verizon network is congested" people would just switch, and I think you would find network connectivity problems get fixed in days rather than years.

> UK people are very happy with their internet service.

I wouldn't go quite that far. There's still lots of grumbling over the speed of fibre rollout. And quality. Despite fibre to the cabinet, I'm limited to 40Mbps due to line quality, for example. This is largely down to OpenReach (the company managing the last mile). But within the constraints of line quality, there's decent room for competition.

I would love to get anything near 40Mbps.

My broadband speed recently dropped from 2.2Mbps down to 1.6Mbps and BT basically said that it was still an acceptable speed for my area and only way I could get a faster speed was switch to fibre. They won't even look into it as the line is working so it can't be logged as a fault!

Apparently there will be fibre in my area in December, but given track record on rolling these services out where I live I won't bet money on it happening that soon.

Seems like a much more 'market oriented' solution. The problem with getting specific is that it ties you in to a specific solution.

>The problem with getting specific is that it ties you in to a specific solution.

Agreed -- would Google have the same incentive to roll out Google Fiber if they could just use (slower) existing coax/cable lines as a middleman? It could limit innovation from that perspective, although it depends if a company like Google has alternative reasons for rolling out their own last-mile fiber (not just data collection en masse).

And after rolling out, Google would then be required to let anyone else use the last mile they built.

> And after rolling out, Google would then be required to let anyone else use the last mile they built.

Either way Google would be getting paid for the cost of providing that last-mile service so it'd really come down to a question of whether someone could profitably outcompete them on either bandwidth pricing or other service value-adds.

I rather doubt that prospect would keep anyone at Google up at night – but it would cause plenty of heartburn in the Comcast/Verizon/AT&T etc. executive suites.

> Either way Google would be getting paid for the cost of providing that last-mile service so it'd really come down to a question of whether someone could profitably outcompete them on either bandwidth pricing or other service value-adds.

When something involves capital investments of billions of dollars, companies don't just want to eke out a small profit. They want enough profit to justify putting the money in that industry as opposed to some other one. If you can make 30% returns over here in social networking, why spend billions of dollars digging holes in the ground for 5% returns?

> I rather doubt that prospect would keep anyone at Google up at night – but it would cause plenty of heartburn in the Comcast/Verizon/AT&T etc. executive suites.

Sure, but only because Google makes 98% of it's money in a very high margin advertising business. Meanwhile, Comcast/Verizon/AT&T actually have to placate their shareholders with the money they make building infrastructure.

> If you can make 30% returns over here in social networking, why spend billions of dollars digging holes in the ground for 5% returns?

This point is meaningless without absolute values for comparison: 30% of small is less than 5% of large.

The other point is the one that Google states every time they do something network related: they're an internet company and they benefit from more people and more activity. If US broadband stagnates, that means things like cloud storage, video, etc. are not going to advance significantly beyond their current levels – and if network neutrality isn't maintained, anyone who doesn't own a network is going to be frozen out. That gives Google a strong incentive to invest in the capacity both so they can expand their network offerings and to provide a clear, obvious example of how much better service can be than what the rent-seeking monopolists choose to provide.

> Sure, but only because Google makes 98% of it's money in a very high margin advertising business. Meanwhile, Comcast/Verizon/AT&T actually have to placate their shareholders with the money they make building infrastructure.

All of those companies are making far more money than they spend on their infrastructure. They're trying to expand into other areas to make more but they're not doing it because they're going broke charging far in excess of the operating costs.

> That gives Google a strong incentive to invest in the capacity both so they can expand their network offerings and to provide a clear, obvious example of how much better service can be than what the rent-seeking monopolists choose to provide.

It gives Google a strong incentive to invest, but what about companies that just build internet infrastructure?

> All of those companies are making far more money than they spend on their infrastructure.

Their profit margins are much lower than Google's.

> They're trying to expand into other areas to make more but they're not doing it because they're going broke charging far in excess of the operating costs.

The irony of your position is that you fault Comcast, etc, for branching out in order to make more money from other areas, but justify regulation that would make building infrastructure minimally profitable on the basis that companies like Google can still make a bunch of money using that infrastructure to push advertising.

You can't have your cake and eat it too. If you make building infrastructure less profitable, companies will invest less in that industry, and the ones that continue to do so will do it with the intent of peddling some other product on top.

> The irony of your position is that you fault Comcast, etc, for branching out in order to make more money from other areas, but justify regulation that would make building infrastructure minimally profitable

You should read my comment again because you're arguing against a claim I didn't make. I don't fault them for branching out but merely noted that they weren't doing so out of financial desperation — the network is hardly a loss-leader.

If you can make 30% returns over here in social networking, why spend billions of dollars digging holes in the ground for 5% returns?

Because a market with 30% returns is going to attract a lot more competition than one with 5%. Look at Jeff Bezos's philosophy. He's deliberately attacked a very well established, low-margin business and he's made Amazon extremely successful at it.

> UK people are very happy with their internet service

Heh. I'm sure its a lot better than what you're working with but personally coming from Sweden I feel like in the U.K. I'm getting worse service (offers, speeds, customer service and support, waiting for and dealing with BT Openreach engineers, hardware not working, getting the run-around between Openreach and utility providers, ..) for more money than we did back home.

That being said I'm sure dealing with Comcast, Time Warner and AT&T is a million times worse.

I second that. I move from Poland to UK and the service is a lot worse in every way.

As an American I find that Poland's service being so great seems so embarrassing to me. How can we embarrass our ISP and our politicians to use that fact to upgrade what we get? No offense to Poland but the US is supposed to lead the world in stuff if for no other reason than our huge size and economic power. Yet our internet service is far below average? What does Poland do that makes this work so well?

On Ookla's NetIndex, Poland is slower on average than the U.S., and significantly slower than U.S. states with comparable density, such as Pennsylvania or New York.

The reason everyone is so impressed by countries like Poland is that their central governments proactively got fiber installed in major cities. In the U.S., that sort of thing is up to state and municipal governments. What's stopping us from wiring up San Francisco or New York or Chicago with fiber? Insane local politics. Right now, people in Chicago are flipping out because of the signage on the Trump tower. People in San Francisco halted FTTN deployment because they didn't like the look of AT&T's network cabinets. Even Google doesn't dare enter these markets.

If AT&T used modern FTTP equipment instead of legacy FTTN equipment, there wouldn't be so many ugly, expensive, unreliable and vandalism-prone cabinets.

Don't get fooled by the phone company, FTTP is the least cost solution in the long term in EVERY situation.

Also this is only a problem in a few big cites; if you go to any normal community in any of the flyover states, Democrat or Republican, they want good broadband so bad they will bend over backwards to get it.

FTTN is both fine on its own right, and a good incremental step toward FTTP. Its the basis for BT's broadband strategy in the U.K., which everyone holds up as great model.

As for this not being a problem flyover country, you're right. But flyover country also doesn't have the density for good bang for buck infrastructure development. The reason why places like Poland and Romania stand out is they combine the density of major American cities with the desparation to stand out. Warsaw wants to be the next San Francisco. San Francisco doesn't.

rayiner could be right. I noticed a few times when one district builds a road and they fail to cooperate with another , so the road ends in the middle of nowhere. Local governments, due to corruption and political reasons cannot cooperate effectively.

I do believe in free markets and little regulations, but in this case central government should act.

Spent nearly 5hrs on the phone with Comcast last week to get them to un-screwup my account and provide the service they agreed to provide. It's working now, but now they're billing me for someone else's account as well. Comcast has fully accomplished being the worst company in america.

> There is a simple solution that would put an end to all this BS, and this is to separate the last mile from the middle mile the same way that the old AT&T was broken up to split local from long distance service

It would be extremely challenging, legally. If we're talking about cable rather than copper, the last mile was built by private companies with private money. BT, in contrast, started out as a government-owned corporation, and when it was privatized, the terms of the separation were written into the prospectus.

> (This is how they do it in the UK and UK people are very happy with their internet service.)

Yet U.K. internet really isn't faster than U.S. internet in areas with comparable density. According to Ookla's Net Index, average broadband speeds in say Pennsylvania, a state with less than half the population density of the U.K. is about the same as in the U.K. (28.3 mbps versus 28.6 mbps). Akamai's data shows the U.S. ahead of the U.K. in average connection speeds.

Sure they can. It was done in the 1990's in the US. Regional Bell Operating Companies were forced to allow other telcos to provision service on their infrastructure. At one point, I had home phone, long distance, isdn and mobile on one bill from Sprint.

That infrastructure was all built while AT&T was a government-sanctioned monopoly with regulated rates and the whole deal. The vast majority of the cable infrastructure was built with private money, post-deregulation.

Maybe it varies by state then.

In Upstate NY, you can buy cable services from Time Warner, Earthlink or a couple of smaller carriers. In the past, this was a good deal, now the 3rd parties don't seem to have access to higher speed services.

There where multi-billion dollar subsidies to build the last mile. More importantly, there is no legal protection from a monopoly being broken up. And local ISP's clearly qualify as monopoly's.

When did the cable companies get multi-billion subsidies to build the last mile? And the local ISP's are not monopolies: almost everywhere the cable companies are in competition with wireless, DSL, and satellite. I know there's a meme on HN that "wireless isn't real competition" but the fact is that wireless is where the demand is right now. Tons of people have no wireline service but everyone has wireless.

competition with an inferior technology outdated product. What a competition. I can see Comcast sweating.

For most people, wireless is a superior product. I don't know if you have noticed, but the 1990's futuristic ideal of a big fat desktop connected to a fat fiber pipe isn't what people actually ended up wanting. They want a thin light tablet connected to wireless internet they can take on the go. Demand for that "inferior technology" is increasing much faster than demand for wireline technology.

And even more people buy toilet paper. And yet toilet paper is also not real competition for cable broadband. Counting customers doesn't establish real competition. I have both cable broadband and wireless. Neither is a meaningful substitute for the other. I can't watch movies over the wireless and I can't take my cable modem out with me, so I'm not sure what you're getting at by pointing out that wireless is growing faster.

The average smartphone bill is like $160 a month, and charges $10 a GB once you go beyond somewhere around 250MB to 10GB.

Contrast to $70-$80 a month for fiber optic that, if capped, is somewhere around 300 GB.

Verizon and AT&T won't give us optical fiber not because it is "too expensive" but because it is too cheap.

If we had universal optic fiber we could have good wireless coverage where everybody lives and works for 1/10 the OPEX of expensive and high latency LTE. Legacy LTE technology is great if you want to play angry birds when you crash your car, but other than that it is a horrible way to provision internet bandwidth.

The average smartphone bill is like $160 a month,

For a family of 4, or for an individual?

And that thinking is how they screw you.

The fastest way I've seen people under 20 end up with a 300 credit score is when they think they are saving money with a "framily" plan. Then things go bad, then they have to pay cancellation fees, pay back handset subsidies and pretty soon it's worse than the average drug deal gone bad.

The cable companies last mile generally runs over copper so you need to look into the initial subsidies when it was first laid down, the most common being a local monopoly.

The sad thing is local, state, and federal 'broadband' subsidies are ongoing. Read up on http://www.ntia.doc.gov/report/2014/nineteenth-quarterly-sta... for some federal info, state level subsides are more complex.

PS: Some of this is hard to pin down, there are 'taxes' designed so cable companies can advertise lower prices than they actually charge. http://www.alternet.org/story/148785/cable_companies'_$46%2B...

> The cable companies last mile generally runs over copper

I use "copper" in this context to refer to POTS lines, not coaxial lines (sorry, it's jargon-y). The two were built under very different regimes, and it's not productive to conflate them.

> you need to look into the initial subsidies when it was first laid down, the most common being a local monopoly.

In general, a "subsidy" is something that yields a return to the company or industry above the market return. A local monopoly is not necessarily, and usually isn't, a subsidy. The monopoly "carrot" is combined with a yoke: universal service requirements and regulated rates. Local utility monopolies generally earn lower returns than unregulated companies. Indeed, it's the height of irony that you call local monopolies a "subsidy" then cite Kushnick further down, whose whole shtick is that deregulation (i.e. elimination of the local monopolies) was a huge "subsidy" to the telecom companies.

Also, the cable industry was deregulated in 1992. DOCSIS 1.0 wasn't released until 1997. I doubt much of the infrastructure that was built under exclusive franchises is still there, because the mid/late-1990's involved widespread network upgrades to accommodate cable internet. Whatever is still there is more like the "last 100 feet" not the "last mile."

> Read up on http://www.ntia.doc.gov/report/2014/nineteenth-quarterly-sta.... for some federal info, state level subsides are more complex.

The BTOP is one of the rare federal telecom programs that's actually a subsidy, and very little of that money has been handed yet. The vast majority of the "subsidies" for broadband are actually transfer payments from the Universal Service Fund. This amounts to billions per year, but it's funded by a tax on the telecom industry. It's as if you taxed Macbook Pro owners 15% and distributed that money to help poor people buy Macbook Airs. Nobody would call that a "subsidy" to Apple.

Your last article is Kushnick's usual Glenn Beck-esque "connect the dots" bullshit. He's stuck in 1970 and think it should be the government's job to set prices for telecom services, and tries to count any increase in price as a "subsidy." Does it matter what a cable company calls its price hike? Is Uber's $1 "safe ride fee" a "subsidy?" It's absolutely inane to call an increase in the price of a product a "subsidy." Subsidies necessarily involve some sort of external transfer, not transfers within the contractual privity of a buyer and seller.

The monopoly bit is a subsidy, sure there are strings attached but they could also be attached to using government land like roads. Afterall good luck building a network without crossing a road.

I don't agree with Kushnick, my point is if a company advertised bubblegum for 50c and charged 90c when you got to the store it's illegal false advertizing. The telecom industry is vary much in favor of being able to do the same kind of price manipulation.

PS: Subsidizing undeserved or low population areas is paying for the last mile, it's not like they don't get to charge the new customers.

> The monopoly bit is a subsidy, sure there are strings attached but they could also be attached to using government land like roads.

The rate regulation and universal service requirement more than outweigh the advantage from the monopoly. At the end of the day, regulated local monopolies make lower returns than unregulated companies.

> PS: Subsidizing undeserved or low population areas is paying for the last mile, it's not like they don't get to charge the new customers.

That money doesn't come out of the government's pocket. It comes from a tax on the company: http://en.wikipedia.org/wiki/Universal_Service_Fund#Backgrou... ("As of the first quarter of 2013, the USF fee, equals 16.1 percent of a telecom company's interstate end-user revenues. As of the second quarter of 2013, the USF fee is 15.5 percent.") Second, the telecom company is not allowed to charge the new customers the cost of actually providing the service.

> It would be extremely challenging, legally. If we're talking about cable rather than copper, the last mile was built by private companies with private money.

What about the government claiming the cable under eminent domain? You can pay the companies "market rate", and when you're done you can lease the lines to anyone, including the original company?

Comcast is buying slightly over 10% of all US broadband subscribers for $45B, so let's say the entire US broadband infrastructure is worth over $400B. That's some serious eminent domain.

True, but they're buying a business. A purchase that's strictly infrastructure would rule out the business relationships and contracts and would be significantly cheaper, especially given that the existing customers are free to remain customers (although who's to say what they'll do when they suddenly are exposed to the competition...)

> If we're talking about cable rather than copper, the last mile was built by private companies with private money.

A significant amount of this infrastructure was built with government subsidies and stimulus money by local monopolies with no competition.

> Yet U.K. internet really isn't faster than U.S. internet in areas with comparable density.

It's not so much about denser urban areas so much as the typical case. I still know people in the US who can barely get 3 megabits on their connections no matter what speed they pay for, and they live in proper cities and not tiny communities.

I tried to get a proper example, but Verizon's website[1] is somewhat inscrutable, and tells me that my friend in Seattle can't get DSL service. Still, their DSL plans they list go 'up to' 15 megabits down and 1 megabit up; in Vancouver, they go up to 50/10, but my plan is 25/5 for $65/mo on its own (no contract). I would be surprised to see that in the US.

[1] http://www.verizon.com/home/highspeedinternet/

> A significant amount of this infrastructure was built with government subsidies and stimulus money by local monopolies with no competition.

No it wasn't, if we're talking about cable and not copper. The cable companies industry is not the phone industry.[1] Almost all the existing cable infrastructure was built after deregulation of the industry in 1992, in which granting exclusive cable franchises was made illegal under federal law. Indeed, almost all municipalities use their cable franchises as a revenue source: imposing a franchise tax as well extracting lump-sum payments to support internet service for municipal buildings, public access television, etc.

[1] And even in the phone industry, the subsidies overwhelmingly go to supporting service in high-cost rural areas, where no profit-minded company would otherwise bother to build infrastructure.

> If we're talking about cable rather than copper, the last mile was built by private companies with private money.

...built on public easements.

Which all these companies pay for.

someone correct me if I am wrong, but the "the Verizon network is congested" was not an indication of the slowness of the overall Verison users, but rather the pipe between Verison and Netflix was congested.

Just as a counterweight to inevitable moaning about UK broadband - I am very happy with what I'm getting from Virgin Media. Rock solid, no traffic shaping, 120Mbit line for not much money is not bad at all. I was happy before as well, with a DSL line. There is competition in the market and it seems to drive the tech and prices in the right direction (unless you live in a rural area...)

Some of my friends leave in Hyperoptic-enabled area of London where they get a stable Gigabit service.

Those of us old enough to remember how awful the local/long-distance split was don't necessarily want to return to it.

The problem isn't that companies are allowed to offer different levels of service for different prices... it's that the existing ISPs have switched to a rent-seeking business model. Once a company does that, corporate culture makes it impossible for them to switch back to the more benign good-service-for-fair-price model.

The existing companies just need to die and be wiped from existence. There are other companies that aren't doing the rent-seeking thing, and we all love them Sonic, Google Fiber, etc.

Sure, let municipal fiber bloom in those places that can manage the trick, but don't mandate it top-down for the entire fucking country.

If you live near enough to an exchange for ADSL over BTWholesale's elderly copper wires to be fast or in a big enough town for Virgin/BT to have laid cable/fibre, your net service is pretty good in the UK - from what I hear better than in the US.

If you live further away so Wholesale has no competition from Virgin they don't lay fibre or maintain the last mile and non-town service is (again, purely anecdotally here) worse. Non-town is not rural, it's maybe 3 miles away from an exchange.

London ain't the UK, basically.

This is legally problematic, as others have posted already. It would amount to nationalization of private infrastructure.

I personally favor a different solution: liberalize frequency allocation and open up quite a bit more of the radio band for last-mile local/regional wireless.

The cost of building out infrastructure rises exponentially as you approach the last mile, for obvious reasons. Allowing the last mile to be wireless opens the field for quite a bit more competition. You'd have a whole raft of local and regional Internet service startups.

Let’s spend a minute to think about political language. Do we want to “ban Internet fast lanes” or “ban Internet slow lanes?” I suspect that most would favor the latter.

It is indeed a poor term. I suggest calling it "bandwidth racketeering" ("Nice streaming service you have here, it'd be shame if some traffic shaping happened to it...")

What I don't get is why the FCC can't simply tell ISPs to provide the service their customers pay them to. Seriously. By paying for internet access, the customer pays the ISP to deliver any packets sent from or to their machine. This is nothing less that a protection racket in order to get paid twice. "Dat's a nice video streamin' site you got there, be a real pity if you couldn' get any customers".

> What I don't get is why the FCC can't simply tell ISPs to provide the service their customers pay them to.

Contract law already does that. The transparency provisions of the previous and proposed Open Internet orders aim to solve the main particular problem of that in the ISP world, which is ambiguity on what exactly the service being paid for is.

(The anti-blocking and anti-discrimination provisions, which are different in the old and new orders, go further and aim to align what the service being paid for is with a certain conception of what that service should be.)

People "paid for" a shared network, with all the benefits and drawbacks of such. You can have a private pipe if you want to guarantee no congestion at all, but expect to pay significantly more for it.

I really think this is a bad idea. If you're a business and depend on a certain type of traffic (say Skype, or maybe email) why wouldn't you consider the option of trying to ensure faster and more reliable delivery of some type of data that is absolutely critical? Maybe as a business, you actually want something like Netflix throttled to some extent so that Skype chats that you depend on are 100% rock solid. Why restrict the possibility for new types of services to emerge from actual needs?

The real issue with this isn't the lack of network neutrality, but all kinds of local and other types of regulations that make it extremely difficult for all but a few connected big companies to take part in. This is why each region has maybe 1 or 2 really shitty ISPs rather than a decent array of choices.

You throttle content (QoS) at the local lan level. If you're paying for service, you should have equal service with all other customers on the same network outside of the lan.

You can do that with QoS on your own network. Even entry level routers can do that. The default verizon one, for example, can do it. You don't need to damage the internet for this.

What you're describing is QoS. Packet discrimination by kind. Bulk file transfer vs streaming video vs VoIP. Necessary because because the impact if your torrent downloads go a bit slow is your torrent downloads a bit slow. If video or VoIP get congested, the effect is the service becomes unusable.

Net Neutrality is about packet discrimination by source and destination.

It's not a problem if VoIP gets higher priority over web surfing and arbitrary file downloads. It's a huge problem if the ISP's VoIP service has higher priority than Skype, despite them being identical traffic.

Net Neutrality is about packet discrimination by source and destination.

Net Neutrality seems to be a No True Scotsman where if you ask 10 people on HN what it really means, you get a dozen different answers.

There are some kinds of NN I support, and some I don't. I would sign up for the definition in your comment.

It's not really confusing unless you ask the ISPs. NN is all about treating all traffic equally. No QoS on the ISP level.

Thank you for the demonstration of "what NN really is".

ISP QoS is fine as long as it's by type and not by source/destination. It's perfectly reasonable to expect all those people torrenting the latest Game of Thrones shouldn't impact other people's Skype calls.

This isn't a value judgement on anything, mind, it's all about what the application needs in order to work. A torrent can still download on a congested network. A VoIP call will have great difficulty on same.

If the ISP can't pipe the advertised bandwidth that's not the user's fault. The torrenter's absolutely should be allowed to fill up the pipes. As for torrenting, there's also a uTP which is supposed to help mitigate some issues at the ISP level.

That all said, if you don't agree that QoS should not be allowed at the ISP level than you don't agree with NN. It's really not complicated.

Advertised bandwidth is always an "up to" number, not a dedicated pipe. If you want that, you want business class service at a significant price premium.

>The torrenter's absolutely should be allowed to fill up the pipes.

And the Skypers should absolutely be allowed to have their calls. QoS ensures that everyone has a minimum usable level of service.

QoS and net neutrality are two different things, as previously mentioned. If you'd like to argue otherwise, you'll have to do better than simply gain-say my definition.

If the network doesn't have the throughput, then QoS makes sense. However, QoS is by definition not NN

Again, you really fail to explain why you think packet prioritization based on type is the exact same thing as packet prioritization based on source and destination. The latter is always anticompetitive at the ISP level, the former can be, but needs to exist to operate a network of any decent size.

Please explain why it needs to exist? All my packets, no matter what they are, should have equal access to another person's packets on the same network.

I myself will employ QoS so that my packets are arranged how I like them in the bandwidth provided by my ISP.

If my ISP cannot give me sustained bandwidth, then I don't receive proper internet service where anything but browsing and maybe streaming will be great.

In that case, I expect to pay less, not to have the ISP implement some technology to decide my traffic isn't as important as my neighbor's skype chat.

The why is simple. Bandwidth is limited on a given time scale and load is unpredictable. QoS ensures that all of your customers receive a minimum usable level of service. If a bunch of people run off to download a given torrent simultaneously, blowing well past average projected usage and you're not packet shaping, anything that relies on real time packet transfers on that network is hosed.

It's that simple. It's ensuring that a minority of customers can't fuck up the service for the majority.

Everything you've mentioned so far is a lot of statement with no backup. You want "sustained bandwidth" - this is not a thing that most ISPs sell to normal customers at rates mere mortals can afford, and for good reason.

All my packets, no matter what they are, should have equal access to another person's packets on the same network.

QoS's entire raison d'etre is ensuring this equal access by making sure sudden high load does not negatively impact the service for everyone else. Would you prefer your call quality/game/video not get shitty when the latest GoT episode comes out? Or would you prefer a complete wild-west free for all where the "bandwidth hogs" can quite literally ruin a service for everyone?

And don't say "expand capacity", either. Nobody in the world, not even Google, has the ability to deliver their max advertised bandwidth to all subscribers simultaneously.

Please, for your own sanity, learn the difference between average speed and max speed.

Sustained bandwidth is not out of the ordinary. I consistently get the same bandwidth day after day, but I realize at times it could have issues, because I don't have a dedicated line.

When I don't use QoS, I can't download a large file and skype at the same time. If I use QoS on my network, it's no problem. Clearly this means my ISP is not doing anything particularly special with my data, and I get on just fine.

I go through Midcontinent Communications, one of the better ISPs in the US.

I am very, very jealous of you. I've got a coworker who used to live in that part of the states who tells me leaving that particular ISP was one of the hardest parts of moving here :)

That said, I bet they got out in front of projected growth so they have a large surplus of capacity. Still, I'd be willing to bet you large and ridiculous amounts of cash that the right combination of growth + popular event would result in contention.

Except how does one tell the difference between Skype and Voip? You don't without deep level packet inspection. Simple QoS based on port could easily piss off a customer, for example, if they want Skype traffic on port 80. The simplest way to deal with this is to treat all traffic equally but serve the bandwidth that you promise your customers.

The customer's network should take that responsibility and pay accordingly.

[ ] Whatever, normal traffic.

[ ] Let this traffic be throttled at will (downloading latest Linux distro)

[ ] I need low-latency or else throw it away (VoIP)

[ ] I'd like low latency, but still do packets in order (SSH session)

[ ] Please give me a channel of X quality for the next 2 hours, or don't provide it to me at all (Netflix)

Exactly. It's information that has to be provided in the IP header.

A great description is here: http://packetpushers.net/how-does-qos-work/

If that information isn't there, then the packets are treated as "best effort" (or at least, they should be in a sane world).

The tricky part is how to handle incoming packets. Since you didn't send them you can't mark them.

WTF? It's none of the ISPs business what I use their network for. If they don't like me using all the bandwidth up, they should either charge more or instill data caps.

> I really think this is a bad idea. If you're a business and depend on a certain type of traffic (say Skype, or maybe email) why wouldn't you consider the option of trying to ensure faster and more reliable delivery of some type of data that is absolutely critical?

Because these 'fast lanes' won't be faster lanes; they'll be the performance we have normally, and other traffic will be degraded.

The solution to the issue that you propose already exists, and it's called Quality of Service; by having your router aware of your upstream/downstream bandwidth and setting simple rules you can prioritize traffic already. You don't need your ISP to charge you extra to do this.

ISPs providing 'fast lanes' would be fine if American broadband issues weren't because of companies with multi-billion-dollar profits not bothering to upgrade their infrastructure to support their customers, or if they hadn't been paid billions by the government to provide universal broadband and then simply not bothered to do so.

History has shown that American ISPs operate as monopolies[1], crush any attempt at competition[2], promise improvements and never deliver[3], and generally take advantage of their customers whenever possible, all the while complaining that the problem keeping them from providing good service is the few regulations under which they operate.

Now they're claiming that their networks can't handle a large proportion of their users using their connections at once, and it's all Netflix's fault (and not theirs for oversubscribing).

So yeah, I think this is a good idea. Force ISPs to deliver the service that people are supposedly paying for, and not pay extra for it.

[1] http://www.washingtonpost.com/blogs/the-switch/wp/2013/10/01... [2] http://arstechnica.com/tech-policy/2014/01/who-wants-competi... [3] http://arstechnica.com/business/2014/06/att-makes-the-same-p...

> Because these 'fast lanes' won't be faster lanes; they'll be the performance we have normally, and other traffic will be degraded.

That's not necessarily true. Consider the possibility of building out new infrastructure and capacity, and specifically using it for that purpose. No degradation of the existing service.

Let's say you're the CEO of an ISP. Netflix pays your company $x million to access a fast lane. You can either spend the money on (a) building $x million of extra infrastructure or (b) giving yourself an $x million raise.

Which do you think will lead to the CEO getting a bigger raise?

Just an example:

I play a 20-year-old video game [0] with the best players in the world. A 1v1 game requires less than 56k (that is, 0.055 megabits) of bandwidth. Yet there are a handful of players who can't play at certain times of day because their 30 megabit broadband connections end up having extreme latency/loss, because everyone on the ISP is streaming Netflix or playing WoW or downloading porn or whatever and the ISP degrades service equally for everyone.

I would love to be able to buy internet service with an option for a "fast lane" guarantee for certain types of traffic where latency and loss actually matter. (How that ties in to "net neutrality" is, as danielweber says in a nearby comment, a bit vague since NN seems to have dozens of different definitions.)

[0] remember Descent? Here's how to get up and running: http://descentchampions.org/new_player.php

So it sounds like this is less of a law restricting tiering and more of direction to have the FCC make regulations to restrict tiering. Also not having seen the bill I can't say for sure but the language "between the customer and ISP" is a little shaky since someone like Comcast is both a customer ISP and backhaul provider so can they claim they are operating as a backhaul company when they change money to Netflix .

To be fair, that is typically how it's done. Congress passes a law that directs the relevant agency to figure out the minutia of implementing it.

The issue I have with the model ( which is true across the board on issues ) is that with so many things by leaving it up to regulators you are putting governance in the hands of unelected career beurocrats with minimal oversight over what is actually being done. That is why the next phase in governance is always legal wrangling over regulations since the legislation has given such wiggle room to regulators to implement policy.

Do you have an example in mind when you talk about regulations that go against the intent of Congress? Not saying it doesn't happen, but it seems like the exception.

What's the alternative? Congress should rewrite FCC rules itself? How else would you bring them into compliance with the new law?

>>"It wouldn't give the commission new powers, but the bill — known as the Online Competition and Consumer Choice Act — would give the FCC crucial political cover to prohibit what consumer advocates say would harm startup companies and Internet services by requiring them to pay extra fees to ISPs."

The title seems a bit off. Wouldn't Congress only be the ones that could ban internet slow lanes? Not the FCC.

The debate is three part:

Part A) Whether the FCC should use the existing powers Congress granted it to push network neutrality

Part B) Whether Congress should force the FCC to use its existing powers to push for network neutrality

Part C) Whether Congress should grant the FCC new powers to push for network neutrality (the FCC reclassifying broadband as a utility probably falls in here as well, given its impact)

The Supreme Court has already weighed in on A.

"Leahy and Matsui's proposed ban on fast lanes would apply only to the connections between consumers and their ISPs"

I don't see how this will address the issue in a practical sense. When I'm having trouble streaming netflix on Verizon, my speed tests still return a solid 25mbps. The connection between Verizon and Netflix is the issue, not the connection between my house and Verizon.

this is why i find it strange people are basically saying netflix is sticking up for network neutrality, when in essence netflix is asking for special treatment because they are big.

I love how the very first paragraph implies that the problem is Netflix (as the example company) getting special, preferential treatment, when in fact, Netflix has been obviously punished for not paying ransom money, and has led the charge on the net neutrality movement that led to this bill.

But this is the Washington Post, the Pravda of the Beltway. What do you expect? Honesty?

"will unveil a piece of bicameral legislation Tuesday"

Is the text of the actual bill up yet? I cannot find it.

I guess this is the text: http://www.leahy.senate.gov/imo/media/doc/The%20Online%20Com...

God Dangit. Don't make this a Red vs Blue Debate.

Its been a red vs. blue debate for a long time. Why is it surprising that significant issues of government policy in which the opposing sides represent conflicts between the deeply-held interests of different groups also turn out to be issues of division between political parties. I mean, isn't that exactly what you'd expect?

What are these so-called "fast lanes"?

That title looks like it was written by a Telco Lobbyist. Immediately it's made partisan (Democrats are regulating!), dire (unveils has a negative connotation), overbearing (forcing the FCC oh noes!), and incorrect (no such thing as an Internet "fast lane").

A better title could read: "Lawmakers Introduce bill to halt ISPs crippling Netflix-type Internet Services"

It's no wonder we can't get anything done for good in Washington.

Anyone have a link to the legislation?

I wonder which lobbying group will kill this at the eleventh hour this time? Will it be the trial lawyers again? Hollywood? Hollywood lawyers? Will it be the NSA, who claim they can not comply with the directive?

Whatever the outcome, rest assured that this Congress is at its most creative when finding a reason to kill any bill that has the support of the American people.

Whatever the outcome, rest assured that this Congress is at its most creative when finding a reason to kill any bill that has the support of the American people.

Phone call works, but we are too distracted to care or that we don't care in the first place.

Cynic here, more likely a move by Congressmen who do not feel they are being appreciated enough by lobbyist for the telecoms.

I did notice it was a last mile restriction, so they are still free to muddle up the lines getting from the provider to the ISP.

Reply on news.ycombinator.com