We learnt about this from Snowdens documents in 2014.
Taking compensation for a peering agreement is perfectly fine if it's imbalanced. What net neutrality should prevent is, ceteris paribus, asking for $X from YouTube and $Y >> $X from Netflix because you figure your dying cable business will benefit from less competition.
Ideally, net neutrality would be FRAND for peering agreements: you can't deny anyone if they are happy to pay market rate, and market rate is what you charge everyone else, plus minus.
The imbalance of the traffic has little to do with what we have here, since the telco is specifically paid to deliver that traffic by consumers. Charging on both ends is very profitable, but is definitely double-dipping.
Your phone provider doesn’t charge you more to receive calls from large customer care call centers when you have a problem and request a call back. The call center may get charged more by their provider if their call pattern is problematic, but your provider doesn’t stop the call from going through. It is really not any different with internet networks and why common carrier rules make a lot of sense.
These same rules also provide a lot of liability protection for the carriers, which I’m guessing they have negotiated to keep somehow.
I would want the authority to choose to peer with who I want, but if someone doesn't want to peer with me at market rate, then they won't experience "degraded" service coming through other networks that I peer with indiscriminately.
If you think this isn't fair, I would like to hear why.
What should be outside net neutrality laws are protocols that leave the customer and applications in control. There is nothing wrong with using Resource ReSerVation Protocol (RSVP), DSCP and IPv6 flow labels to get QoS that differ from the dumb pipe. Many future applications require it.
- Online games might want to request low lag.
- Streaming services might want constant bandwidth with less strict lag requirements.
- File downloads might be OK with bursts and pauses as long as the best effort throughput is good.
- Safety-critical applications like, robot cars, remote surgery or emergency messaging, need guarantees that discriminate against other uses.
Except that NN does not prohibit it for traffic engineering purposes. And I'm here, just doing traffic engineering:
if (you-pay-me-for-PNIs) { set-local-preference really-high } else { set-local-preference default }
Oh look, your traffic is going to the non-congested path which happens to be the path to my PNI locations, directly into your network. Jack on the other hand, Jack gets to use the standard transit. "Hey Jack, you know, we will drag all your traffic to SJC, even if you are on the East Coast and hand it off to TATA because TATA gives us the lowest transit price. Man, TATA seem to be congested from 5pm Eastern to about 2am Eastern. Oh well, maybe you guys need to yell at TATA. Oh, do you want to buy a dozen of 10Gs to me all over the country? That would avoid those dolts at TATA."
Not outright but it does restrict it specifically to prevent it from being a backdoor for this kind of edge provider discrimination.
1. Terminate transit on transit-edge. All routes form transit edge get preference X.
2. Terminate peering routes on peering-edge. All routes from peering edge get preference Y.
Oh wait ma, I have traffic engineering and it matches $$. Aint life grand? NN rules are what happens when dilettantes who have never seen a 20U sized router are in charge of writing rules about networks.
TCP/IP is optimized for neutral nodes in a network, sure you're doing traffic engineering, throttling is already happening. At a network level this makes sense, doesn't make sense that we give ISPs carte Blanche to openly start filtering traffic based on any rules they see fit. The people who want to repeal NN are the suits, not traffic engineers and techies like yourself. Obviously
If we are to stop pretending that NetFlix does not pay Comcast and Verizon directly to get content to the customers of Comcast and Verizon without sucking in addition to paying others to deliver the same content to customers of Comcast and Verizon ( like an upstart MyFlix.com ), we can have a conversation about Net Neutrality 2.0, traffic engineering and settlements.
A) Traffic shaping is done in the public's best interest.
B) It doesn't discriminate between packets of the same type but different providers.
C) Carriers are transparent about how and why they shape traffic.
D) All affected parties retain ultimate control over how shaping happens.
ISP sells service with fixed price that states what kind of QoS requests it can fulfill. Net neutrality prevents the ISP from setting them on their own.
ISP should just provide enough capacity to pass all their traffic without being dropped. Don't oversell what you can't provide.
Tech Policy Podcast: #206: The Future of Internet Regulation w/ FCC Chairman Ajit Pai http://podcast.techfreedom.org/e/206-the-future-of-internet-...
But, I think Robert Graham summed it up pretty well when he said "you believe the FCC should allow companies to bend the rules when it's in the interests of customers, then you don't believe in NetNeutrality." Lots of politics and emotion in the hysteria, very little sound reasoning.
You’ve (the FCC) seen abd addressed seen and acted against selective slow lanes, (intentionally inhibited traffic) which are the same thing as fast lanes, just with which side is the “default” switched. And you've seen paid prioritization fast lanes in foreign markets without the FCC’s 13-year history of pro-neutrality policies. So this is either misleadingly selective or just plain outright lying, but then that's true of basically everything Pai has said about neutrality since joining the FCC.
But even with murder, the law doesn't stop murders before they happen, but rather only acts retroactively (and that is for good reason).
> "Fast lanes DO exist in other countries right now."
But this is ignoring the whole debate: proponents of net freedom don't consider fast lanes to be a problem.
I think that this is the part that I can't comprehend. The railroads misbehaved so badly that the government had to break rail transportation as an industry to kill them. Ninety years later we're still using long-haul semi trailers for everything. Bell's abuse of its monopoly was similarly severe. How do people hold this belief without becoming laughingstock?
No. There's actually a very healthy cargo rail system and relatively little trucking is truly long haul.
So I'll admit: fair enough, the rail industry is still healthy and carrying a good amount of cargo. I'll withdraw the wording; we do not in fact use long-haul trucks for everything. The fact that trucks are competing with rail at all, however, given the geographical extent of the US and how horrible long-haul trucking is, I believe makes me stand by my claim that the US rail industry was broken and never recovered. It frankly astonishes me that driving something down the highway is even remotely financially effective in comparison to rail freight.
And, of course, none of this is relevant to my original point: the railroads abused their powers to an astonishing degree and the government was forced to smash them for the good of the economy. Is it obvious to me that today's ISPs are following in the footsteps of yesterday's railroads and telephones and I don't understand how people convince themselves otherwise.
[1] https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/pu...
[1] https://www.treehugger.com/cars/rail-versus-trucking-whos-th...
With Semi's, each separate truck has to endure its own wind resistance independently.
With a train, each rail car effectively drafts the car in front, producing significant reductions in wind resistance for all cars behind the locomotive.
This site (https://www.energy.gov/eere/vehicles/parasitic-loss-reductio...) quotes wind resistance, braking, and rolling resistance as responsible for a 45% reduction in efficiency.
This site (https://www.nap.edu/read/13288/chapter/7) has a table that separates the factors out. Of the parasitic losses, aerodynamic losses (i.e., wind resistance) is responsible for consuming 53% of the motive power.
I do believe there's some benefit from the tracks themselves as well, though. Surely there's a lot less friction (per ton of weight) between the train wheels and steel rails, than there is between truck tires and the road?
Except all freight trains in the US have 6600 HP diesel engines, usually ~6-8 engines per 100 cars from what I've seen (which is mostly coal).
I understand there are a lot of lobbyists, but there are many people who sincerely believe the "less regulation and more competition will solve the issues" mantra.
I place a lot of the blame on outlets like the Wall Street Journal giving credence to arguments that strip out all context and use fancy words to justify these positions. Doesn't help that other people are afraid to call out their lies because they came from the same social circles.
Yes, the "free market + competition makes things cheaper" argument works well when you remove all the reasons why a specific case does not fit the theory perfectly.
If there's just one infrastructure, or maybe two (cable and DSL) and the ISPs own it then there's barely any competition. That's when you get anti-competitive behaviour.
Such behaviour would be punished in a free market even if it were allowed. But we don't have a free market in this segment. We got monopolies and duopolies.
There are different ways to limit the power of monopolies and duopolies. One is legislation to force them to treat all data equal. Another one could be to break up the large mammoths who are monopolies and duopolies. The latter ain't happening, so why revert the former?
Either due to sheer stupidity or because of heavy lobbying from certain interest groups who'd profit greatly from this (ISPs is my guess).
Yeah last time we tried this we got Amtrak so I don't really see it as a workable solution. Size limits and market choice requirements might work, though. They did, pretty well, with media companies, for quite a long time.
Of course, those also only last until we get an FCC that's either so corrupt it'd be right at home in the Gilded Age, or so out of touch it'd be right at home in Versailles. I tend to the latter suspicion, myself; DC people are long known in general to make a habit of thinking of 495 more or less as the end of the civilized world, beyond which lies only a terrifying wasteland full of barbarians and howling ghosts. That's why lobbyists exist - just as in Versailles, there's a market for paying people to present one's case and pray intercession on one's behalf at court. I suppose we'll see whose lobbyists pray most effectively this time. And next time, and the time after that, and...
E.g. the way it works in the UK (it works similarly, but not entirely identically, in all EU countries), which is absolutely not ideal but gives reasonable competition is that BT OpenReach (by virtue of BT's past monopoly) is regulated and is required to offer any ISP access to their network on equal basis. This means the ISPs have a few choices:
* They can put equipment in the exchanges, and "just" get access to a raw connection to the subscriber. OpenReach maintains the lines.
* They can contract OpenReach for "backhaul" of IP traffic to central locations.
* They can build their own last mile network.
I can pick from dozens of providers because of that, but most of them mean I'll pay a basic line rental either directly to BT or via the provider, and that is the same for any of them. Others are free to build their own network without the same restrictions, but this for natural reasons happens very slowly (predominantly fibre providers in dense urban areas, and a cable provider).
In that model net neutrality is less of a concern. It's not entirely free of issues, as major ISPs certainly still could offer preferential access to certain content, but at least we have real competition beyond the last mile.
The biggest problem with this is that the way it is structured does not give BT a strong incentive to invest in improving the last mile network. I'd have preferred if there was an additional clause restricting them to only take out dividends from OpenReach proportional to the investments made in improving the network. But overall it's working reasonably well.
Networks are a great example where a regulated monopoly on one level can allow for a healthy free market on the upper levels.
If you forget about the money, having a single operator makes the most sense. Thus we should model our economic system to take this into account
Not that it worked completely, because everyone has nearly merged back together, but it is evidence that Republican administrations aren't necessarily pro-monopoly.
A large part of the issue is that, through lobbying and franchise agreements, ISPs have locally secured their status. Even if you break them up they only become more local monopolies (which is better but not competition and still requires regulation).
At least in Germany planning a murder is illegal. Just like preparing a terrorist attack is - otherwise the police wouldn't be allowed to stop it.
Also Wiki: "In some jurisdictions, the very act of preparing for a crime is a criminal offense in itself" (https://en.wikipedia.org/wiki/Crime_preparation)
For the US, interesting read: https://www.lawyers.com/legal-info/criminal/criminal-law-bas...
This type of flow management is already how VoLTE flows are prioritized on LTE (see https://en.wikipedia.org/wiki/IP_Multimedia_Subsystem); I wouldn't be surprised to see land-line ISPs implement it.
Not so fast. It's zero-sum if and only if call Internet subscribers have the same ability to pay for the service. If any user has a very high willingness to pay, but does not have the resources to pay it, then this amounts to a transfer of "wealth" (i.e. bandwidth) from that user to another who might not value it as much. It's not hard to construct utility curves in which this decreases total welfare.
That said, QoS is probably not the concern here. This is very much not a zero-sum proposal.
Dropped packets are a sign of poorly designed networks at which point you can't assume QoS will improve things. The added complexity might make things worse for all traffic.
All that being said, that's why cell networks, until 4G, used actual circuits for voice, rather than having it be packet-based; and that's why 4G+ networks use IMS (a layer 1+2+3 protocol suite that all middle boxes are aware of for QoS purposes) rather than just relying on layer-3 https://en.wikipedia.org/wiki/Differentiated_services QoS fields (which are handled by full layer-3 switches within a given LAN, but get dropped the moment the packet crosses onto the public Internet and are just ignored by layer-2 switches.)
(And even on large LANs—like your cable provider's ATM network—DiffServ tags aren't relied upon, because QoS is just so much easier when layers 1 and 2 are aware of it. If you get a IP phone line from your ISP, it'll be using https://en.wikipedia.org/wiki/Multiprotocol_Label_Switching tags instead, to give its packets QoS priority across even the dumbest middleboxes.)
Voice is very low bandwidth as long as the minimum bandwidth per device is over 100 kpbs and the device to prioritize voice data towers don't need to bother with QoS for traffic. They still do this for various reasons, including the fact they are sending the voice data to a separate network not the internet.
TLDR; Total bandwidth / devices > 100kbps and it's not actually over subscribed.
EDIT: Example: Free Music Streaming by T-Mobile [1] is active in EU as well. So if you stream from Spotify its "free" yet when you stream from your home server, via a VPN, or an unsupported service it isn't "free".
[1] https://techcrunch.com/2014/06/18/t-mobile-stops-counting-da...
This is sick (the lack of net neutrality is more likely to reduce freedoms than to promote them). If you remotely support net neutrality, please never use that term again. This is self defeating.
Existing regulations are a restriction on the freedom of service providers. That is undeniable. There many cases where we restrict freedoms for the greater good, but I don't like to pretend we are not doing so in those cases.
I'd also add that freedom is not the right to demand other people give you something you want. I wouldn't say that I've lost my freedoms if there are no good ISPs left after the regulatory change. Everyone involved can still either choose an existing provider or set up their own. We have monopoly/anti-trust laws when that ceases to be the case. Of course, this may not be the efficient path to a world that a huge majority of people want - which is why I slightly side with net neutrality.
Then go ahead and call it "corporate freedom". The network isn't getting any more free without neutrality.
In practice, they are. There are exception, but they remain exceptions.
As in, for real, in practice? I seriously doubt it, the hurdles are huge, and most deals favours gigantic corporations (bulk often being cheaper than detail).
Try to get a peering agreement with some big provider. Most wouldn't even talk to you if you're not big enough.
Murders are punished after they happen. But a plot to murder is in itself a crime.
Also, the fact that punishments come after the fact has nothing to do with this. Murder isn't banned after it happens. It's banned before it happens.
Lastly, you say "the law only acts retroactively". The main purpose of punishing past actions is to disuade people from doing them in the future. If I hate my boss but I decide not to murder him because I don't want my child to grow with me in prison, the law has stopped a crime before it happens.
And they literally wouldn't be if internet service was delivered in a free market. The problem is that the market isn't free. ISPs have spent years rigging the market in their favor, politically. The landscape is already hostile to the consumer--they intend to make it more so.
I believe your suggestion is already practiced with phone carriers. The infrastructure is finite & there are physical bandwidths limitations, at least currently.
It's called shaping and you easily pay 2x the price to avoid it. Everyone above the poverty line pays.
In my mind, this is very different to allowing YouTube load faster than Vimeo
Speed differences will happen though, e.g. I notice that Netflix sometimes tends to be a bit slower but that doesn't improve using a VPN so that I guess it's not BT's fault.
Can't remember who my ISP was at the time, but definitely have hit times when (especially gaming) would be fast & then would hit a wall suddenly. That said, seems to happen less recently, was more in early broadband days.
"With peering, there are two methods that connectivity is formed on. The first is where direct connectivity is established between individual networks routers with multiple 10Gigabit Ethernet or 100Gigabit Ethernet links. This sort of connectivity is known as “private peering” or PNI (Private Network Interconnect)."[0]
[0] https://aws.amazon.com/blogs/architecture/internet-routing-a...
“network freedom” was the FCC’s original name for net neutrality before it shifted to “open internet”; because it protected end-user freedom. Using “net freedom” to refer to the abolition of that protection is, well, about as honest as the rest of the anti-neutrality effort.
Don't care whether there will be a problem with how the internet works as they will become filthy rich. That is why they want to kill net neutrality; dollar dollar bills. They couldn't care less about the user experience.
In a way the law tries to accomplish this by limiting ways like guns, bombs and sometimes doing checks randomly or based on suspicion.
But the reply was in reply to Ajit's position that "we don't know what will happen because we don't have them".
No thanks. It’s amazing what a little bit of polishing a turd and renaming it will do.
EDIT: I do not claim everything I say is right. But I will stand the courage of my convictions and am happy to be proven wrong. So far this thread is a total disaster of emotional pleas, doomsday predictions, and bad reddit-like behavior (such as yours).
No it doesn't. The FCC's net neutrality orders have always included an exception for "reasonable network management", most recently defined as:
> (f) Reasonable network management. A network management practice is a practice that has a primarily technical network management justification, but does not include other business practices. A network management practice is reasonable if it is primarily used for and tailored to achieving a legitimate network management purpose, taking into account the particular network architecture and technology of the broadband Internet access service.
(emphasis mine; source: https://www.law.cornell.edu/cfr/text/47/8.2)
Prioritizing gaming latency is clearly a "primarily technical" concern rather than a matter of "business practices" - at least as long as the ISP didn't try to charge game publishers for the service. So it would be fine.
That isn't a thing. ISPs don't actually do that.
Latency on any reasonably provisioned network is determined by distance and the speed of light. There is no magic thing routers can do for special packets to make it go down, it's just the laws of physics.
They could only do something when the network isn't reasonably provisioned and the ISP's uplinks are fully saturated, but that should never normally happen unless the ISP is either incompetent or creating artificial scarcity for anticompetitive reasons.
See Verizon and Comcast. Today. With Net Neutrality on the books. Want to bypass this congestion? Buy some PNIs.
So what we have here is dilettantes talking about a how they think the world should be or is today rather than acknowledging that net neutrality stopped existing in 1995 when AS1239 started asking for $$ to get peering routes.
That absolutely is a problem with the existing rules. Some people pointed that out when they were proposed.
But the existing rules solve some problems, just not all of them. And you don't go forward by going backward.
NN does not live in reality. In reality there are so many line cards that one can put into a router/switch/linux box, there's so much space in a cabinet and there are only so many people who know how to operate in a physical world. Everyone with a little money and clue can generate a hundred gigabit per second traffic towards eyeballs behind Comcast but only those with serious money and serious clue would be able to do it with reasonable performance - Comcast/Verizon/ATT/NewFiberToEveryoneOnTheEastCoast ISP will need to prioritize physical world deployment and it would go to the one with serious money and serious clue. And since NN cannot address this, it is toast and it has always been toast.
The solution is regional ISPs and regional networks - breaking up franchise agreements with Comcast and Verizon that your little town has, getting ROW via cities/towns, leveraging small electric companies ROWS, etc.
Which is why we still need a rule prohibiting that. But the existing rules are still helpful. They address the consumer side of things. At least you don't have Comcast charging its customers money for a Netflix package and charging Netflix for peering.
> The solution is regional ISPs and regional networks - breaking up franchise agreements with Comcast and Verizon that your little town has, getting ROW via cities/towns, leveraging small electric companies ROWS, etc.
That could be a solution, but it would certainly take at least a decade if not more, and until that has actually happened we need something else.
ISPs do not care about charging customers for tiered packages - policies are too difficult to implement. It is much easier to charge Google, Netflix and Facebook. Single point of contact, single large payment and if Google/Fb/Netflix don't pay ability to deliver real pain.
> That could be a solution, but it would certainly take at least a decade if not more, and until that has actually happened we need something else.
That is the solution and it is something that needed to be started twenty years ago. If we started it at that time, we would not even be having this conversation because ISPs would be highly competitive utilities. The problem is that building networks, digging ditches, mounting routers and dealing with customers is not as sexy as launching HotOrNot or SnapChat.
I need to buffer that 50mbit of traffic, which could be 100,000 packets. It also introduces delay.
I can still use qos to ensure packets I want with less jitter get through in front of the 100k queue.
The latency over the second will vary from zero to 50ms, even with 20% network utilisation (or I get lost packets)
Sending at gigabit for 100ms implies a 12.5MB buffer. A lot of routers sensibly won't (or just can't) buffer that much data, so sending in such large bursts is a good way to get your packets dropped. Which is why sensible programs (and the TCP stacks in major operating systems) don't do that. Instead they'll either use a smaller time slice or use the event-based TCP "ACK clock" where packets are sent in response to receiving acknowledgements of previous packets, which spreads the transmissions more uniformly.
Moreover, notice that the thing you're describing still shouldn't happen within an ISP's network, because it's caused by an individual stream sending faster than the path bandwidth. That can happen when you go from 1Gbps LAN to <1Gbps WAN, which is why QoS can be helpful on the customer's edge router, but that isn't part of the ISP's network.
It shouldn't happen when you go from a 500Mbps customer link to an unsaturated 40Gbps peering link. The 500Mbps link can't saturate the 40Gbps link, and if some combination of them do, it's not because of the sending pattern, it's because the uplink is too over-subscribed.
No it doesn't. Having every customer use 100% of their connection at the same time is something that could happen in theory, but it never happens in practice. Literally never.
A certain amount of over-subscription is perfectly reasonable. As long as you have the capacity for the actual peak load, there is no congestion, regardless of whether you have enough capacity for the theoretical peak load.
Pre 2015 people were worried about the government actively monitoring and controlling what could be said on the internet. It was a fight to protect free speech on the web.
The current issue about the actions of the FCC focuses on net neutrality. While also being essential for free speech on the web, this issue focuses on the prevention of monopolization and antitrust practices by internet service providers.
Or did I miss some point you made?
No, but it does outlaw it before it happens. The question isn't "is it happening at the moment", but "should it ever happen".
And I'm sure some people think murder is not a problem, but most of the rest of us think they're also crazy.
You know, if the ISPs were better at selling this particular turd sandwich, or turd polishing it at least, what they'd do is say they intend to charge astrometric tons more per unit of spam, email and crappy ads, that slow down people's machines and disproportionately suck up internet resources. But they aren't. So why do they want this? To do exactly the same f'd up shit they've been doing with cable TV.
Pre-2015, ISPs weren't classified under Title II. The unspoken (and occasionally spoken, when disagreements came up) agreement was that the ISPs wouldn't violate net neutrality principles, and the FCC would leave them alone and not classify them under Title II (which introduces more pain for ISPs than just net neutrality concerns).
In 2015, Verizon decided not to play ball anymore, and start violating net neutrality principles. They won a court case saying that the FCC couldn't regulate them in this way without Title II. They thought they had the political/lobbying clout to avoid the Title II classification despite that. Turns out they were wrong.
So clearly Pai has his agenda and is just lying to support it. ISPs have absolutely tried to do fast-lane-style bullshit in the past, and were stopped first by the threat of Title II, and then later by Title II itself.
If I had to start a new society tomorrow and I needed a quick list of things that should be illegal, I think I could figure out things like "Don't rape, murder, or steal" before having to actually witness someone rape, murder, or steal.
Which countries?
You pay your base fee ( to access anything ) and you can then add particular bundles to get flat-rated access to certain websites. Or you can just continue to pay by the MB.
I don't see how that changes anything about the situation. Just because mobile providers have already been far more ruthless doesn't suddenly excuse these practices, some would argue that mobile providers have pretty much normalized this in the very first place.
I consider the vast majority of mobile plans, with data volume caps, straight up rip-offs: The moment you go past your volume cap you might as well not have any plan at all, as the throttled bandwidths usually ain't even fast enough for just regular browsing on the www.
Yes, I realize there are differences in the medium, mobile being shared and all, I still can't shake the feeling that the vast majority of mobile providers use this as an excuse for not expanding capacities and instead nickle&dimming everybody trough data volume.
One could draw a comparison between what these packages mean - they only bundle the big services, making it even harder for small alternatives to compete. Same would be possible if ISPs would start to play the big money game the other way round - serving faster access on cable lines, which I would view essentially as the "landline" version of free traffic on mobile connections.
This is the exact same problem.
Don't you understand that the capacity issue was artificially created to push this agenda? There was blog article from level3 showing the capacity issue was coming down to enabling another port[1], perhaps purchasing extra transciever that costs just few thousand (level3 offered to even purchase one for them if they can't afford it)
[1] https://web.archive.org/web/20140907021805/http://blog.level...
The fact that he is trying to sell this as a repeal of regulatory overreach makes me kind of sick to my stomach. And, it's apparently an overreach because it solves a problem that didn't have a chance to come about? I feel like he's bullshitting me. Regardless, I like my internet access right now, so what problem is he solving for me or the rest of the public by doing this? If it's that these fast lanes will save us money it better be an enormous chunk of money to merit a such significant shift in the way I experience the internet.
Ajit Pai: The answer to the first is we're not sure. We've never seen them before and that's part of the reason why I thought the rule in particular was, that was adopted in 2015, was very premature banning something that's simply didn't exist.
I took the time to read Ajit's argument here. It's completely mislead. If murder hadn't happened in a certain area yet, would it be "too preemptive" to outlaw it? Fast lanes DO exist in other countries right now. This is outrageous.