The consumer can no longer choose an ISP on their individual performance separately from various internet services and their performance. The bundling of agreements will increasingly reduce the ability of the market to deliver efficiently.
- Comcast $20 - ATT $15 - Other $10
Edit: It also a gatekeeping hurdle that smaller businesses would have much more difficulty crossing than large established players.
But right now the ISPs are paying for Netflix's bandwidth and that's being passed to customers in their ISP's bill. I'd rather only Netflix customers have to fund Netflix's service.
Or maybe you can mix streaming access. For $1 each, you can pick up to eight packages on the basic plan: Amazon Prime, Spotify, Netflix, XBoxLive, etc. etc. or For $12, get all 20 of them. Or for $20, real Internet access.
Why would ISPs lower prices? They have monopolies in most markets so there is virtually no force driving prices down.
When people can charge whatever they want, its a free market right?
I don't think their monopoly would last very long, if cheaper options became available.
My rural town will likely not see the benefits from competition for a long time.
It seems to me like people who would (and can) otherwise buy internet will continue to buy internet, but people who couldn't afford it or didn't want to afford it might jump on cheaper packages. This is especially illustrated in e.g. third-world countries where we have majorities still not connected to the internet (where a super-cheap Google/Facebook/whatever package would go far), but there are still many, many people in poverty here in America that don't have access to basic things online. Offering them a cheaper plan to access just the parts of the internet they want to use seems like a win-win here.
As long as it doesn't go the other way.
At present the internet is more or less a neutral marketplace in which everyone pays their way to get packets from source to destination. Much of America is served by only one to two members of a small group of huge isps who have more to gain from bilking me and you than from competing with each other.
If we let a cartel of isp who are the only way you can get internet access in a reasonable fashion decide who is allowed to do business on the internet and in what fashion you will find that your basic package ends up costing more than your present open internet and any perceived discount whatsoever is outweighed by the money that people like netflix pay per subscriber for access and must necessarily recoup by charging you for.
If the media companies even let players like netflix survive instead of slowly strangling them this sucks for you and netflix but as for future players they aren't part of the in group that is allowed to do business at a reasonable rate and they just never get off the ground.
The system you are ok with would have strangled kids like youtube and netflix in their cribs.
The basic economics of it is allowing them to decide what is served and at what individualized price changes the economics of the cost of internet companies serving customers from what a competitive market can charge (internet backend) to the most a market can bear under a monopoly (the last mile).
1. Peak bandwidth usage
2. The cost of building (and maintaining, see above) the cables
Killing net neutrality helps with neither of those, so any significant cost reductions (if there are any) will come from bribes from content providers. And guess who'll end up paying for those...
So, it's not an invented hypothetical.
As everyone else has pointed out - we'll be paying for the full cost (and probably quite a bit more based on ISPs previous behavior) of the connection one way or another. Either we pay for it through ISPs charging service owners for the privilege on being on the limited packages (thus upping the cost of the subscription indirectly for everyone) or ISPs forcibly create monopolies by pushing users on to their own web platforms, then raise prices later (as is the end goal of any push to fabricate a monopoly). It's pretty much strictly lose lose from the customers' point of view, at least in aggregate over the long term.
I don't think it's unreasonable to assume that some large internet companies would subsidize plans like "Facebook-only" to connect more people to the Internet (see: to their service). This in turn, well, connects those people, gives them opportunities and access to knowledge they wouldn't otherwise have, and eventually opens them up to the rest of the Internet. It's a stepping stone: obviously, all internet is better than some internet, but some internet is better than no internet. There's an alternative of just giving them the full internet for free/cheap (which I'd love to see), but is obviously way less likely.
In that sense (and, to bring things back to the comment my parent is responding to), I think that providing a cheap plan that allows access to some subset of the Internet targeted at people who wouldn't otherwise have internet is preferable to those people just not having access at all. Is that wrong of me?
I think it's not preferable. I would rather see prices driven down through demand so that full access is affordable to all, then for an artificially low price that allowed access to a Facebook internet to kill the market for cheap plans.
The problem here is that the last part -- the part where these minimal plans lead to a path where the full internet is available in these underserved areas -- doesn't have to happen.
All it does is (incorrectly) teach those users that the internet is just a means to access Facebook and a few other partnered sites. You're teaching them to expect the cable TV model for their internet access, which is precisely what we want to avoid.
It's not hypothetical. It has actually happened.
> I don't think it's unreasonable to assume that some large internet companies would subsidize plans...
Again, you don't have to assume. This has happened.
> I think that providing a cheap plan that allows access to some subset of the Internet targeted at people who wouldn't otherwise have internet is preferable to those people just not having access at all. Is that wrong of me?
Considering they would be losing a cheap plan that allows access to all of the internet in the process, yes, it is wrong of you.
Repeal of Net Neutrality makes his a possibility. Yours not so much.
People, by default, will consume stuff that enables exploitation (Ads, Facebook and its friends, porn etc). Very few ll actively seek out stuff that empowers them, even when provided access.
But when you provide only, say, facebook, in the name of giving Internet, to a population, you open a new channel to that population, that is solely meant to exploit. This is probably what facebook tried to do with India and failed.
How could anyone successfully create a new competing product once the majority of the market uses the cheap Netflix+Hulu-only package, locking them out from any potential competitors?
I mean, I get that it is bad now. Feels like it could get much much worse.
The base rate hasn’t changed and the category exemptions were open to anyone in a market rather than being negotiated behind the scenes.
I’m expecting that’s going to look really good compared to what we’re about to see.
FCC liked what T-Mobile was doing with Binge on because it didn't favour particular services or alter consumer choice significantly. It was fair and open or any website to participate in
> When it comes to T‑Mobile US Binge On services, FCC staff decided that the service "did not compel edge providers or consumers to participate in Binge On, and did not charge them anything if they opt to do so" – and therefore was legit.
> However, AT&T's Sponsored Data program "presents competitive problems" because it "imposes hefty per-gigabyte charges on unaffiliated third parties for use of Sponsored Data" – charges that "far exceed the costs AT&T incurs in providing the sponsored data service."
In Netherlands T-Mobile has a unlimited music streaming offering. To participate, a contract needs to be signed. Plus the organization/company signing such an offer needs to sign a paid support contract.
Meaning: this option within Netherlands is not free, not fair, not open. Apple refused to join that scheme in the Netherlands. Due to that new information there's going to be a new legal case about it.
Just asking, because, you know, what site is this?
Yes, some towns have created their own municipal fiber optic networks, but it's done on a small scale.
Let's get down to the practical.
Laying fiber is expensive you say? What about spacecrafts/sats? Let's phone Elon wether he has time to negotiate a deal…
Recap: 1. Everybody wants Internet. 2. Not everybody (who could pay for it) has it.
Conclusion: there is an opportunity.
Which means, yes, it is a perfect opportunity for entrepreneurial activity. Extremely high, prohibitive costs means a new, cheaper, scalable way of doing things would be disruptive and have high reward
Let me know the second you see one so we can both not buy it.
https://en.wikipedia.org/wiki/FCC_Open_Internet_Order_2010
The key about the 2010 rules is that they did not reclassify ISPs. These rules were similar (actually much weaker) to the 2015 rules, but they were issued under the authority the FCC claimed to have over Title I services. As most here know, from 2005(ish?)-2015, cable/fiber ISPs were regulated under Title I.
ISPs sued and argued that since they were Title I services, the FCC's 2010 regulations were illegal. The judicial system agreed (Verizon v. FCC 2014). The solution to this problem -- the court explicitly told them this -- was to reclassify ISPs as Title II services; FCC would then have the proper legal authority to re-issue their NN regulations. Wheeler's FCC did this in 2015.
This is wrong, but oddly not quite as wrong as the FCC chair’s description as it at least as some connection with reality.
Enforcement of “net neutrality” rules (the FCC name was always “open internet”) has gone on for quite a long time, first under a casd-by-case regime without specific regulations and from 2010 under Title I regulations that were substantively similar to the Title II regulations adopted in 2015 (though they had looser rules for mobile broadband providers.)
The court struck down the case by case approach in 2010, and the Title I approach in (IIRC) 2014, specifically pointing to Title II as the authority that the FCC needed to invoke if it was going to apply rules of the type it adopted in its Title I regs in 2010. So, in 2015, the FCC turned around and adopted rules under Title II.
Why would no not what a specialized network option for 'super low latency' for things like games?
Or if you don't use any kind of streaming, then why not get cheaper rates for 'super high latency'?
This is good, it's their job.
Where it gets ugly is if they start discriminating between customers, i.e. Netflix has to pay $X while their parent companies other asset, HBO only has to pay $Y.
Or if you have a 'Facebook package' that optimizes for FB or something like that.
So QoS, yes, content discrimination, no.
So you have an issue connecting to netflix... oh sir for just $14.99/mo you can get our Netflix QoS package added on to your connection and have no further troubles. A few weeks later... oh sir your having issues connecting to youtube, well we have our youtube QoS package for just $9.99/mo extra...
It's content discrimination under disguise.
From time to time we get a customer who calls up to order a connection specifically for a VPN. They know exactly what they need. They won't listen. I try to steer them to a less expensive package that is perfectly adequate for a VPN. They'll have none of it. They want a bunch of special add-ons because they have done this all before with other ISPs and they know the ropes. They have no idea how a VPN even works, and they've never even heard of our company before, but they know exactly what they want.
I try to politely point out that I could help them save money and still guarnantee it will work without problems. But I won't argue with the customer. If they want to pay 5x for a bunch of custom stuff they don't need, I can do that.
In the end, they are very happy. They have outsmarted the greedy, conniving ISP, and got themselves a rock-solid service. I make a mental note to try to find a local contact for them with whom we can build a relationship of trust and perhaps persuade down the road that they could actually pay less without compromising their service.
This is not always the case, but it is the most common case.
So when a customer demands a dedicated static IP address, I wonder if they really need that. We charge extra for that because we have a limited pool of public IPv4 addresses. We can do it, but it is usually unnecessary, and better to avoid unless you really need it.
I'd assume they're restricting access further by only allowing certain ips thru a firewall in front of the VPN server, or requiring extra auth factors from unknown ips, or something.
There are other valid reasons for restricting the IP. I ask about those. I'm not arguing with the customer. I'm trying to be helpful.
Unless the customer doesn't want help.
When you have a corporate IT rep who is clueless about the options, they are not in a great position to build the optimal solution.
When you have an ISP rep who is clueless about the requirements, they are not in a great position to build the optimal solution.
When you have an ISP rep who is not willing to explain the options, you have a problem. When you have a corporate IT rep who doesn't know the requirements, is unwilling to explain the requirements, or is unwilling to learn about the options, you are unlikely to achieve the optimal solution.
When you have a helpful and capable ISP rep and a cooperative and capable IT rep, the world is a brighter place. But some levels of trust require either naivety or time and shared experiences. That's difficult when every contact between the customer and the ISP involve different people each time.
So, again, just like everything else, try to work with as small and local of an entity as possible to get the job done. This is true for government, auto dealers, ISPs, or anything else. Only scale up when there is an advantage to doing so.
It's twice the money for half the bandwidth, but I have enough bandwidth and I sleep better at night. Shouldn't be necessary, but . . .
This is a goldmine for ISPs. Over the last five-ish years they decided to think of themselves as "advertising" companies. FCC repealed the internet privacy rule last April, and what are they doing now? They're functioning as data brokers. We should expect this trend to continue and assume that ISPs will take advantage of Pai's NN repeal as soon as possible.
Comcast currently extracts almost $100 a month from me, in addition to the ~$10 I pay for a VPN exit node to hide traffic from their surveillance machine. Why wouldn't they want another $10 or so to "allow" VPNs, given regulators who are loudly, publicly giddy about screwing consumers?
I think the most efficient way to hide your service would be to tunnel it over HTTPS given that it's not really realistic to drop that when you're an ISP. Still, a relatively simple traffic analysis tool won't have too much difficulty differentiating between proper web browsing and something fishy over HTTPS even it it just looks at the shape of the traffic.
It's a 15 minutes task for good admin to start dropping VPN connections (not to block, but to interrupt every 5 minutes, but blocking is even easier).
One important info: - ISP can't see your VPN traffic. But they can see that you are connecting to specific IP (VPN service's IP) at specific port. Once you're connected to VPN you probably pass ALL the connection through it. So ISP can see only single connection to single point. That is more than enough.
So how can they block VPNs: 1) Most VPNs uses standard VPN ports. Just dropping connections to that port is a good start. 2) Dropping connections to a servers when this is your (almost) only, long standing connection. This is a good moment to store that IP as VPN service IP. 3) Reverse DNS on IPs you're connecting to. A lot of services will reveal that they belongs to VPN services. 4) Public list of VPN IPs. Some provides them. Some are already noted somewhere on the internet.
ISPs can even share their databases with each other to be better at VPN dropping.
This is very easy thing to do.
It would be REALLY hard to hide your connection to a VPN service. One thing that came to my mind, to hide your VPN servers you would have to: - use VPN services at random ports - change VPN IP every few minutes at random factor, so dropped connections will occur anyway - in the background you should keep direct (un-VPNed) connections to random servers at random ports, passing random data around (it will look encrypted just like VPN connection)
This would require huge infrastructure though. Really huge.
But still - if ISP notice behavior like that - it means you're probably trying to do something shady, like walk around their VPN dropping rules - so they could just throttle your connections even more, just in case :)
So in general - you're fu^Hat very bad situation. I'm from Europe. I'm "safe" now, but we are also going to have bad law for net neutrality soon.
If we can open up these rules while radically increasing enforcement of quality of service then this might be a good thing. It might be good to replace "up to 75Mbps" with "guaranteed 50-75Mbps and 5-15ms ping to nearest public internet junction, 95% of the time during peak hours." I'm sure some folks might even sign onto world-beating five-nines performance guarantees too.
But this is probably just a dream and we'll see the ugly product differentiation you're describing.
Hell yeah. I would happily pay extra for a decent service that advertised it’s minimum speeds instead of its maximum because in reality, the theoretical maximum speed of my connection is really not a useful metric unless I’m hitting it 99% of the time. I have a fast connection but I still typically get a good 50 Mb/s less than advertised (which luckily still leaves me with 200 Mb/s or so) making the advertised amount completely inconsequential and useless. Telling me the lowest I should expect, on the other hand, is actually helpful and allows me to do some rudimentary capacity planning to see if it’s suitable for me or not.
No.
A 'Netflix QoS package' is not a QoS package. That's against Net Neutrality.
A 'Video Streaming' package that offers high quality video streaming with low latency, but doesn't care about things like picking up dropped packets - and doesn't discriminate between content sources - this would be 'net neutral'.
"Because in order to sell you onto these bullshit packages they will end up making your regular connection hellish"
This is a completely separate issue and has nothing to do with NN really. They can (and probably do) this already today with their regular package offers.
Let's not confuse general ISP stupidity with actual NN.
Net Neutrality doesn't exist any more. Did you not read the article? This isn't hypothetical, it was allowed to expire as of today. It's no longer the law, ISPs can legally do whatever they want now, including a "Netflix QoS package".
I am admittedly not a lawyer, but I'm reasonably well informed on the topic and completely open to being corrected if you know better.
Yes, the US law requiring a certain kind of Network Neutrality did expire.
The point is: service providers charging for differentiated QoS (such as different packages for latency) is well within the concept of Network Neutrality.
What is 'out of bounds' for NN is charing for customers based on the content of those packages.
So:
> A low latency service (designed for gaming in mind), is NN
> A low-latency service 'for games only' - is not NN
> If the ISP charges you money to access 'Steam' games, but not for others - that is not NN.
I didn't.
You are saying what I am saying: You pay for 'Quality of Water Service'.
What if you need 5x the flow that the water company normally provides?
What if you needed, for medical purposes 'unflourinated' water (whereas all of their other customers use flourinated).
With water:
+ If you want more water monthly - you pay for that. + If you want higher water pressure - you pay for that. + If you want different kinds of water (hard, soft) you pay for that.
That's 'Water Neutrality'.
What you do not pay for is a higher rate because you're a hospital. Or because you are wealthy and the water company knows they can scam you.
If you do not allow the water company to charge for 'high pressure' or 'hard / soft' water types - then they cannot justify in the investment and no such services will exist
As for networks:
+ If you want a dumb pipe, you pay for that. + If you want a faster pipe, you pay for that. + If you want more data, you pay for that. + If you want higher service level guarantees - you pay for that.
It's like buying AWS, or a new computer, or whatever.
What you should not have to pay for is: $X for Netflix data and $Y for HBO data. Or $X for Facebook and $Y for Linked in data. That's raw price discrimination and it's a problem.
Internet providers have massive equipment and support costs to support streaming video, so if you want amazing streaming at 4K super high quality, well, they can provide that if you're willing to pay. But - they should charge the same irrespective of content (i.e. HBO, Netflix etc)
Do they though? Over and above just providing a link with a sufficient bandwidth and contention ratio?
It doesn't matter.
Apple can charge you whatever it wants for it's iPhone. You can pay or not.
ISP's can charge you whatever they want for whatever QoS. You can pay or not. That's Network Neutrality.
What they cannot do under NN, is charge you based on who you are, or based on what you use your service for. So, charging you for Netflix usage, vs HBO usage is totally out of bounds.
If they sell you a 'streaming service' with whatever QoS - you should be able to it for whatever on earth you want.
My point is: there is usually an underlying cost-based reality to most of these things, and some of the people on this thread are not actually arguing for 'Net Neutrality' - they are arguing for 'ISP's should only be allowed to bill me in a very, specific narrow way, by law, irrespective of what consumer needs are or what their costs are'.
So if you buy a QoS from an ISP, they can't throttle you if it's for P2P or for Netlflix or HBO or whatever. That would be against NN and of course outside the bounds of any normal QoS.
It's just like buying form AWS: you pick from any number of configurations, but AWS can't charge you if you are doing 'gaming' vs 'news' on your hosted service. Amazon doesn't bill you based on what you are doing, just how you use the service.
That's NN.
So, under NN, ISP's can definitely sell you a 'low latency' service, as long as customers want that.
Eh, this isn't true in most places.
You pay a flat rate since you most likely have a 1"-1 1/2" pipe ran to your residence. Even if you cut the pipe off after the meter and let it run full blast you can only burn up a relatively limited amount of water (compared to what the system can carry). If you have a much larger line it is highly likely that your contract has usage conditions.
But the water company should not bill me differently if I choose to water tomatoes, but not chillies.
If turning on your faucet took an average of twenty minutes before the water showed up you'd probably want a high-cost low-latency spigot in case of a fire.
Nobody cares if an ISP wants to offer 2 "packages":
1. High latency, high bandwidth
2. Low latency, low bandwidth
Just like the metaphor of a water pipe, an ISP could offer "normal" where the water takes 20 minutes, or "fast" where it takes a second, but you have to pay 10x or have a much lower limit.
What NN is all about is preventing the ISP from saying HOW you are using the data. In the "water" analogy it would be like your water company saying you can ONLY use their "fast" water in their proprietary "soup making system", then they slow down all other water if they detect you are using it in any non-compliant pots.
I'm more than fine with an ISP that wants to offer a special low-latency VOIP option, but when they slow down all their competitors VOIP services and only allow you to get their VOIP service with low-latency, then it's a problem. Or the all too real possibility of Comcast allowing something like HULU to not apply to the data-caps, while not offering that option to Netflix, Youtube, Amazon Prime Video, or even your own startup or self-hosted option.
That's what I want my government protecting me against, that's the kind of thing that should be heavily regulated.
I know this forum is full of people who want total control of everything but I would bet this would provide a better experience since it doesn't have to be managed by the customer.
> Or the all too real possibility of Comcast allowing something like HULU to not apply to the data-caps
This doesn't technically fall under NN but I agree it's pretty scummy. I think it's a tough argument to make the ISPs shouldn't be allowed to do this though. If Hulu wants to make a deal with an ISP where they pay for their customer's data usage and pay more to be exclusive I don't see why that should be forbidden.
An example outside of ISPs. Spotify used to integrate Musixmatch into their app which would otherwise cost users $12/year. I wouldn't feel comfortable saying that Spotify can't make an exclusive deal or that Musixmatch must allow GPM to integrate with them as well.
I'm perfectly okay, as long as you are using the word "application" as a synonym for "type" and not as another word for "service or platform".
In other words, it's okay if VOIP service is prioritized via QoS rules, much of the internet as we know it won't work without this kind of thing, and it's a good thing for everyone involved.
What's NOT okay is when ComcastVOIP is allowed low-bandwidth, low-latency, but GoogleVOIP isn't. Or when each VOIP company needs to "apply" to get approved for the special consideration, or there are additional "rules" that have to be followed that deal with where the data is going, what it's doing, or how it's being used (like saying "VOIP that is used for business must pay more" is not okay in my book).
>This doesn't technically fall under NN but I agree it's pretty scummy. I think it's a tough argument to make the ISPs shouldn't be allowed to do this though. If Hulu wants to make a deal with an ISP where they pay for their customer's data usage and pay more to be exclusive I don't see why that should be forbidden.
It may not, but it absolutely should. At this point nothing falls under NN as it's not a law any more, so I'm not buying this "excuse" any more.
un-limiting one companies data is the same as limiting all other companies data in my opinion.
Again, if that data-cap-bypassing stuff works on a protocol level (as in VOIP data doesn't count toward your limit), then i'm perfectly fine with it, but when it deals with who owns the server, what content it's displaying, and who you are paying to get the service/data, then it should be heavily regulated.
>I wouldn't feel comfortable saying that Spotify can't make an exclusive deal or that Musixmatch must allow GPM to integrate with them as well.
neither would I, but that's not what i'm worried about. Bundling and deals including other services is fine, but what isn't is when those bundles bleed over into the actual data being sent. Just like how i'm fine if my water company provides a free water filter service with their water service. But when they start interfering with other companies water filters, or start making their core service (the water itself) cheaper when used with their provided water filter, then it's in what I consider anti-competitive territory.
And I guess that's the core of the problem I have with things like "unlimited for one service" deals. It's not just including another service with your internet, it's the ISP actively changing the price of their own core service when you use another unrelated service.
While I understand that a policy prohibiting this might end up slowing or stopping some innovative or new ways of providing internet access to people how they want it, I truly feel that the upsides of not allowing the ISPs to play gatekeeper (whether intentional or otherwise) will greatly outweigh the lost innovation.
I also want to point out that i'm not calling for these "shady" practices to be outright illegal, just that they should be "heavily regulated". Meaning there should be quite a lot of oversight, time should be spent determining if there is a conflict of interest or if the specific practice will end up harming competition or innovation.
I have a 3/4" pipe entering the house that I pay for. I can upgrade it to 1" for $50 more a month (and a ton of capex replacing the spur to the service line). I can upgrade it to multiple inches for for $100's more a month as well.
My Internet connection is capable of considerable more bandwidth than I'm paying for. Generally speaking, it's just a software setting on my modem.
I guess you could sort of compare it to having the water provider bringing in a 6" line to my house and tapering the line down to 3/4". But even then, their isn't a way for them to force me on to a specific water well or source, or directly force pricing based upon what I use the water for.
Its really just a tough comparison.
It's like if a football stadium was located near a tollway, and the tollway decided to start charging the stadium itself during sporting events.
That's not a legitimate business transaction, it's a cash grab.
Yes, I do as the consumer of that water. In this analogy Netflix would be the dam operator, so why would a particular dam pay more to the pipe company than another dam?
You pay your utility bill every month for electricity (probably a base rate, and a fixed consumption rate). Does your utility then go to the provider (A local green power source) and demand they pay the utility as well, since 60% of the requests in the region are for their power? (users opting into to green power on their bills?)
So do I.
Sounds like it is already paid.
What they want though, is to charge data coming in, and data going out.
Wouldn't segregating VPN traffic, or gaming, or streaming traffic be its own form of content discrimination? They are picking winners and losers in terms of tech that consumers use right there.
I live in the Bay Area and there is only one ISP (Comcast) that offers me more than 5Mbit.
Package 1: Fortnite, CS:GO, Hearthstone
Package 2: Rocket League, Dota2, World of Warcraft
Package 3: Overwatch, PUBG, Borderlands 2
Everything a gamer needs!
Well, the vendor shouldn’t dictate the class of my packets. Offering me “super low latency” or whatever isn’t being a dumb pipe—emphatically my preference—it’s upselling me something I never asked for.
If it costs the vendor 2x to provide a super low latency network, then they can charge you 2x for that.
Do you understand that there are actual costs associated with providing differentiated QoS?
Selling based on latency is no different than computer manufacturers pricing on GhZ, RAM and any number of other factors.
They should definitely be able to discriminate on the QoS that you want.
They should not be able to discriminate inherently on the content type.
If you buy 'low latency' you can use those packets for whatever you want. Gaming. Otherwise. And of course there can be no discrimination based on content vendor, i.e. game 1 vs. game 2.
But to suggest that vendors can't discriminate based on the service they provide you ... completely defies the notion of what business and entrepreneurialism is.
I have one ISP in my area that offers speeds above 5m consistently :(