(1) We provide speeds "up to 100mbps" (actual max throughput in any 24h period being 10-60mbps)
(2) We provide mean speeds of 200mbps, easily beating the definition (mean is measured over time, not from the customer's perspective, so if most people get on at the same time then most browsing sessions see throughput of 10mbps or less due to poorly handled congestion)
(3) We provide mean speeds of 100mbps (upload is 0.01mbps)
(4) We provide mean speeds of 100mbps, over 99% of the time (every 3 months an 8h maintenance block drops speeds to 10kbps, rendering the network totally unusable that workday)
(5) Mean throughput is 100mbps, mean latency is 2000ms (totally unusuable for real-time _anything_)
(6) Mean throughput is 200mbps, mean latency is 20ms, every 5min a few packets will have up to 10s of latency (still totally unusable for most real-time purposes, but naive metrics make it sound good)
(7) You're allowed "unlimited" "100mbps" (but if you use that for more than a contiguous day, or a few long evenings in a row, then 100mbps is redefined to mean 100kbps)
And on and on and on. More useful metrics might measure actual consumer impact:
(a) Can you rely on this every workday to shuffle documents back and forth?
(b) Can you rely on this every workday to have low-resolution video calls?
(c) Can you usually play a quasi-realtime game for 30min without lag killing your character?
(d) During peak hours, can you reasonably expect to be able to stream buffered 720p video?
A raw 100mbps floor doesn't guarantee your network suffices for any of those use cases.
Now all of those speeds feel like they're dial-up modems.
- 10mbps is the bare minimum to not have frequent major problems with all kinds of things but it still just a minimum (symmetric, if with reasonable latency and reliable and always actually 10mbps, i.e. not a contract which says 10mbps but is often much less or supper low latency)
- 100(down)/~50(up)mbps is "good enough" for quite many (but then bluray dvds are still make some profite here, soo...; also again actual speeds not "marketed speeds")
- 500+mbps down are what you really want if you play a lot of AAA games or similar, else download times for updates and similar become quite in-convenient and can infer with you playing with other people (e.g. due to you waiting for a unexped download). Some jobs need speeds like that, too.
- there is little point in more then 1gbps down or 250mbps up (where I live; outside some niche cases). It's partially a chicken and egg problem: Without good internet you can't sell services which need/want such speeds but without such services why should anyone spend (larger amounts of) money to buy such internet and if very few buy it where available why should a teleco make it available in other areas
Through that's private usage, not b2b, e.g. remote controlled agriculture machines.
What's hurting me is the advertised 12Mbps upload, or in reality, 9-10Mbps upload.
Customers are so focused on synthetic download speeds, but meanwhile the upload rug is being pulled from them. Get a few devices on your network with iCloud Photos enabled and you're hosed, because traffic can't ACK/get out.
I have 1 GBit up+down at home, I cabled the whole house with CAT7 cables and set up the best access points money can buy for mobile devices. Everything else is wired. It's an amazing experience and internet everywhere sucks compared to home.
When I lived in Romania this costed me €10, now that I am back in the Netherlands it's a bit more pricey at €52, but it's 100% worth it.
Gotta love the late stage greed in this world
Upload and latency increasingly matter more and more and is essential to many work from home endeavors. Whether you are uploading content for work or play, or have a family video conferencing, or a host of other things.
Perhaps the more interesting part of the doc is the future proposal for 1gb/500mb. I understand the engineering tradeoffs and reasoning for the asymmetrical bandwidth on coax/docsis but perhaps it's time to rethink some of that. That would be a good thing.
No.
This lacks all historical and practical understanding of why upload speeds are limited.
Back in the 70s/80s when coax was being planned and rolled out for cable, they needed to figure out what frequencies they would use. Cable providers knew they needs a ton in one direction and only a tiny bit for things like buying pay per view in the other.
This has been fine traditionally because even up to Netflix and streaming we haven’t needed the upload at residential. No one complained. That may start to change a little.
As far as I am aware, all DOCSIS (cable) providers in the USA that are using cable-tv coax are limited to a max of 35Mbps upto DOCSIS3.1.
I am writing this on 1250/35 right now. The local tiers for me are 100/15, 400/25, 1000/35, 1250/35.
It isn’t a “cash grab” at all. If you need more upload you need a new line that isn’t divided so asymmetrically, which means fiber. And once you fiber, there is zero reason to go go symmetric, as you have two lines with the same equipment on either side, to a point where the ISP will charge differently if you cut into their bandwidth too much.
Over 35? You are remembering wrong.
Everything else you said could be true, yet not the part about a cash grab. Anything worse than $10 for 100 Mbps down and 100 Mbps up from a large ISP is a cash grab now that fiber exists [1], DOCSIS 3.1 can offer up to 1 Gbps (to say nothing of DOCSIS 4) [2], and large ISPs like Comcast keep committing subsidy fraud [3].
I wrote more about pricing in a different comment [4].
[1] https://www.eff.org/deeplinks/2019/10/why-fiber-vastly-super...
[2] https://en.wikipedia.org/wiki/DOCSIS#Comparison
[3] Pick any one of these links. https://www.techdirt.com/2020/10/06/mississippi-says-att-too... https://www.techdirt.com/2023/02/15/report-shows-comcast-con... https://www.techdirt.com/2023/02/16/verizon-t-mobile-oversta...
I thought you were serious for a sec. You should open an ISP, it would be wildly successful with your background knowledge.
https://www.xfinity.com/networkmanagement
I do wish this was required to be easily accessible from the marketing pages. Usually it takes some digging to find it. But at least they do have to make these numbers public.
>Latency is typically measured in milliseconds, and generally has no significant impact on typical everyday Internet usage.
I took that from their networkmanagement page. Their statements seem to conflict with the consensus of the HN crowd, but I could be wrong.
They are god awful expensive, but I feel the quality of their product is excellent. And I'm not even a Comcast employee or shill.
Note they're not saying 100% of customers got their rated speeds, but the average customer got over 100% of their rated speeds. So, say the 50th percentile customer pays for 100M, they actually got 102M, boom, that metric is met. However easily 25%+ of customers didn't get their rated speeds during that peak period.
Some people ranting on an internet forum isn't high quality data of overall network metrics.
For a lot of residential customers, once latency is good enough it doesn't really matter. Once you get down below like 30 or so milliseconds, a lot of consumer applications just won't be any different. I could have 80ms latency to HN and my experience of the site would be roughly the same. Not saying all residential applications don't care about latency <30ms, but generally speaking the vast majority have very little difference at the moment. Watching a streaming movie, loading a web page, doom scrolling social media, etc are all going to be the exact same experience at 2ms or 10ms or 30ms. This is changing though, and I do prefer having such low latency with my residential fiber connection.
Most people that need any less are likely gaming, heck even I'd like less for gaming but 30ms is more then enough to be able to play FPS games well.
I appreciate there is more to latency than just gaming, but I hear more people wanting less ping for those reasons than anything else.
The monopoly situation I read about in the various areas of the US really sounds like the FCC or another regulator needs to break up region exclusive deals, or something. It feels like Australia, of all places might actually leap frog the US once fibre is more widely available, and that is pretty bad since we have had some of the worst average internet speeds in the developed world at very high prices. Regulatory capture is a plague , and this FCC “proposal” feels like a crumb to delay what sounds like much needed broader disruption of monopolies.
Did you also have to pay for fibre to be installed?
* https://en.wikipedia.org/wiki/Open-access_network
Ideally OSI Layer 1/2 would be a neutral third-party that doesn't do an IP but rather only makes sure physical plant is ubiquitous in good working order:
* https://en.wikipedia.org/wiki/Municipal_broadband
It was possible to wire up countries and continents with electricity and telephone in the early 1900s, so I don't see why is not technically possible to do the same with fibre in the early 2000s.
I paid for 200mbps for a long time and only typically saw 20-30mbps. When I moved, Comcast gave me a deal for 400mbps, I’m now seeing 50-60mbps typically. Never bothered looking at 1g plans because why pay the price to get just 100mbps in reality.
This is all well and good, but they keep increasing the price, and refuse to offer lower-tier service. I probably wouldn't want 12Mbps anymore, but 20 or 25 would suffice. I wish I could pay a fraction of what I'm paying for 75Mbps and get a lower tier of service. It should suffice that x% of customers are offered "broadband" for a reasonable price, even if some choose a lower speed.
Ultimately, this comes down to competition. There are no other wired options where I live (in the heart of Silicon Valley). I have considered using a cellular hotspot, but it's not competitive in terms of pricing.
I'm all for prodding companies to move to faster speeds, but there should still be room for consumer choice.
There certainly should be consumer choice, but in your case another problem is that the amount you pay Comcast for 75 Mbps download now should be no more than what you paid for 12 Mbps before. As I wrote in a different comment [1]:
> It's not reasonable for big ISPs like Comcast to offer me 300 megabits download 15 megabits upload for $70 a month (might've been $90, but assume $70) while EPB of Chattanooga [1] offers 1 gigabit symmetrical for $67.99 a month. What speed any individual actually needs doesn't have to come into the picture. In matters of consumer protection, the principle of the thing matters just as much as actual consumer needs.
(Ignore the [1] inside the quoted paragraph above. I'm leaving it in for Ctrl-F purposes.)
Imagine it costs $2000 per household to connect a fiber network. With 70% of households signing up, that produces a certain cost. Now imagine there was a second fiber network that also cost $2000 per household that splits the customer base. By definition, this is going to be more expensive.
Capitalism created this problem. It won't solve it. This is capitalism working as intended: doing the absolute least possible for the highest possible price while lobbying to exclude competition (eg by making municipal broadband illegal).
Expecting markets to solve this problem is neoliberal brain rot. Let's imagine that you do get 2 or even 3 ISPs in an area similar to Comcat, Spectrum, AT&T or Verizon. Now what? They just collude to keep speeds low and prices high [1].
Build municipal broadband. Failing that, regulate ISPs like utilities. And this would include unbundling TV services.
[1]: https://www.legitreviews.com/american-isps-collude-to-keep-b...
I'm just not a big fan of their "security" features. Other than that, I haven't had many issues here in Oregon.
This reads like a joke.
Sure. They could apply the line sharing rules from the 1996 Telecom Act. Afaik, that's still law, but when courts said the FCC couldn't apply it to telephone companies and not cable companies, the FCC opted to declare the market was competitive and end mandatory line sharing.
Help from congress would be great, though. There was a lot of shenanigans under the line sharing regime, like ILECs pricing retail internet lower than wholesale prices. After a transition period, the line owners shouldn't offer retail service at all, make them restructure it into actually separate companies providing the internet service vs the lines.
The problem is that most of the costs of providing broadband are fixed costs. To run a coaxial cable from their headend to your house, and to provide employees and crews to repair it and install it all costs the same whether you buy 10 megabits or 2 gigabits
The variable costs--things like backbone bandwidth, paying for transit, doing node splits, etc are a smaller part of the overall cost than you might expect
This lower end of the market, though, is being taken up by Fixed Wireless (T-Mobile and Verizon primarily) for example you can get Verizon home wireless internet for as low as $25. My mom has it, and it's decent 50 megabit service for a low price
The TL;DR is that the cost of cell towers is mostly the same irrespective of bandwidth (e.g., hiring a technician to break-fix as parts fail from weather conditions), so carriers figured out they could "rent" their leftover bandwidth.
Most of the "cheap" carriers are simply slinging together MVNOs from all the carriers (e.g., Boost Mobile, MetroPCS). It creates a weird situation where you get plenty of signal on off-peak hours and (often) the middle of nowhere, but at the expense of having no signal in larger metro areas at peak hours.
But is it still required though? Isn't this how Sonic uses AT&T lines? Perhaps it only applies to DSL, which is basically obsolete now?
Internet has yet to be regulated that way and fiber lines aren't considered the same as those copper lines.
Also, fiber to the home is done using most often a PON system purchased from a specific vendor and is a time shared medium. You might only get 1Gbps service but you also aren't getting 100% of the fiber to yourself. The home side device will filter information from your neighbors for instance.
There's always ways to do this of course but, it wouldn't be as straightforward as patch the fiber from your home in the same way that they did with copper lines.
It did make sense at the time of the law: carriers kept clustering around the major metro areas and wouldn't expand to the boonies. Nowadays, that law has outlived its purpose.
Not sure if this is completely the case when it comes to older networks that might need upgrades because lots of folk are sharing the same coax line. In an HFC[1] network, bandwidth is shared on the coax branches, and a variable number of units can be sharing the same fiber link, which can serve a neighborhood, block, or even unit. And, so upgrades can make a network more "fiber-rich."
Where I live roads are private (part of the lot) with reciprocal easement to all other lot owners to travel on the easement. It's not a toll and it's not 'public' or 'government'. No worries about government making crazy rules on their roads, when we own it.
Closer to topic, I want to get neighbors and everyone onboard with installing fiber in road and leasing to ISPs. Way too many hassles, so we remain dependent on Comcast.
A real upgrade for me would involve running a cable through a beaver meadow. There is an electric line and crews do go in there to limb up trees and they are planning to replace a pole, if there was some rational coordination of services and a commitment to universal service they’d get the fiber in when they do a maintenance cycle on the power line.
We used to go out every day to take apart unwanted dams near our driveway. If you go at it in a disciplined way, taking out the sticks from downstream, then digging out the mud, and never letting the water level to get high enough that they can easily float materials to the dam site, it's not that hard. It was awful, however, if they dammed 3/4 of the culvert pipe in which case you had to work inside the pipe with low clearance and (I think) high CO2 levels and no really good place to put the debris. At least they would never reuse materials that they used once.
I don't think anyone has done that in the last year, I certainly haven't.
I hope this leads to wired ISP price reductions, or at least slower price growth. I happen to live in a VZW dead zone, so unfortunately it won't work for me...
I had the opposite experience. I was a very happy Comcast customer for a year or two (their curfew-for-kids-computers feature on their router was great, and the mobile app integration was very handy), but the 5G Verizon offering was cheaper and faster (where I live). I have never felt so bad canceling my account, as everything about the experience was great, but VZ had double (or more) the upload speed.
Unlike most folks my experience with fiber has been dreadful but at least I'm not paying out the nose for a connection with less upload speed than phone.
I even made my own router because I didn't like the consumer options available: https://danangell.com/blog/posts/10gbps-router-for-fun/
Unfortunately, I’ve since moved to another part of the country, with much less enlightened ISP options.
The actual reason in almost all cases: non-techies are clueless about their slow and shitty WIFI. Picking empty or less busy channels, using 5GHz, using gasp cables, all black magic and never done.
Sometimes the upselling actually helps because the better plan comes with a maybe-better AP...
The guy strongly insisted I not do it, but then when I persisted I had to sign two waivers acknowledging that my wifi will probably be shit and the basement is a terrible spot.
I was just using their router/modem as a pass through, but man, they really don't want people putting routers in their basement. I can totally understand why too.
That'll teach me to stay consistent.
ok.
Next month honest
I'm still embarrassed about the time I called my ISP and complained that the new Xmbs package they just installed is capped at only Ymbps only to realize I had run the speed test over wifi (back in the wireless-g,maybe n days)...
Maybe people are actually using their 1 Gbps bandwidth if they have multiple 4K TVs and do serious gaming? I don't know, I doubt it.
I tell everyone to look at their actual cellular data usage too. I think all middle class Americans just buy unlimited plans now, but I do just fine with 500 MB per month for $5.
I shouldn't have to look at my actual bandwidth usage. As I wrote in a different comment [1]:
> You are unknowingly accepting being ripped off. It's not reasonable for big ISPs like Comcast to offer me 300 megabits download 15 megabits upload for $70 a month (might've been $90, but assume $70) while EPB of Chattanooga [1] offers 1 gigabit symmetrical for $67.99 a month. What speed any individual actually needs doesn't have to come into the picture. In matters of consumer protection, the principle of the thing matters just as much as actual consumer needs.
> Today's internet technology (particularly optical fiber [2], paired with hardware implementing DOCSIS 3.1 or 4 [3]) is fully capable of providing 1 gigabit symmetrical for "the majority of people", even in rural areas. Moreover, in the long term, transitioning to fiber would be less expensive to the big ISPs like Comcast [4], but Comcast keeps raising prices on broadband over decades-old copper wires and committing subsidy fraud [5]. Don't let big ISPs define "good enough" to be much lower than technology and the price of the technology allow.
(The [] citations within my quotes refer to links in my other comment. I'm leaving them in for Ctrl-F purposes.)
My issue in the past with cable gigabit service was the relatively tiny upstream bandwidth of 20-50Mbps. This made for some challenges a few years ago when suddenly everybody in the house was on Zoom all day long.
500 for $5 would cost me about $50 - unlimited makes perfect sense.
I'm admittedly a minimal user of my phone. I average about 10 unlocks per day. Especially when I'm on-the-go, it's just calls and text messages, some driving directions, and maybe looking up a phone number or business hours. If I use it to listen to an audiobook or podcast, I download those ahead of time on WiFi. Using only 200 MB per month is normal.
Continuously streaming video? Not at all, might watch something here or there but most of my device usage is text or just reading. I barely use the phone or texting or social media.
I think we forget just how data-hungry modern apps are. A 1080p "Full-HD" YouTube video will consume 4GB/hour of watch time. Even going down to 480p is still 1GB/hour.
True. I really try to avoid using the browser for anything unless I'm on WiFi because a single page (really all the other crap that gets loaded besides the actual content that I'm looking for, even with uBlock Origin) might eat dozens of MB. For a while I used the NPR One app to stream podcasts, but I gave it up when it started using way, way more data than streaming audio should ever require.
That's about 5 - 10 times the amount of unlocks I do. Not what I'd consider minimal, at all. :D
This plan will definitely come to fruition just like all the other ones since the 90s
For reference, I live in a small town of ~6,000 people, and we're getting fiber laid. For the longest time we only had 25mbps, and then a new provider came in with coaxial and offered 200mbps (with a data cap), and now a third is laying fiber.
Our average income is $34k/year, so it's not because we have more money than the opulent Bay Area.
So it probably does fall into rich vs poor in the same way that renters may be less likely to be rich.
The lowest plan that mi ISP offers is 100Mbps and is at 21 USD.
[1] https://www.movistar.com.pe/hogar/internet/solo-internet (Link in Spanish)
> Broadband penetration as of June 2017: 23.5 broadband connections for every 100 people. > Distribution of broadband connections by type, as reported by Ancom, is as follows 94% FTTx (FTTH/FTTB/FTTC/FTTN) internet access connections, 4.8% Coaxial cable, 0.2% other.
Now guess the country. Answer in ROT13 at [1]. Hint: it ain't a first world country. Another hint: Latine loquitur. Oh and it ain't a recent development either. They've been at it like forever.
[1] uggcf://ra.jvxvcrqvn.bet/jvxv/Vagrearg_va_Ebznavn
Consolidation is the cause - US had 2-se as many listed companies 30 yeats ago.
Now US capital is buying up independant businesses across easterm Europe and developing countries, killing competition and the local economy.
The big upfront cost is trenching. The rest is potatoes. If you think the "big heroes" of an ISP are the router wizards, oh boy. It's the permit people. A ransomware gang has nothing on the uppity council of a town of 4000 people.
My sense is there is some cost savings at these lower tiers, but more importantly it would prevent ISPs from jacking prices up while using higher speeds as the rationale ("Look, we just tripled your speed, and it only costs double!"). ISPs would be less likely to double prices without any service improvement. But since many people don't care about the service improvement, it's an illusory benefit for them.
At least they're giving more speed instead of just doubling the cost with no increase in service
Because you keep paying for it. It's not like you're going to not pay for internet, and you probably don't have many alternative options.
The cable company now needs to both increase speeds _and_ lower prices to even keep their existing customers. This is a good place to be, as a customer.
If pricing stayed even close to fixed my gigabit internet would cost well over $10,000mo, considering back in the day I was paying SBC for less than a meg for more than I'm paying for a gigabit today. And that's before adjusting for inflation, those 90s dollars are worth way more than 2023 dollars.
Additionally there should be a over subscriber limit defined based on used capacity based on something like a 95th percentile calculation.
It doesn't help if you have a gigabit internet connection when it's unusable because of the lack of backend capacity.
Although if you had an open infrastructure the free market every one loves so much would kill bad service.
I highly doubt the argument itself was relevant whatsoever.
Maybe I need to find a different bank?
I have the opposite experience with Mastercard. Of the dozen or so disputes across 5 cards I had over the years I essentially lost them all. Visa is somewhere in the middle. They tend to side with customer but also will investigate thoroughly.
(Also, a bit of a side-point: The chargeback notices always had a VERY short window during which you were allowed to contest them. In most cases we had about 5 days to respond, but sometimes the notice wasn't delivered until after the respond-by date.)
$500 for items stolen off my porch. So they WERE delivered, but I didn't receive the item.
Over $1,000 for a hotel I booked and they wouldn't let me reduce my 5 day reservation by 1 day (the first day because I realized I couldn't get there on the expected day). I went back and forth between the booking website and the hotel 3 times each with them both telling me only the other could change the dates. So, I charged back the whole stay and found a different hotel for my updated timeline.
Nor should you quite frankly. Kids buying stupid things is not the same as fraud.
"You inaccurately described the merchandise or services at the time of purchase." (billed as high-speed, it wasn't)
2003: “Does everyone really need 5Mbps?”
1994: “Does everyone really need 28.8Kbps?”
1967: “Does ARPAnet really need 2.4Kbps?”
- Gaming.
- Other large downloads/uploads for WFH purposes.
- High-bitrate 4k UHD streaming.
So for a majority of people, the ISPs have a de facto monopoly. No need to improve service or customer support.
Also need to overturn legislation at state level which forbids municipal ISPs from being created. This would help break the mini monopolies that ISPs.
Defining standards is fine. But the real issue needs to be fixed.
> The Sherman Act broadly prohibits 1) anticompetitive agreements and 2) unilateral conduct that monopolizes or attempts to monopolize the relevant market.
Broad band does not mean speed. it means it uses a lot of frequencies at the same time.
OK getting dizzy here from too much excitement, pedant time is over for this week. We are all allowed to go back to using the non-technical meaning of broadband where it means speed.
It is modulating a ~100THz carrier wave, so is clearly not base-band.
Without mandated illegality of data caps and other Comcast data and modem games this does nothing to change the reality of boots on the ground unfortunately. These speeds only really allow you to close your data cap sooner in the month for the vast majority of customers who use Internet for modern things. A fixed data cap locks in an “era” of technology so that when the era ends (HD > 4k, 4k > 8k for example) every one of the customers ends up being a heavy user. It is a scam that should be illegal.
We’ve already paid for this with multiple rural broadband bills. It’s time for change.
This is woefully (and shamefully) inadequate action by the FCC but at least we don’t have to deal with Ajit’s basically sellout behavior anymore and can at least talk about solutions beyond the current regulatory captured standards.
I commend them for moving, but I can’t condone their proposal as it’s truly and utterly inadequate and far behind global standards as well.
If you want your country to slip into uncompetitive mess, go ahead and hamper the backbone of the new Industrial Revolution by giving away all the value to your corporate friends and leaving the innovation a generation of true broadband performance could generate laying on the shop floor.
Did I mention we have already paid for broadband (ready, was supposed to be fiber rollout)? Twice? Low standards might as well be low standards FCC.
Barely. Most costs to telcos are fixed costs, and upgrading the speeds of their lower speed plans shouldn't lead to a big increase in traffic.
What I'm trying to say is that many of these devices regular people have don't even have the capability to receive 1Gbit/s streams. It's great for us nerds or people who (also) work from home and need blazing fast connections to work servers etc, but as of 2023, it's overkill for many regular people.
Datacaps should be illegal though.
Wireless is a convenience feature but it gets treated like it should act exactly like a direct wire does.
It’s also about latency.
But the thing people miss is that this is mandating projects for future infrastructure projects. Whatever we set at a baseline is what we will standardize on for the next generation. People donut understand these are leading values for project acceptability not minimum standards of throughout as they are being presented. Massive difference.
It’s not about whether you can make it work today, it’s about what kind of projects are acceptable speed for tomorrow and whether we can call it broadband when it’s not.
Just look at gaming. When it came on a CD people complained that gaming would never work on a t1 line, and 1.5 mbit was supposed to support at 25 person office (1999). Then 10 mbit was mind blowing with bonded DSL and you could support everything. Now we are having an argument about 100 and 300, but the same logic applies. What works for today isn’t the standard we should be applying to qualify infrastructure minimums for buildout (which is what this is).
I mean imagine when games at 2-3TB each (it’s coming) and drives are 50GB. It’s not a question of throughput, as you said it becomes a question of time to complete — aka productivity — and that means waiting around for stuff to finish while you download. That’s what broadband standards are working against — what we think is acceptable to be installing today — not whether you can max the link with your current mix of devices and two 4k TVs.
Good point.
I imagine the 20 Mbps upload target is useless for most of us, even those on asymmetric cable, but there's a good chance that if the ISPs were forced to provide 20 Mbps up at their lowest broadband tiers, it would raise the ladder quite a bit. For example, on Comcast I get 800 / 20, but there's also a 400 / 10 tier (and maybe a low-cost tier below that?). If everyone was getting 20 up, maybe I'll get upgraded to 40 Mbps.
1gig/500meg would be a huge step forward. I'm currently on the highest most bestest plan that my ISP (Astound) offers. And it's 1.2gig down, 50 mbps up. That's not a typo, FIFTY up. I'm a techy and even I know I don't need 1.2gig down. My house is wired with CAT6 because I dragged it around, but my main switch is still gig. So... I can't even enjoy the increased download. The only reason I stepped to that highest plan was to get more upload than the 20mbps they previously offered.
Personally I'd trade my current speeds (800 / 20) for 100 / 100 in a heartbeat. There's just ... not that much I can do with more than 100 down. That's fast enough to stream a Blu-ray disc directly off of a web server. It's fast enough that package updates on Linux take longer to install than they do to download.
Can I theoretically use more? Sure, it's nice to be able to download files or update Steam games a little faster. But 20 Mbps up is killing me. Backing up a drive to cloud server can take hours or days.
I'm in rural middle of nowhere in central Europe, at a latitude with very dense Starlink coverage and I'm almost certainly the only Starlink user in my hex - the speeds I'm getting are pretty much unaffected by network congestion.
Even in those ideal conditions upload is 10-15Mbps 95% of the time, even when download is closer to 300 than 200Mbps
https://arstechnica.com/tech-policy/2022/10/want-faster-comc...
Unfortunately, though, there are currently only three customer-owned modems approved for higher upload speeds: the CODA and CODA 56 from Hitron and the CM2050V from Netgear. The first two are from a company noone has heard of, and the last one is $350 (!) and includes pointless phone support. None of the near hundred other DOCSIS 3.1 modems on the market are supported, for no apparent reason.
What's more scummy (IMHO) is that it's actually cheaper for me to rent their modem—which includes unlimited data—than it is for me to pay the fee to remove the data cap.
Already starting the rollout in Colorado.
The lack of high quality, high bandwidth symmetric connections to the home (and ipv6) is a huge boon to the entrenched big tech companies.
I am not following this one. Why would upload bandwidth at home inhibit backing up your phone/data to a NAS at home? Or are you referring to remote restoration / access?
Wouldn't the NAS be using its download bandwidth to receive backups when your phone is remote? and your phone would be on the local network when you're at home?
But it could be possible to do this with your own device at home with sufficient upload.
It might be too late though, since if sufficient people are used to the big tech companies doing it, it might not leave enough of a market to develop sell a device to do it (although I feel like it should be doable with software and existing NAS vendors like Synology).
So cable at least will always be a bit asymmetric. A lot of fiber deployments (GPON) are also but the bandwidth limits are much higher so it does not really matter.
Coax has to manage the single (or few) headend transmitter vs many CPE transmitters, you get better multiplexing to the CPE than from the CPE; and while the newest equipment may be able to handle CPE uploads across most of the spectrum, older equipment had a narrower range of frequencies available to upload with. If there's spare frequencies and downstream capacity isn't using it, then yeah... if not, there's trade offs.
For DSL, everybody has a dedicated pair to the DSLAM/remote terminal whatever, but crosstalk between pairs impacts the signals and again, the DSLAM has more ability to mitigate that towards the CPE than the CPE can when sending upstream. Anyway, VDSL/VDSL2 are specified with asymetric up and down bandwidth, as was ADSL. Of course, DSL ISPs tend to cap sync rates based on profiles, rather than running the lines to their maximum capacity, so a lot of lines do have more upload capacity than is allowed to be used, and to the extent that's the case, it would just take a profile update to get more upload for many users.
I wish this were true. I'm in Seattle and Xfinity will only sell me 5 Mbps up, even at 480 down, unless I pay out the ass for a symmetric gigabit connection I don't need. Very much looking forward to the FCC forcing at least 20.
I had fiber installed this summer and until they finally sent a staff engineer out it was going down 2-5 times a day. I work from home so I had to tether to my phone for Teams meetings to make sure I'd stay connected. It was actually completely fine and I only get 2 bars of LTE at home.
The worst part is that the much larger condo right next to us has Google Fiber :(
Thankfully there are alternative ISPs here that sell symmetrical gigabit, but that’s definitely not the case everywhere.
Instead we have monopolies that do what all monopolies do in a capitalist organization of the economy: they rent-seek, spend their time and money on lobbying to legislatively exclude competition (eg [2]), do the bare minimum and take public money to roll out broadband but then just not do it (eg [3].
The Internet is the US is objectively awful for most people by the standard of other developed economies [4].
It's ridiculous that people still defend this system.
[1]: https://qz.com/1996234/the-best-broadband-in-the-us-is-in-ch...
[2]: https://arstechnica.com/tech-policy/2014/02/isp-lobby-has-al...
[3]: https://arstechnica.com/information-technology/2015/06/veriz...
[4]: https://broadbandnow.com/report/2018-fcc-international-data-...
But you don't have to run a municipal ISP like a commercial ISP. If pricing goes low enough, just make it an amenity, like parks and sidewalks, and fund it that way.
And then you can choose a totally different architecture than a commercial ISP could operate: no network admission barriers, mesh networks, etc.
I find that latency rather than throughput is what determines my perceived quality of an Internet connection. But am curious how others think.
Upload speed probably makes more sense for more use cases though. I used to have symmetric 1Gbps fiber and never bothered to setup QOS as my upload was never saturated.
I moved and am stuck with "1Gbps" Comcast. Which really means 25Mbps upload. I had to setup qdiscs on my gateway and split my network into tiers to get acceptable upload speeds and latency for the workstations in my home. I maybe have more uploads than 'normal' people, as I have automated backups that store data off-site, but normal people have "backups" in the form of cloud storage I think.
Uploading videos (to YouTube, for example) is painfully slow. I'm simulating living in Australia when I upload a video.
Then again I'm not a networking guy. Someone who knows more about this stuff than I do can give you more comprehensive advice and talk about stuff like buffer bloat, etc.
Most people could get away with less, but sooner or later 1Gbps will go from "excessive" to "good" to "acceptable" to "slow".
Or there's three people in your house all on video calls.
IMHO we need to address the "last meter" experience before mandated gigabit internet speeds mean anything.
Furthermore, 4K video is < 25mbps each per stream, usually.
Sure, if I download a torrent, it is much faster. But is not the kind of upgrade that I experienced from ADSL (7Mbit) to VDSL. Since most of the time I use the PC under Wi-Fi anyway, that doesn't go over 600Mbit/s near the AP, but really not over 100 in the location where I usually have my PC.
What I've seen instead is a much more stable connection. Giving that the network is entirely fiber and passive there weren't (so far) any interruption of service in roughly one year, while with VDSL there where time to time that the connection did not work, in one occasion for nearly a week. Also since it was copper lines in case of bad weather, or crosstalk with other users, the performance did vary a lot.
Most complaints that I've seen in various neighborhoods are from people that were not getting anywhere near their full speed. Usually the cause is their wifi router.
Sounds like instantaneous update is too, but that's a relatively infrequent use of your connection.
even stuff like backing up computers, or running security cameras that you can view remotely or that record to the cloud (encrypted) are really hampered by upload speed. there's a lot of cool stuff one can do if only one had the speed to do it with.
My current internet connection uses DOCSIS 3.1, but I can't recall any resource contention even if I have several computers trying to upload simultaneously. Not a big fan of asymmetric speeds, though, but apparently DOCSIS 4.0 (which is being finalized this year) gets us closer to symmetrical speeds.
I believe it's false advertising when ISPs claim their price is $X per month if it's only for 1/240th of a month. I understand and support the idea of overprovisioning, but a 240x ratio is insanity.
I think somewhere in fine print they claim they get to round up to the nearest KB or MB every minute or some asinine made-up scale that only benefits them.
So cell service, internet, trains, and cheap flights are all easier to provide there.
Whereas here in the much more sparsely-populated North America, those things are all pricier, and it’s easier to accommodate big trucks, big highways, giant Costco’s, etc.
Location, location, location.
I do agree the over the cap costs imposed by residential ISPs are many, many multiples of their true cost, but they're mostly there to discourage use not be a reasonable price. If you want business level usage hop on their business networking where pricing models are more designed for more average load.
My home internet has no data caps.
Assuming that for most consumers the end price averages to be the same, the benefit of metered is it puts an incentive on the data provider to make their speeds as high as possible. Higher speeds will result in more bandwidth usage (i.e. Youtube/Netflix will default to a higher data rate), which will result in more profit for the company. The desires of the customer and the desires of the provider are the same - faster and more stable internet leads to better customer experience and more profit.
One of the issues I see with fixed-speed-unlimited-data options is that providers are incentivized to oversell the fixed speed bandwidth, leading to "up to 100mbps" type connections where it's a theoretical max, not the average. Under this model the desires of the customer are misaligned with the desires of the provider. The provider wants the customer to use as little data as possible - the less they use the more they can sell plans on the same infrastructure. This can be addressed with rules and regulations, but I prefer a situation where I know both the customer and provider have aligned desires.
> Network congestion arguments in this debate operate on an unsubstantiated and uncontextualized assumption of scarcity — there is only so much bandwidth, and a few people are going to use it all. Yes, some network congestion arguments have prevailed with wireless broadband because of the atmospherical and technical limitations of the medium. However, congestion is not as consequential for fixed broadband. Broadband providers reason that instead of limiting these internet “super-users” (which would be discriminatory and litigated accordingly), they need to implement data caps for all consumers, which they claim is fair and unbiased. However, scarcity is not a reality for broadband providers, even with super-users. Internet functions by “statistical multiplexing” meaning that bandwidth is dynamically allocated and reused without a limit “[u]nlike other utilities such as water, electricity, gas or oil.”
> This means that no super-user is consuming bandwidth at the expense of other users; you’re not going to receive less internet because your neighbor runs a Twitch channel, for example. Rather, super-users are simply using more internet more often, and the “super-user discount” (the fact that broadband providers aren’t able to charge them, specifically, for this “above normal” usage) angers providers more than anything else. This brings us to the real reason broadband providers apply data caps to consumers: money. Applying data caps to all of us, therefore, enables broadband providers to pat themselves on the back for devising clever “overage” fees.
As time goes on the same will become true for the majority of wireless usage, if it isn't true already.
Sending twice the data at a time doesn't cost the ISP twice the money. But large ISPs with regional monopolies continue overcharging you [2] while committing subsidy fraud [3] and subsidy obstruction [4] at every opportunity (sometimes with help from corrupt legislators [5]).
[1] https://publicknowledge.org/no-cap-the-truth-about-data-caps...
[2] https://news.ycombinator.com/item?id=38105873
[3] Pick any one of these links. https://www.techdirt.com/2020/10/06/mississippi-says-att-too... https://www.techdirt.com/2023/02/15/report-shows-comcast-con... https://www.techdirt.com/2023/02/16/verizon-t-mobile-oversta...
[4] Pick any one of these links. https://communitynets.org/content/monopoly-providers-mire-nt... https://www.techdirt.com/2023/02/15/report-shows-comcast-con... https://www.techdirt.com/2015/04/16/alec-threatens-to-sue-cr...
[5] Ditto. https://www.techdirt.com/2021/02/19/new-bill-tries-to-ban-co... https://www.techdirt.com/2022/04/11/illinois-missouri-and-ne...
IP transit costs per GB are a fraction of a hundredth of a penny.
It barely makes any sense to bill consumers per GB on any terrestrial network.
In Northern Europe a typical 4G SIM with a monthly data quota is $9 per month. An unlimited 4G SIM with all you can eat data is $18. How many price points do you need in between?
Heck, international 4G/5G data roaming fees are capped at $2/GB and by 2027 the cap will be $1/GB.
But caps are different and don't work the same.
I'd obviously rather have a useful 500/200 than a 100/100 or 50/50, even though the latter is symmetrical. (I think my service now is around 400/20. I'm way happier with that than I would be with 50/50.)
Please stop spreading such misinformation as its an unnecessary expense. It costs zero dollars for upload stop cucking it because you don't understand its use.
If "good definition of minimum broadband speed", then "applies to the majority of people". The converse is not true. It is not the case that if "applies to the majority of people", then "good definition of minimum broadband speed."
1. Whether a speed applies to a majority of people is necessary but not sufficient.
2. You are unknowingly accepting being ripped off. It's not reasonable for big ISPs like Comcast to offer me 300 megabits download 15 megabits upload for $70 a month (might've been $90, but assume $70) while EPB of Chattanooga [1] offers 1 gigabit symmetrical for $67.99 a month. What speed any individual actually needs doesn't have to come into the picture. In matters of consumer protection, the principle of the thing matters just as much as actual consumer needs.
Today's internet technology (particularly optical fiber [2], paired with hardware implementing DOCSIS 3.1 or 4 [3]) is fully capable of providing 1 gigabit symmetrical for "the majority of people", even in rural areas. Moreover, in the long term, transitioning to fiber would be less expensive to the big ISPs like Comcast [4], but Comcast keeps raising prices on broadband over decades-old copper wires and committing subsidy fraud [5]. Don't let big ISPs define "good enough" to be much lower than technology and the price of the technology allow.
Going by the Chattanooga metric, 100 mbps symmetrical should cost no more than $10, and anything lower is not a good definition of minimum broadband speed for the present (never mind the future).
[1] https://www.eff.org/deeplinks/2019/10/why-fiber-vastly-super...
[2] https://en.wikipedia.org/wiki/DOCSIS#Versions
[3] https://epb.com/fi-speed-internet/gig/
[4] https://www.eff.org/deeplinks/2020/06/why-slow-networks-real...
[5] Pick any one of these links. https://www.techdirt.com/2020/10/06/mississippi-says-att-too... https://www.techdirt.com/2023/02/15/report-shows-comcast-con... https://www.techdirt.com/2023/02/16/verizon-t-mobile-oversta...
--------Everything below this line is bonus.--------
The following excerpt is from an article about data caps [6] but applies equally well to the state of broadband pricing in general.
> Network congestion arguments in this debate operate on an unsubstantiated and uncontextualized assumption of scarcity — there is only so much bandwidth, and a few people are going to use it all. Yes, some network congestion arguments have prevailed with wireless broadband because of the atmospherical and technical limitations of the medium. However, congestion is not as consequential for fixed broadband. Broadband providers reason that instead of limiting these internet “super-users” (which would be discriminatory and litigated accordingly), they need to implement data caps for all consumers, which they claim is fair and unbiased. However, scarcity is not a reality for broadband providers, even with super-users. Internet functions by “statistical multiplexing” meaning that bandwidth is dynamically allocated and reused without a limit “[u]nlike other utilities such as water, electricity, gas or oil.”
[6] https://publicknowledge.org/no-cap-the-truth-about-data-caps...
>It is not the case that if "applies to the majority of people", then "good definition of minimum broadband speed."
I don't see why this wouldn't be the case. This isn't a physics problem with a right/wrong answer, it's a balance of the wants/needs of corporations vs. individuals. You're only talking about half of the equation to begin with, so forming logical statements on it is incomplete to begin with.
And yeah, I could go for a gigabit minimum.
Thankfully synchronous gigabit fiber came this year.
asymmetrical is as antiquated as dial-up. there's no reason for it in 2023.
They have increased the channel allocation for upstream in recent years and DOCSIS 4.0 allows symmetric connections, but the demand is for downstream, and the fewer channels they allocate to upstream the bigger number they can advertise.
It really was just the odd teenager torrenting here and there.
There are technical reasons for offering asymmetrical speeds when it comes to coax connections. There's only so many channels on the wire, a lot are still dedicated to television, people mostly only care about download, so they prioritize it. PONs don't carry television and are just fundamentally different in design, so its easier for them to support symmetrical bandwidth.
ISPs that evolved from cable companies might still be using cable hubs with a common collision domain, but only a relatively small number of subscribers will be riding the same copper — it’s just cheaper these days to convert the signal to fiber as early in the signal path as possible. Plastic wires are cheaper than metal wires, and you need fewer of them (and so fewer switches.)
If I drink less water than you, am I subsidizing your water usage. I'd say so!
Not the hill you want to die on, trust me.
While I agree that grandma doesn't need 1Gbps, neither do I and neither do most people. So why mandate such a high speed for the label anyway? It seems like a way to disqualify things like Starlink that will have difficulty providing extreme speeds to large numbers of users at once.
The only time I'll use such a speed is when downloading some data - an ISO file or something similar. Streaming video does not require such (gbps) speeds even for multiple streams in the same house.
I cannot believe how anyone with a straight face can claim that the infrastructure and maintenance required to serve "n times 1MB" is exactly the same as the costs for "n times 1TB". This is so obviously not the case that I genuinely don't even know how to explain it.
You can serve your thousand subscribers on a dingy Pentium 1 if they're using 1MB, because it adds up to just 1,000MB. You wouldn't be able to serve just a single customer with the same hardware. Never mind the cabling etc.
They over-subscribe because they know not every customer is going to be using up that full 2gbps 24/7.
Now, you can argue that an ISP SHOULD provision that way, with the expectation that their customers are going to be using the full connection 24/7, but that would raise the cost per customer a LOT, and I don't think people want to be paying for an ISP that provisions like that. It would be a lot more expensive per customer, and it would end up with a lot of the bandwidth going unused most of the time.
If the average bandwidth used by customers goes up, it will require the ISP to pay for and maintain more circuits and cost more money. We can argue there are better, fairer, ways to limit usage, but it is simply not true that every customer downloading a lot more doesn't cost the ISP more money in the long run.
Say your ISP gives you 100mbit/s = 1.08TB/day ≈ 30TB/mo. On gigabit that’d be 300TB. While you do have some heavy torrenters they are outliers.
Now I assume everything but TV/movie streaming is a rounding error for average Joe. Netflix says 1-7GB/h depending on quality. Average user watches ~3.2h/day (wtf is wrong with people!) but that’s ~100-700 GB/mo. Now that’s between 0.033%-2.3% of downstream bandwidth.
Of course, people generally watch TV at the same time of day, so it gets more complicated to provision resources. But there’s also no question that pooling bandwidth (over-provisioning) makes sense to reduce costs. The question is more about how much congestion is acceptable, and I wouldn’t trust shitty monopolistic companies to behave. But if you can handle eg Super Bowl or a World Cup final without degradation you’re probably good the rest of the year?
No, but bandwidth is. Particularly wireless where you can't just lay down another cable or fiber.
Which is why they already charge for bandwidth
They either have infrastructure that can handle 1gbps per user, and then they have bandwidth available, and data caps are nonsense... or they don't have the infrastructure, and are banking on their service being idle 90% of the time for most users, and this is why after 5 oclock, my internet goes to crap when everyone is getting home and starting to use the internet. And now suddenly Im drastically overpaying for non-broadband internet because my ISP underprovisioned way too much.
This goes back to the point that everyone is making that ISPs shouldn't be able to underprovision to the point that it degrades performance during high usage times.
My ISP saw $800 million in profits on $2 billion in revenue in 2019. That's an AMAZING profit margin. They can afford to not underprovision, and when they don't under provision, then no, bandwidth isn't a limited resource.
That's only true in a very abstract sense. If everybody went from using 1MB/day to 1TB/day there would be massive congestion issues and costs would increase as ISPs rush to install higher-bandwidth equipment. Put a different way: It is cheaper to construct a network where all subscribers consume only 1MB/day than a network where all subscribers consume 1TB/day, because the former can be done with much lower end equipment.
Go to a grocery store in Florida before a hurricane and try to buy bread.
If you want to know how much more expensive the connection would be if you expect to use the full bandwidth 24/7, just look at the cost you pay for transit in a datacenter. It is multiples of the cost home consumers pay at a per-bit level.
Most users are best served by being told the speeds they will usually get if they stay within average usage patterns, because that is what most people do.
But that's true of almost every industry. Your bank advertises that you can withdraw your money at any time, but if everybody withdrew their money simultaneously there would be issues. A store advertises next-day shipping, but if everybody ordered simultaneously there would be issues. A house might have 100A power service, but if every house started burning 100A simultaneously there would be issues.
Put another way, what you're proposing is dramatically slower speeds for most users. ISPs are profitable, but they're not that profitable. Actual average usage on most lines is probably low single digit percentages, if not sub-1%. I just checked my line - my average use is about 200kB/s (for a total of about 500GB per month), or about 1% of my speed cap.
It's not without its faults though. Fuel usage isn't directly related to cost of road maintenance, it's just a very rough approximation. Fuel usage has mattered less and less over the past couple of decades with hybrids and EVs – though this is addressed in some places by imposing an extra EV tax (since EV drivers would pay no fuel tax but would still cause wear on the roads).
[^1]: https://en.wikipedia.org/wiki/Fuel_taxes_in_the_United_State...
The additional problem with this is that road wear scales a lot faster than fuel usage. https://en.wikipedia.org/wiki/Fourth_power_law
WFH workers likely pay more per mile travelled in taxes than someone who super-commutes.
Similar thing probably goes for internet access.
We'll set up electronic tolls on your cul-de-sac and every light and stop sign, or toll you by mile driven using an odometer beaconing out to the tax authorities per drive and registered to your tax ID number (US) or equivalent. And then cap you on the number of miles you're allowed to drive per year on all cars unless you pay an additional fee.
Older cars must be retrofit at the driver's expense.
Also the auditing and anti-fraud infrastructure must be instituted for all of this.
---
Or we can just do flat taxes and chill.
The streets in my neighborhood cost way less per mile for maintenance than giant highways with massive bridges and interchanges. It makes sense they'd cost more to use.
Showing the real cost of these things to consumers might make them change their minds about riding the train. These days a lot of people don't even think about the cost of driving their car across town, but gosh that metro day pass is $5? How expensive!
Nah, metro day pass should cost as much as howevermany miles the buyer traveled. Otherwise people who buy the passes for just one stop would be subsidizing the costs for people who travel from one end of a line to the other.
It'd be pretty shit to meter one and not the other.
This partly why gas tax is even a thing.
Right now we have this weird sunk cost and lack of obvious incremental charges to driving, that encourages driving when there are other better options for some trips.
It’s sad seeing people on a forum called hacker news having no idea how networking infrastructure works.
To think otherwise is to falsely believe they support unlimited throughput. Furthermore the bandwidth itself literally is a function of the transmission material and infrastructure, which also isn’t free…
Bandwidth caps not only have a weak fiscal basis but also clash with net neutrality tenets, go against what made the internet successful in the first place.
Which is algebraically identical to a monthly charge and data cap with overage charge. The main issue is the overage charge is too high, it should be like 1 cent per GB (Comcast is charging 20x that).
The average US residential customer uses a bit over 500GB/mo in data. You're doing 2x that in 1/15th the time.
https://www.allconnect.com/blog/report-internet-use-over-hal...
I don't think the actual cap really matters if the per-GB and base pricing reflects the true costs. If it's low it means heavy users pay more, if it's high, light users pay more.
I had to live under an ISP that charged something on the order of $10+/GB after a certain amount and after they were forced to offer unlimited internet it was an actual breath of fresh air just to be able to use the internet and not worry about literally everything in the house, or if I could update my OS, or play this game, or watch this Netflix show, etc.
i'm honestly curious. why not a standard margin on your 'per second' allowances (100/100Mbps, etc) or other 'addon' services?
why should someone with a ring doorbell (or a homekit secure video whatever) be punished over someone who doesn't?
Aren't we effectively arguing CapEx versus OpEx?
Sure, an ISP's OpEx is tiny. But CapEx is huge, especially when you've got an old neighborhood that's all wired up with copper and you gotta replace it all with fiber going into the house.
This is actually relevant to my ISP day job, we have a pipe to a well known backbone provider, and dual pipes to a statewide provider. We ran on the dual pipes for a long time before pushing our ISP traffic over the single backbone provider. These days, if that main ISP pipe goes down, it fails over to the dual pipes to keep everything running. But now that we are offering packages higher than 25/3 for DSL, terrestrial wireless, and cable, the dual pipes don't have enough bandwidth to keep up.
If we offer gigabit service over fiber (and we do) just a couple of customers actively using their whole pipe is an enormous chunk of the network compared to the 40 or so customers that same gigabit of bandwidth would serve on previous 25/3 packages. We don't have data caps or even contracts (small town benefits :P ) but there is a lot more to it for smaller ISPs than adjusting the rate limiting and packages we allow people to use.
Of course that's not all that different from restricting bandwidth, just that you are restricting average bandwidth while allowing for some burst. Still violates the spirit of having fast bandwidth, unless the caps are quite generous.
Also - just look at LTE operators. They're also no-cap at this point (though subject to QoS at certain breakpoints, but they're typically reasonable breakpoints ime?). And that's in a SIGNIFICANTLY more capital intensive market - you have most of the concerns with terrestrial fiber, plus the joy of having to own RF spectrum and maintain towers etc.
So I think no-caps-and-required-speed requirements would just make sense and be required to keep ISPs from trying to backtrack on being forcibly dragged into the modern area like balky calves.
The cost to track and bill people for caps and let them pay overages though..
Some combination of caps, overage charges, traffic shaping, and edge CDN appliances is needed (Netflix, Akamai, etc). Off peak unlimited is also a potential strategy to drive low priority transfer to low utilization periods. Pick your poison. There is a reason most of Netflix global infra is monster CDN boxes closest to customers.
Because it would be ridiculously expensive to egress video streams directly from AWS. They've determined that it's cheaper to pay ISPs for rack space.
If you want to learn more, lurk on the NANOG (North American Network Operators' Group) mailing list. It is common to monitor AS traffic flows to understand whether you can offload with CDN appliance relationships or how to improve your network architecture (which networks to peer to directly or which internet exchanges [IXs] to meetup at). These agreements sometimes happen at a bar over beers during NANOG events. Call Verizon and ask them how much a 1Gbps circuit is. Consider why Comcast built their own national backbone.
Also, some important history to remember. Fast.com measures your speeds to Netflix servers because some ISPs wanted to charge their customers and Netflix for that transfer (or to upgrade peering points) because of the revenue destruction from cable customers fleeing for streaming (net neutrality debate).
https://archive.nanog.org/meetings/nanog45/presentations/Nor... (Peering 101, Bill Norton [Equinix cofounder], NANOG 45)
https://drpeering.net/white-papers/Video-Internet-The-Next-W... (Video Internet: The Next Wave of Massive Disruption to the U.S. Peering Ecosystem (v1.7))
https://drpeering.net/HTML_IPP/ipptoc.html ("The Internet Peering Playbook")
https://arstechnica.com/features/2008/09/peering-and-transit... ("How the ‘Net works: an introduction to peering and transit")
https://arstechnica.com/information-technology/2014/07/how-c... ("How Comcast became a powerful—and controversial—part of the Internet backbone")
https://openconnect.zendesk.com/hc/en-us/articles/3600361636...
> How much does the appliance cost my organization?
> Appliances, including replacement appliances, are provided to qualifying ISPs at no charge when used within the terms of the license agreements.
Cable at an apartment building I visit is often 300mbps, but Sunday evening it can have horrible jitter and lag and much slower speeds. It's incredibly frustrating; common sites can be basically unusable.
This also feels like it's setting a bar where most DSL systems wont have any chance of competing. In cities I can be more sympathetic (put in more pops) but in rural areas, this feels like a barrier that is extremely cost prohibitive to make happen, for not great reasons.