There's a lot of comments on this thread pointing out typical bandwidth usage of popular streaming services, games etc and claiming that there's no need for higher bandwidth connections.
There's an important point that I think people are missing, which is that the bandwidth usage of internet services is going to be carefully tuned to balance quality and stability at typical internet speeds. When a large portion of consumers have < 100Mbps down of course all services are going to use < 100Mbps of bandwidth - you don't know what kind of stuff you're missing out on because of these slow internet speeds, because nobody is making it available, because it wouldn't work.
The most obvious thing that we already know we're missing out on is video streaming quality. You've possibly already noticed that low contrast dark scenes and scenes with snow/rainfall don't look very good on Netflix/YouTube/insert streaming service here - this is because these are the kind of scenes with the most obvious sacrifice of quality from video compression. (It's also possible you haven't noticed this, because streaming services generally avoid producing content that doesn't play well with their video compression).
There is still a market for Blue rays for precisely this reason, compression does a lot of harm to the image and streaming services often run at 1/4 or less the bitrate. There is also the sound compression as well which is very heavily done on places like Netflix, presumably because their intended audience is watching on a tablet or a low end smart TV that has terrible in built speakers but if you have reasonable addon speakers and especially surround sound the average has clearly gotten worse.
The effect is very dramatic on Youtube and Twitch. Twitch has its ~6mbit/s limit and Youtube limits 1080p to around 10 mbit/s and 4k to about 18mbit/s which is well below the blue ray standard bitrates for these resolutions. You can't even override it with a better quality upload as it always compresses the Video to its standard bitrates. It shows up a lot in game footage where grass often becomes a smudge. The somewhat manipulative use of this recently has been DLSS and FSR. These two technologies effectively hide really well within Youtube's compression, while using them they introduce noticeable artifacts but once compressed by Youtube they largely disappear and appear almost identical to the original image. Soft images due to Temporal Anti Aliasing and reduced resolution also appear relatively normal on Youtube, they just hide in the impact of the compression.
These defaults for ADSL levels of performance of the internet are quite harmful to image and sound quality across a range of services and while those companies might not want to pay for more bandwidth we as customers should want better bitrates for when we play it on devices capable of showing and producing better.
You really see how lame our so-called HD is on streaming services when a lot of pixels change at once. Images of rough seas and fast-moving live sports are good examples.
And “4K” under 100 Mbps? A joke.
My parents are stuck with AT&T (which peddles glorified DSL), and Netflix barely works.
>There's an important point that I think people are missing, which is that the bandwidth usage of internet services is going to be carefully tuned to balance quality and stability at typical internet speeds.
...or the fact that there can be and often are more than one person in a single household.
This was super noticeable for a few years whenever I went back home to visit my parents. I would struggle to load basically anything on my laptop most evenings due to the 5 mpbs or so being split between 3+ devices (although the weak signal in most rooms due to the poor placement of the router that they were not interested in moving somewhere more central didn't help either). Eventually when the pandemic hit and my father started working from home, they quickly discovered that he would not be able to get his work done if my mother was streaming something at the same time, so they ended up upgrading to a more expensive plan, and the next time I visited, I was pleased to find that my issues had vanished.
i can provably determine whenever someone in my household is using youtube. We have "fixed wireless" as primary internet access right now, and the upstream is measurably between 300 and 800 kilobits. However, the "router" assumes perfect connection (which i am guessing is somewhere around 5mbit) so any sort of QOS is out the window. And it's behind CGNAT, so i'm getting a class C on a class A private network, and all gaming stuff that doesn't like NAT to begin with balks. I've worked around that (check my github or whatever), but the fact remains, without literal triple-NAT my household can't actually use the internet as it exists, even with potentially 40mbit downstream speeds.
In video many things become possible with higher bandwidth availability:
Youtube Recommended video bitrates for 60fps HDR UPLOADS:
8K: 150 to 300 Mbps,
2160p (4K): 66 to 85 Mbps,
1440p (2K): 30 Mbps,
1080p: 15 Mbps,
Some example uncompressed video bitrates:
1080p60: 3 Gbps,
2160p30: 6 Gbps,
2160p60: 12 Gbps,
2160p/8k@60: 24 Gbps and upward,
It might be hard to imagine using uncompressed video over the internet,
I know it is for me.
But all the applications we can only do locally start to become possible
over large geographic areas. The applications could be amazing relative
to what we know and understand today.
I think applications we can't imagine would become possible with large jumps
in low-cost available bandwidth.
One of the things that frustrates me is that ISPs have successfully lobbied broadband surveys to exclude "inconvenient" plans.
AT&T asked that their DSL plans were not counted towards their "average bandwidth" because they were "obsolete and no longer being marketed".
To be clear - they were still _available_, and in many cases, your only option, but they were an anchor pulling down the average, so AT&T just "didn't want them to count".
> The most obvious thing that we already know we're missing out on is video streaming quality. You've possibly already noticed…
I only pay for the lowest tier of Netflix which does not give me HD, and I have literally never noticed a quality problem watching on my TV (which I have no idea the resolution of, honestly).
Apparently I don't care or notice much about quality. I actually have no idea if i'm the majority or you are.
We do the same. Subscriptions and rentals are 1080p. 4k versions of content usually cost twice as much. A few simultaneous compressed 1080p streams easily fit in the slowest broadband service I can get.
> There's a lot of comments on this thread pointing out typical bandwidth usage of popular streaming services, games etc and claiming that there's no need for higher bandwidth connections.
You said it, not me. You said there's shills on this thread. That's the only possible motivation to argue there's no need for more bandwidth, there's another reason I came up with but I don't want it to be plagiarized, so I won't say it, they're stuck rehashing the "really sincere" focus group maneuver. It's also because actual bandwidth meaning through a wire has more rights and is better protected from tapping than wireless or radio or what is it now 5G? So the whole game is to funnel people to depending on this little walkie talkie that never turns off and which has no buttons, and is so sophisticated you can't fix it yourself, and has a baseband (that's a whole nother conspiracy theory but it's boring), and it gets GPS readings like every time you don't consent for it to, and that's why they're attacking the pipe, the wire, the cable, the fiber, they want everybody getting a cellular antenna for everything. The wire means having rights, photons it's like, everything goes. Surveillance in fact, they want us all to literally broadcast our eg 911 calls instead of calling them on a landline. They don't like landlines because they're well-litigated, have rights, martyrs died over getting their line tapped or being unable to call 911 just by pressing 911 (there's a case over that, they were meant to dial 9 first, 9911, and victims couldn't get through to emergency service and one died), 5G turns back the martyr clock forty years.
Yeah totally want to get all my needs met by a nagging machine.
Like these guys run their own ISP you think they can't log into a site and tell them both sides of the story? They've been doing it since the 1910's. Or shill robots by this point. I wonder if it counts as public speaking if you're only speaking to robots. Just trying to pass the Cussen Test here. https://news.ycombinator.com/item?id=31111806
Regarding your Cussen Test, did you write this comment using something like GPT-2?
I got curious so I checked a few of your account’s other comments. Some of them are very similar to this one [0], but others are short and concise [1]. I can think of a few other explanations for this, but since you mentioned the test…
You know what? I've been diagnosed as spectrum autism, believe what you want. Suppose this can be reverse psychology, there is nothing I can say that can't be plagiarized. I don't have a really good way of proving it digitally. At least I'm getting replies, which I could not on the initial Cussen Test post. Any reply is progress, but I think I passed the test already, wow.
With how slowly these proposals roll out, 100/20 won't be enough. As others said, there is a lack of both competition and pricing to worry about too.
In a major tech hub like San Francisco, where I live, it's possible to get 1gb/1gb in some areas for under $90, but which those are isn't consistent nor wide spread. Home cell coverage at 100/20 is spotty at best, and there is only one flat rate vendor (T-Mobile Home, no Verizon Home yet). For those outside of the Bay Area exurbs, I can only imagine the horrors of Comcast/Xfinity or only DSL people must deal with.
Looking forward, the FCC is assuming that people are just watching Netflix, Tiktok, and making angry Karen posts on Nextdoor. We're not sure where web3 will lead, and how much bandwidth will that take: Distributed file systems/data privacy? Fast crypto transactions? Metaverse/VR/AR glasses/contacts? And what about latency?
I am insanely embarrassed by the third world telecom infrastructure many Americans have to live with.
100Mbps symmetrical is just on the edge of what telcos can jam into dsl over their existing install base. The goal of the telco is to keep the FCC speed goal at copper limits for as long as possible to avoid paying the fiber they were paid to build in the 90’s.
Om Malik wrote a fantastic book entitled: “Broadbandits” which covers the $750B telecom heist (way back 30 years ago when a single billion was a lot of money for a business).
I live in Kazakhstan and am paying just a little over $10 a month for a 500 Mbit/s symmetrical FTTH link (no caps either). It's even faster and cheaper in Russia (at least it has been until now, I don't know how long it's going to be true since they're sanctioned to the moon). Both countries pretty much have (their own) monopolistic internet providers, and yet. Why do Americans tolerate this bullshit?
When comparing local production prices between different countries it is important to keep local wages in mind. The ISP employees in Kazakhstan are not getting paid US salaries.
To do some quick math, let's take the median monthly salaries of both Kazakhstan and USA, 364,000 KZT ($755) and 3,500 USD per month respectively.
That means that the $10/month 500Mbps symmetrical fiber is about 1.32% of median monthly wage in Kazakhstan. For the US, 1.32% of 3,500 USD is about $46/month.
Looking at the AT&T website, they offer 500Mbps symmetrical fiber for $65/month + taxes. So yes, the US internet is more expensive by about 50%.
However one should still ask - well, maybe AT&T provides better service? I don't know whether AT&T is a better provider than your Kazakhstan ISP, but some points to consider are:
- Can you have a connection uptime of 6 months without interruptions and it's just business as usual and not an achievement? (In Estonia I have router & connection uptime measured in years. When I lived in the US, my Xfinity connection dropped for a few minutes every week.)
- Does the ISP have excellent global peering? Poor peering can result in much higher latencies to international destinations, and also much lower bandwidth. (In Estonia I know all the budget ISPs buy (as opposed to build) their international traffic and oversell it, which means connections across oceans are at quarter of the advertised speeds at peak hours.)
- Is the last mile cabling cost eaten by the ISP, or will they demand you to pay for its building?
Sure, that was implied. All hardware and software used to build networks is imported though, and is bought for 'hard' currency (and a lot of it is needed to build networks in a country with an extreme climate and such low population density).
> Can you have a connection uptime of 6 months without interruptions
I've been using this provider since 2008 (and switched from ADSL to fiber-optics in 2013). It's been pretty great actually, I believe we had two service interruptions in all that time, roughly half-hour each. I don't remember the exact time since the last one happened some years ago.
No idea about 100% stability across many months of use since I don't host anything at home. This sounds like something you'd want if you were getting remotely operated on.
The router keeps the same IP for months, and I've never had the connection break from under me (besides two occasions mentioned above).
> Does the ISP have excellent global peering
Well, it's definitely not going to be as good as what you have in Europe. I mean, look at our geographical location, it's right in the middle of freaking nowhere. I guess it's pretty similar to what they have in NZ, only they are surrounded by water, and we by a thousand miles of steppe.
Last year, I looked at what they paid in Russia to get similar connection speeds. Prices were significantly lower in Moscow and SPb than in Siberia (roughly $5 for 1 Gbit/s versus $5 for 200 Mbit/s IIRC), which makes sense when you think of it. And we're way further down.
Both up and down speeds to US and EU datacenters are stable and close to 100% advertised.
Right now, ping to HN is at 220 ms. Going by the most direct route possible at the speed of light in vacuum, I get 100 ms to that same datacenter (us-west-2). So it's 2.2 times worse than the unreachable ideal. Don't know how much better it could be if we had direct cabling to the US since it's not going to happen for financial reasons.
We have a couple more providers (who seem to have been allowed to the market to make the main one look like less of a monopolist) that are not as popular. They definitely do oversell bandwidth. You'd be lucky to get 10% of what you had paid for, especially on weekends.
> Is the last mile cabling cost eaten by the ISP
You don't pay anything when signing up for the service.
So no, I think US customers are simply getting shafted for their money.
Mostly agree but add a point: Kazakhstan (or say all country except US) ISPs needs broad international connection to US because many services hosted in the US despite it's CDN era, so transit fee is added, while US ISPs just need national connection.
> 100Mbps symmetrical is just on the edge of what telcos can jam into dsl over their existing install base.
I'm not sure that's going to cover much of their base. I've got a pretty good line, I could probably hit 100M down, it's around 90M now, but not more than 40 or 50M up, it's capped at 16M though, cause fixed profiles :/
Elsewhere near me, there's like a chance of 10M. New technology isn't going to help that, unless it's fiber to the curb; but you might as well go the rest of the way to the premises if you can.
I’ve never had good internet in America, but my internet has been so good in east Asia. Even japan, which is supposedly worse than other countries, is worlds better than what I ever paid for in the states.
The one exception is China. I could never live there because of their shit internet due to the great firewall
Telecom act of 1996 paid them to build fiber but they set the speed limits low enough that they could satisfy it with DSL.
There’s new DSL tech that’s ~100mbps, so this is that technique again.
The fatal flaw in 1996, like today, is not pushing for gigabit, and is one of the major reasons the US is a declining global power. Faster internet access is core infrastructure for participation in the 21st century.
In fact, the text of the act doesn't mention fiber -- which part involves paying companies to build fiber? -- and one part is very explicit in not referring to a specific form of technology.
> The term `advanced telecommunications capability' is defined, without
regard to any transmission media or technology, as high-speed,
switched, broadband telecommunications capability that enables
users to originate and receive high-quality voice, data,
graphics, and video telecommunications using any technology.
In fact, for a simple sanity-check on the factualness of your statement, we need just one question: If FedGov paid companies to build fiber -- which companies, and how much did it pay each of them?
Completely based on the lens through which you are viewing it. I myself am American and have seen the quality of life for the average citizen declining in my life time. If I wasn't a part of the new middle class provided by technology, my children would have a worse life than me, combine that with the recidivism displayed by our judiciary recently and some could definitely argue we are declining. But with a user name like that, you already know all that. You might say our gdp and other metrics still show us as a giant on the world stage, I'm sure there are other metrics we can crow about, Yet it feels like little of that is trickling down to our citizens.
> In a major tech hub like San Francisco, where I live, it's possible to get 1gb/1gb in some areas for under $90
As a point of comparison, I live in a small city in Europe (80k population) and we have 10gb/10gb for $60/month. My internal network is only capable of 2.5gbit, but I can upload at that speed indefinitely. I could honestly not imagine going back to ADSL speeds.
San Francisco is an interesting case (and I say this as an SF native). The city government is so unfathomably dysfunctional in large part due to London Breed (the current mayor) bringing cronyism back as fast as she can.
Parts of the city are covered by MonkeyBrainz (microwave p2p), parts by Webpass (fiber), and most of the city is covered by Comcast (cable) or PacBell/SBC/AT&T (DSL and adjacent). Up until Sonic started rolling out 1 Gbps fiber a few years ago all I could get at my last place in the city was 3/1 Mbps ADSL or Comcast.
PacBell was always reluctant to build out RTs compared to GTE/Verizon (who has a significantly smaller presence in the Bay Area) so DSL speeds in large chunks of SF were always shit. Even when they did build out RTs the ILECs were never required to resell RT access so speeds remained shit. I'm not entirely sure why AT&T's U-Verse (FTTN, VDSL last mile) was rolled out so slowly but I expect the need for intrusive sidewalk boxes was part of the problem. Sonic got boned because SF's supervisors were/are reluctant to allow micro trenching. For a while there was some buzz about muni fiber but that never got off the ground (and Sonic certainly helped put a few nails in that coffin).
Honestly it's mostly a novelty for me, I did it for bragging rights and out of curiosity.
1Gbps to 10Gbps has some tangible benefits like downloading large games in a fraction of the time, or doing super fast backups.
However when you have 25Gbps, you have a pipe twice as large as is common in standard hardware. You usually see yourself limited by 10Gbps peering links or by endpoints with only 10Gbps NICs.
And of course 25Gbps is faster than most conventional storage, you need very high-end NVMe or RAID to keep up.
If I'd have had to pay a bit extra every month, I probably wouldn't have done it. But since it was just the price of the hardware plus installation, I thought it was worth trying.
I'd pay $200/mo for that if it didn't come with a transfer cap.
That's the problem with making bandwidth the only metric that we grade broadband connections on. You end up with "Gigabit" connections that you can only use at that speed very occasionally, because otherwise you'll hit a monthly transfer cap. (E.g. Cox's "Gigablast" service, which pairs a 1Gb connection with a transfer cap worth about 20 minutes per month at gigabit speeds.)
Serious question: what do you want that much bandwidth for? I could upgrade from 100Mb/s to 1Gb/s but I haven’t because honestly can’t think of a use case for myself.
I upgraded to 1Gbps a while back, mostly because the price was the same. Also though I don’t really need it. Then I realised I can now download AAA steam games in the time it takes to make a cup of tea. No need for a separate storage drive anymore.
I rarely need that much bandwidth, but it's convenient when available. Example: when I do an OS update. The lower latency and reliability of fiber is actually more important. DOCSIS / Cable networks are very flaky, subject to interference. There will be small periods of packet loss that literally come and go with the weather (or perhaps a neighbor messing around with their wiring.)
95% of consumers won't notice, but if I'm in the middle of uploading a docker image and my upstream bandwidth drops from 30 megabits to 3 megabits due to increased packet loss and TCP re-transmissions, I definitely will. I've had this happen before. Massive packet loss in the neighborhood. It took months to get it fixed.
Can't reliably run my personal cloud with just 100 Mbit/s. The best fiber service I had access to was Frontier out in Texas, and my location capped out at 500 / 500, which was great at the time. I could easily run my personal cloud... Plex Media Server, Nextcloud, etc.
Nowadays, with my UltraHD Blu-ray rips on the Plex server, it's not unreasonable at all to require more than a 2 Gbps connection to reliably stream those. I'm often away from home, so it's nice to just be able to log into my own personal server to watch stuff at it's native and highest available quality.
Mostly for my personal cloud... Plex Media Server, Nextcloud, etc., because any service that's offering a $300 connection will also let buy a static IP for a nominal amount.
Los Altos Hills is one of the wealthiest zip codes in the country, filled with tech millionaires and billionaires.. whatever the price may be, it’s heavily subsidized by extremely rich people who became wealthy due to tech companies wanting it in the first place. On Zillow, the cheapest house there is $4M and will likely be purchased to be torn down, the most expensive property is $40M and looks like a boutique hotel. Not really fair to compare them to the normal town.
When I moved in to my house the only available option was AT&T DSL at 18/1 mbps. Comcast quoted $210k to run cable despite everyone else on my street already having a connection. It took me a year, but I organized with a bunch of my neighbors on the two streets behind my house and we trenched to get ourselves connected.
Being in a wealthy area is no guarantee of service. I currently pay $155/mo. It's symmetric 10 gbps. My router isn't grunty enough to deliver that.
It sounds like Sonic, but Sonic's 10 Gbps offering is AFAIK symmetric. Of course Sonic's gotten a lot more cagey about specifics and I don't have any 10G equipment so it's hard for me to tell.
I can’t actually think of any where the client is the limit. However for some having faster server nodes that are hosted at peoples homes would help speed up some currencies such as nano.
> Why is broadband so shitty in the heart of Silicon Valley? Why doesn’t Google just wire the whole place up as a tech demo?
Google couldn't figure out how to get on poles. When they announced their batch of 20? or so real Google fiber cities, AT&T rolled out fiber to a good number of them before Google got city council approval to do anything. Comcast announced a loss leader 2Gbps symmetric for $300/month program [1], etc. It's not too hard to roll out fiber if you've got access to the poles and enough clout to put your fiber boxes wherever you need to. (sure, you gotta pay the property owner a little bit, but not it's not too bad)
[1] IIRC, it was always delivered as 10G metro ethernet, originally limited to 2G, but I think they increased that over time (I thought I had seen 6, but you're reporting 7). You've got to be close enough to their existing fiber that they're not losing too much on the buildout with their $1000 installation fee cap.
Google wired my city up as a tech demo. Their microtrenching approach was atrocious and resulted in frequent service interruptions and significant traffic disruption as they patched the same spots multiple times.
Eventually they decided the network was so damaged that they’d have to do a total rebuild, and that wasn’t interesting. They threw up their hands up and left, giving the city a few million to clean up the mess.
That's because Google has shitty leadership that doesn't understand playing the long game. This is why Google's only real source of income is search, and by that, I mean of course, ads.
Google kills any project that's not profitable within a year. They kill any project that doesn't become a 100x'er in a few years.
Microsoft lost $4,000,000,000 to $7,000,000,000 on their Xbox division until the Xbox 360. That's the kind of loss that Google executives and directors can't stomach, for whatever reason, but it's the kind of loss you have to suffer through to get fiber out to a whole state. Google could slowly conquer state after state with fiber, which would be an enormous benefit to them, but they're too cowardly and short-sighted to see it through.
Pole ownership isn't a clear cut thing, and cities are cagey around erecting new poles. So using existing poles requires paying someone money, or is impossible due to local telecom monopolies. Either way, microtrenching is very cheap and attractive, though it has maintenance issues in areas with large climate swings.
A grab bag of federal/state/local rules and widely varying ownership of those poles.
First, are you in a FCC regulated state, or one of the 21 that have passed their own pole regulations and have very different rules?
Have you negotiated a pole access agreement with the owner of the poles - which might be an electrical company, a municipality, or a Baby Bell? Some of those entities would love to see fiber internet in their community; others will be sure to make your life difficult because you are a competitor. And in some cases the ownership dictates the regulatory framework: a rural electric coop’s poles are handled differently from a municipalities, which are different from a telco.
Does your agreement/regulatory body allow you to do make ready work, or are you depending on the pole owner to do site surveys & line up engineering crews for the other utilities attached to the pole - likely telco, cable, maybe some existing fiber too? Are those poles also running electrical, in which case you’ve got even more headaches to deal with? Is the pole owner or anyone attached to the pole going to try to slow you down by claiming your changes are complex and need additional oversight?
Are you willing to wait, if you are required, for each other utility on the pole to send their engineers out to move their lines so you’re ready to place yours?
Alternatively: you negotiate an agreement with a city to trench in their right of way. One set of local regulators to work with. Looks appealing in comparison, no?
Fiber was installed here, and they provide three tiers - 50Mb/s, 100 Mb/s and 1000 Mb/s, for 45, 50, and 55 euro each. We took the 100Mb/s.
I have no idea what I would use 1 Gb/s for, using it as definition for broadband would be premature. Than again, Bill Gates said that: '640 kB RAM ought to be enough for anybody'.
At 1Gbps the internet becomes close to indistinguishable from a local network, opening up interesting possibilities like a "NAS" that's actually just the cloud. That extra bandwidth, combined with the low latency of fibre, can make it possible to easily interact with a remote PC as if it's sitting in front of you, so you could do away with a home PC or workstation and use a thin client to access a beefy machine on-demand in some datacenter.
But you can dial it down pretty damn well. When I had Frontier in Texas, I rarely saw latency over 19 ms to any location in the United States. 35+ ms and I was fussy... amazing how quickly fiber can spoil a man.
The problem, in the USA for most people, is Comcast.
Downspeed hardly matters. It's been plenty for decades. Upload speed has stagnated. I uploaded as fast in 2003 as I do now in 2022. And as others have said, it's not so much about 1 fast provider as many providers and choice. In my comcast monopoly my plan started at $60/mo in 2011 and without any changes on my end and no change in my upload speed that price has increased to it's current $105/mo. In addition to the price increases they've implemented a new system of total data transfer limits ((up/down)+internal overhead) meaning even if I could have a faster connection it'd mean nothing if I used it in the first week and then got hit with multiple $10 overage "fees".
But I don't have a choice. It doesn't matter how good, or bad, Comcast is. They're the only game in town.
The problem is not just Comcast but everything of its ilk, and by that I mean Charter (Spectrum/former Time Warner Cable), RCN/Grande/Astound and similar giant size coax based DOCSIS3 operators. In Canada, Shaw and Rogers.
They are absolutely determined to squeeze every last dollar of ROI out of that existing coax cable plant by doing asymmetric RF channels for downstream/upstream. And not overbuilding themselves with GPON FTTH.
There's a limited amount of RF spectrum in a given coax segment and they bond it together in such a way that 85% of it is used for downstream, and some paltry amount for upstream. On a heavily loaded DOCSIS3 segment it can also be 12-13ms latency to the gateway which is an eternity in metro area networking fiber terms. So you end up with something like a 200 Mbps down x 15.5 Mbps up cable connection.
Obviously the abusive bait and switch customer acquisition and retention tactics are also a huge problem. Now they're doing additional bullshit like making special offers to bundle your cellphone service and cable together (Comcast is a MVNO on some nationwide cellular carriers), for a "discounted" price, so that it makes it more painful for you to switch your ISP service in the future if you're unhappy.
Eventually they will need to spend a spectacular amount of money to overbuild themselves because in some places the local phone company (centurylink, ziply, verizon, others) is building 1 strand of dark fiber to each house, running it to some neighborhood level splitter, and going 1G or 10GPON on it for service that approaches real symmetric gigabit to the customer. To individual single family houses.
Comcast is just the most obvious example of the problem in the incumbent cable operator industry.
I always thought the upstream speed was an artificial limitation so they could upsell "business internet" to people who wanted to host servers - and that they doubled down by reassigning your IP periodically if they saw a lot of incoming connections on layer 4. I did try hosting a server on my comcast cable internet for a while and the IP which I had for seven years prior did start changing every day - sometimes more frequently. Just a coincidence? Or is that really how they play things? When I set up Tor so I could reverse proxy into my server as a hidden service my IP stopped changing - though I did notice my IP got reassigned again a few days after I got a 4K tv and started streaming the UHD content.
> I always thought the upstream speed was an artificial limitation so they could upsell "business internet"...
No, really not that at all. As noted in a few comments here, they rolled out their local infrastructure in the era of "on demand TV", when it made sense for both the cable operators and their customers to prioritize download spectrum over upload. And that meant installing expensive analog filters all up and down their lines.
I bet they would love to offer symmetric service, but at the relatively low frequencies that coax cable operates compared with optical fiber, and the cost of replacing all those filters, there is only so much they can squeeze out of it.
I guess you could call this The First Mover Disadvantage.
At this point they've gotten plenty of money from those filters, and they should be replacing them instead of doing their best to pretend upload doesn't exist.
It's a disadvantage vs. somehow knowing the future, but in the absence of future-knowing I bet they still made more money by deploying earlier.
On the other hand a cable tv incumbent in a given metro area has a huge first mover advantage of having established the right-of-way in various places (aerial pole to pole, underground duct, etc). And is usually treated by the local government as a default utility service same as water and power so they can build whatever they want almost anywhere they want, anytime they want.
RF channels actually are a finite resource on a coax segment because the aggregate capacity of one port coming off the CMTS is considerably less than properly implemented singlemode fiber. Because most residential customer traffic is inbound, like people watching netflix or browsing the web, they allocate the capacity asymmetrically.
In addition to what everyone else is saying, even if you allocate the spectrum for upstream and downstream, downstream can be run with more data efficiency because the multiplexing is handled in one place, whereas upstream is multiplex among many modems and you can't get them as tightly synchronized. When I was looking at GPON, it would typically run at half the bitrate upstream vs downstream to make upstream multiplexing more doable.
I dunno about static vs dynamic ip. In ye olde days, there was a real capacity savings with dynamic, but I don't know very many people who don't have a 24/7 device leasing an IP anyway. It's certainly a way to get a little more revenue though.
I have cable "business internet" and I get more upstream bandwidth for a similar plan, compared to a regular residential subscriber (30 megabits vs 15 or something.) However, it all shares the same infrastructure. I still only have 4 upstream channels. Total DOCSIS upstream bandwidth is limited to about 100 to 120 megabits for the entire node (neighborhood.) If you have a couple of bandwidth hogs near by, business or residential, it's going to suck!
While there is plenty of things that can be blamed on cable providers, the allocation is asymmetric RF channels it's not something that they decide to do specifically to mess up with clients but actually part of DOCSIS standard that is made by CableLabs (yea.. with input from cable companies, yet...) . A lot of todays restriction are derived from dark ages when cable was tv and not internet. IIRC till DOCSIS 3.1 (or 3.0?) data downstream was actually framed inside mpeg segments (or something like this) .
They did manage to incrementally squeeze much more out of their infra (coax) compared to what is achievable with twisted pair.
Last house that I rented had 30+ years old coax drop that were handling 1000/30 just fine and at my current house i get 1200/30.
Yea, it sucks that upstream is limited but sometimes legacy tech is legacy tech and decisions must be made in order to properly serve all the clients on plant with all the gazillion of different devices that they have
I got lucky, the subdevelopment I live in has been upgraded from classic Centurylink DSL to “Spectrum” fiber (1gbps fiber line to the house for $60/month, no contract and no equipment or setup fees. Growing up a few blocks away, it was a comcast subscription or a 1mbps DSL line(~10yrs ago).
It’s awesome when a competitor can enter the market.
> But I don't have a choice. It doesn't matter how good, or bad, Comcast is. They're the only game in town.
To extend that, Comcast is on my street but refuses to run a line up to my house because they consider it to be too cost prohibitive. At the same time, they block anyone else from making money in my region. I'm stuck with whatever cell signal I can acquire and that is as good as 5mbps during the day and might double that at night on a clear evening.
Look into wireless internet service provider gear and mesh networks. It won't be faster than Comcast. 50Mbit maybe. Make flyers and talk to your neighbors. Call around and ask about pricing for fiber backhauls.
Of course you don't want to run your own ISP. But, you could. It's not really that hard. Your neighbors could chip in for gear and be part owners. or a co-op or whatever.
If you can create a plausible threat to the $5000 a month they're collecting around you, I suspect they'll greatly enhance performance in your area, and might even offer to connect you up.
What a nightmare. What if we had to do this for everything. Hey everyone hook up to my rainwater collector. Don’t forget to refuel the generator. Sorry it’s so loud. Keep digging, we have to put the trash somewhere!
I suppose their monopoly extends to offering service for money, not for having it for yourself (or a collective of neighbors who own the gear and share costs). Were it nit so, they could sue you for sharing the cellular internet around your house.
Talk to your local government. We were in the same situation, and it turns out Comcast was legally required to service our house, they just didn’t want to (because of the expense). A few days after getting the government involved we had a trenching crew on-site.
In our case googling “Comcast service <town name>” pulls up a page from the city government’s site that has contact info for our Community Connectivity Manager. That person, when contacted, got us hooked up.
Often if they're the only option in your area AND they're on your street they've signed an exclusivity contract with the town. Part of that contract is they must service you. I know multiple people they've done this to and in all cases getting the city involved swiftly got them an installation.
> Downspeed hardly matters. It's been plenty for decades.
Until the power company ran fiber earlier this year, the only wired option here was 3Mbps DSL from AT&T. AT&T buried fiber all over the area (including in front of my house) years ago, but it sits there unused.
Same where I am. Most of the city has competition because it’s most new developments out here and a few fiber companies were able to get in the ground early. My neighborhood is older so we only have Comcast. But Comcast buried fiber here to “future proof” ages ago, they just haven’t lit it up. And they’re the only company who got their hardware in the ground. So I’m stuck on their cable.
As it turns out the service here is good, so I shouldn’t complain. But it’s still frustrating that they are just sitting on fiber and not using it. And also that there’s no competition where I am.
If I ever move, then I’ll know better and pick a house with fiber already. :/
And no need to be in the countryside to experience this. I am living in Manhattan, Upper West Side, one of the most affluent neighborhood in the richest city of the country. It is Spectrum cable or nothing. There seem to be fiber around, but it did not make it to my street (lined with small buildings so maybe not interesting enough?). Meanwhile in France where we have spent some time this year, you get fiber in the most rural and isolated areas... Go figure.
and people wonder why folks in not-very-far-out suburbs are buying and using starlink or other options, if all they have is ultra long loop length copper DSL and no DOCSIS3 available and no GPON or active ethernet fiber based service.
i've seen starlink terminals in places they have absolutely no logical need to be if the local terrestrial incumbent phone company and cable companies with rights to the ROW (right-of-way) would get off their asses
lmao one of my mates has a starlink terminal in the outskirts of los angeles just to get work done because spectrum's gigabit plan is like 1.2MB/s upstream most of the time and in the subkilobytes during peak times
it's comical how bad this is
edit: how we found out about this was also amusing: he sent a group snapchat of the arrival and setup in backyard and we were all "did you move or go traveling?" "no, spectrum is godawful"
My parents live 4 miles from the nearest city in a rural area. Someone paid for a fiber line and my parents got a free install because it crossed the property. Before this they were using a microwave antenna and getting 10 megabit.
Hah, my internet in San Francisco is worse that it was... in San Francisco in the 90s.
My symmetric DSL then was faster than downstream-DOCSIS is now. I also have to keep a backup line, because Comcast here (SOMA) goes out about every other month, and I work from home. So I pay another $35 for MonkeyBrains, which is much slower, but more than nothing when Comcast is dead.
To paraphrase Gandhi, I think American capitalism would be a very good idea.
This is such an artificial/politically generated problem too. I used to live in Tennessee and there were several cities with gigabit internet that was created by public funding. Chattanooga did this but was blocked from providing access to those that were in unincorporated areas. Even though the city still provided these people other utilities (electricity, water, gas) and mail. Same thing happened in Tullahoma (city next to where Bonnaroo Music Festival is). Then I remember Google Fiber trying to come in but the state had a law where lines needed to be placed in certain areas (which were of course in use) and that one operator could not lay down lines unless there was a representative from those that already had lines in place (supposedly to protect property). But this obviously led to shenanigans. Similarly when I moved into an apartment I chose the faster city internet but the previous tenant had AT&T. When the utilities company came to turn on my fiber (as they had provided for the apt previously) they found that AT&T had slashed their line, which caused me to have to pay an instillation fee. For a state where the people were so proud and supportive of libertarianism and the free market there sure were a lot of laws that supported or enabled monopolies to run a muck.
There are clearly available solutions and there's room for markets to exist and provide competition but it is very clear that we've set up a system that enables the unregulated monopolistic actions of ISPs.
The problem is lack of competition. My father gets outstanding service at his house in RI because he can switch to about 3 providers. When he threatens to cancel they know he means it. I get fisted by Cuckcast up here in NH because they're the only choice.
This is the key - capitalism only works when there is viable competition, otherwise it's just a monopoly.
They got around this in the UK by splitting the monopoly (well, except for Kingston upon hull) into BT and OpenReach. OpenReach provides the infrastructure, and are bound by government regulation to provide open access to last mile infrastructure for a fixed fee per product - this allows a load of ISP's to offer their own value-add services.
AAISP is pricey, but you get legendary support
Zen is less pricey (and unlimited) and good support
PlusNet is cheap as chips
The one are this is different with is FTTP - because of the slow roll out on OpenReach's part, plenty of smaller providers have stepped up and built their own infrastructure and cover various parts of the country.
Is it perfect? Of course not, but I'm paying 54 GBP a month for 900 down and 110 up FTTP.
That's how DSL was supposed to work in the US. The telcos sabotaged it by not maintaining their copper lines and forced the independent ISPs out of business.
Not really; they just lobbied the Bush II Administration into crippling the Clinton-era rules that required them to allow competitive DSL providers.
The requirement was called "Local Loop Unbundling" and it required the RBOCs (legacy local telcos, from when the Bell System was broken up) to lease "dry pairs" to various competitive DSL providers at reasonable and nondiscriminatory rates. And for a few years it worked! The dialup ISPs started to move into DSL services and it looked like we would have real competition (admittedly DSL, but this looked decent enough at the time).
But they lobbied the shit out of Congress and got LLU killed in the early 00s.
Have you heard from Fidium yet in your community? I know they haven't got coverage everywhere yet, but its surprising how quickly they have been backfilling. Plus, fully symmetrical 1 gigabit/sec under $100/mo has been nice.
I was curious about the upload issue recently and learned that it is because of limitations of the old cable infrastructure. You presumably receive your internet over coax (which can still provide really good download speeds, I was surprised when my gigabit internet package came and it only had a coax port instead of fiber). Cable TV is pretty much all downloads, with only a small portion of the bandwidth reserved for uploads, since uploading is only needed for authenticating your subscription, pay-per-view purchases, etc. Unfortunately there isn't an easy way to re-allocate the upload bandwidth, because this splitting is implemented with actual hardware frequency filters that are buried and on poles all over the US.
As far as I have heard, they could choose to use N out of M channels for either upload or download and they are choosing to keep it the way it is. Probably they need to keep it that way for other reasons (their own cable-TV signal goes through it). But I know nothing about the electronics/signal-processing involved, so I could be totally wrong.
Also so they can upsell business internet. I'm pretty sure most ISPs still have a clause in residential service agreements that you won't "operate a server" and everyone knows that the only reason you'd be uploading a lot is if you're running a server. /s
One might say that the state of the art has been chained up with awful upload speeds. Where you can think of one possible application, I can think of more, and undoubtedly there are even more that don't occur to people because it doesn't enter their experience. Being able to do practical off-site backups from home, even snapshots on an hourly basis? That'd be the dream. Oh wait, I'm doing that now because I have 1 gigabit/sec upstream at home!
It's also worth mentioning that Comcast is one of the largest "traditional" media corporations on Earth, and is therefore heavily motivated to act against media sharing (piracy).
Everything about their business outside internet service is built to be a vertical integration, so it's only natural to use their ISP business arm to support that.
Most of the TV channels are shit, so it makes no sense to reserve so many channels for them, especially as people are cutting TV service and switching to streaming.
Yes, you have it exactly right. The cable infrastructure was built out in the era of "on demand TV", and the coax cable and frequencies they used required some hard choices. Hard in the sense of hardware. Perhaps quite sensibly for the time, they installed hardware filters that allocated most of the spectrum for download.
In Denmark I had 300Mbit/s coax/cable broadband. I think the upload was about 30Mbit/s.
It was then upgraded to 1000Mbit/s down, 250Mbit/s up, with no change to the cost, IIRC. A technician installed a frequency (?) splitter of some sort, one side for TV and the other Internet.
He remarked that my socket was nicely accessible, so I think he'd been doing this hundreds of times.
(I may have paid an engineer call-out fee for this, I can't really remember. It was 2 years ago.)
And! When you lease a Comcast modem + wifi equipment, you then become an Xfinity wifi hotspot provider. Mind you, they don't tell you this. But to charge the customer money to expand their hotspot availability (and marketing bragging rights) is shite.
But like you said, often there's no choice. But even when there is - in my area of NJ - aside from new customer deals both Verizon and Comcast are wildly expensive.
And in the new world of a lot more people working remotely, upload speeds are suddenly a lot more important. Mostly because of viseo conferences, but transferring other kinds of data too.
So far it appears that only early beta customers of Starlink are getting anywhere close to useable service and as soon as it opens up to more customers the links quickly saturate and you're back to whatever everyone else is offering - except with higher latencies. Something about Shannon's Law. It's better than nothing for folks who don't have any other options, but it's not a solution to this particular problem.
The best solution I've seen to date is city-run fiber.
Or 4g/5g. I bought a place way out in the boonies and I’m getting 70 down with a 4x4 MIMO antenna, no data cap. Without the antenna it would be <5 down.
Neither starlink nor most terrestrial telco offer real internet service. I don't really care about speed. I care about participating in the internet and consistent latency. Some rare terrestrial telcos will offer ipv4 addresses and so use of ports for a much higher fee but it is the general rule that most don't. Consistent latency is impossible to achieve with wireless' random round trip time and tcp back-off.
They're fine services for consuming media but they aren't very good ISPs.
FCC needs to care about stability through statistics like packet loss. In my hometown area, the sole DSL provider in my area oversold capacity and it demolished the DSLAM. Ironically, they also took advantage from the Connect America fund to do this. I did connectivity tests to various hosts throughout the days, pings were 500ms+ 40% of the time and packet loss was 1%.
Yup. Same with one of the local WISPs around here. On top of it all, I think they used some external grant funding to upgrade the last mile links, and cut upload by 90%.
I haven't run a ping test since then, but last time I checked, it was > 1% and 1 second to the internet.
In fairness, they're lobbying for a govt bail out to subsidize their upstream fiber provider.
If these are considered to be the correct way forward, the constitution needs to be amended then. When it was written Internet didn't really exist, so it would make sense there aren't many provisions covering it.
I'd rather have two providers that offers 25Mbps than one that offers 100Mbps. I don't need 100, or even 50. But Comcast has been jacking up their prices, along with increasing the minimum speed available, for years. They respond to government regulations like these. I'm now paying 2x what I used to (from $25 to $50/mo), and getting 2x the speed. But I don't want to be forced to buy 2x the speed (or even faster, as this proposal would require).
I would much rather have the govt provide incentives for there to be more than one provider. I live in the heart of SV (Menlo Park), and Comcast is the only option. 5G is not well-served where I live, and the pricing is also not competitive. We need more competition, not more redefinition of "broadband".
Bear in mind that connection speed sold “up to xx mbps” is largely irrelevant to the actual cost of service. It’s just the sales model that most ISPs have adopted over the last few decades when bandwidth really was constrained. Comcast’s costs are its people, truck rolls, equipment and infrastructure outlay, and advertising.
Convincing or requiring you to pay double for 2x the speed is just the way they can increase cash flow.
Ideal world would be seeing the laws preventing municipally owned ISPs struck down. I'd also like the see the right of ways municipally owned as well, to better facilitate competition with new providers.
> I would much rather have the govt provide incentives for there to be more than one provider.
In the UK, ISPs are forced to lease their last-mile lines to anyone who wants to provide internet service. This results in a decent amount of competition.
Yeah we have MVNOs that do this for mobile. But it can't currently compete with wired internet on price, and we apparently don't require it for cable internet.
ISP customer service is never fun, but my problem is not really with who picks up the phone. It's with the physical network that's provisioned for my neighborhood and the way it's maintained. The lessee has no more power over this than I do, only now I can get stuck in customer-service hell between two different companies instead of two different departments.
POTS. ILECS were required for a while (still are?) to sell access to their central office DSLAMs. For a hot moment in the early 00s there was a fair bit of competition for DSL service (at least in the Bay Area).
If memory serves – as technology moved on to higher speed DSL that worked over shorter distances (and required more RTs as a result) legislation stayed the same. Essentially the only things that ILECS were required to provide access to were outdated products that nobody wants. Sonic finally shut down their resold AT&T ADSL (G.DMT) service this year in large part because it was so unpopular. They still offer ADSL2/VDSL2 service out of their own DSLAMs as well as fiber and resold AT&T whatever (not sure how that came about).
> ILECS were required for a while (still are?) to sell access to their central office DSLAMs.
It was just the copper, not the DSLAM access. What you could do is buy colo space at the CO and install your own DSLAM, and the ILEC was required to cross connect the copper to your cage. What you did from there was your business, as was all of the CPE.
I worked at a CLEC for 10 years during the good times. Even set up a few of our colos myself back in the day. I still think we (the US government) should have kept going and required incumbents to lease access to any last-mile wiring technology. It should be part of the price for enjoying what amounts to a government-sanctioned monopoly. Even better, just make the last-mile network a regulated utility like water & electric.
Agreed that the FCC should've continued requiring open access. However I'd go further and say that much of the infrastructure should be state owned and leased to private companies.
Or outlaw exclusive franchise agreements, require timely access to utility poles, etc. Regulate like an electric company? Well I'm from California and given how bad our electric companies are and how toothless the state regulator is I'm not sure how much of an improvement that would be.
If, say, Comcast had to lease their last mile, why wouldn’t they charge an arm and a leg to whoever was trying to sell internet? I don’t see how this would bring prices down meaningfully.
You can practically regulate how this works by determining cost and markup. I mean most places have one power company for example and they aren't simply allowed to charge you whatever they please.
In fact this was done years ago successfully. It was discontinued so cable companies could make greater profit.
This is by design. Utility service regulation in the US is famously corrupt.
The system is working exactly as intended.
Internet access should now be close to too cheap to meter, yet it is a nice cash cow that can be milked, so the regulators have an arrangement with the highest bidder.
You might be eligible for ATT [gigabit] fiber? Could be worth a shot. (Unless you're in far-east Menlo)
IME the fiber service is faster more reliable than Comcast, but AT&T support is the absolute worst, super incompetent (across the entire company, not just U-verse internet).
It's fine as long as you don't need any support and don't try to deactivate anything, but if your account gets messed up it can literally take a dozen calls and 20+ hours of being on the phone to get seemingly simple issues resolved. </rant>
The Comcast ATT duopoly is pretty atrocious. Given the regulatory climate, how can a new player even succesfully enter into the market?
One trick for better AT&T support here on the Peninsula may be to sign up through Sonic instead of directly with AT&T. In some areas like SF, Sonic runs their own fiber, in others they are an AT&T reseller - with the wrinkle that Sonic handles all support with their own local staff. I have been told that they have better insight and connections into the internal AT&T systems than AT&T's own support team. I haven't had much need for support, so I can't vouch for that, but it wouldn't surprise me.
I'm in Flood Triangle (between Menlo Oaks and 101), and AT&T ran fiber though our neighborhoods last year.
I had good Comcast service, at least after I reported some intermittent outages and they finally sent someone out to test the connection and he said "I'm surprised you're getting any service at all! Your drop has been spliced a few times. I'm calling my buddy with a bucket truck and we will run you a new drop."
(A "drop" is the cable that connects your home to the main cable line.)
After that it was solid. Had gigabit downloads, but uploads were limited to 30 megabit. Comcast is in a bind here, as they built out their infrastructure in an era of "on-demand television" where download speed was the important thing and upload was just a simple control channel. So they have hardware filters everywhere to separate the upload and download frequencies, and naturally they gave more spectrum to the downloads.
In my work I often need to upload large files of 15GB or more. So when AT&T came through I signed up and it is a huge difference: ten minutes to upload one of those files instead of two hours (with people waiting on it).
> One trick for better AT&T support here on the Peninsula may be to sign up through Sonic instead of directly with AT&T. In some areas like SF, Sonic runs their own fiber, in others they are an AT&T reseller - with the wrinkle that Sonic handles all support with their own local staff. I have been told that they have better insight and connections into the internal AT&T systems than AT&T's own support team. I haven't had much need for support, so I can't vouch for that, but it wouldn't surprise me.
My experience buying DSL from Sonic about a decade ago in San Francisco is that if you run into a problem with AT&T equipment (e.g. lines upstream from the point where POTS enters your building) you're just done because Sonic won't actually get AT&T to fix their stuff. I was never able to get more than 1Mbps down and eventually just quit and switched to Comcast.
When I first tried to get Sonic ADSL in San Francisco there was a multiplexer on the line (no 56k, no DSL). Lots and lots and lots of calls with Sonic later and they threw their hands up and told me to deal with AT&T myself. In the end what worked was going into the PacBell/SBC/AT&T direct support forum on dslreports.com and having one of those techs prod the bitchy bell into disconnecting the multiplexer. Nothing I've seen since then suggests that Sonic has any more sway over AT&T these days. In fact if you get the resold AT&T service from Sonic you still have to rent the AT&T dongle and use the AT&T DNS servers.
Atherton Fiber is a local institution, a true old-school ISP in the finest tradition, but with modern technology.
I live very close to their service area in Lindenwood, and I remember walking down Ringwood a few years ago when they were stringing fiber on the poles and I stopped and asked them, "Hey! I live just a few blocks over that way. You must have a few hundred extra feet of fiber you could sneak over to me?"
They knew I was kidding, but seemed like nice people.
Hang in there, my friend, someone will get you fibered up soon.
Unless they're in far-east Menlo or like 99.99999% of the rest of the world. While bay area folks are certainly heavily-represented here, the commenters come from all corners of the globe.
have you considered looking into the 5g home offerings from Verizon and TMobile? In my area Spectrum was having a fine time jacking up rates unfettered and laughing at me when I asked to reduce them. I switched to Verizon 5G Home and when I called Spectrum to cancel, they practically begged me to stay (edit: I didn't) and have now adjusted their rates to match Vzw/Tmo; of course these reduced rates are only good for one year because they are Spectrum, and there is no cure.
Where I live, there is neither Verizon or T-Mobile service, unfortunately. This is a stone's throw from Stanford; I cannot understand how we only have cell provider that works at our house. Fortunately there are MVNOs, so we're not stuck paying AT&T pricing for mobile.
Here in Baltimore, the only option for any kind of broadband is Comcast XFinity, and it costs $85/month for 50 Mbps service. (And there is no plan available for under $65/month).
What does it mean to say the "minimum" is 100, are they just gonna charge me a minimum of $150/month or something?
The fact that I see other people in this thread saying they're already getting gigabit(+) service for $40(!)-$80/month, while I pay $85 for 50Mbps from the single monopoly provider that does not offer gigabit/fiber....
makes me think the biggest need for regulation is not about minimum speeds, but about price gouging from monopoly providers. I wonder what Comcast charges for the same 50Mbps service I pay $85/month for (with again no plan avail under $65), in markets where they have a competitor. i would love to pay $40/month for my internet, and don't even need any more than 50 Mbps -- comcast increasing the speed of my $85/month plan (or doing so and raising the price!) doesn't actually help me much.
Has anyone done a national broadband price map? They intentionally don't make their geographically disparate prices very transparent of course. It would be interesting to see the variation in what comcast charges for the same speeds, and how it correlates to having competitors in a market vs monopoly.
I'm trying to figure out what the likely outcome of this proposal will if enacted be, mapping to my tenuous but sufficient home internet service.
Is my current best-choice provider (TMO 5g Home internet) going to boost my speeds from current 10-100/1-10 Mbps dn/up to a solid floor of 100/20Mbs or lose Federal funding?
Or, will TMO receive more Federal funding to bring me up to a solid 100/20Mbs floor?
Or, will TMO classify my service as something other than "broadband" and continue to provide unchanged QoS?
Or, will TMo drop customers such as myself "on the performance margins" of their service (instead of boosting tower capacity)?
100/20 is definitely an improvement, and is probably a reasonable minimum bar to set. I'm sure most of us have much higher speed connections, but I know I rarely take my 1Gb pipe over 10% of that... Heck, 2 TVs streaming 4k during a zoom call would be fine on 100Mb...
> We're on 400Mbps, and even then I manage to block internet for others when I download a large file at full speed.
That's more of a router/QoS issue. A large download shouldn't negatively affect quality for other uses, especially VoIP. If the place you're downloading from has a pipe bigger than yours and can saturate your bandwidth you're going to need to implement some kind of rate-limiting/queue management.
I've seen a lot of people complaining that their 100M+ connections are "slow" for that reason, and they upgrade to 1G and they complain again. It's not like they're doing much on that 1G line either, just say a Steam game download at full speed followed with a stuttering Zoom call. You'll be surprised how far you can push a 10M DSL connection with good QOS (assuming you don't have multiple users streaming video and such).
I don't think I've ever seen a 4K YouTube stream over 15 Mbit, where do you find a 70 Mbit stream? Even the king of streaming bitrate, Apple TV, is said to peak at 30 Mbps
I'm not sure this is accurate, maybe a sufficiently cherry-picked small # of frames in a VBR recording?
It, at the very least, hides more than it illuminates in that 68Mbps is an order of magnitude higher than an average 4K video's bitrate over it's duration
I don't think it's reasonable if it means my provider's offering start at 100 Mbps vs my current 25 Mbps, at what I'm sure will be at least double the cost to me.
Higher minimum upload speeds will be more impactful than cranking download speed even higher.
100/100 is way, way better than 500/20.
Upload speed is what enriches the two-way internet - servers, content uploads, high quality videoconferencing, cloud backups. Prioritising download speed is keeping us in the broadcast era.
True, though I'd guess a low single digit percentage. Those are the exception, not the rule. ISPs could upsell them on higher end plans. Though I doubt that will happen. A real problem with broadband is lack of choice, both in the plans themselves and the providers.
The amount of bandwidth you need depends on the number of people. 50Mb or 25Mb is probably fine for 1 person. As long as it can do 4k, and zoom calls, it's fine for most people.
100Mb might not be enough for six people, who might need to download, zoom, game, and stream all at once.
I think you're somewhat overestimating the amount of bandwidth used by those services. TO watch Netflix in 4k, you only need 15Mbps, zoom uses about half of that for a decent connection and will downgrade nicely.
I agree with you, a lot of these services aren't as high bandwidth as people imagine but I will note that part of the reason is because that's what people routinely have access to. Netflix or whoever could be pushing higher bitrate content or potentially exploring other more bandwidth intensive ideas if people had access.
It's obsession with technological advancement. There was a time when people were resistant to even 720p, saying that 480p is more than enough and they can't even see the difference.
I just hope the FCC will make it 100 symmetric. I can already hear the army of Comcast lobbyists calling their friends in the government. Look closely at who opposes that.
Am surprised to see the pushback asserting this will cause prices to rise. This speed was practical on comcast's cable system back in 2007 (when we got it) and probably earlier. There is a wealth of evidence that higher bandwidth has minimal to marginal costs for typical residential usage.
On to the actual proposal, it raises the minimums from:
* Down: 25mbps --> 100mbps
* Up: 3mbps --> 20mbps
As a head of household of 5, I carefully monitored and managed our internet usage during covid, and upgraded from comcast to AT&T fiber (1g symmetric). I can say with high certainty that the previous minimum (25/3) would be severely inadequate for a family of 5. This bump to 100/20, while not amazing, is a good step in the right direction. It would make 5 people working/learning remotely at least _possible_.
TLDR: The tech is cheap and decades old. There should be minimal cost to ISPs. The best arguments in this thread, imho, are that we should strive for even higher.
I pay $40 for a Verizon FIOS connection. It's marketed as 300Mbps but my speed test just now showed 120Mbps. But it also consistently exceeds 100Mbps on the upstream. It's easily the best home internet connection I've had since the university dorms in the late 90's. They also offer 1Gbps for $70, which is tempting but really unnecessary for my needs.
I don't understand why I'm getting such an amazing deal! It's a fixed price, not a special. And this is in Manhattan, where most things are noticeable more expensive than in the rest of the country!
Previously, I lived in SF and was forced to use Comcast because they were the only option in my old, rent-controlled building. I paid a lot more, both for the service and for modem rental, but received much slower speeds. I hope the FIOS model is finally making it to more of the country. But 3 years ago I couldn't even get it in downtown SF!
Presumably, they'll be measuring by taking the minimum available speed (up or down) over 5 minute windows, on a quarterly basis, and disqualifying anyone with > 1 second latencies for one of those windows, right?
Yeah. Didn't think so.
Politicians gotta politic, I guess.
Edit: At least I'll have decent speeds at 4:38AM now...
I haven't read this whole thing yet but there should also be a limit on latency. All the viasat services have stupid high latency. Starlink is so much better, also cheaper.
There needs to be much more upload speed and much higher data caps. The current structure is stunting the economy for the profits of some not awesome people.
Here in the portland area we get Centurylink and I have to say, having been forced to use comcast in the bay area for years it's a breath of fresh air. It's a large company, sure, but 1gbps symmetric is very reasonable (around $50/mo) and they let you directly connect your hardware via IPOE (using a VLAN tag).
What resources does a 100 Mbps connection give access to that a 25 Mbps connection does not?
You don't need 100 Mbps to access educational content. Khan Academy looks fine in 1080p. And even if it didn't, Netflix says their 4K streams only need 15 Mbps.
You don't need 100 Mbps to be a remote knowledge worker. Video calls need a stable connection with low latency—I wish more people used ethernet—but that connection does not need to be particularly fast. An upload speed of a few megabits is sufficient.
You also don't need 100 Mbps to apply for jobs, read the news, chat with friends, or participate in the political process. Any noticeable improvements in page load times above 25 Mbps will be marginal at best. It's fun to hate on how bloated modern websites are, but the situation isn't anywhere near that dire.
I'd posit that if a non-technical person with a 20+ Mbps connection thinks their internet is too slow, they probably aren't actually getting 20 Mbps, most likely because of spotty wifi. And if they really are getting 20 Mpbs, they're probably reacting to latency rather than bandwidth.
I don't understand how this is the top comment here. This is a modern equivalent of "640K of memory should be enough for anyone." You're also making your assumptions off of a single device on the network. I don't even know a home that has only one connected device at any given time. It also doesn't consider things like how speed can help an unstable connection as a high speed but with moderate variance can beat out a low speed that is stable. There's just so many factors at play that are ignored here that this comment is odd to be placed so high in a forum of tech literate people.
> This is a modern equivalent of "640K of memory should be enough for anyone."
If you have a 4 bit CPU, it probably is enough memory for anyone!
I think we're focusing on the wrong thing. Download bandwidth is fine; stability, latency, and to some extent upload bandwidth are not. I think a forum of tech literate people should understand the importance of optimizing for the right thing.
But bandwidth does matter when you have multiple devices. On a typical day it is not uncommon for multiple video calls to be going on at the same time in my apt. My roommate and I used to schedule our meetings around one another so that we could have a call with our respective advisors that wasn't choppy. We'd also make sure not to be streaming video at these times either. Bandwidth is 100% part of the problem.
Now I agree that there are other issues like stability, latency, and upload, but it is just naive to dismiss download bandwidth.
> On a typical day it is not uncommon for multiple video calls to be going on at the same time in my apt. My roommate and I used to schedule our meetings around one another so that we could have a call with our respective advisors that wasn't choppy. We'd also make sure not to be streaming video at these times either. Bandwidth is 100% part of the problem.
I don't know your situation, but fwiw, if a friend was having this problem and they came to me for help, I would assume it was their wifi network.
Up until a few years ago, I only had 3 down / 1 up in my apartment. My building had FIOS, but it was much more expensive than my absolutely dirt cheap cable connection. The internet was stable and reliable, and Facetime calls went fine. I did live alone, and large downloads were a massive pain, but 3 Mbps is much slower than the speeds we're discussing.
A tip for anyone having such issues with wifi on linux: Is it a 2.4GHz network and is anyone using bluetooth (such as headphones)? They might be interfering with each other.
The iwlwifi kernel module has an option that tries to allow them to coexist, but it seems totally random whether or not you should have it on. Different bluetooth headphones also affect it in different ways (as I discovered a few weeks ago when I bought a different brand; with my usual the bluetooth would drop for about a minute then reconnect with no effect on the wifi, but with the new ones they would go crazy maintaining the connection, causing a lot of audio stuttering and extremely slow wifi).
Yeah, that's not what was happening. It was bandwidth issues over peak usage times. You could run speedtests (to the router) and find this out very quickly. I know you want to assert your original answer as still being correct. I'm not saying you're wrong, I'm saying that there are other important variables here that you are just throwing out. Anyone that has done debugging with network issues knows that there's a thousand things that can go wrong. Yes, there's common problems, but there's a lot of breaking points. Certain ones became more common during the pandemic as we stressed networks in different ways than most people were doing previously.
What if you have a family with four kids, all of whom need to engage in remote learning on occasion?
20Mbps is also entirely noticeable for bulk downloads, whether it's a Steam download or a Windows update. By going from 20 -> 100 Mbps, you save nearly an hour on a 10GB download (1h6m -> 13m).
In the unlikely scenario where all four children are simultaneously using the internet full-throttle, they're still each getting 5 Mbps down, which Netflix says is sufficient for 1080p video. Upload speeds could start to be a problem though, and IMO that's a much more significant issue.
> 20Mbps is also entirely noticeable for bulk downloads
It is, but downloading Steam games quickly is not a critical need. It can be done in the background.
5 Mbps down is certainly more than adequate for a single 1080p stream that's been compressed ahead of time. When you have to stitch together multiple real-time streams as in a classroom environment, then the actual numbers can vary significantly.
If your kids are forced into remote learning, and you're working from home, then we might be talking more like 3-4 Mbps per person if your network is split 5 or 6 ways. And unless you're on a business internet plan, the only guarantee you have when you buy a 20Mbps plan is that your ISP will throttle you to that rate if you exceed it.
To be clear: I think most people would be happy if they could get a guarantee of 50 Mbps download speeds. But the fact of the matter is that during peak hours, if you pay for higher download speeds, you get a higher fraction of the available spectrum. Your big-name cable companies can have oversubscription rates of 100x or more. That doesn't translate to only getting 5Mbps out of your 500Mbps connection at peak, but you will see _some_ contention. If you're paying for 20Mbps under the assumption that you'll only need 18Mbps at peak, you'll be frustrated when you end up only getting 15Mbps and all the video calls stutter / cut out intermittently.
> If you're paying for 20Mbps under the assumption that you'll only need 18Mbps at peak, you'll be frustrated when you end up only getting 15Mbps and all the video calls stutter / cut out intermittently.
But isn't that the real problem here? Companies shouldn't be able to advertise a 20 mbps connection if they aren't able to maintain that speed. Dropping one megabit on occasion may be reasonable, but five is at the point where you should be compensating customers for the outage.
My thesis is not "everything is fine and everyone should shut up and accept what they've got." My thesis is that we're focusing on entirely the wrong thing.
There are certainly a lot of problems with incumbent ISPs that many other commenters have pointed out: deceptive marketing practices, anti-competitive behavior (ranging from lobby-driven legislation to lock out new entrants, to outright sabotage / destruction of cables), wildly asymmetric up/down speeds driven by economies of scale from 20 years ago, just to name a few.
Yes, even 4k video streaming at 20 Mbps is viable (assuming that your connection is stable). Yes, poor wifi performance / latency / bufferbloat / squirrels eating your coax are more common underlying causes of end-user pain points. But that doesn't mean that there aren't common use cases that benefit significantly from adding more bandwidth, nor does it mean that adding more upstream capacity is useless in the context of these other factors.
Also, let's be realistic: most regulators don't have anywhere close to the level of technical expertise that you'd expect from, say, a network engineer. It's a lot easier to reduce everything to the single dimension that is download speed, and it's a lot easier for a layperson to understand said changes.
It's also a straightforward way to place pressure on ISPs to upgrade their networks if they want to keep saying that X% of their customers have access to broadband plans.
> But that doesn't mean that there aren't common use cases that benefit significantly from adding more bandwidth, nor does it mean that adding more upstream capacity is useless in the context of these other factors.
I agree! I just don't think it's what we should be pushing for right now.
> Also, let's be realistic: most regulators don't have anywhere close to the level of technical expertise that you'd expect from, say, a network engine
Well, I have higher standards than you do. I think regulators should have a very strong understanding of what they're regulating, or should talk to people who do.
> It's a lot easier to reduce everything to the single dimension that is download speed, and it's a lot easier for a layperson to understand said changes.
To me, this is a bit like saying we should still be measuring all CPUs according to clock-speed, because it's easier for a layperson to understand a single number. (And make no mistake, clock-speed is important!)
People here seem to think I'm letting the ISPs off the hook somehow. On the contrary, I think this fuss over bandwidth is great for the ISPs. People pay more for faster speeds because they think it's making their Zoom calls more reliable, when the actual culprit is just the piece of crap wifi router they're renting for $5 per month...
> I agree! I just don't think it's what we should be pushing for right now.
While I agree that it's not the only problem at hand (and perhaps not the single most useful change the FCC could make), I think any movement is better than none. If nothing else, the WFH / distance learning situation brought about by the pandemic demonstrated that the current standard for upload speeds is entirely inadequate (and so the proposal increases that from 3 Mbps to 20).
It's also possible that these government agencies are nervous about trying entirely new things, lest the current Supreme Court swat down any such efforts as perceived overreach (as they've done recently to the EPA).
> I think regulators should have a very strong understanding of what they're regulating, or should talk to people who do.
Should != do
It certainly would be nice, but in the last 20 years, the only FCC chairs with relevant prior industry experience were Ajit Pai (Verizon counsel) and Tom Wheeler (CEO of CTIA).
As far as talking to people with experience, guess who's most likely to make sure they're represented at the table? That's right, the incumbent ISPs and their lobbyists.
> To me, this is a bit like saying we should still be measuring all CPUs according to clock-speed, because it's easier for a layperson to understand a single number. (And make no mistake, clock-speed is important!)
Well, we sort of still are, except that one number is Geekbench or whichever other benchmark is popular these days. (Okay, there's also single-threaded vs multi-threaded perf, but there's a strong correlation between the two for consumer-grade CPUs.)
(and I'd expect that your typical consumer doesn't really care much about the CPU details in large part because performance is so opaque now. They probably just buy something with an i7 or the latest generation Macbook and call it a day.)
Tbf, the updates can just take longer to download. Higher speed does not enable them as such, which is what the op asked. Similar for steam - spending extra hour for a download ahead of time is... fine for me.
You've got strong opinions here, but... chill. I'm not enjoying the wait, I'm not pushing it on others and some wait time to play a game is not suffering. The question was what does the higher speed enable - of course it makes lots of things better, but waiting less is not enabling.
> The question was what does the higher speed enable - of course it makes lots of things better, but waiting less is not enabling.
It enables more productivity, if it doesn't take me x time to upload photography to a client, or push a bunch of Docker images to a remote, but x/2, x/3. That cycle time being reduced is useful.
Everyone's needs are different. Yes, a lot of people can get along just fine with 20 megabit up and down, and you are spot on that latency may be the worse issue.
In my case, my work requires me to frequently upload files of about 15GB in size. On my old cable connection - which did provide gigabit downloads - the uploads were capped at 30 megabits. So it took about two hours to upload one of these files, with people waiting for it.
Now that I have a symmetric gigabit fiber connection, these files take only 10 minutes to upload. That's a game changer.
It's because I'm uploading to https://www.box.com/ where IBM has an Enterprise account.
I am amazed that Box has nailed a good portion of the Enterprise market with a substandard product: poor transfer speeds and an astoundingly bad UX on their web access.
I do get the expected speeds when uploading to or downloading from other servers.
You never lived in a modern house with kids I take it? We got >10 devices online at all time. Phones,4k TV's, between 1 and 6 computers, up to half a dozen tablets. Live video feeds coming in (up to 4k in res), online gaming, live video chat connections going out, system updates, vpn into work, constant content streaming, software downloads, web apps, web games. Can easily fill a 100mb connection. Hell we do most of the day.
On the other side...you know once you have fiber pairs in the ground in good network layouts...its about 80 bucks per user for the infrastructure on each end to get the user 10gb symmetrical connection. In rural areas you can roll fiber cheap as chips, lil vibratory plough + conduit. air blow fiber from poi straight to peoples front door so to speak. Its actually quite a cheap process given the infra lasts for about 20 years.
ISP's already have the ability to switch everyone to 100mbps. It's a simple remote config switch. They don't, because they are milking money. They are all publicly funded through city contracts and federal subsidies.
I can imagine very easily why 25mbps isn't enough - people have families. No, you cannot actually do business or school work when media content is now the vast majority of consumed bandwidth.
Also don't advocate against yourself. Your city boards have already been doing that for decades by giving exclusive access to specific ISP's and making the public pay for it.
> What resources does a 100 Mbps connection give access to that a 25 Mbps connection does not?
a 25 Mbps connection might be at the absolute top end of what's possible on some old degraded infrastructure that the local phone company has held together with the metaphorical equivalent of duct tape (like a VDSL2 DSLAM in a street cabinet attached to your copper). It can never get any better until replaced with a totally different technology.
100 Mbps at least requires something like reasonably well built DOCSIS3, if on old coax, or it could be a low tier service plan on something that is actually capable of gigabit. There's people out there right now buying cheap residential 100 Mbps accounts on GPON ONT that are actually capable of full gigabit.
by requiring providers to meet a metric of 100 Mbps down, it is one method of ensuring that whatever tech they use to provide the 100 in the last mile is at least somewhat future-proof.
In a modern household you may have multiple people on the same connection. Two people streaming Netflix at 4k from the same household would not work in your example of "sufficient" 25mbit/s speeds.
Additionally, higher speed provides opportunities which lower speeds don't (e.g. building a live streaming business) which you otherwise don't have.
For one person, I completely agree. But if you have two remote workers on a conference and a kid doing something else at the same time, 20mbps is not going to be comfortable anymore. (Yes, it's probably still usable with degraded quality.)
Aren't they more like "minimum maximum speeds"? If your local service is fast enough then you'll only manage to saturate your connection to nearby servers that aren't heavily loaded. Those with gigabit Internet can probably relate.
Faster speed will mean that providers will charge more. I am perfectly fine with 25 Mbps. I was fine with 10 Mbps too until they eliminated that level and I was forced to pay more.
This is not forcing providers into faster speeds. It's forcing them to stop marketing crawling speeds as "broadband." Of course if they are competing for contracts for deploying internet, they will have to meet the higher standard if "broadband" is specified in the contract.
Additionally, it's far from a foregone conclusion that "faster speeds will mean that providers will charge more." I've been paying roughly the same amount for the past 20 years starting from DSL speeds all the way up to 1Gbps. There were some years that my speed doubled but the bill was lowered.
If you have a monopoly, they can arbitrarily raise their prices without increasing the speed at all. From the situation you described in another post, it sounds as if they were going to raise your bill no matter what, and pushing you to a higher speed tier was just the rationale they gave. At some point it becomes more expensive for them to maintain old equipment than to replace it, and if the replacement equipment has much higher rated speeds, they're not saving anything by continuing the lower speed plan. To the contrary, a lower bill just means a lower profit margin.
Consider: Let's pretend that broadband speed worked like a typical commodity so that the producer cost for 10Mbps was roughly ~10x as 1Mbps. And let's say your bill for 10Mbps was $30 (producer cost $20). Would they offer a 1Mbps plan for $3 (producer cost $2)? No, because they would be cannibalizing their own customers. They don't want $1/month profit per customer, they want $10+/month profit customers. And it real life, they have even less incentive to offer low speeds. If gigabit plans are on the table, there's virtually no difference in the maintenance cost for the lower end consumers using 1Mbps, 10Mbps or 100Mbps. It's really only the top few percent of customers who use enough bandwidth to affect QOS. Hence, almost all of these companies offer just a few pricing tiers, the top tier for high bandwidth whales, the middle tier for almost everybody else, and a bottom tier for the super-budget conscious (and also to make the middle tier users feel like they are paying for a premium service.) As the top tier offering gets ratcheted up by technology, so will other tiers. And so will your bill, if you're suffering under a monopoly.
So, in areas where a provider only offers 20 or 50... will they just stop serving that area until it's financially worthwhile to upgrade that equipment?
No. The minimums set by the FCC are just the minimums for the service to count as broadband. There's no requirement that a given internet service actually is broadband internet service. The worst that would happen to the provider is that they might not be eligible for some government programs that promote broadband.
The law requires the FCC to regularly check on the availability of broadband in the US and to determine if it is being deployed in a reasonable and timely fashion. If they determine it is not they are required to take action to accelerate deployment by doing such things as removing barriers to infrastructure investments and promoting competition.
I do I suppose. I just want my files to download, send, and my websites to load. I don't play games or operate some kind of server from my house, so for me it works as well as the wired internet at my work.
What's the point? It's not even possible for me to get what's currently defined as broadband where I live. Changing some definitions isn't going to give me faster internet.
Correct me if I'm wrong, but I'm pretty sure the term "broadband" is connected to things like subsidies (i.e. giving an ISP $x if they can provide y people with broadband) intended to increase internet connectivity in poor or remote areas. Increasing the definition also means that ISPs can no longer cheap out and pocket the subsidies.
I'm sure it's also connected to incentives to bring 'broadband' to 'unserved communities'. Now, CenturyLink can get a new incentive to let my sync rate float up a smidge. Or, maybe they'll decide since it doesn't qualify for broadband that they'll just bow out, it's not like they enjoy being a phone company anyway.
There's an important point that I think people are missing, which is that the bandwidth usage of internet services is going to be carefully tuned to balance quality and stability at typical internet speeds. When a large portion of consumers have < 100Mbps down of course all services are going to use < 100Mbps of bandwidth - you don't know what kind of stuff you're missing out on because of these slow internet speeds, because nobody is making it available, because it wouldn't work.
The most obvious thing that we already know we're missing out on is video streaming quality. You've possibly already noticed that low contrast dark scenes and scenes with snow/rainfall don't look very good on Netflix/YouTube/insert streaming service here - this is because these are the kind of scenes with the most obvious sacrifice of quality from video compression. (It's also possible you haven't noticed this, because streaming services generally avoid producing content that doesn't play well with their video compression).