Comcast and AT&T set data caps at 1 TB per month, knowing that we will all soon be doing much more traffic than that. Every year bandwidth requirements go up and it's accelerating.
Home internet access in the US went from unlimited usage to capped with $200 overage fees overnight.
We need an "Effective Sustained Rate" law that says something like:
If you sell internet access with a 1 terabyte cap, you have to market it as a 3 megabit connection. The effective sustained rate. Or preferably, just outlaw data caps entirely.
It actually kind of is related to data caps. There's many telcos who use a cap, but make it so certain websites (HBO, Facebook, etc) won't count towards your cap. This means that traffic is no longer being treated equally. It also means you're more likely to use services that don't count towards your data cap, so entrenched players are going to stay entrenched longer.
I wonder if this will lead to capped content providers to mask their data as online gaming or other different traffic. A mod on the protocol both at servers and clients side would allow that just fine.
That’s different. The GP was referring to global caps. What you mention is treating data differently(count vs non-count) which is exactly what NN is all about.
> If you sell internet access with a 1 terabyte cap, you have to market it as a 3 megabit connection. The effective sustained rate. Or preferably, just outlaw data caps entirely.
That's like saying that laptops must be marketed with their "effective sustained" battery life. The MacBook Pro can't be marketed as having 10 hours of wireless web, but rather just 2 hours of "effective sustained" performance. Regardless of whether it's a sensible measure for your typical user, who is actually likely to get the 10 hours because they're using their laptop for office and web browsing.
Residential ISPs are selling burst connections, which is an entirely sensible thing to do in a world where the top use case for a home internet connection is sucking down a 5MB Facebook page as quickly as possible then going idle. It's not like they're hiding the data cap in fine print: https://dataplan.xfinity.com.
Note that outlawing data caps doesn't address the phenomenon of companies with market power being able to charge monopoly prices. It just redistributes the burden of that upcharge away from heavy data users to ordinary users. E.g. consider if you outlawed Apple's practice of charging $100, $200, etc., more for $15-20 worth of flash memory on an iPhone. Do you think Apple would just lower those prices and leave it at that? No, they'd raise prices on the base model. All you'd do is hurt the users who keep everything in the cloud in order to favor those who store a lot of data locally. (Unsurprisingly, price discrimination tactics that benefit ordinary users at the expense of power users are unpopular among power users.)
So then all the measurements are relative but at least there is an equal and measurable baseline? I’m not for regulating technology though because it’s by definition not set in stone. What might have been reasonable to measure in the 80s is out of touch with how it might be measured today. Ideally we outlaw deals that restrict competition that allows monopolies in the first place
> So then all the measurements are relative but at least there is an equal and measurable baseline?
But the baseline measure may not be useful to the customer. In the laptop example: much of the advance in "web browsing battery life" in the last several years has come from lower idle power usage, or being able to reach idle more quickly. An Ivy Bridge Core i5 laptop and a Kaby Lake Core i5 laptop may actually get similar "full load" battery life, but the Kaby Lake model might last twice as long for web browsing or general office use. Reporting the "full load" battery life will actually mislead the consumer about the typical battery life of the two laptops.
Likewise, if you sell a gigabit cable connection with a 1 terabyte cap as a "3 mbps" connection, then you're misleading the customer about whether that's fast enough to do things like download an iTunes movie in a reasonable time, use a VPN into their work network, etc. Ironically, you might do more consumer harm by effectively upselling consumers to buy "faster" connections than they need.
1. Exactly. There are no usage fees associated with a laptop. You don't pay for recharging your battery over a certain number of times per month. They are marketed and sold in a technically optimal way. What country would permit $200 fees for charging your laptop too many times?
2. These corporations sell "burst" connections because they want to screw users out of their hard earned money. They could just as easily cap the peak bandwidth, throttle past a certain amount of usage, or shut down the connection if they were operating in good faith. They want to send $200 surprise bills to poor people.
3. People are watching YouTube/Netflix, etc not just browsing Facebook. It's not 2004. Many millions are also doing 24/7 home monitoring, software updates, nightly backups, etc. Bandwidth caps restrict US technical competitiveness and make life suck for citizens.
4. There is no need to keep charging more and more money. The costs of fiber optics, switching, routing, etc have all dropped dramatically even as our usage increases. We're being charged more than anyone in the world already. These corporations have also received hundreds of billions in government subsidies.
5. A 3 megabit connection is not "broadband" and that should trigger some other law. They should be forced to sell 10 megabit sustained (3 TB/mo) connections at a minimum perhaps. The minimum could go up every year based on average usage.
Any country with a functioning democratic government would ensure all citizens have access to low cost unlimited broadband. Look around the world, they mostly do.
I pretty much just use the internet to browse the web (where I care about burst speed a lot) and to watch the occasional video online. I come nowhere near the kind of caps you are talking about and don't care at all about sustained speeds.
Your plan would mean that I end up paying more money while those that need the sort of sustained speeds you're talking about pay less.
Why do you want to be forced to pay for something I don't want or need?
You're accepting as fact that rising bandwidth usage = rising costs. This is not how the technology works.
A 10 gigabit switch a few years ago used to cost $1,000,000 and now it costs $10,000. CPUs, optics, everything is just printed plastic and metal boxes. There's nothing fundamentally expensive about any of it.
Engineers and scientists can drive the price of bandwidth down to nearly zero. Comcast and AT&T simply want to extract as much money as they possibly can.
None of it would be possible if they didn't bribe politicians and run massive disinformation campaigns against the population that convinces them that bandwidth is rare and expensive.
People are defending the idea that we should be happy to have the scraps Comcast and AT&T see fit to give us. We should be demanding to exercise the full power and freedom that the technology allows.
Saying you don't use it so you don't care is just shorted sighted and selfish. The poor kid in Chicago educating himself by watching YouTube videos is going to end up getting his parents water shut off with $200 overage fees.
Internet bandwidth usage is growing 25% per year: https://www.cisco.com/c/en/us/solutions/collateral/service-p.... The amount of performance you get out of $1 of Juniper or Cisco gear is not going up 25% per year. And it's not even just about upgrading to the latest switch or router. You've got a cable network where each node used to support up to thousands of users, and each user might be several layers of amps away from the node. To accommodate newer generations of DOCSIS, you've got to split nodes and eliminate the intermediate layers of amps: http://www.lightreading.com/cable/ccap-next-gen-nets/cable-n.... Replacing 1 node with 10 isn't cheap, nor are the (often unionized) technicians that have to drive around neighborhoods burying fiber and installing nodes.
1. Juniper and Cisco aren't nearly as important as they used to be and their importance is growing less every day. Networking gear is headed towards commodity pricing.
2. Network technology almost certainly improves by 25%+ per year in terms of price-to-performance. Generational upgrades will make it uneven but that's not an unrealistic number to hit on average.
3. The "last mile" is the hard part but sticking with cable (DOCSIS) technology is not a requirement. There's no reason we couldn't have a massive federal Fiber To The Home program that cuts through current regulations.
The one-time costs are undeniable but they can be easily and predictably amortized over a long period of time. The theory that networking technology can't keep up is technically baseless.
> 1. Juniper and Cisco aren't nearly as important as they used to be and their importance is growing less every day.
If I were running a multi-billion ISP, I'd use Juniper and Cisco gear too.
> 2. Network technology almost certainly improves by 25%+ per year in terms of price-to-performance. Generational upgrades will make it uneven but that's not an unrealistic number to hit on average.
It took 12 years to go from 10 gigabit ethernet to 100 gigabit (about 21% per year over the decade), but I'd be very surprised if the price was the same (i.e. that 100 gig switches cost the same now as 10 gig switches cost back then). Carriers and data centers moved directly from 1G to 10G. But it seems like with 100G, there is significant pricing back pressure creating the need for intermediate speeds: http://www.mellanox.com/blog/2016/03/25-is-the-new-10-50-is-....
But let's say a 100G router today costs the same as a 10G router cost 12 years ago. 21% per year is just about what you need to keep up with projected demand growth over the next five years: https://www.cisco.com/c/dam/m/en_us/solutions/service-provid.... Even if 100% of your cost was scalable at the same level, you're talking about keeping up with demand for the same money, not less money. Of course, 100% of your costs aren't scalable. The last mile DOCSIS part of your network is not scaling 20% per year.
> 3. The "last mile" is the hard part but sticking with cable (DOCSIS) technology is not a requirement.
Verizon tried that. It's barely breaking even on FiOS even though the average FiOS customer brings in more than $100/month in revenue.
> There's no reason we couldn't have a massive federal Fiber To The Home program that cuts through current regulations.
I can think of way better uses for hundreds of billions of federal dollars than subsidizing Netflix's business model: http://www.afr.com/technology/web/nbn/australias-highspeed-i.... For the forseeable future, fiber is a luxury product, and to the extent we want government-funded broadband to address "Internet access as a human right" issues, we should do it using cost-effective 5G wireless technology.
> The one-time costs are undeniable but they can be easily and predictably amortized over a long period of time. The theory that networking technology can't keep up is technically baseless.
It's not that networking can't keep up. It's that upgrading the existing cable network to keep up with demand growth is expensive, and replacing everything with fiber is even more expensive.
You're talking about changes in absolute costs. I'm talking about relative costs. It doesn't matter how good networking equipment gets. It will always cost providers more money to service users saturating their connections than more typical consumers that don't saturate but care about burst speed.
For every poor kid in Chicago that happens to be using a highly above average amount of total bandwidth per month there will be 50 poor kids in Chicago with more typical usage patterns that your plan would hurt.
None of those facts contradict my point as they, again, speak towards absolute rather than relative costs.
Providers have a finite amount of bandwidth in their trunk connections. More bursty users can use a single trunk than saturating users. Therefore saturating users are more expensive.
It's just like CPU. If you're bursting up to capacity but on average only using, say, 10% of available CPU then Amazon can pack more of your workloads on a single physical machine compared to a different user that is using 100% of CPU all the time. The fact the CPUs are always getting cheaper and faster doesn't negate this point.
> Residential ISPs are selling burst connections, which is an entirely sensible thing to do in a world where the top use case for a home internet connection is sucking down a 5MB Facebook page as quickly as possible then going idle. It's not like they're hiding the data cap in fine print:
Go idle? Since when?
Steam auto updates my games, Windows downloads multi-gigabyte updates in the background. My Android phone syncs files in the background to cloud storage, and on a nightly basis finds a good dozen or so apps to auto-update. For many people, Alexa now streams music in the background whenever they are home. Multiplayer online games nowdays have voice chat, increasingly richer and more populated worlds, and support video streaming of one's exploits.
Even that 5MB Facebook page is actually an endlessly scrolling multimedia fest, that starts streaming videos as they come into view.
Heck Hacker News is one of the few paginated sites left!
It's not "endlessly scrolling", it's scrolling for the few minutes you're looking at it. Nobody scrolls FB pages - or even downloads updates - for 24h in a day, and it would be absurd to limit speeds as if they did.
Even if you’re reading FB 50 minutes a day, you’re probably only downloading 5-10 minutes of that time. In the last two months, The FB app has used 557 MB of data on my iPhone. That’s about 1% of my soft data cap over that time with VZN.
Checking my Laptop's network usage, over the last 30 days, Chrome has downloaded 88 GB of data.
To be fair, I stream movies using the Netflix website, 3 movies, so approximately 18GB to video.
The rest? Who knows. Youtube is certainly part of it, though I rarely watch videos full screen so whatever 480p takes up.
Then there are mobile apps like Snapchat which, for heavy users, can eat up an easy 5GB or more per month.
I had the Chrome bandwidth monitor open while researching that number. The 3 or 4 articles I read to figure out the average Snapchat data use ended up themselves using 36MB of data! (and this is with an ad blocker running...)
Figure it took me 3 minutes (not really) to do the research. that is 12MB per minute of data consumption just browsing the internet for references!
My overall point here is that data usage has exploded, and not just for consumption of streaming media.
This. Even when I'm sleeping, my internet connection is literally never idle. I wouldn't consider myself to be a heavy internet user - demanding and impatient, maybe, but I'm always clocking in at < 1000GB.
Well if they were selling "burst" speeds, then why are the bandwidth limits so low? My cable modem is locked at >600Mbits of bandwidth of which I get a tiny fraction, against a small subset of the channels available (totaling a few Gb/sec), and the node is split about two blocks away. The upstream numbers are even uglier.
Worse, the cable companies have had actual burst traffic shaping abilities for at least a decade. They could sell you 100Mbit, and run it at 600Mbit when the line was idle for a few seconds. That would make people pretty happy at long as they were getting the advertised minimum...
In fact if someone were to actually put real regulation down, I would expect that the plans were sold with a minimum guaranteed/sustained bandwidth (up and down, because I can't tell you how many advertisements I've seen that never mention the connection is asymmetric) and an "up to" burst metric.
It would be pretty simple, buy a (50/20 plan, that is up to 500/200 during off peak). What it would do though is guarantee that the company isn't selling 1000000x oversubscribed plans because they haven't upgraded the infrastructure in 20 years.
> Well if they were selling "burst" speeds, then why are the bandwidth limits so low? My cable modem is locked at >600Mbits of bandwidth of which I get a tiny fraction, against a small subset of the channels available (totaling a few Gb/sec), and the node is split about two blocks away. The upstream numbers are even uglier.
Because the medium is shared with hundreds of other houses? Because regardless of the amount of bandwidth on the medium, users are sharing a limited number of ports on upstream routers? Because the ISP wants to price discriminate, and make more money from people who want faster speeds?
> In fact if someone were to actually put real regulation down, I would expect that the plans were sold with a minimum guaranteed/sustained bandwidth (up and down, because I can't tell you how many advertisements I've seen that never mention the connection is asymmetric) and an "up to" burst metric.
Regulation has to have some sort of economic point. What's the economic point of preventing ISPs from practicing price discrimination, which is generally beneficial to less intensive users?
> What it would do though is guarantee that the company isn't selling 1000000x oversubscribed plans because they haven't upgraded the infrastructure in 20 years.
The factual premise of this is false. There's been massive upgrades over the past decade to support newer DOCSIS versions: http://www.lightreading.com/cable/ccap-next-gen-nets/cable-n.... Your desktop CPU is maybe 2-3x as fast today than it was in 2008 (quad core Nehalem to quad core Coffee Lake). Less than 2x for single-threaded workloads. Your internet connection is probably 20-50x faster, at least in a large-ish metro area.
#1: Price discrimination, because you know you have to reserve >75% of the bandwidth to assure that the guy with the 100mbit plan isn't being pushed out by the dozen or so users with the 20mbit plan, on a docsis 3.0 network with 16 channels bonded... Really, ever wonder why the cable industry won't actually tell you the line utilization on your node? What they have been really successful at is selling people service and modems that barely give you line quality information, knowing what your modem is actually capped at, or what current activity on the line is privileged information available to only the select few.
And finally, the article you linked mentioned that the amps haven't been upgraded in 12-15 years... you may be in a market getting a docsis 3.1 upgrade, but in many markets docsis 3.0 is barely rolled out, and it rolled out at speeds achievable with docsis 2, a standard that is 17 years old.
Never mind the insane garbage peddled by the DSL providers in many markets. So I maybe I exaggerated a bit with 20 years, 15..
If the caps apply equally to any and all data, I really have no problem with them, and neither should you.
We're just asking for a content-agnostic utility. Not "as much as you can eat". Consider if electricity was "unlimited". "Unlimited" is part of what got us into this mess in the first place. Delivering lots of data to each of many subscribers is just not that easy and cheap. At least with explicit caps companies can compete by offering higher caps. Instead of mysterious limitations on "unlimited".
There's no "mess" with unlimited bandwidth. Bandwidth is not like electricity itself. It's a fundamentally cheap thing to flash LEDs down glass pipes.
There's no technical or financial reason why every major American city shouldn't have unlimited gigabit for $50/mo per household. The only expensive part is running conduit and it's a one-time cost.
It's disappointing how many people buy into the talking points of these shitty corporations as they work hard to fleece and permanently damage an entire country.
> There's no "mess" with unlimited bandwidth. Bandwidth is not like electricity itself. It's a fundamentally cheap thing to flash LEDs down glass pipes.
Electricity is actually a lot like Internet. Sending electrons down a wire isn't as cheap as sending flashes of light down optical fibers, but in both cases variable costs are vastly lower than the fixed costs.[1] In both cases, there is a total system capacity you can achieve for a given fixed cost. And in both cases, usage is such that what you're really trying to manage is the amount of fixed capacity in relation to usage peaks. It's fundamentally about managing congestion rather than delivering units of a product.
> every major American city shouldn't have unlimited gigabit for $50/mo per household
Other than the fact that it's not economically possible? Chattanooga charges $60 for 100 megabit and $70 for gigabit, and its just about breaking even. And Chattanooga is a very favorable city for fiber buildout: power lines are overhead, not buried (the same reason Google built in places like Kansas City, not San Francisco), and its a southern city with relatively low cost of labor. New York, D.C., etc., are all going to be more expensive. Nor is $50 for gigabit common in western countries for fiber, at least in places where fiber isn't subsidized. Gigabit fiber in the U.K. is $70-100/month. In Sweden and Denmark it's $120+/month (and typically have multi-thousand up-front install costs).
[1] The typical American pays 12 cents per kilowatt hour for electricity. The fuel cost for a fossil fuel plant is about 2.5 cents per kilowatt hour, while for nuclear it's less than 1 cent. Almost all of the cost of electricity, just like with broadband, is operations and maintenance of the infrastructure, and fixed cost recovery.
I live in Japan where "1 Gbps" connections are offered at around $50/mo over a nationwide open access fiber network. But to prove your point - the consumer ISPs here are comically oversubscribed, to the point where with few exceptions you're lucky to exceed 100 Mbps during peak hours.
I want a law requiring ISPs to post how much time per month you're allowed to use a connection at full speed. For instance, if you have a 100Mbps connection and a 100GB cap, then you're only allowed to use it for 10,000 seconds ~= 3 hours per month. Does your cell phone's LTE give you 50Mbps for up to 10GB? They're really selling you about 2000 seconds ~= 33 minutes per month.
I think any advertisement that doesn't include this basic fact is misleading to the point of being a lie.
And, my own metric, of how many 56k modems do you need to get the same amount of data (@~14GB/month download rate).
AKA, most cell phone plans this year, can't even download as much in a month as my 56k modem could in the late 1990's.
In the end, the big scam here is that if anything netflix/etc traffic is probably less expensive per packet than most other services because netflix is more than happy to scatter their little caching boxes everywhere, so the ISP's backhaul/peering arrangement has less traffic...
Before anyone gets ahead of themselves, this bill only protects net neutrality and does not enforce net neutrality as defined here https://en.wikipedia.org/wiki/Net_neutrality .
ISPs are still allowed to discriminate traffic under this bill. They are allowed to block and throttle content such as illegal streaming downloads, some torrents, spam, and other traffic that degrades network performance. For example, this bill does not prevent Comcast, At&t, Verizon et al. from blocking port 25 as they have been for the past decade. The bill also does not preclude zero-rating.
This bill is a good thing yes, but let's not get ahead of ourselves and call this net neutrality.
I'm not surprised at all. I bring it up because it's an example of a blatant net neutrality violation that many people agree is acceptable. Same goes for IP ingress filtering.
Net Neutrality doesn't prevent network operators from stopping malicious activity on their network.
Blocking Port 25 to specific providers unless they pay up would be a problem, but blocking port 25 globally (except, say, to their own SMTP servers) is perfectly okay.
What the parent is talking about is the more generalized "net-neutrality violation" of assuming that users on residential plans have no legitimate reason to be hosting any public-routable services. It's the ~30-years-back ISP coup of dividing the flat space of Internet "peers" into separate categories of "servers" and "home PCs", where default-deny policies are enforced on "home PCs" to prevent them from doing most of the things "servers" can do.
Theoretically, true "net neutrality" (in quotes because it's not really the same thing that the law that got struck down protected) would require that ISPs not discriminate what types of traffic a customer can use their pipe for, any more than a bank is allowed to discriminate what you use your checking account for.
Under this true-Scotsman net neutrality, ISPs should be required to let people host web servers, or mail servers (if they're not open relays), or whatever-else servers, on their residential Internet. If that causes uplink saturation, then price uplink bandwidth in your plans to match your costs! (Which basically would eliminate the difference between residential and business Internet plans anyway.)
Please direct me to a source explaining why malicious network traffic is exempt from net neutrality. Under the net neutrality definition linked above, blocking specific ports is a clear violation.
Also, the port 25 ban directly affects legitimate customers. At least for Verizon, port 25 is blocked for residential customers, but it's not blocked for commercial customers. Over the past decade, many residential customers have called to enable port 25 so they can run their own mail server, and I've read Verizon refuses to unblock the port. Verizon's "solution" for them is to upgrade to their commercial package where it's not blocked.
> Under the net neutrality definition linked above, blocking specific ports is a clear violation.
I don't see how you can read that from the linked wikipedia page. I assume you're conflating blocking ports with blocking/slowing protocols.
> Please direct me to a source explaining why malicious network traffic is exempt from net neutrality.
It's not an exhaustive policy document of every possible reason why network access might be terminated/restricted. eg no discussion of terminating/filtering BGP peering announcements from a misconfigured device.
Few people, if any, have ever said that network management activities are would be prohibited under NN. If someone is actively disrupting the network, or is otherwise acting maliciously, then clearly as a network operator you're going to be permitted to block that traffic.
> Over the past decade, many residential customers have called to enable port 25 so they can run their own mail server
Sure, but even if they did permit it - the chances you will be able to run a mail relay from an IP that's clearly in the same network block as a bunch of other residential networks is near zero.
Most major mail providers will outright block you for existing in that range, and most others will block you for being unable to demonstrate control over the IP (by setting Reverse-DNS).
> I don't see how you can read that from the linked wikipedia page.
Here's how net neutrality is defined.
"Net neutrality is the principle that Internet service providers must treat all data on the Internet the same, and not discriminate or charge differently by user, content, website, platform, application, type of attached equipment, or method of communication."
Blocking all TCP/IP packets with first two bytes set to 25, and not blocking TCP/IP packets with the first two bytes set to 24, is a clear violation of the Net Neutrality principle of no discrimination based on application. If I send a TCP/IP packet with source port 24, my ISP does not block it. However, if I change just one bit in that packet, my ISP refuses to send it.
An ISP blocking an entire class of applications from the internet (in this case, SMTP servers) is most certainly a net neutrality violation.
> If someone is actively disrupting the network, or is otherwise acting maliciously, then clearly as a network operator you're going to be permitted to block that traffic.
Where specifically in the definition of net neutrality does it permit ISPs to block such traffic? Blocking such traffic is a violation of the net neutrality principle that "Internet service providers must treat all data on the Internet the same".
> Most major mail providers will outright block you for existing in [a residential IP block]
ISPs have blocked residential customers from using SMTP for over a decade now, therefore it's no surprise mail providers have little incentive to support residential SMTP servers. They've been forced out of the market.
ISPs should be dumb pipes. They should not be inspecting nor blocking packets and should transmit data regardless of its content.
The alternative, that ISPs should not route unlawful traffic, would require ISPs to search packets and block the ones they find unlawful. That is censorship.
Using weasel words like "standard network management" doesn't imply those standards conform to net neutrality principles. It's clear that that some of those standards, such blanket blocking port 25, are a direct violation of net neutrality.
I did notice you tried earlier to change the definition of net neutrality by claiming it "doesn't prevent network operators from stopping malicious activity on their network", when really there is no exception for this. Like netflix tried to do, you can't decide what is and what isn't net neutrality just to fit your viewpoint.
I fully understand why you're trying to avoid this topic by describing at as "extremely unhelpful" because I know it's a sore point many people try to avoid discussing. But I have no qualms pointing out flaws in other people's arguments :).
the negation of "operators must block all unlawful traffic" is "operators are permitted to not block all unlawful traffic", not "operators must not block any traffic on the basis of it being unlawful."
> Please direct me to a source explaining why malicious network traffic is exempt from net neutrality
All the Net Neutrality legislation proposals I've seen contain explicit wording to the effect of "lawful network traffic". If malicious traffic violates the CFAA or CAN-SPAM, they would be in their right to block it.
I don't see any real point in including zero rating under net neutrality, for several reasons.
1. The usual argument is that if an ISP can zero rate service X, that makes it harder for competitors of X to compete, and so we need to ban zero rating to encourage competition.
This argument fails because a last mile ISP like Comcast can easily side step any internet neutrality rules in offering X by making a deal with the X provider to serve the service from Comcast's own servers. Then it is not being provided over internet and so is out of scope for net neutrality.
They could even go farther, and not even use IP at all to deliver it from Comcast to the home. They run IP on top of lower level protocols, and they could use one of those lower level protocols to deliver to an app running on the set top box, which could then serve on the LAN using normal IP protocols (which would be out of scope for net neutrality because they are just running on a LAN in this case).
If we want to stop this for competition reasons, we'll need to use antitrust, not net neutrality. Antitrust cannot be evaded by simply changing the protocols.
2. We accept zero rating in every other or nearly every other common carrier service. In telecommunications. for example, I've never seen any objections to zero rated telephone service, whether paid for by the phone company (such as free calls to government offices), or paid for by the company called (commercial toll free numbers). In transportation, we don't get our feathers ruffled when a bus or cab company offers free rides (possibly paid for by a sponsor). What is different about internet?
Generalizing this, in general we allow nearly every provider of goods and services to give you MORE than you have paid for. As long as they deliver the service that you paid for, if they want to give you a gift of more service to promote a particular offering of their own, or to promote (possibly for pay) some sponsor, we allow it. Regulation is concerned with making sure they don't under deliver, not preventing over delivery.
It is hard to make a case to the typical consumer that there is something fundamentally different between, say, Safeway giving a discount on Pepsi, and an ISP giving a discount on bandwidth used for Spotify.
3. Suppose I'm a company selling something that uses a lot of bandwidth, such as a game that requires a large download. I'd like to pay major ISPs to zero rate it. If I cannot directly do that, but the ISP sells prepaid bandwidth cards that you use by entering a code from the card, I could simply buy prepaid bandwidth this way, and when someone buys my game I could give them a bandwidth code for their ISP to cover the costs of the download. Are we going to ban prepaid bandwidth to stop this?
If we want to get net neutrality back, and have it STAY back instead of just flipping every time we change which party is in control we need to keep it to just involving things ON the network. That's basically throttling, blocking, and paid prioritization. Billing is OFF the network. I think it should be reasonably possible to get bipartisan support for those.
The argument for prohibiting zero-rating under NN generally goes like this:
If provider X provides a very limited (or no) quota for regular internet traffic, and provides access to sponsored zero-rated services - then this inhibits competition, and (depending on how agressively it's implmented) can inhibit free speech.
This has been seen seen in some countries such as India where facebook.com was zero-rated traffic, and data prices for regular internet access were extremely high.
It's also seen today in Australia where Apple Music, Spotify and Google Play Music services are zero-rated traffic on certain mobile (cellular) plans - many of these plans also have zero-rated access to various popular social media platforms too (facebook, instagram, etc) - and data quotas are generally fairly limited (2-10GB/month is typical for mid-range plans). A competitor to these services is going to have a very hard time entering the market, since their customers will all have the data usage come out of their regular quota.
Continuing with Australia - Most ISPs provide zero-rated access to Netflix (or provide 'unlimited' quota).
Telstra does not zero-rate Netflix traffic, but they do zero-rate Foxtel and their own movie/tv services. At the same time, having rather low data quotas.
So those who're stuck with Telstra (eg because they live in an apartment/development that is a Telstra-exclusive estate, or just because they're on a 24-36 month contract, having been unaware of the problems) - are unable to use Netflix.
T-Mobile zero rates in the US as a technical measure, and all content that adheres to their technical requirements is eligible for zero rating (even adult content).
With Australia, submarine transport costs to Singapore (nearest 'major' Internet hub) or all the way to North America or Europe for the largest hubs is an order of magnitude more expensive than continental terrestrial transport within North America or within Europe. So that Netflix offers free caching boxes to ISP's meeting minimum traffic levels allows the Australian networks to save substantially on their overall bandwidth costs. It's reasonable for them to zero-rate that traffic as they're effectively passing their cost savings on to their customers, as opposed to trying to double dip by charging content networks for access to their subscribers. This may violate more pure definitions of net neutrality, but should be desirable behaviour as it's pro consumer.
This goes back to the distinction between the real problem, the lack of competition in many markets compared to net neutrality which mitigates situations where there's a lack of competition, but it's not necessarily a win all across the board. What's needed is to separate last mile access, from eyeball network service, from carrier network service, from content network service, from actual content. This would ensure competition across the board, and prevent excessive power accumulated by such a small number of large corporations as is currently the case. However, the nearly religious reverence to capitalism in the US would make such a solution nearly impossible to legislate.
> With Australia, submarine transport costs to Singapore (nearest 'major' Internet hub)
There's very little direct bandwidth to Singapore - there's only one cable going west (SEA-ME-WE-3), it's old, and constantly has breaks.
Unless you live in Western Australia, your traffic (even to Singapore) doesn't go via that - it goes via the US over SXC or Endeavor to the east, or to the north to Guam (PIPE-1) or Japan (AJC)
> It's reasonable for them to zero-rate that traffic as they're effectively passing their cost savings on to their customers, as opposed to trying to double dip by charging content networks for access to their subscribers. This may violate more pure definitions of net neutrality, but should be desirable behaviour as it's pro consumer.
I agree, however I've had (heated) arguments with people on this aspect. Not everyone agrees, and thinks that all traffic should be zero-rated if any traffic is going to be.
If that were the cause, then all ISPs would provide zero rating for netflix. Instead, only certain ISPs zero rate netflix, suggesting that those ISPs are being payed to do this.
Then it is not being provided over internet and so is out of scope for net neutrality.
This needs to be explained. If the zero-rated service is being provided as part of the same subscription, I don't see why it matter where the servers are located.
Comcast of course blocks incoming port 25 as well. I pay to have easydns deliver over SMTP to an alternate port. This means I can't really do effective spam filtering / denying at the connection level as the connecting server is always easydns. (Although denying spam that then bounces often is misidentified as originating spam... so...)
What best practices do others do when still hosting SMTP at home?
> However, state laws can only restore network neutrality for some Americans, and only a federal rule can ensure that everyone in the country has access to a neutral net.
To accurately reflect the scope of the problem that should read "some Americans some of the time," no?
I doubt California businesses and residents restrict their internet traffic to stay within the state.
Relevant anecdote: I've recently switched to full-time hotspotting with Verizon, and I'm still adjusting my Netflix habit to match.
Verizon, as usual, sent me my "you're at your last 10% before we throttle you" notice just as I went over my limit, and at the exact same time, DNS stopped working for netflix.com.
Everything else worked just fine, but none of the netflix.com properties would resolve.
I had lazily been using Verizon's DNS for my hotspot. Changing to 8.8.8.8/8.8.4.4 immediately brought it back online.
I can't prove it wasn't just a glitch, but it was pretty damn suspicious anyhow, and exactly the sort of underhanded behavior I'd expect from today's telecoms.
Oh! I thought I had lost this terminal session due to another recent forced restart (there's a VERY annoying bug in udev, Debian Stretch, or KDE Plasma 5 ... but another rant for another day).
But I haven't!
The server was fe80::188e:1b74:da0:42e9%3.
Here's some of my terminal output from when it happened. I was on Netflix when the notification came in and Netflix stopped loading less than a minute later.
...huh, that's weird. I doubt this is a Netflix issue. What does Google say?
$ dig www.netflix.com @8.8.8.8
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 48982
;; flags: qr rd ra; QUERY: 1, ANSWER: 10, AUTHORITY: 0, ADDITIONAL: 1
;; QUESTION SECTION:
;www.netflix.com. IN A
;; ANSWER SECTION:
www.netflix.com. 824 IN CNAME www.geo.netflix.com.
www.geo.netflix.com. 1314 IN CNAME www.us-west-2.prodaa.netflix.com.
www.us-west-2.prodaa.netflix.com. 28 IN A 52.25.169.106
www.us-west-2.prodaa.netflix.com. 28 IN A 52.39.122.191
www.us-west-2.prodaa.netflix.com. 28 IN A 52.35.217.11
www.us-west-2.prodaa.netflix.com. 28 IN A 52.26.64.50
www.us-west-2.prodaa.netflix.com. 28 IN A 52.33.220.134
www.us-west-2.prodaa.netflix.com. 28 IN A 52.33.159.26
www.us-west-2.prodaa.netflix.com. 28 IN A 52.35.162.191
www.us-west-2.prodaa.netflix.com. 28 IN A 52.39.27.188
;; Query time: 50 msec
;; SERVER: 8.8.8.8#53(8.8.8.8)
;; WHEN: Sat Jan 06 01:26:00 PST 2018
...Okay. What about YouTube, Reddit, HN, Google, Slashdot, my servers? Can I reach those? (Yes.) Okay, let's try reconfiguring this connection's DNS then to use Google public DNS.
$ dig www.netflix.com
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 45369
;; flags: qr rd ra; QUERY: 1, ANSWER: 10, AUTHORITY: 0, ADDITIONAL: 1
;; QUESTION SECTION:
;www.netflix.com. IN A
;; ANSWER SECTION:
www.netflix.com. 704 IN CNAME www.geo.netflix.com.
www.geo.netflix.com. 1194 IN CNAME www.us-west-2.prodaa.netflix.com.
www.us-west-2.prodaa.netflix.com. 26 IN A 52.39.175.187
www.us-west-2.prodaa.netflix.com. 26 IN A 52.40.134.13
www.us-west-2.prodaa.netflix.com. 26 IN A 52.37.33.45
www.us-west-2.prodaa.netflix.com. 26 IN A 52.40.109.21
www.us-west-2.prodaa.netflix.com. 26 IN A 52.39.98.186
www.us-west-2.prodaa.netflix.com. 26 IN A 52.37.69.124
www.us-west-2.prodaa.netflix.com. 26 IN A 52.40.112.84
www.us-west-2.prodaa.netflix.com. 26 IN A 52.39.34.178
;; Query time: 33 msec
;; SERVER: 8.8.4.4#53(8.8.4.4)
;; WHEN: Sat Jan 06 01:27:40 PST 2018
$90/month plus all their various fees and whatnot, works out to around $103 I think for a single line. Still cheaper than a reasonable Verizon plan with data plus Comcast.
I did it for much the same reason. Had to move anyway, don't expect to spend as much time at home over the next year.
20GB/mo coupled with the website obesity crisis (http://idlewords.com/talks/website_obesity.htm) coupled with it being nice to have a dumb show on in the background while doing dishes (about the extent of my Netflix habit) makes things uncomfortable for the last few days of my billing cycle. I'm currently browsing the web at a stunning, 21st-century America speed of 600kbps -- aka, I actually get better speed by going to a crowded Starbucks.
If you get sprint coverage in the areas you'll be you cant beat calyx institute. $400 a year for 4G LTE Mobile Hotspot + 1 Year Unlimited Unthrottled Wireless 4G / LTE Internet Service + T-Shirt https://www.calyxinstitute.org
Why is this article hidden and not on frontpage anymore? Only dumb people would choose to select their hangman themself aka not want to stay such an article visible. So oviously no private person would do that.
I don't want to be a bummer, but video streaming might stress the hell out of the internet infrastructure.
Before YouTube and Netflix, video broadcasting meant that you send data once, but a per view service sends a copy each time.
I'm all for net neutrality, but to be honest I think that some use cases might add a little pressure to the whole network when you compare them to more classic uses (HTML, images). I don't think that Snapchat, YouTube and Netflix can really be sustainable models if they keep growing and growing forever.
At some point, transmitting video online, on a large scale, at a good resolution, at peak hours, is going to be expensive.
Netflix is popular because the real cost is on the ISP, and there never are any guarantee of quality of service, it's not up to Netflix.
Netflix is already paying their fair share of bandwidth usage and infrastructure upgrade costs via their peering agreements. Ending net neutrality is double-dipping.
If you pay for 250 GB of data from your ISP, then they're on the hook for the cost of delivering that 250 GB. That's what they sold you - or actually what you should be able to buy.
The problem is when they'll deliver 250 GB of $PARTNER_SERVICE but not 250 GB of Netflix or YouTube, and none of the 1 or 2 ISPs in your area will give you a reasonable deal on your GBs from Netflix. (And to be clear, Netflix is not charging them anything.)
(The other problem is when they say "unlimited" but of course it's never really unlimited.)
I agree with you but then I believe that with the recent developments in video streaming (becoming mainstream) it would then be totally reasonable for ISP's to start charging a lot more for packages that give you the data to stream that much video.
It just seems like underlying the whole NN debate is the very large uptake in the average user's data usage over the last 10 years that is just sort of ignored when it comes to how do such upgrades get paid for and before such upgrades take place, who or what gets slowed down in the case of congestion?
Moore pays for that. My internet hasn't gotten 10 times as fast, but networking equipment did. So much exponential growth and here you are discussing if we can really deliver 10 Mbit to the population?
ISPs want to pay so little industry is still spending R&D money on how to make that fucking copper cable go all the way to Shannon capacity. G.fast is now doing over 100 Mbps on 500m copper.
The radio spectrum isn't the bottleneck for wired Internet. The bottleneck is in the routers, switches, etc. that sit in between the pieces of wire. Moore's law is highly relevant to how much bandwidth they can handle.
I agree that ISP's may need to charge a lot for a lot of data. Like say $200 for 1TB if you really need that much (made-up numbers for sake of argument). But the fact that you use it for Netflix vs any other service or reason should not affect the price.
There's also the issue mentioned in the sibling comment of the fact that internet use is naturally over-sold to consumers - the cost to guarantee a speed "at any time" is way higher than offering that speed "most of the time". I think that's also fair: if everyone in a city starts streaming a few different 4k shows at the same time one evening, it's fair if service slows down, as long as the advertised speed really is available the rest of the time. SLA (service-level agreement) is just another orthogonal aspect of the service that is sold (and businesses are generally willing to pay a lot more for consistent speed than a consumer, and this is also fair).
It just comes down to unfair treatment of competing services/uses by a monopoly or duopoly ISP. Bytes should be bytes. Like the electrons you get from your electricity provider.
I think the whole point OP is trying to make is that at the moment bandwidth works sort of like gym memberships or even airline tickets - they count on not everyone showing up at once.
In the even everyone shows up at once, sort of like in your scenario, infrastructure will of course have to be built out, but prices will rise as a result.
> I don't want to be a bummer, but video streaming might stress the hell out of the internet infrastructure. At some point, transmitting video online, on a large scale, at a good resolution, at peak hours, is going to be expensive.
This is really frustrating, since Netflix is usually not served over the internet, but directly from the ISP's head end. Netflix usage largely has no major impact on the world wide internet.
ISPs don't even have to serve you internet access for Netflix videos. Netflix already hand delivers the data to your town and then pays cash to Verizon and Comcast to serve their already-on-local-network videos to you. That's all on top of the money you are already paying the ISPs to receive that exact same traffic.
All ISPs really do for Netflix is take traffic from a private local field office to your home. Frustratingly, they are often too cheap and lazy to even manage that.
Netflix is already paying for a lot of infrastructure, local CDN boxes, to the point where ISPs only have to stick the fiber in - but they are unwilling, out of allegiance to their cable business or because they realize that if people start to use their 50 Mbit internet, it will become painfully obvious how oversubscribed, underdeveloped and underinvested their infrastructure is.
The ISPs are not keeping up their part of the bargain.
Netflix using bandwidth doesn't harm anything. IPSs are free to offer users all the bandwidth they have available. Running out of bandwidth is not harmful to anyone.
Putting pressure on the network is the only way infrastructure gets updated. We haven't reached the point where every user has the highest reasonable amount of bandwidth. I posit that we never will. ISPs want you to believe we have so you will believe people are using the internet "wrong" when the real problem is that ISPs don't want to invest in infrastructure. I won't hesitate to call them out in their BS.
I think you are overestimating the cost of bandwidth. Non-video web surfing as the bandwidth determining application has been surpassed a very long time ago.
A prime reason broadband stopped getting faster after around ADSL2 is that nobody knew what to do with all that bandwidth, and that was 15 years ago. In fact you can watch today's Netflix 4k on 2003's 24 Mbps ADSL2+ connection, if you're reasonably close to the DSLAM.
ISP backbone bandwith cost has continued dropping on that curve even if home broadband speeds have stagnated due to telcos investing all their chips into 4G coverage. (A silver lining is that 4G is symmetric so you can reasonably run video calls and home servers off it..)
I wish ISPs billed like a utility. Something like $10 connection fee + 10-15 cents per GB.
This will force bandwidth hogs to self-regulate. If you have the ability to purchase 4k/8k TVs, then you should be responsible for the associated streaming costs and not have others subsidize it.
Also, NN does nothing to foster competition - the real problem. All we have are a bunch of local monopolies/duopolies.
That's not actually what bandwidth costs. The cost to the ISP is based on the connection speed (eg, bits per second) not on how much is transferred.
The ISP needs to size for peak usage, and they over-subscribe because that's the only way to be profitable. This is why your speed will often be slower at night than during the day. It also means when you pay for data transferred at non-peak times, you're not actually costing anything since the bandwidth was sitting there unused anyway.
Sorry, poor wording on my part. I meant slower in the evening (peak time), and you're right: late at night is typically the least busy and thus fastest time.
Regulate the use of the term "internet service" to include network neutrality principles. If the provider violates these principles, they cannot call themselves an "internet service provider".
They can all themselves "network provider" or "mobile data provider" but they can't use the term "internet" anymore.
104 comments
[ 4.0 ms ] story [ 188 ms ] threadHome internet access in the US went from unlimited usage to capped with $200 overage fees overnight.
We need an "Effective Sustained Rate" law that says something like:
If you sell internet access with a 1 terabyte cap, you have to market it as a 3 megabit connection. The effective sustained rate. Or preferably, just outlaw data caps entirely.
That's like saying that laptops must be marketed with their "effective sustained" battery life. The MacBook Pro can't be marketed as having 10 hours of wireless web, but rather just 2 hours of "effective sustained" performance. Regardless of whether it's a sensible measure for your typical user, who is actually likely to get the 10 hours because they're using their laptop for office and web browsing.
Residential ISPs are selling burst connections, which is an entirely sensible thing to do in a world where the top use case for a home internet connection is sucking down a 5MB Facebook page as quickly as possible then going idle. It's not like they're hiding the data cap in fine print: https://dataplan.xfinity.com.
Note that outlawing data caps doesn't address the phenomenon of companies with market power being able to charge monopoly prices. It just redistributes the burden of that upcharge away from heavy data users to ordinary users. E.g. consider if you outlawed Apple's practice of charging $100, $200, etc., more for $15-20 worth of flash memory on an iPhone. Do you think Apple would just lower those prices and leave it at that? No, they'd raise prices on the base model. All you'd do is hurt the users who keep everything in the cloud in order to favor those who store a lot of data locally. (Unsurprisingly, price discrimination tactics that benefit ordinary users at the expense of power users are unpopular among power users.)
But the baseline measure may not be useful to the customer. In the laptop example: much of the advance in "web browsing battery life" in the last several years has come from lower idle power usage, or being able to reach idle more quickly. An Ivy Bridge Core i5 laptop and a Kaby Lake Core i5 laptop may actually get similar "full load" battery life, but the Kaby Lake model might last twice as long for web browsing or general office use. Reporting the "full load" battery life will actually mislead the consumer about the typical battery life of the two laptops.
Likewise, if you sell a gigabit cable connection with a 1 terabyte cap as a "3 mbps" connection, then you're misleading the customer about whether that's fast enough to do things like download an iTunes movie in a reasonable time, use a VPN into their work network, etc. Ironically, you might do more consumer harm by effectively upselling consumers to buy "faster" connections than they need.
2. These corporations sell "burst" connections because they want to screw users out of their hard earned money. They could just as easily cap the peak bandwidth, throttle past a certain amount of usage, or shut down the connection if they were operating in good faith. They want to send $200 surprise bills to poor people.
3. People are watching YouTube/Netflix, etc not just browsing Facebook. It's not 2004. Many millions are also doing 24/7 home monitoring, software updates, nightly backups, etc. Bandwidth caps restrict US technical competitiveness and make life suck for citizens.
4. There is no need to keep charging more and more money. The costs of fiber optics, switching, routing, etc have all dropped dramatically even as our usage increases. We're being charged more than anyone in the world already. These corporations have also received hundreds of billions in government subsidies.
5. A 3 megabit connection is not "broadband" and that should trigger some other law. They should be forced to sell 10 megabit sustained (3 TB/mo) connections at a minimum perhaps. The minimum could go up every year based on average usage.
Any country with a functioning democratic government would ensure all citizens have access to low cost unlimited broadband. Look around the world, they mostly do.
Your plan would mean that I end up paying more money while those that need the sort of sustained speeds you're talking about pay less.
Why do you want to be forced to pay for something I don't want or need?
A 10 gigabit switch a few years ago used to cost $1,000,000 and now it costs $10,000. CPUs, optics, everything is just printed plastic and metal boxes. There's nothing fundamentally expensive about any of it.
Engineers and scientists can drive the price of bandwidth down to nearly zero. Comcast and AT&T simply want to extract as much money as they possibly can.
None of it would be possible if they didn't bribe politicians and run massive disinformation campaigns against the population that convinces them that bandwidth is rare and expensive.
People are defending the idea that we should be happy to have the scraps Comcast and AT&T see fit to give us. We should be demanding to exercise the full power and freedom that the technology allows.
Saying you don't use it so you don't care is just shorted sighted and selfish. The poor kid in Chicago educating himself by watching YouTube videos is going to end up getting his parents water shut off with $200 overage fees.
2. Network technology almost certainly improves by 25%+ per year in terms of price-to-performance. Generational upgrades will make it uneven but that's not an unrealistic number to hit on average.
3. The "last mile" is the hard part but sticking with cable (DOCSIS) technology is not a requirement. There's no reason we couldn't have a massive federal Fiber To The Home program that cuts through current regulations.
The one-time costs are undeniable but they can be easily and predictably amortized over a long period of time. The theory that networking technology can't keep up is technically baseless.
If I were running a multi-billion ISP, I'd use Juniper and Cisco gear too.
> 2. Network technology almost certainly improves by 25%+ per year in terms of price-to-performance. Generational upgrades will make it uneven but that's not an unrealistic number to hit on average.
It took 12 years to go from 10 gigabit ethernet to 100 gigabit (about 21% per year over the decade), but I'd be very surprised if the price was the same (i.e. that 100 gig switches cost the same now as 10 gig switches cost back then). Carriers and data centers moved directly from 1G to 10G. But it seems like with 100G, there is significant pricing back pressure creating the need for intermediate speeds: http://www.mellanox.com/blog/2016/03/25-is-the-new-10-50-is-....
But let's say a 100G router today costs the same as a 10G router cost 12 years ago. 21% per year is just about what you need to keep up with projected demand growth over the next five years: https://www.cisco.com/c/dam/m/en_us/solutions/service-provid.... Even if 100% of your cost was scalable at the same level, you're talking about keeping up with demand for the same money, not less money. Of course, 100% of your costs aren't scalable. The last mile DOCSIS part of your network is not scaling 20% per year.
> 3. The "last mile" is the hard part but sticking with cable (DOCSIS) technology is not a requirement.
Verizon tried that. It's barely breaking even on FiOS even though the average FiOS customer brings in more than $100/month in revenue.
> There's no reason we couldn't have a massive federal Fiber To The Home program that cuts through current regulations.
I can think of way better uses for hundreds of billions of federal dollars than subsidizing Netflix's business model: http://www.afr.com/technology/web/nbn/australias-highspeed-i.... For the forseeable future, fiber is a luxury product, and to the extent we want government-funded broadband to address "Internet access as a human right" issues, we should do it using cost-effective 5G wireless technology.
> The one-time costs are undeniable but they can be easily and predictably amortized over a long period of time. The theory that networking technology can't keep up is technically baseless.
It's not that networking can't keep up. It's that upgrading the existing cable network to keep up with demand growth is expensive, and replacing everything with fiber is even more expensive.
For every poor kid in Chicago that happens to be using a highly above average amount of total bandwidth per month there will be 50 poor kids in Chicago with more typical usage patterns that your plan would hurt.
I don't think it's an accident that so many people believe this myth. It sounds logical enough but it's completely false.
Fact #1 You have to replace technology because it fails over time.
Fact #2 The cost of networking technology is decreasing every year.
Fact #3. The performance of networking technology increasing every year.
Fact #4 The costs of running fiber to premises is a (mostly) one-time expense.
There's no reason that users couldn't have more bandwidth for less money every year. Technology is magical like this.
Providers have a finite amount of bandwidth in their trunk connections. More bursty users can use a single trunk than saturating users. Therefore saturating users are more expensive.
It's just like CPU. If you're bursting up to capacity but on average only using, say, 10% of available CPU then Amazon can pack more of your workloads on a single physical machine compared to a different user that is using 100% of CPU all the time. The fact the CPUs are always getting cheaper and faster doesn't negate this point.
https://www.cloudyn.com/blog/analyzing-aws-ec2-price-drops-o...
This is exactly what would be happening if Comcast and AT&T had competition or smart regulation.
Go idle? Since when?
Steam auto updates my games, Windows downloads multi-gigabyte updates in the background. My Android phone syncs files in the background to cloud storage, and on a nightly basis finds a good dozen or so apps to auto-update. For many people, Alexa now streams music in the background whenever they are home. Multiplayer online games nowdays have voice chat, increasingly richer and more populated worlds, and support video streaming of one's exploits.
Even that 5MB Facebook page is actually an endlessly scrolling multimedia fest, that starts streaming videos as they come into view.
Heck Hacker News is one of the few paginated sites left!
Snapchat has 30 minutes. Reddit has some ungodly number, but is much less multimedia heavy.
And that is average, I know multiple people who average ~2hrs a day on Facebook.
To be fair, I stream movies using the Netflix website, 3 movies, so approximately 18GB to video.
The rest? Who knows. Youtube is certainly part of it, though I rarely watch videos full screen so whatever 480p takes up.
Then there are mobile apps like Snapchat which, for heavy users, can eat up an easy 5GB or more per month.
I had the Chrome bandwidth monitor open while researching that number. The 3 or 4 articles I read to figure out the average Snapchat data use ended up themselves using 36MB of data! (and this is with an ad blocker running...)
Figure it took me 3 minutes (not really) to do the research. that is 12MB per minute of data consumption just browsing the internet for references!
My overall point here is that data usage has exploded, and not just for consumption of streaming media.
Worse, the cable companies have had actual burst traffic shaping abilities for at least a decade. They could sell you 100Mbit, and run it at 600Mbit when the line was idle for a few seconds. That would make people pretty happy at long as they were getting the advertised minimum...
In fact if someone were to actually put real regulation down, I would expect that the plans were sold with a minimum guaranteed/sustained bandwidth (up and down, because I can't tell you how many advertisements I've seen that never mention the connection is asymmetric) and an "up to" burst metric.
It would be pretty simple, buy a (50/20 plan, that is up to 500/200 during off peak). What it would do though is guarantee that the company isn't selling 1000000x oversubscribed plans because they haven't upgraded the infrastructure in 20 years.
Because the medium is shared with hundreds of other houses? Because regardless of the amount of bandwidth on the medium, users are sharing a limited number of ports on upstream routers? Because the ISP wants to price discriminate, and make more money from people who want faster speeds?
> In fact if someone were to actually put real regulation down, I would expect that the plans were sold with a minimum guaranteed/sustained bandwidth (up and down, because I can't tell you how many advertisements I've seen that never mention the connection is asymmetric) and an "up to" burst metric.
Regulation has to have some sort of economic point. What's the economic point of preventing ISPs from practicing price discrimination, which is generally beneficial to less intensive users?
> What it would do though is guarantee that the company isn't selling 1000000x oversubscribed plans because they haven't upgraded the infrastructure in 20 years.
The factual premise of this is false. There's been massive upgrades over the past decade to support newer DOCSIS versions: http://www.lightreading.com/cable/ccap-next-gen-nets/cable-n.... Your desktop CPU is maybe 2-3x as fast today than it was in 2008 (quad core Nehalem to quad core Coffee Lake). Less than 2x for single-threaded workloads. Your internet connection is probably 20-50x faster, at least in a large-ish metro area.
#1: Price discrimination, because you know you have to reserve >75% of the bandwidth to assure that the guy with the 100mbit plan isn't being pushed out by the dozen or so users with the 20mbit plan, on a docsis 3.0 network with 16 channels bonded... Really, ever wonder why the cable industry won't actually tell you the line utilization on your node? What they have been really successful at is selling people service and modems that barely give you line quality information, knowing what your modem is actually capped at, or what current activity on the line is privileged information available to only the select few.
And finally, the article you linked mentioned that the amps haven't been upgraded in 12-15 years... you may be in a market getting a docsis 3.1 upgrade, but in many markets docsis 3.0 is barely rolled out, and it rolled out at speeds achievable with docsis 2, a standard that is 17 years old. Never mind the insane garbage peddled by the DSL providers in many markets. So I maybe I exaggerated a bit with 20 years, 15..
We're just asking for a content-agnostic utility. Not "as much as you can eat". Consider if electricity was "unlimited". "Unlimited" is part of what got us into this mess in the first place. Delivering lots of data to each of many subscribers is just not that easy and cheap. At least with explicit caps companies can compete by offering higher caps. Instead of mysterious limitations on "unlimited".
There's no technical or financial reason why every major American city shouldn't have unlimited gigabit for $50/mo per household. The only expensive part is running conduit and it's a one-time cost.
It's disappointing how many people buy into the talking points of these shitty corporations as they work hard to fleece and permanently damage an entire country.
Electricity is actually a lot like Internet. Sending electrons down a wire isn't as cheap as sending flashes of light down optical fibers, but in both cases variable costs are vastly lower than the fixed costs.[1] In both cases, there is a total system capacity you can achieve for a given fixed cost. And in both cases, usage is such that what you're really trying to manage is the amount of fixed capacity in relation to usage peaks. It's fundamentally about managing congestion rather than delivering units of a product.
> every major American city shouldn't have unlimited gigabit for $50/mo per household
Other than the fact that it's not economically possible? Chattanooga charges $60 for 100 megabit and $70 for gigabit, and its just about breaking even. And Chattanooga is a very favorable city for fiber buildout: power lines are overhead, not buried (the same reason Google built in places like Kansas City, not San Francisco), and its a southern city with relatively low cost of labor. New York, D.C., etc., are all going to be more expensive. Nor is $50 for gigabit common in western countries for fiber, at least in places where fiber isn't subsidized. Gigabit fiber in the U.K. is $70-100/month. In Sweden and Denmark it's $120+/month (and typically have multi-thousand up-front install costs).
[1] The typical American pays 12 cents per kilowatt hour for electricity. The fuel cost for a fossil fuel plant is about 2.5 cents per kilowatt hour, while for nuclear it's less than 1 cent. Almost all of the cost of electricity, just like with broadband, is operations and maintenance of the infrastructure, and fixed cost recovery.
I think any advertisement that doesn't include this basic fact is misleading to the point of being a lie.
AKA, most cell phone plans this year, can't even download as much in a month as my 56k modem could in the late 1990's.
In the end, the big scam here is that if anything netflix/etc traffic is probably less expensive per packet than most other services because netflix is more than happy to scatter their little caching boxes everywhere, so the ISP's backhaul/peering arrangement has less traffic...
ISPs are still allowed to discriminate traffic under this bill. They are allowed to block and throttle content such as illegal streaming downloads, some torrents, spam, and other traffic that degrades network performance. For example, this bill does not prevent Comcast, At&t, Verizon et al. from blocking port 25 as they have been for the past decade. The bill also does not preclude zero-rating.
This bill is a good thing yes, but let's not get ahead of ourselves and call this net neutrality.
Are you really that suprised they block it? It's an exceedingly effective anti-spam measure.
Blocking Port 25 to specific providers unless they pay up would be a problem, but blocking port 25 globally (except, say, to their own SMTP servers) is perfectly okay.
Theoretically, true "net neutrality" (in quotes because it's not really the same thing that the law that got struck down protected) would require that ISPs not discriminate what types of traffic a customer can use their pipe for, any more than a bank is allowed to discriminate what you use your checking account for.
Under this true-Scotsman net neutrality, ISPs should be required to let people host web servers, or mail servers (if they're not open relays), or whatever-else servers, on their residential Internet. If that causes uplink saturation, then price uplink bandwidth in your plans to match your costs! (Which basically would eliminate the difference between residential and business Internet plans anyway.)
Also, the port 25 ban directly affects legitimate customers. At least for Verizon, port 25 is blocked for residential customers, but it's not blocked for commercial customers. Over the past decade, many residential customers have called to enable port 25 so they can run their own mail server, and I've read Verizon refuses to unblock the port. Verizon's "solution" for them is to upgrade to their commercial package where it's not blocked.
I don't see how you can read that from the linked wikipedia page. I assume you're conflating blocking ports with blocking/slowing protocols.
> Please direct me to a source explaining why malicious network traffic is exempt from net neutrality.
It's not an exhaustive policy document of every possible reason why network access might be terminated/restricted. eg no discussion of terminating/filtering BGP peering announcements from a misconfigured device.
Few people, if any, have ever said that network management activities are would be prohibited under NN. If someone is actively disrupting the network, or is otherwise acting maliciously, then clearly as a network operator you're going to be permitted to block that traffic.
> Over the past decade, many residential customers have called to enable port 25 so they can run their own mail server
Sure, but even if they did permit it - the chances you will be able to run a mail relay from an IP that's clearly in the same network block as a bunch of other residential networks is near zero.
Most major mail providers will outright block you for existing in that range, and most others will block you for being unable to demonstrate control over the IP (by setting Reverse-DNS).
Here's how net neutrality is defined.
"Net neutrality is the principle that Internet service providers must treat all data on the Internet the same, and not discriminate or charge differently by user, content, website, platform, application, type of attached equipment, or method of communication."
Blocking all TCP/IP packets with first two bytes set to 25, and not blocking TCP/IP packets with the first two bytes set to 24, is a clear violation of the Net Neutrality principle of no discrimination based on application. If I send a TCP/IP packet with source port 24, my ISP does not block it. However, if I change just one bit in that packet, my ISP refuses to send it.
An ISP blocking an entire class of applications from the internet (in this case, SMTP servers) is most certainly a net neutrality violation.
> If someone is actively disrupting the network, or is otherwise acting maliciously, then clearly as a network operator you're going to be permitted to block that traffic.
Where specifically in the definition of net neutrality does it permit ISPs to block such traffic? Blocking such traffic is a violation of the net neutrality principle that "Internet service providers must treat all data on the Internet the same".
> Most major mail providers will outright block you for existing in [a residential IP block]
ISPs have blocked residential customers from using SMTP for over a decade now, therefore it's no surprise mail providers have little incentive to support residential SMTP servers. They've been forced out of the market.
The alternative, that ISPs should not route unlawful traffic, would require ISPs to search packets and block the ones they find unlawful. That is censorship.
It's also extremely unhelpful to bring it up in the context of NN.
I did notice you tried earlier to change the definition of net neutrality by claiming it "doesn't prevent network operators from stopping malicious activity on their network", when really there is no exception for this. Like netflix tried to do, you can't decide what is and what isn't net neutrality just to fit your viewpoint.
I fully understand why you're trying to avoid this topic by describing at as "extremely unhelpful" because I know it's a sore point many people try to avoid discussing. But I have no qualms pointing out flaws in other people's arguments :).
All the Net Neutrality legislation proposals I've seen contain explicit wording to the effect of "lawful network traffic". If malicious traffic violates the CFAA or CAN-SPAM, they would be in their right to block it.
1. The usual argument is that if an ISP can zero rate service X, that makes it harder for competitors of X to compete, and so we need to ban zero rating to encourage competition.
This argument fails because a last mile ISP like Comcast can easily side step any internet neutrality rules in offering X by making a deal with the X provider to serve the service from Comcast's own servers. Then it is not being provided over internet and so is out of scope for net neutrality.
They could even go farther, and not even use IP at all to deliver it from Comcast to the home. They run IP on top of lower level protocols, and they could use one of those lower level protocols to deliver to an app running on the set top box, which could then serve on the LAN using normal IP protocols (which would be out of scope for net neutrality because they are just running on a LAN in this case).
If we want to stop this for competition reasons, we'll need to use antitrust, not net neutrality. Antitrust cannot be evaded by simply changing the protocols.
2. We accept zero rating in every other or nearly every other common carrier service. In telecommunications. for example, I've never seen any objections to zero rated telephone service, whether paid for by the phone company (such as free calls to government offices), or paid for by the company called (commercial toll free numbers). In transportation, we don't get our feathers ruffled when a bus or cab company offers free rides (possibly paid for by a sponsor). What is different about internet?
Generalizing this, in general we allow nearly every provider of goods and services to give you MORE than you have paid for. As long as they deliver the service that you paid for, if they want to give you a gift of more service to promote a particular offering of their own, or to promote (possibly for pay) some sponsor, we allow it. Regulation is concerned with making sure they don't under deliver, not preventing over delivery.
It is hard to make a case to the typical consumer that there is something fundamentally different between, say, Safeway giving a discount on Pepsi, and an ISP giving a discount on bandwidth used for Spotify.
3. Suppose I'm a company selling something that uses a lot of bandwidth, such as a game that requires a large download. I'd like to pay major ISPs to zero rate it. If I cannot directly do that, but the ISP sells prepaid bandwidth cards that you use by entering a code from the card, I could simply buy prepaid bandwidth this way, and when someone buys my game I could give them a bandwidth code for their ISP to cover the costs of the download. Are we going to ban prepaid bandwidth to stop this?
If we want to get net neutrality back, and have it STAY back instead of just flipping every time we change which party is in control we need to keep it to just involving things ON the network. That's basically throttling, blocking, and paid prioritization. Billing is OFF the network. I think it should be reasonably possible to get bipartisan support for those.
If provider X provides a very limited (or no) quota for regular internet traffic, and provides access to sponsored zero-rated services - then this inhibits competition, and (depending on how agressively it's implmented) can inhibit free speech.
This has been seen seen in some countries such as India where facebook.com was zero-rated traffic, and data prices for regular internet access were extremely high.
It's also seen today in Australia where Apple Music, Spotify and Google Play Music services are zero-rated traffic on certain mobile (cellular) plans - many of these plans also have zero-rated access to various popular social media platforms too (facebook, instagram, etc) - and data quotas are generally fairly limited (2-10GB/month is typical for mid-range plans). A competitor to these services is going to have a very hard time entering the market, since their customers will all have the data usage come out of their regular quota.
Continuing with Australia - Most ISPs provide zero-rated access to Netflix (or provide 'unlimited' quota). Telstra does not zero-rate Netflix traffic, but they do zero-rate Foxtel and their own movie/tv services. At the same time, having rather low data quotas.
So those who're stuck with Telstra (eg because they live in an apartment/development that is a Telstra-exclusive estate, or just because they're on a 24-36 month contract, having been unaware of the problems) - are unable to use Netflix.
Either way, there's no additional charges on the customer or provider side.
[1] https://www.t-mobile.com/offer/binge-on-streaming-video.html
This goes back to the distinction between the real problem, the lack of competition in many markets compared to net neutrality which mitigates situations where there's a lack of competition, but it's not necessarily a win all across the board. What's needed is to separate last mile access, from eyeball network service, from carrier network service, from content network service, from actual content. This would ensure competition across the board, and prevent excessive power accumulated by such a small number of large corporations as is currently the case. However, the nearly religious reverence to capitalism in the US would make such a solution nearly impossible to legislate.
There's very little direct bandwidth to Singapore - there's only one cable going west (SEA-ME-WE-3), it's old, and constantly has breaks.
Unless you live in Western Australia, your traffic (even to Singapore) doesn't go via that - it goes via the US over SXC or Endeavor to the east, or to the north to Guam (PIPE-1) or Japan (AJC)
> It's reasonable for them to zero-rate that traffic as they're effectively passing their cost savings on to their customers, as opposed to trying to double dip by charging content networks for access to their subscribers. This may violate more pure definitions of net neutrality, but should be desirable behaviour as it's pro consumer.
I agree, however I've had (heated) arguments with people on this aspect. Not everyone agrees, and thinks that all traffic should be zero-rated if any traffic is going to be.
This needs to be explained. If the zero-rated service is being provided as part of the same subscription, I don't see why it matter where the servers are located.
I know I'm in the minority, but I've never seen zero rating as part of NN.
What I worry about is people playing games where services are slowed or blocked.
What best practices do others do when still hosting SMTP at home?
To accurately reflect the scope of the problem that should read "some Americans some of the time," no?
I doubt California businesses and residents restrict their internet traffic to stay within the state.
Verizon, as usual, sent me my "you're at your last 10% before we throttle you" notice just as I went over my limit, and at the exact same time, DNS stopped working for netflix.com.
Everything else worked just fine, but none of the netflix.com properties would resolve.
I had lazily been using Verizon's DNS for my hotspot. Changing to 8.8.8.8/8.8.4.4 immediately brought it back online.
I can't prove it wasn't just a glitch, but it was pretty damn suspicious anyhow, and exactly the sort of underhanded behavior I'd expect from today's telecoms.
What specific configuration did you change from 8.8.8.8/8.8.4.4 - i would like to know for future reference if it does happen.
But I haven't!
The server was fe80::188e:1b74:da0:42e9%3.
Here's some of my terminal output from when it happened. I was on Netflix when the notification came in and Netflix stopped loading less than a minute later.
...huh, that's weird. I doubt this is a Netflix issue. What does Google say? ...ok. And Verizon? ...Okay. What about YouTube, Reddit, HN, Google, Slashdot, my servers? Can I reach those? (Yes.) Okay, let's try reconfiguring this connection's DNS then to use Google public DNS. ...Neat.(edited to shorten it up.)
I did it for much the same reason. Had to move anyway, don't expect to spend as much time at home over the next year.
20GB/mo coupled with the website obesity crisis (http://idlewords.com/talks/website_obesity.htm) coupled with it being nice to have a dumb show on in the background while doing dishes (about the extent of my Netflix habit) makes things uncomfortable for the last few days of my billing cycle. I'm currently browsing the web at a stunning, 21st-century America speed of 600kbps -- aka, I actually get better speed by going to a crowded Starbucks.
Great write up from boing boing expalning how and why they can do this: https://boingboing.net/2016/09/22/i-have-found-a-secret-tunn...
Before YouTube and Netflix, video broadcasting meant that you send data once, but a per view service sends a copy each time.
I'm all for net neutrality, but to be honest I think that some use cases might add a little pressure to the whole network when you compare them to more classic uses (HTML, images). I don't think that Snapchat, YouTube and Netflix can really be sustainable models if they keep growing and growing forever.
At some point, transmitting video online, on a large scale, at a good resolution, at peak hours, is going to be expensive.
Netflix is popular because the real cost is on the ISP, and there never are any guarantee of quality of service, it's not up to Netflix.
The problem is when they'll deliver 250 GB of $PARTNER_SERVICE but not 250 GB of Netflix or YouTube, and none of the 1 or 2 ISPs in your area will give you a reasonable deal on your GBs from Netflix. (And to be clear, Netflix is not charging them anything.)
(The other problem is when they say "unlimited" but of course it's never really unlimited.)
It just seems like underlying the whole NN debate is the very large uptake in the average user's data usage over the last 10 years that is just sort of ignored when it comes to how do such upgrades get paid for and before such upgrades take place, who or what gets slowed down in the case of congestion?
ISPs want to pay so little industry is still spending R&D money on how to make that fucking copper cable go all the way to Shannon capacity. G.fast is now doing over 100 Mbps on 500m copper.
There's also the issue mentioned in the sibling comment of the fact that internet use is naturally over-sold to consumers - the cost to guarantee a speed "at any time" is way higher than offering that speed "most of the time". I think that's also fair: if everyone in a city starts streaming a few different 4k shows at the same time one evening, it's fair if service slows down, as long as the advertised speed really is available the rest of the time. SLA (service-level agreement) is just another orthogonal aspect of the service that is sold (and businesses are generally willing to pay a lot more for consistent speed than a consumer, and this is also fair).
It just comes down to unfair treatment of competing services/uses by a monopoly or duopoly ISP. Bytes should be bytes. Like the electrons you get from your electricity provider.
In the even everyone shows up at once, sort of like in your scenario, infrastructure will of course have to be built out, but prices will rise as a result.
This is really frustrating, since Netflix is usually not served over the internet, but directly from the ISP's head end. Netflix usage largely has no major impact on the world wide internet.
https://openconnect.netflix.com/en/
https://gizmodo.com/this-box-can-hold-an-entire-netflix-1592...
ISPs don't even have to serve you internet access for Netflix videos. Netflix already hand delivers the data to your town and then pays cash to Verizon and Comcast to serve their already-on-local-network videos to you. That's all on top of the money you are already paying the ISPs to receive that exact same traffic.
All ISPs really do for Netflix is take traffic from a private local field office to your home. Frustratingly, they are often too cheap and lazy to even manage that.
Is that neutral?
The ISPs are not keeping up their part of the bargain.
Netflix using bandwidth doesn't harm anything. IPSs are free to offer users all the bandwidth they have available. Running out of bandwidth is not harmful to anyone.
Putting pressure on the network is the only way infrastructure gets updated. We haven't reached the point where every user has the highest reasonable amount of bandwidth. I posit that we never will. ISPs want you to believe we have so you will believe people are using the internet "wrong" when the real problem is that ISPs don't want to invest in infrastructure. I won't hesitate to call them out in their BS.
A prime reason broadband stopped getting faster after around ADSL2 is that nobody knew what to do with all that bandwidth, and that was 15 years ago. In fact you can watch today's Netflix 4k on 2003's 24 Mbps ADSL2+ connection, if you're reasonably close to the DSLAM.
ISP backbone bandwith cost has continued dropping on that curve even if home broadband speeds have stagnated due to telcos investing all their chips into 4G coverage. (A silver lining is that 4G is symmetric so you can reasonably run video calls and home servers off it..)
This will force bandwidth hogs to self-regulate. If you have the ability to purchase 4k/8k TVs, then you should be responsible for the associated streaming costs and not have others subsidize it.
Also, NN does nothing to foster competition - the real problem. All we have are a bunch of local monopolies/duopolies.
The ISP needs to size for peak usage, and they over-subscribe because that's the only way to be profitable. This is why your speed will often be slower at night than during the day. It also means when you pay for data transferred at non-peak times, you're not actually costing anything since the bandwidth was sitting there unused anyway.
Fixed that for you
Regulate the use of the term "internet service" to include network neutrality principles. If the provider violates these principles, they cannot call themselves an "internet service provider".
They can all themselves "network provider" or "mobile data provider" but they can't use the term "internet" anymore.