Utah Locals Are Getting Cheap 10 Gbps Fiber Thanks to Local Governments | Hacker News Reader

Utah Locals Are Getting Cheap 10 Gbps Fiber Thanks to Local Governments

techdirt.com

365 pointsWarOnPrivacy220 comments

Comcast lobbied so hard to get our local mileage for fiber expansion voted down. When it passed, they suddenly decided they’re going to spend 6 figures worth of money to expand their own fiber backdown (they’re not even to the house). Its amazing when corporations will spend more money fighting something than actually implementing said thing.

when you can simplify things down to a number on a spreadsheet, it's a simple math equation. It doesn't make sense if you step back and look at what's actually best, but money warps everything it touches.

> money warps everything it touches.

I wonder what would happen if money transactions frame dragged transaction metadata along with the money. Today, the money is public data, but the far more interesting transactional metadata exchange is a hyper-fragmented market.

I’ve had fiber at my house from day one provided by a local ISP. Comcast came in and buried their own fiber lines (years later) for all the houses in my neighborhood. Soon after, my ISP bumped my speed and lowered my monthly bill.

We desperately need more competition. It’s the only thing that actually lowers prices.

Also, I will happily pay higher prices to never pay for Comcast again. So they won’t see a penny from me.

> We desperately need more competition. It’s the only thing that actually lowers prices.

Unless it's a deep pocketed behemoth coming and stomping out local competition because they can afford to lose money on price dumping.

Though I agree with the general idea of the competition.

This is all covered in microeconomics. Competition is very good for society. Things that stop competition are bad and require government action. The scenario above is describing the behaviors of a “natural monopoly” due to the barriers to enter the market. As with all monopolies, regulation or some government action is normally required to promote competition and stop bad behavior (some examples of government actions are described in the link below / the root article is an example).

https://en.m.wikipedia.org/wiki/Natural_monopoly

I really wish microeconomics was a required course in high school or primary. I find it to be one of the least understood of the well-established fields, and one that matters when we get older and vote or debate on these topics.

I hate how corporate lobbying is described as "government action", as if the government here is the actor and not the condom.

The government is the actor. Do you think that corporate lobbying gives corporations the right to write their own laws, or something? Lobbying is paying someone to speak to a government representative. Every government representative has the full authority to respond, "No? That's stupid," and if they respond any other way they should be the one held accountable for it.

>Do you think that corporate lobbying gives corporations the right to write their own laws, or something?

Given how common it is for politicians to pass bills directly written by corporate lobbies, yes. I think exactly that.

People worry about price dumping waaay out of proportion to how often it is a problem.

Usually when there is a deep-pocketed entity stomping out local competition they are using regulatory capture as their main tool (this is what Comcast seems to do) and the problem is competition has been made illegal.

In situations like this, my default assumption is that it is illegal to lay competing cables until proven otherwise. We had a similar situation of massive monopoly provider in Australia back in the early 2000s, the economics didn't stop new companies laying new cable to compete with the incumbents. That all died when the government launched a new national network though.

That new national network ("the NBN") is so hit-or-miss that it's not funny.

My current connection in Brisbane is 1GbE download (incredibly fast for Australia), but only 50mbps upload. That 50mbps upload is the fastest commonly available speed in the whole country. Most ISP's don't offer anything faster than 20mbps. :(

Reliability isn't great either. Had an 8 hour (!) outage overnight about a month ago, then a ~2 hour outage last week. Both caused by NBN Co, rather than my ISP.

The concept of hosting anything actually useful or important from home infrastructure is just laughable when there are commonly multi-hour outages. It's just fucking stupid. :(

My favorite part of NBN was always data caps. Data caps on internal traffic thru fiberoptic backbone!

Do data caps still exist with the NBN in recent years?

I somewhat remember them from years ago, before I moved overseas for a while.

6 figures? For a single house or?

He explained himself in the last sentence

>"Its amazing when corporations will spend more money fighting something than actually implementing said thing."

Corporate ISPs take the same approach to community ISPs as they do to unions. Any hint of their existence must be crushed immediately and forcefully regardless of cost. If community ISPs (or unions) succeed, it shows what is possible.

I recently bought a house in Utah. Utopia is the best internet provider experience I've had so far (I've had Xfinity and Verizon previously).

Speeds are typically 10-25 Mbps _faster_ than the advertised rate I'm paying for. Being able to have multiple (~5 or more) provider options to choose between is a welcome change from past monopolies I've been trapped in.

Granted this all could change down the road (I hope not), but so far so good.

Utopia provides an active ethernet connection to each subscriber with the specified bandwidth. This is superior to most fiber deployments, which are passive optical networks (GPON) where groups of 16-32 homes share a connection (usually 2.4 Gb down and 1.2 Gb up).

I live in Utah and actively worked to get fiber in my small city. Utopia requires cities to guarantee a take rate (40% if I recall correctly) and taxpayers must make up any deficiency. I've never heard of taxpayers actually doing that but it was enough to scare people from adopting Utopia in my city.

Ultimately what brought fiber to my city was covid. The monetary tsunami unleashed by the federal reserve and the government made it possible for a local private company to build a fiber network. I just got hooked up last week. It is great.

The other thing that happened is that the wireless companies offer high quality internet service. For a time I had both Comcast and T-Mobile home internet. I regularly got speeds of 1.2-1.4 Gb down on T-Mobile (100 Mbps up). Verizon also offers home internet but it was too unstable when I tried it.

All this is to say that I have three high quality Internet connections at my home (fiber, T-Mobile, and Comcast), one marginal connection (Verizon), and one garbage connection (Century Link; however, they have also been installing fiber in the area). Competition is great!

With active its just the last mile that is not shared, everything after that is shared. So it can suffer from the same oversubscription problem, just depends on how good the ISP is. But adding more capacity on those shared links is easier than fixing capacity issues on a PON deployment.

> Ultimately what brought fiber to my city was covid. The monetary tsunami unleashed by the federal reserve and the government made it possible for a local private company to build a fiber network.

Same. Federal broadband funds led to fiber deployment here. We went from one ISP to ~10.

We went from 40Mb up w/ 40ms latency for $119 to 1Gb up w/ 4ms latency for $49.

Over 20+ years, I've read dozens of articles about major ISPs receiving government money/incentives (billions in total) to improve consumer options.

This is the first time I've witnessed an internet company keep that promise. Predictably, it wasn't a major ISP.

Verizon home, where available, isn't too bad... it's definitely a good backup connection for WFH scenarios.

T-mobile business has a backup plan for $30/mo ($15/mo on a promo a month ago) that can be used for up to 7 days of internet per month. I signed up for it so I will still have internet if my fiber goes out (my router will automatically switch over in an outage).

Yeah, Verizon is similar... I really need to setup the automatic failover on my router, I'd been unplugging one an plugging in the other for the couple times I've needed it.

Aside, really want to run a script to reboot the cable modem daily... it seems to build up a number of errors on a couple of the channels, but a reboot has it working fine again. It slows down a bit while it happens, and I only notice when streaming starts buffering a lot.

I also have UTOPIA and I'm getting 1Gbps up and down for $50 a month. And I think there's a slower option (250 Mbps or something like that) for under $40. Works like a dream. My only complaint is that my ISP (XMission) didn't add IPv6 support until relatively recently.

I live in the bay area and just got sonic fiber to my house.

49 bucks a month for 10gbe.

I have only seen it get up to 8gb in TESTING... I might be able to get to 9.

That is a stupid amount of bandwidth.

I give it 6 months before I get a call and they fire me as a client. Cause im going to try to use the hell out of it.

I LOVE Sonic. I also have the 10Gb internet for $50/month and I can’t even use that much bandwidth. Also their customer service is amazing. I was caught off guard when I called in for an issue and a local person answered when I expected it to go to a machine. They also support net neutrality.

> I can’t even use that much bandwidth.

I built an opnsense box (part of my proxmox cluster) and passed through a dual 10gbe nic. That whole PC was cheaper than any off the shelf routing solution I could find.

The rest of the network is going to take time... I suspect that im gonna end up with a few fiber runs in the house before all is said and done.

What networking gear do you use inside your home? What router? Do they terminate an actual fiber connection (SFP+) to your home or you get 10GB Ethernet?

You get a 10GbE port off the ONT.

Sonic 10G rocks. I use Asus RT-AX89X router, with 2 10G ports, one input from Sonic, one output to 10G switches (Ubiquiti Networks UniFi Switch Flex XG 4-Port 10G).

I’m not the op, but I can recommend the Unifi link aggregation switch and cheap Mellanox connectX3 cards. They are about $20 and are amazing. Fibre runs aren’t expensive but the transceivers are so copper DAC is good for short runs (as the connectors are built in).

Thunderbolt sfp adaptors also work.

You can thank early HFT and other early customers for those juicy cheap Mellanox cards. They used to be thousands when they first came out.

Fibre transceivers are not expensive. You can get 2km 1310nm single mode 10G SFP+ modules for US$15 here: https://www.fibermall.com/sale-420055-sfp-10g-ir-1310nm-2km....

Their ONT terminates to 10gb copper.

I bought a dual nic intel card off amazon (its from china) works a treat. There is an OK switch after that to get the network working...

It's going to take me. a minute to build out the rest of the network. Switches and cards are NOT cheap for 10gb. I'll probably end up running SFP+ fiber to a few places around the house and putting in a drop or 3 for some 10gbe love in other rooms.

I once called Sonic on a weekend because my modem kept rebooting. They answered immediately, and the debugging felt like magic.

"My modem keeps rebooting."

"Put the modem to your ear. Is it hissing?"

"Um...? Yes. Yes, it's hissing."

"Bad capacitors. We'll send you a new one."

This was during the Capacitor Plague https://en.wikipedia.org/wiki/Capacitor_plague , so clearly he'd seen this before. Still, fastest support resolution I've ever experienced.

Sonic is reselling AT&T fiber in most situations, but they provide their own support which is nice if you need it. If the problem is with the fiber, you are still relying on AT&T.

That doesn’t appear to be the case any more.

https://forums.sonic.net/viewtopic.php?t=17580

Nope, they have laid a ton of fiber. Not giving AT&T a dime and loving the fast connection.

Not "most situations" for years and not at all for a bit now. AT&T fiber reselling in an area was always a stop gap until we could build out our own offering at that locatuon and offer a better and cheaper product of our own which is more profitable for us to people that were already our customers, which is a win for them and for us.

Any plans to resell AT&T fiber in locations where you don't have service? (Asking not-so-transparently if you will come down to San Jose where I live…)

That was the plan until AT&T pulled the rug out from the resellers, so it's not an option anymore, as much as we'd like to.

That said, I believe we are in San Jose, just mostly for aerial (telephone pole) locations. Below ground infra is much more expensive to serve, and historically we've focused on aerial for that reason, but I believe plans are moving forward to fill in the gaps in the areas where we already provide service with micro-trenching, so hopefully we'll be able to offer you something in the future.

I think our fiber comes from telephone poles. Would you be able to run another line on those?

Not an individual line, but if the poles aren't full then theoretically we could serve the area. Until the poles are full its free to use with an accepted usage plan and engineering plans to show pole stress and load/tension, but each tenant has a section with a buffer between them and there's only so much room.

If you want to contact me through my email in my profile) and are willing to accept that my crazy personal inbox situation might mean I easily miss it), I'm willing to do some closer looking with your actual info.

It might not be "most situations", but its certainly mine. I've been filling out the sonic form yearly for at least 5 years now. I'll say i do appreciate that it stopped offering me verizon. It always felt like a bit of a con for the frontpage to advertise gigabit, ask for my address, then give me a "success" page that only offered verizon. Unfortunately i suspect that with verizon not being offered it means sonic has decided this area isn't worthwhile and has stopped accepting customers (not going to fully out myself, but we have a bart stop and are not end-of-line).

Depending on where you live, if we already serve some houses in the area (very possibly if we're advertising locally), then it's likely that your specific location is more costly to serve and is being delayed until we reach that phase of the rollout for the area.

Specifically, aerial deployment through telephone poles is an order of magnitude cheaper than underground unless you get in when the neighborhood is being built (which AT&T and Comcast often do, and get pre-wired to each house). My understanding is that we are trying to fill in the areas between that aerial neighborhoods we initially built out where we provide service, it's just slower. If the availability form shows service for houses in your city that seem to have service delivered through telephone poles then it's possible we'll get to you in the semi-near future.

If you live in the bay area have service through a telephone pole and we can't service you, then I'm not sure what's going on. My understanding is we've covered most the bay area, but honestly that's not my department...

This was in reference to Silicon Valley

Yes. Sonic is not currently selling AT&T at all anymore as a new service, but we have existing customers on that product which can continue to use it. Where we are currently selling fiber, in much of the bay area and in Los Angeles now, it's our own product. Also, the vast majority of our current customers are on our own product and not resold AT&T fiber or any resold service from another provider.

Sounds like XGS-PON which is limited to around 8Gbit/s and those are shared with up to 64 other customers. If you constantly use it at prime time they could get angry way sooner when the other customers on your PON tree start complaining.

Sonic (I'm 95% sure that's your provider) won't ever bother you. Bandwidth is cheap once the lines are in.

Got a sub $700 Mac with 10Gb port just to test their redunk speeds.

The future is fun

Nah, they don't do that. You can even run a server on it if you want. Or that used to be the case. Email the CEO, if you'd like. He does respond.

As long as you aren't consuming most of your bandwidth through abusing one of our servers used to provide ancillary services like email or or a website we host (and you would have to abuse it likely to do so), I sincerely doubt anyone here would even notice.

You don't deploy a 10gig network to consumers as your only speed tier and design so it can't be used. Even if there ended up being problems if a few people used their full bandwidth (again, there shouldn't be), it's uncommon at this point that people have the hardware to be able to and if they do that it would be more than a brief burst. Honestly, your problem will be getting the other side to keep up.

What hardware do you use to not immediately down step the bandwidth coming off the modem?

When I subscribed a few years ago there was no reasonable home network hardware that exceeded 1gbps so I finally gave up and “settled” for the lower speed…

We’ve had symmetrical 1gb fiber owned and installed by our public utilities for almost two decades here in central Washington. Was $80 a month until a year or two ago, then it went down to $60.

Just got gigabit in Knoxville for $65. Which is actually a tough sell since AT&T already offers 300 megabit for $55 and I’m not sure what I’d do with a faster connection.

But $65 from AT&T would only get me 500 megabit, so for people who do want faster speeds the KUB option is looking good.

There's more to a connection than speed, there might be data limits, there might be latency differences and it might just be unavailable or expensive.

I'm not from the US, but if it's somewhat economically viable, go for the underdog.

Speed is consistently higher than advertised (330-350ish), latency is fine, 1 TB/month (it says this, but then also says X GB of unlimited used), and I don’t think I’ve had an outage in the two years that we’ve been here. I’ll be considering it still (just became available last week in my neighborhood) but since I work remotely I have to be risk averse on switching internet providers.

In the US it’s also not uncommon for connections to be not only asymmetrical, but downright lopsided. Comcast sells “gigabit” in my area that’s 1.2Gbps down but only like 20Mbps up which is just silly.

Thankfully a competitor offers true gigabit for about the same price, and so that’s what I subscribe to.

AT&T currently doing 390 down / 340 up on my symmetric 300 plan, can’t complain

Except you are paying the company that built the NSA's domestic spying apparatus, paid to flood the FCC with millions of fake anti net neutrality comments, and paid Michael Cohen $600k to fight net neutrality and anti-trust.

Other than than that, totally great

That is probably not a terrible configuration for a lot of users. It is disappointing that the internet didn’t remain a network of mostly peers where everyone might host something. But, realistically, lots of people approximately just use their bandwidth to stream video, right?

Netflix themselves say that 15Mbps is enough for 4k.

1.2Gbps downstream is enough for >50 such 4k video streams without saturation -- something that will probably never happen in any household, even in Grandma's house on Christmas Eve with a bunch of grandkids around.

Meanwhile, I've had single-player games of Factorio of medium complexity that would saturate 20Mbps upstream.

Nobody needs a connection that is so heavily lopsided as 1.2Gbps/20Mbps is except for the marketing department, and the DOCSIS sunk-cost-fallacy department.

Why is single player factorio phoning home so heavily?

It's about video streaming. The challenge with Factorio is that there is always lots of very tiny little changes on the screen from all the tiny boxes and trays of stuff moving around on belts, which means if you want acceptable video quality, you need a lot of bitrate as the usual h264/h265 compression magic of just encoding differences between frames doesn't work out any more - too much change between frames, you're (effectively) forcing the codec to transmit full frames all the time.

Strange that when one talks about a particular application using a lot of upstream bandwidth, we are meant to assume what they actually mean is that a separate application responsible for recording the UI of that first application and uploading it to a server uses a lot of upstream bandwidth.

Perhaps I wasn't clear. It isn't phoning home to Mother like a Windows box might. :) Factorio runs, by default, completely locally.

I was using Factorio, as a single player, with the thing split into two parts: A server component, and a client component.

The data betwixt my headless server and my client was just getting too heavy at times for the 20Mbps upstream of that server.

Why so much data? Thousands of bots, constant wars, lots of things being moved around, interplanetary mining happening on a decent scale... The data transferred gets bigger as the game becomes more complex.

"But, like -- why didn't you just play it remotely like everyone else does using Steam's remote play or something?"

The latency of Remote Play sucks in ways that are distracting to me. The h.264 video sucks and has artifacts, presumably because of tradeoffs made to improve latency. My 10Mbps upstream at home works OK with Plex and hardware encoding where latency isn't important, but it doesn't work well at all for me with Factorio.

Splitting Factorio into two parts (client and server) eliminates video encoding, but it isn't completely without cost.

"But you said you had a server with a 20Mbps connection. What happened to that?"

I do. But it's under my desk at work (as is tradition), not at home. And that box doesn't have enough GPU oomph to deal with rendering Factorio.

"Right. OK, weirdo. So why aren't you just playing Factorio at home on your presumed gaming PC, then?"

Sometimes, I'm not at home or I just don't want to sit at my desk. My laptop has enough GPU to run Factorio smoothly as a client, but it does not have enough CPU to keep a moderately-complex game running smoothly on its own. And that's where splitting the game into two parts becomes worthwhile for me.

(Or, you know: Maybe I just want to play it online with a friend, like a normal person might -- but someone still has to run the server, somewhere. Some options include running that server on available hardware with available connections [for "free"], or renting hardware from Hetzner or something [not even a little bit free].

The point is: I've found that it sucks running a Factorio server with a 20Mbps upstream, even with just a single player, and thus a top-flight residential Comcast connection won't keep up. If I had 5 friends who wanted to play instead of 0 friends, that problem would be even more pronounced.

I mean. It's 2024, my dudes. What's up with this 60:1 down:up ratio?)

Interesante. Is this server/client mode something factorio supports natively or is it a mod of some kind?

Well... I mean, it's Factorio. Everything including the base game is a mod. :)

Silliness aside, it's a built-in part of "Multiplayer," which is included with the base game. But my late-night brain refused to use the word "Multiplayer" in an instance where only a single player (me) has ever been involved.

Like Quake was back in the day, online/multiplayer play of Factorio requires servers, and those servers are all run by regular-ish people. Some servers are public and are easy to find (and as with Quake, sometimes this is unintentional), and some [like my own] are private, unlisted, and don't even exist on the public Internet.

> Netflix themselves say that 15Mbps is enough for 4k.

Aggressively compressed 4K, it's convenient for them to have convinced everyone that resolution is the one measure of video quality.

4K content on blu-ray UHD is H265 at 72, 92, 123, or 144 Mbps.

It would be a lot more data than Netflix will send you, but still doesn't saturate a fiber connection.

Is that enough for just the TCP ACKs?

Can you even achieve 1200 down with only 20 up? Is it enough bandwidth to send all the ACKs back and get full speed down?

It depends on the window size. A pure TCP ACK packet is going to be 64 bytes (the minimum ethernet frame size). If you were to ACK every incoming packet, that would then mean the incoming packets have to be at least 3804 bytes, which puts them in jumbo frames territory, which are unlikely to work over the Internet.

However, you only need to send an ACK when the TCP window is full, not for every incoming packet. If the TCP window is set at 64k, for example, and assuming an MSS of 1460, then you can receive a total of ~66K bytes for each ACK you send ((64k/1460)*1500, assuming the provider counts every byte in every packet, not just the TCP payload). The window size can actually scale much higher though, up to 1GB. So, in principle, yes, you can comfortably get this much download speed with only this much upload speed.

Of course, this would only apply for downloading huge files with a very efficient implementation. If you are using HTTP to download lots of small resources, your uplink would be saturated with GETs far before you saturate your downlink.

That's pretty much all DOCSIS providers worldwide. I'm not sure if it is a technical limitation or just what the customer wants, but you pretty much always see ~ 1/5 ratio <200mbits and a 1/10 ratio >200mbits. So, 5 up / 20 down or 60 up / 600 down.

It is a technical limitation. Depends on number of shared downstream and upstream channels in a given network.

Its not just the US as most last mile standards are not symmetric because they only use one line for both directions and use different frequencies for each side (most DSL, cable, PON). So the spectrum available is limited and since most users will benefit from more download rather than upload the split is not even by choice. But like in your example the ISP can sometimes tweak how much is available in each direction and you end up with those insane differences.

Where I live we mostly have symmetrical ethernet fiber, and some cable. We started using fiber long before PON came out and it never became popular here so pretty much all fiber here is 100/100, 1000/1000 or even 10000/10000.

> I’m not sure what I’d do with a faster connection

I have a 500 mbps and a 1 Gbps fiber into my home. For redundancy (WFH 100%) and because it's cheap (Romania).

The only time I notice which one I'm on is when I install games from gog/steam/psn. Even operating system isos (which are only a few gigs, not 50-100 G like modern games) don't make a dent any more. Streaming is a-ok, but I'm stuck in the last century with an 1080p TV.

Both of my providers^H^H^H one of my providers now offer 2.5 Gbps I think. And I wonder why :)

Meanwhile over here in Bellevue, home of T-Mobile and (close enough) Microsoft, I’m stuck with Comcast or really crappy DSL. How is it that central WA gets way better internet than Bellevue? Oh right. Comcast lobbying.

The big telcos ignored us so we took it into our own hands. Had it been a bigger market I’m sure they would have lobbied our local ISP out of existence. They still do some questionable marketing around here.

This reminds me of Longmont, Colorado's city-managed ISP (NextLight).

They fought the Comcast / Verizon duopoly for a while before founding it. Turns out when good people who live next door are the ones responsible for your basic utilities, you get better, faster, more reliable service with better customer support.

Who knew.

I used to work in one of the participating ISPs. The organization of Utopia was a PITA, but the idea was really good and generally well implemented, with some portable exceptions. My biggest issue is that they take huge subsidizations from the state to deliver internet, then turn that around and outbid competitors for state circuits using the subsidized rate. Not really saving them any money, but it looks good on a quote.

Good on them. They are using the same tactics ISP's have used to screw the general public but instead they are at least providing the services they promise.

Official release:

https://www.utopiafiber.com/2024/05/08/utopia-fiber-adds-thr...

Utopia Fiber:

   1Gbps for $45/mo
  10Gbps for $150/mo

AT&T Fiber:

   1Gbps for $80/mo
   5Gbps for $250/mo  *** note: slower speed

I think there should be a $30/month charge for Utopia added to their service.

https://www.utopiafiber.com/residential-pricing/

So it's slightly cheaper than AT&T for 1Gbps and a more significant difference for their 10Gbps service which AT&T doesn't offer, as you point out.

> So it's slightly cheaper than AT&T for 1Gbps

Utopia is:

  77% cheaper than AT&T (1 Gbps)
  66% cheaper than AT&T (5 Gbps) yet delivers 2x the speed than AT&T

That doesn't seem "slightly cheaper" to me.

You're using the wrong pricing. AT&T has one price. Utopia has 2 components, the ISP and Utopia itself, making up the total price. If you go to the website link I posted, it's not $45, it's $45 to the ISP and $30 to Utopia. That makes $75, which is slightly cheaper than $80.

"Service from UTOPIA Fiber is billed in <two parts>. One part is for services provided by the Internet Service Provider (ISP) you select, and the other is for the physical network connection provided by UTOPIA Fiber. Check out our price list to help you compare service options from our various ISPs. "

https://www.utopiafiber.com/residential-pricing/

Emphasis mine. The cheapest service available from Utopia is $30 to them + $35 for 250Mbps service, totaling $65/month. There is no $45/month service.

How have you calculated that price for 1Gbps? I'm looking at the numbers and it seems like the cheapest 1 Gpbs connection would be $70/mo?

In Singapore there's a telco intending to launch 10 Gbps fiber for 30 SGD (~22.35 USD) per month. likely to be asymmetric though, since their 2.5 Gbps service at ~16.24 USD per month is advertised as 2.5 Gbps down and 1.25 Gbps up.

The things you could do with 1.25gbps. Crazy stuff. Vastly different infrastructure/density situations but here in Australia I'm paying ~56USD per month for 0.04gbps up.

Australia has Abbott and friends to thank for fucking the whole NBN plan up royally.

We would have had fiber to the premises everywhere and affordable gigabit by now, but Abbott decided Australia shouldn't spend money on what he saw as merely a "video entertainment system" and binned it in favour of the current Mixed Technology mess.

Which then actually cost more to implement.

Pretty sure they intentionally screw it up. It means that the populous will generally think that the government can't provide effective utility at cost and so private equity can move in and provide equally bad utility at monopoly rent prices.

There’s a bunch of stuff there but I think they were also trying to protect Foxtel so Murdoch could avoid having to compete with Netflix etc.

Assholes all round

I'm getting symmetric 10 Gbps for 49CHF (~54$) per month in Switzerland, and this is one of the most expensive countries in the world.

Besides Swisscom, internet is very reasonably priced in Switzerland. I had 1gbit symmetrical without contract or install fee and free equipment (including a standalone fiber to Ethernet converter) for 49.- a month.

Many cities or communities have built a municipal fiber network that any provider can lease and they’ve also wired up all buildings (inside) at the time. Only Swisscom (owned by the govt) is doing everything they can to get in the way of these municipal rollouts and keep the prices high.

I'm in NYC and I have a local ISP called Honest Internet that gives me gigabit for $50/mo. Comcast/Xfinity, Verizon, Spectrum all wanted $90/mo for this. I also have a direct email address to get support rather than some useless "chatbot" and endless phone queue. Support local ISPs.

Cox is more than that in Phoenix... I've got 2gb down (100mb up) currently. I do wish they didn't charge 2x or more to host a server at home though (business account).

> I do wish they didn't charge 2x or more to host a server at home though (business account).

How is this enforced? Is it an IP address (instead of NAT) you're paying for?

The cable company blocks common inbound server ports for residential connections... I've considered just using DNS for letsencrypt and a non-standard port for http/https, but would rather not.

For less than the cost difference, I have a good sized VPS on OVH that I use.

Our CenturyLink internet in Salt Lake City went down and it took them almost two weeks to come fix it. These companies have done a tremendous job of driving demand for municipal broadband with their unconscionably terrible service.

> Two decades later and the coalition just announced that 18 different ISPs now compete for Utah resident attention over a network that now covers 21 different Utah cities. In many instances, ISPs on the network are offering symmetrical (uncapped) gigabit fiber for as little as $45 a month (plus $30 network connection fee, so $75). Some ISPs are even offering symmetrical 10 Gbps fiber for around $150 a month:

Congratulations, you've invented European style internet access with prices that almost match, super proud of you Utah.

Personally our local fibre network was built as a joint venture between 8 or 9 companies who then openly compete to run the service to customers so its mostly down to customer service and perks.

A 1gb symmetrical unlimited connection (sold as 900mb due to some legal shenanigans) is about $38 with a 250mb package unlimited going for $24

For however conservative Utah is, they are very community oriented so in practice, they seem to implement some fairly socialist policies. They were real early in housing first and they have a pretty nice built-out public transit system, a bunch of coops, 3x the average of credit union membership ... things like that.

There’s a communitarian-libertarian axis that’s distinct from the conservative-liberal axis. While libertarianism tends to dominate both the left and right in the U.S., I suspect most American conservatives outside the coasts are more communitarian, closer to Christian democrats: https://en.wikipedia.org/wiki/Christian_democracy. Conversely, while American liberals have putatively communal ambitions, they’re often too fragmented and individualistic to accomplish them.

If you're referring to the political compass, that thing is just an astrological star chart for politics. It's just about as accurate.

Go to Wikipedia for that, about 70% of the word count is calling it nonsense https://en.m.wikipedia.org/wiki/The_Political_Compass

Those are the numbers you get with topics like homeopathy

I don’t know what the political compass is. I’m talking about polls that show a large share of republicans support things like social security and Medicare expansion: https://nymag.com/intelligencer/2018/10/poll-a-majority-of-r....

Thanks for the clarification

> Congratulations, you've invented European style internet access with prices that almost match, super proud of you Utah.

Isn’t Utah like 3M in an area the size of Western Europe?

It is an accomplishment considering

No, Utah is less than half the size of Spain or France. That's vastly less than the size of Western Europe.

So, while it's an impressive accomplishment, it's not "size of Western Europe" impressive.

Semi-rural residential Texas Triangle needs this. Some places have Spectrum (Charter) and nothing else. Google Fiber is fantasy outside of Austin. One notable exception is GVEC. https://www.gvec.net

Google Fiber is a fantasy for some people in Austin too. I was right outside their initially planned extent, literally by a block, and they have never bothered to extend out to my subdivision despite being within Austin limits.

I got AT&T Fiber instead and it has worked great for a decade. I just upgraded to their 2Gbps because they gave me a pretty solid deal on it.

Yep. I had 1 Gbps for 2 years and 2 Gbps for 2 years. (Can get 5 or 8 Gbps, and possibly 20 with a beta program.) GF excludes most of the downtown core except for Google buildings. I've speculated it's some sort permitting cost or dark fiber cost issue. I don't understand the logic they use to select cities or wire some customers but not others while leaving holes randomly. https://broadbandnow.com/Google-Fiber

AT&T Fiber is expanding quickly in North Texas. Got 2.5 gigabit a couple years ago to my suburban home in the edge of DFW where it transitions to rural. Their headquarters are in Dallas so they are keen on expanding locally.

Makes sense in the Metroplex where there is already tons of dark fiber from the dotcom days and right-of-way already in place from cable, DSL, and other distribution networks.

Meanwhile I live 20 minutes from Google's heardquarters and I can't even get more than 25 Mbps uplink.

Maybe these big internet companies should dig up all the infrastructure in their home turf and replace it?

But they are evil so why will they do it?

They're not allowed to. Regulatory capture is running wild.

They're fucking Google. If they dig the streets up and give me 10 Gbps fiber I will be more than happy to show up to city council and yell the shit away from any NIMBY old fart that says no.

It’s more that CA law makes it easy for comcast to tie it up in the courts forever, regardless of what your city council says.

And more NIMBYs will show up to scream about Google overrunning local government.

It is extremely easy to make hugely powerful corporations the villain, for obvious reasons, and this doesn't stop being true in the case where they want something actually good.

Sorry to have creeped into your bio, but what is your bandwidth in Munich?

And yet all these mega corps can spend 100s of millions on lobbying and get favorable law passed or regulation taken care whether labor, FTC or taxes or EPA .

The local council is apparently the last bastion of democracy that can ward of influence of their largest tax payers and the wealthiest residents who are their employees

The notion that any ISP has more pull than Google, Facebook, and Apple in the Bay Area is facially absurd. Cupertino literally has a deal with Apple where it pays back 1/3 of Apple’s sales taxes: https://appleinsider.com/articles/23/11/04/california-wants-...

That’s Cupertino. MV and Google have had much more adversarial relationship going back more than a decade. Look up their failed free wifi project as an example of this

People really need to cool it with the 2024 buzzphrase "regulatory capture".

I don't know which podcast, and I'm almost certain it is some dumbass podcast somewhere, is promulgating the phrase which has been popping up with ludicrous regularity in 2024 to describe every situation where someone is pissed off about something, but the primary reason people don't have 10gbps internet isn't regulatory capture, it is that the deployment costs are too high unless either subsidized by the government (as is the case in many parts of Europe and this instance) or there is a combination of high density and eager market.

"20 minutes from Google's Headquarters" is San Jose. San Jose is served by over 20 ISPs offering services from wireless to 10-gig fiber depending on location. The locations aren't determined by "regulatory capture" they are determined by a cost-benefit analysis.

You, yes you, can start your own ISP anywhere in the country, right now, today. Anywhere.

Remember 20 years ago when NYC gave Verizon exclusive franchise rights and then everyone got upset that they didn't fulfill their end of the bargain? That's gone. Everywhere, it's gone. Anyone, including you, yes you, can get a permit to microtrench fiber to any building in NYC unless there is excess capacity already in which case the party holding the excess capacity MUST sublicense it to you.

Same for California. Same for any state.

Rural Texas? Vexus Fiber is diggin shit up in Tyler TX. Middle of nowhere WA? Ziply is taking the risk.

Doesn't matter if it is the densest or most historically protected city or loneliest prairie town you, yes you, can start an ISP right now, today.

Reforms to the Pole Attachment Act have made this shit is easy now. Easy, but not cheap.

Just gather $100 million so you can pay the $30k-50k per mile (due to material and labor NOT "rEgUlAtoRy CaPtuRe") to microtrench or pole-attach, get all of the permits to trench on public rights-of-way and negotiate the few (if any) non-public crossings you have to make, post the bonds needed to ensure that if you go bankrupt mid-construction there's cash to clean up YOUR mess, hire the network/civil engineers needed, pay for peering, staff an office and support center, and you, yes you, can start your own fiber ISP--today.

Hell, many states will give you money to offset the costs.

If I had $100 million I could wire all 2,300 homes in my Maryland suburb with fiber and no regulation could stop me. If I had $100 million I wouldn't waste it on that (we can already choose from multiple gigabit+ ISPs) I would retire.

> "20 minutes from Google's Headquarters" is San Jose.

It's also Burlingame, San Mateo, Fremont, Union City...

Or, if the traffic is bad, it's Redwood City, Menlo Park, Stanford, Cupertino...

Or, if you're taking public transit (or walking), it's pretty much just Mountain View.

It doesn’t matter in which cardinal direction you travel for 20 minutes— regulatory capture ain’t stopping anyone from starting their own ISP.

> Reforms to the Pole Attachment Act have made this shit is easy now. Easy, but not cheap.

AFAICT, the pole attachment regulation changes only happened in mid-January 2024. They're probably good changes, but

1) They happened just yesterday, speaking in "policy time".

2) If you don't think an absolute assload of time and money can be wasted with doomed-to-fail appeals and spurious orders for documentation and impact analysis and the like, you've spent all your life in a shockingly well-run city (or maybe a place with next-to-no municipal services and associated power grabs).

> It doesn’t matter in which cardinal direction you travel for 20 minutes...

Well, it does, actually. See below for commentary on your particular selection of SJ as your exemplar city.

> ...regulatory capture ain’t stopping anyone from starting their own ISP.

And yet for some weird reason, there are precious few non-telco/cableco ISPs in places like San Francisco that are doing something other than reselling rotting DSL, or running small microwave dishes on rooftops (that is, being a WISP).

San Jose may be remarkably well-managed (though, I expect WISPs make up the lion's share of your 20 ISP figure (and they're playing on easy mode in regards to permission-seeking)), but other places in the Bay Area are definitely not.

I've been living in SF for more than a decade, and I've watched Monkeybrains notice that building a city-wide fiber network was completely impractical and resorted to being a WISP. I've also watched Sonic's fiber installation program in the city stall for roughly a decade. While I'm glad that starting in mid 2019 [0] they've been able to string and light up fiber to not just small sections of the most western portions of the city but a significant fraction of the city, much of Sebastapol got lit up quite a while back. You'd expect a big city with a high density of customers would get lit up before a fairly small, sparse town if it was just a cost/benefit analysis doing the location selection.

Whatever you want to call it, the political and regulatory situation on the ground is obviously getting in the way of folks providing civilized-speed data transit services to potential customers in many major cities. You don't want to call it "regulatory capture"? Okay, sure. The on-the-ground political resistance and obstructionism is still rooted in the same sort of backroom dealing and under-the-table kickbacks, so you're just using a different phrase for the same behavior.

[0] Notice that this substantially predates the pole attachment rule changes that you've brought up.

>You'd expect a big city with a high density of customers would get lit up before a fairly small, sparse town if it was just a cost/benefit analysis doing the location selection.

The costs are so high that a reasonable person would never expect that, not without government subsidies. Like what's happening in the small towns.

Nothing is stopping you from starting an ISP-- except the fact that you'll go broke because most people are happy with Comcast. Sorry if I didn't make that assertion clear enough in my original comment.

That's why Google Fiber pitted cities against each other in mortal combat to see who could get the most people to sign loyalty pledges promising to subscribe and they turned on the electrified subsidy extraction machine to "high" before making a deployment decision.

> The on-the-ground political resistance and obstructionism is still rooted in the same sort of backroom dealing and under-the-table kickbacks

This is not true. Literally and actually you can start an ISP and run 100-gig fiber to every single block in San Francisco and no government or cabal in a smoke-filled room can stop you. Only your creditors can.

> That's why Google Fiber ... turned on the electrified subsidy extraction machine to "high" before making a deployment decision.

If by "subsidy extraction" you mean "Required that the only objection incumbents could make to GF's pole attachments were objections rooted in real safety issues, that all objections must be resolved rapidly, and that incumbents could not charge GF outsized pole attachment fees (IIRC, often defined as 'not more than the incumbent had been historically charging themselves').", then, yeah, I agree they did demand a level playing field free from unreasonable interference from incumbent operators.

> That's why Google Fiber pitted cities against each other ... to see who could get the most people to [promise] to subscribe...

Yeah, that's part of any big, expensive rollout. You don't do it unless you have enough people who commit to use the service for a period of time to make it worth doing the rollout.

> ...most people are happy with Comcast.

The difference between the rates and restrictions imposed by the incumbent ISPs in areas where there is healthy ISP competition and areas where there is not suggest otherwise. Even credible rumors that an independent ISP might serve a significant portion of an area with a mon- or duopoly cause the incumbents to scramble to offer freebies, rate reductions, and gratis data transfer cap and speed increases.

Given the all of the previous, and that the FCC continues to update their pole attachment regs to add more avenues for appeal and impose more and more "playing field leveling" rules, I suspect you know less about this topic that you claim to.

You live here - you know that that the NIMBYs would never allow that. It took 20 years to run a power line to electrify Caltrain.

The last NIMBY lawsuit on electic caltrain was settled back in like 2017. Hate to break it but that’s just your standard Cali corruption

The actual NIMBYs here are those who can afford lobbyists and lawyers - might not be who you're imagining.

This came up at lot at google when they launched fiber. CA regulation and environmental laws make it to easy for the monopolies to stop or slow you down.

If you look at how they pick cities the main criteria is friendly regulations like “one touch make ready” requirements.

Meh, they could have bought one the monopolies if they wanted to, like they did with Motorola when it is was in their interest to do so.

It is a matter of will and persistence to run a this kind of business, Google does not have the DNA to run a serious enterprise business that needs decades of commitment to a product .

In 2012 google paid like 12B for Moto, comcast had a market cap of like 80B

Google did a $70B stock buyback last year. Every year they are doing similar numbers for the last 5 years.

In 2012 they had $50B cash on hand. Today despite the buybacks every year they still have $100-150Billion cash on hand, while Comcast is only worth $150 Billion.

They could absolutely afford it now and most likely 10 years back as well without even considering financing the deal.

---

Google is just very bad at acquisitions, they simply are too risk averse and don't have the DNA.

Compare with a peer like Microsoft in the same time frame: bought Linkedin($24B), Github ($8B) Activision($70B) invested in OpenAI ($10B), or Facebook : Instagram($1B), Occulus($3B), WhatsApp($19B).

Mandiant is the only acquisition of note since Motorola even that is pretty small at $5B.

> Google is just very bad at acquisitions

Youtube, Android, Where2 (== Google Maps), Writeley (== Google Docs), reCAPTHA, Fitbit ?

We were talking about last 10-12 years

Any reasonable observer wouldn’t compare the first 15 years of google to the next 15.

Also the examples(baring fitbit ) you cite were great product ideas but fairly small companies, it takes a very different kind of DNA to assimilate large organizations and still be somewhat functional, this is what a comcast acquisition would be .

Microsoft or even SAP have practiced this for years Google has in the last 10 years done of note only Fitbit and Mandiant you could maybe consider also Nest in outside edge of the window in 2014 , all small bets for google’s size

Apple is also similar, both companies believe they will do better doing everything in-house rather than acquire expertise in areas of interest (fiber, cars etc ) so far they have failed and now are content in giving money back to shareholders rather than find useful areas to invest .

I think fiber has worked exactly as planned

all the technology in the world ain't gonna fix our social issues.

Oh, I think we’ve seen it make the issues a lot worse.

Do you live in a donut of Sonic coverage? There definitely is Sonic 10gbps service in Mountain View.

Ugh, it's a little annoying that I cannot get faster than 2 gigabit internet in NYC, and it cost me like $120/month. 2 gigabit is more than fast enough for pretty much anything I do, especially since most of the stuff in my house is on WiFi anyway, but I did get my house completely wired for 10 gigabits about a year ago to try and future-proof everything.

It's frustrating that NYC is a super expensive place to live, and it's behind the times with internet speeds.

Worst internet I ever had was at a Cupertino apartment…in 2013. Walked a mile to the office where we had blazing fiber. Manager said a lot of the valley outside of company offices was way behind cause of the cost of installation. Hope it’s improved.

There are parts of NY that are getting 8 gig fiber apparently, but it doesn't appear that Brooklyn is one of those parts, which annoys me. 2 gigabits if fine, at least for now, but it would be nice to be able to increase the number of concurrent downloads I could have.

Just moved out of Cupertino and the best speeds we could get was 1 Gbps down/50 Mbps up on Comcast. The other option was AT&T DSL. So…not really.

I doubt much has changed. The Bay is lagging in terms of infrastructure — it’s hilariously sad.

My only options currently are Comcast or DSL. I’m anxiously waiting for Sonic to become available.

0.10Gbps down & 0.02Gbps up…

FCC updated just 2-months ago the minimum “broadband” as defined to now be 0.10Gbps up and 0.02Gbps down.

https://docs.fcc.gov/public/attachments/DOC-401205A2.pdf

EDIT: updated typo (on "up/down")

Other way around, 100 Mbps down and 20 up.

The private ISPs also tend to have under the table agreements with each other to avoid competing with one another in some regions. Keeps the prices high while spending the bare minimum to keep the network running (which means low bandwidth and asymmetrical options).

Occasionally you may live in an area with more than 2 options, but that’s rare. I have moved around a dozen times and only recall a single instance where I was able to choose between 2 providers.

I would absolutely love to build out municipal internet. We have plenty of good examples to build off of (Utah and Chattanooga, TN). The only thing holding a majority of us back is: state bans on municipal internet.

This would have the added benefit of creating local high paying jobs as well. Keep the money circulating within the region for as long as possible. Money doesn’t escape via temp out of state workers or corporate bonuses to C level executives

Can you point to any evidence of these "under the table agreements" or is that speculation?

Reading about all these >= 1 Gbps connections, I'm envious.

The fastest I've been able to get quoted to my office in the centre of a UK town is 17 Mbps, and to home it's 34 Mbps.

The speed limits are both due to old infrastructure that's outside my control unless I move. Landlords aren't going to upgrade. In the central library where they have "superfast broadband" in this "superconnected city", it seems to be capped at 10 Mbps per user.

None of those products are worth it - my mid-range phone is faster and gets 20 to 100 Mbps down at different times of day, so I use that for internet connectivity instead. I was surprised at the phone connection speed recently, when I downloaded 30GB of LLMs in 30 minutes on the phone. But the latency is noticably poor.

Lehi, UT opted to build their own infrastructure instead of using UTOPIA. Not sure if that was a good decision, but I'm not happy about the rates they're charging. I'd stay with Comcast if it wasn't a matter of principle.

Lehi offers 1Gbps fiber for $80/mo, which is about the same as Utopia. The pricing seems pretty good for dinner in Utah.

This is such a misleadingly narrow headline. Utopia's a decade old and has been a mess from the business side pretty much the entire time. Even Google fiber in Provo's somehow giving better rollout service than they are. I was excited when it was announced as a teen but it's been a long road of failing to live up to expectations for all but a tiny fraction of our residents in the rich parts of our cities. And it's supposed to be a public-good program!

Provo built a municipal fiber network and then sold it to Google for one dollar, so I'm not sure that's a great counter example

> Provo taxpayers still will have to make $3.3 million in bond payments on the system each year for the next 12 years, even though Google will own the network.

Did I say it's a "counter-example"? Both are operational dumpster fires. I obviously brought it up in a negative tone with the words "even" and "somehow". Utopia's a business, too. Why are you defending them? Do you even live here? They even did basically this same thing to their cities years back, which you'd know if you did.

UTOPIA frustrates me.

I live in a part of Utah where Xfinity is the only gigabit option. Despite a huge number of the UTOPIA providers claiming to serve my area, none of them actually do. My home is in a stupid weird bandwidth dead zone, with 10Gbps options available all around me. The house across the street from me has Google Fiber, yet Google Fiber won't install it at mine a mere 50ft away. I'd be a lot more upset if my Xfinity connection wasn't rock solid. It provides the exact advertised speeds, and in the 5 years I've had it, it's gone down maybe once. Still, I want 10Gbps. It's maddening.

You are this https://www.sunbeltrentals.com/equipment-rental/earth-moving... away from fast internet

> Cheap 10 Gbps Fiber Thanks to Local Governments

> $150/month

Meanwhile I'm here in Sweden rocking 10 Gbit symmetrical from Bahnhof (<3) for 349 SEK/month (~$33/month).

As someone who has 1/1000th of that bandwidth to the Internet, and yet still only saturates that connection when transferring from some nearby sites, I'm curious how useful having a 10Gbps local connection will actually turn out to be in practice. Being able to download Linux ISOs faster is the prominent use case that comes to mind.

Large families.

Is that a Jab at my friends in Utah?

{ Multiple people in household each having multi-party high resolution video calls | band practice | data manipulation collaborations | 4K TV on demand } can suck up the bandwidth when the priority is smooth real-time with low interactive latency.

It's not always going to be peak demand, but it'll be noticable when everybody piles on at particular times of day.

TBH I have relatively low bandwidth most of the time and don't notice download speeds at all as I queue the things I want and return to working on what's in front of me.

You absolutely cannot saturate a 10 Gbit link with today's consumer's needs, let alone with today's consumer hardware.

That's hundreds of Netflix 4k streams.

Sure, but the OP talked about being happy with a 10Mbps link.

I think a lot of modern households might find that constraining, even if a 10Gbps link was massive overkill.

There are two orders of magnitude in between of course :)

Personally I found that going back from a 1Gbps link to ~100Mbps has felt a little sluggish, particularly when downloading games. It's not a huge thing, but being able to get that new game pretty much now was great.

Never download software updates or games?

Band practice?

Edit to clarify: I cannot fathom live collaboration over the internet when even 15 ms lag is annoyingly noticeable (10 ms latency with a midi keyboard seems like the cutoff for me personally at least).

Forget ISOs, those are only a few Gb each. LLMs, on the other hand, can easily go into triple digits at full fidelity.

Video games. AAA titles, even not brand new ones (like Red Dead Redemption 2, Flight Simulator X) can run in the 100GB+ range.

My household of 5 adults maxes out a gigabit connection routinely. Streaming video, cloud based gaming, video calls, downloading game updates, backing things up to cloud storage. It adds up. I anticipate some might argue "that's not right" or "that's not possible, your network is wrong" - all I can say is that it happens and that it'd be less likely to happen with a 10gbps connection.

Assuming they allow server use-cases without booting you off, I could see it being a boon for indie web services. Even fairly old (5-10 years) computers are already mind-bogglingly powerful (I've been playing with a toy myspace clone that can do ~15k profile pages rendered per second including db queries on an 8 year old computer with an i5. Throw caching nginx in front and it saturates my 1 Gbit adapter so I'm not sure what it can do), so if you had uncapped data transfer, you could e.g. try to start up a reddit competitor out of your home (legal risks of user-submitted content aside), and at least on the infrastructure side, could probably scale to 1-10M users on 10 Gbit/s. At least in the US, uplink capacity and legal risks seem like the biggest blockers for techies to be able to bootstrap services from home at more-or-less $0 operational cost and few thousand upfront cost.

I had symmetrical 10GB fiber in Lugano Switzerland 5 years ago. It was ~$30 / month. No commitment.

It's funny. For eons, Americans on the Internet could be relied on to say that the reason they don't have fiber is because their country is big. But their more rural areas get it first and best, revealing the lie everyone tells.

But forget that, these ISPs absolutely rule https://www.brigham.net/residential-pricing/

That's so much nostalgia. I adore it!

Brigham City is not really that rural (more suburban), and it's maybe 0.05% of all suburban / rural areas in the US. Not really evidence that "rural areas get it first and best." Most truly rural areas have absolutely awful internet options, or none at all save for Starlink or satellite.

Brigham City isn’t rural.

The Utopia rollout has been so poorly done. Where I live, the entire area was supposed to get it and they missed houses here and there. Ours is one of them.

Gladly forking out $$$ to Comcast for their coax gigabit until the city gets it right.

Wi-Fi speeds

Most consumers will connect via WiFi.

And that will be typically capped at < 1 Gbps

Wifi 7 is coming and is rated to 11.5-46.1 Gbit/s (depending on channel bandwidth).

https://en.wikipedia.org/wiki/IEEE_802.11be

(In real use probably a gigabit or so)

As a Libertarian, I've always felt that govt should be limited to ensuring (not necessarily providing) essential infrastructure. At this point, internet access as part of communications should be considered essential infrastructure.

I always find it kind of wild how Utah tends to be a Progressive Republican state. Projects like this are definitely nice to see.

Best: Open Pit and Pipe Very Good: ULL Ok: Government Wholesale Shit tier: Any Monopoly.

ijust came here to abe simpson it about att. may they rot in hell

This is another failure for capitalism. Why is it that socialism is working better here?

Utah realizing people need extra VPN bandwidth to watch porn without a drivers license...

I know people disagree with me about this. I certainly hate it that governments gave local monopolies to companies like Comcast

That said, I'm pretty confident municipal internet is bad thing in the long run (compared to actually competitive markets, not local monopolies)

Problems:

(1) Unlike say, water, sewage, power - internet tech changes 10gig might seem fast today. In 15 years it will seem slow. Will the government upgrade? My guess is they'll be slow to upgrade where as places with truly competitive markets will upgrade quick. I watched this happen in Japan. There were 4-5 different providers. One upgraded (twice the speed of competitors for same price), Competitors were forced to upgrade or lose their business

(2) Government run internet is bound to eventually be lobbied to be safe (what do you mean my taxes are being spent so people can download adult content! Stop that! ... etc)

(3) With Government run internet there is no competition so no one else can enter the market and compete.

Come back to this comment and in 10 years. See what speed competitive markets are at and compare to municipal and non-competitive markets.

My understanding is that none of the municipal fiber networks are monopolies, so Comcast and co. still play ball in these markets. If this is truly bad then people will just leave.

In reality I think that enough local governments recognize that high quality Internet is a competitive differentiator when trying to attract business.

> I know people disagree with me about this. I certainly hate it that governments gave local monopolies to companies like Comcast

They didn’t do that. Exclusive franchises have been illegal under federal law for over 30 years: https://www.congress.gov/bill/102nd-congress/senate-bill/12 (“Prohibits a cable television franchising authority from granting an exclusive franchise or unreasonably refusing to award additional cable television franchises.”).

This is a pretty silly comment given the "private" options that could have been at gbps speed for a decade+ are still at DSL and Cable speeds. Blind faith in 'the market' to solve these things after it's failed to do so seems odd.

It has given us starlink. The market makes interesting decisions.

Praise the Market.

Parent's argument wasn't that the existent private market was great, but rather that a public utility monopolizing the market might not be great in the future.

Many of these remote areas are barely being served by companies at the moment because they’re not profitable (enough). Why would anyone object to the people just doing it themselves instead? That way at least they’re not at the mercy of some company that needs to earn a dime.

Objecting to people providing their own infrastructure where private companies have failed to do so, because at some undefined point in the distant future those same companies hypothetically could offer better service is the most worm-brained thing I've read in awhile. If the 10gbps network isn't fast enough in 10 years (already a hilarious premise), the same citizens who voted to invest in muni fiber to begin with could choose to increase their speed/capacity.

Who says they’re monopolizing it?

I’m not aware of any municipal deployment that is a monopoly, and the one in the article is open-access.

> The inter-local agency collaborative venture then set about building an “open access” fiber network that allows any ISP to then come and compete on the shared network.

Comcast et. al can still build all the fiber they want to in these markets.

The market works when there's at least 4-5 competitors. That's enough to make monopolistic backroom deals impractical.

That is what driving me nuts about these threads.

These aren’t free markets. They’re government regulated into being something else entirely, then people stomp their feet and complain about capitalism.

Uh, how did the government forbid there from being more than one ISP in a neighborhood?

The regionalization of ISPs is entirely their own doing to avoid competition.

> The regionalization of ISPs is entirely their own doing to avoid competition.

And what mechanism did they use to do that?

Explicit contracts [1] and implicit collusion [2] [3].

There aren't that many large ISPs in the US. They know that if they compete they have to lower prices and if they don't compete they don't; so rather than compete and all of them accept lower prices (and worse margins) they don't.

[1]: https://www.techdirt.com/2014/08/14/parallel-conduct-how-isp...

[2]: https://arstechnica.com/information-technology/2014/09/comca...

[3]: https://arstechnica.com/information-technology/2016/05/chart...

Ok, so charters, regulation, bribery, local laws, and city policies…

None of that sounds like free market to me.

"Free market" ends the moment government starts issuing corporate charters. We can debate what it should look like, but either way it is an artificial construct at that point.

But that seems more like an US issue? In places which actually allow or even encourage free market competition providers tend to be able to keep their infrastructure up to date while still offering reasonably low prices.

> compared to actually competitive markets, not local monopolies

...and I want a pony!

Last-mile infrastructure tends towards monopoly unless you provide it through the government or regulate it so heavily you might as well have. There is too much cost in redundant buildout to support competition and it's too difficult to tell dumping from price discovery to just fix it with antitrust. Not to mention for the last 40 years the USA has pursued antitrust with less enthusiasm than a decaffinated sloth, so leaning on antitrust would be an idea so terrible that only a corporate lobbyist should ever propose such a thing (and we shouldn't allow that either, but let's talk about one evil at a time).

Last-mile was the "network effects" before "network effects" -- the game theoretic cheat code for capitalism that all the rich investors dreamed about owning. There's a reason why we don't let the free market provide roads or utilities, and that same reason absolutely applies to internet service. This has been proven again and again and again across the nation by letting the game play out and seeing the inevitable result. Why do you think the game will end differently on the 4,952nd playthrough?

Wouldn't (1) create an opening for competitors to enter, thereby negating (3)?

I have exactly one ( 1 ) option available to me in term of internet provider. It’s cox, and it’s sucks.

And I live in a real city (Does New Orleans qualify ? Let’s say yes. )

Where is the free market in my situation ? I would throw money at anything else than cox.

It does exists in other part of the city. Not where I live. I’m within city limit, in a relatively affluent neighborhood.

It’s just weird.

I think the core problem is that these markets often aren't truly competitive. Last-mile wiring is expensive to do so you typically find whole neighborhoods/towns with a single provider.

I don’t think we need to wait 10 years. There are already countless examples of private providers with out of date speeds, unreasonable bandwidth caps, etc and no competitor to place any kind of meaningful challenge.

The article isn’t talking about a “government run” internet, it’s talking about government forcing ISPs to compete. And UTOPIA has been providing faster-than-corporate-internet for 15 years, as the article mentioned. Why would waiting another 10 change anything?

Points 2 and 3 seem imagined, and points 1 and 3 contradict each other. Either the municipal net is slow and will be replaced by for-profit competition, or it’s fast enough to shut out rent-seeking, but you probably don’t get to claim both.

Lol, wow, I'd put 90% of my net worth into a bet against this take if I could. I don't even know where to start, gestures around, reality is here, now.

The top end of residential internet right now is a 2.5G GPON link split 32 ways and sold as 1G which in practice gets you about 300-600M.

10G symmetric being common even oversubscribed to hell is 20+ years out from most metro areas with entrenched IPSs. And for $150/mo is baffling, we could have done this the whole time? There has to be a catch, right? Folks outside the nice areas still don't get better than 100M advertised.

This might be a case of "nobody needs more than 640k of memory", but well deployed fiber has a very very long upgrade path.

As I understand it, switching out optics and equipment on each end for faster versions is relatively simple (vs digging it up and replacing the cable).

If it's active fiber, then you should be able to put 100Gb+ optics and equipment and be good to go.

If it's PON fiber then it might be a little different, but not drastically.

As for the government vs corporate thing... Physical infrastructure is hard to get good competitive environments in. They're pretty rare, and become increasingly so as companies merge or get acquired.

It's very difficult to get new competitors into a market because of how capital and red-tape intense the process is.

Few people are happy with roads and lawns being dug up, existing competitors have a motive to swing legislation to their advantage.

Running physical infrastructure as an open platform that anyone can access is IMO a better approach.

Sorry homie, utopia is ten years old. Check the date. People freaking love it. Still.

Tell me you haven’t read the article without telling me you haven’t read the article.

This article is talking about something that was created 15 years ago. So your “come back to me in 10 years” comment is 5 years too late.

And turns out if it wasn’t late it would have been dead wrong.

You have fully failed to understand the concept that is described here, which is still private internet service, with the government simply providing the open access infrastructure.

Also, thanks to the open access concept the customers have a choice between what appears to be 18 different ISPs. I doubt there’s a single other area in the U.S. that provides that much choice.

These complaints can be levelled against private net companies as well, particularly in the USA it seems.

> Come back to this comment and in 10 years. See what speed competitive markets are at and compare to municipal and non-competitive markets.

The sleight of hand here is amazing - "See what speed competitive markets are at" - the US doesn't really have competitive markets, it has monopolistic markets, and they seem to be a prime cause of the very issues you're concerned about. The article you've replied to is talking about encouraging competitive markets by opening up locally-owned last-mile infrastructure to ISP competition.

> (2) Government run internet is bound to eventually be lobbied to be safe

This would be a clear cut first amendment violation. These kinds of things would be less likely to be restricted by a government ISP than a private party ISP.

You mean a 1A violation like the government pressuring media companies to push their narratives?

Because we have seen this year no one seems to care enough.

If you are unable to exercise your first amendment rights, you sue, and you win. This happens over and over again. Media companies push whatever narratives they want to.

Ok, who was sued over the twitter files?

Twitter can do what they want on their platform. If they thought the government was infringing their rights, they would sue the government. If you think your rights were infringed, then you would sue the government.

> Come back to this comment and in 10 years.

We don’t have to. Municipal broadband initiatives have been around for more than 10 years.

I’ll pick the first one I find on Google for a search of “2014 municipal broadband install”. The first result I got was Chattanooga, which was installing 1gig in 2014 (https://amp.theguardian.com/world/2014/aug/30/chattanooga-gi...). Not the fastest thing available, but certainly still quite respectable in 2024. Not many people with a gig today cry that it’s unusably slow.

But, let’s just check on Chattanooga’s current speed.

Apparently, in 2022 they were rolling out 25gig symmetric (https://epb.com/newsroom/epb-news/25-Gbps-internet-service-n...). That’s light-years better than I can get in a suburb outside of Seattle.

Dunno, municipal broadband seems like it can be just fine.

As I understand it, the fiber would be capable of much higher bandwidth in the future, it's the devices at the end that are the bottleneck. They'll be able to keep upgrading for a long time without replacing the fiber that was laid down.

> Come back to this comment and in 10 years. See what speed competitive markets are at

The promised Utopia which for one reason or another will never come to be. This world where monopolies are not a thing anymore and all that’s there is pure clean competition is, just a dream.

Reply on news.ycombinator.com