I would argue different - we need those folks to have 4+ viable ISP options that will compete for them ferociously, rather than making double sure no one else (but another telco megacorp that can afford the compliance) comes and they're stuck with 1-2 greedy ISPs forever.
If TCP/IP followed the ISO model I would phrase it as "the city is responsible for layer 2, the ISP for layer 3 and content providers for layer 4 and up".
Suppose we did as you say 30 years ago. The government would install phone lines and use DSL to carry internet traffic for competing ISPs.
The performance of DSL was fine 30 years ago, but now it's slow, and no ISP is allowed to install anything faster because in your system the government has a monopoly. The government could upgrade it, but that costs money and getting the government to spend money upgrading infrastructure has been a recurring problem. So now you're stuck with DSL.
Whereas if the government just runs conduit, and then Verizon is using DSL from 30 years ago, Sonic can come in at any time and install fiber. Which spurs Verizon to install fiber because now they have competition.
You want the monopoly to be made as narrow as possible. But the natural monopoly isn't layer 2, it's not even the entirety of layer 1. It's the road, and the high cost of digging the trench. Once you have the conduit, the cost of having a hundred ISPs string fiber through it is minor, so doing that should be open to competition.
That's a major point of my analogy -- having the road system being anything but a monopoly is stupid, yet the road system has largely destroyed the railway companies which are not a monopoly.
They do a pretty crappy job of it, in general. The relevant metric here would be congestion, i.e. are they maintaining adequate capacity to prevent people from being stuck in traffic?
There are a lot of arguments people will make for how they should be preventing traffic congestion (e.g. add lanes vs. facilitate more housing construction so people aren't driving as far), but on the question of whether they've succeeded in preventing traffic congestion, the answer is no.
They also spend rather a lot of money on it.
> That's a major point of my analogy -- having the road system being anything but a monopoly is stupid, yet the road system has largely destroyed the railway companies which are not a monopoly.
But the main reason for this is politics. Trucker unions see rail as competition so they lobby for laws that keep people using trucks. Rail has a significantly lower cost per ton but projects to build new rail lines etc. are opposed because they would compete with truck routes.
Something like this, although I'm not sure this is a complete list - I'm not particularly aware about intricacies of starting an ISP in the US so there's probably something I'm missing.
That’s all that’s required to be Neutral. What other kind of expensive “compliance” are you talking about?
I should’ve been more clear, as I mean that NN imposes expensive compliance. I heard that there are various barriers to running an ISP (like exclusivity agreements - TIL FCC had actually prohibited those in 2022, hope this stands unturned), different from Net Neutrality, and that they are the real issue, while NN is less relevant long-term.
As for the NN itself - I can see the need and even argue in favor of it (as a band-aid), but it’s not a solution and it becomes unnecessary (and possibly even harmful) long-term, if that future has a presence of a really good competition. I’m saying “harmful” only because it may theoretically be weaponized at genuinely benign (or at least honestly meant so) QoS situations (which is crappy, but in such cases the choices are typically between working for most and not working at all - nobody does this stuff for fun).
Until one ISP wins, and then we're back at square one.
Seriously, I've seen it myself - if some ISP starts to do shitty things to their customers, if the market is healthy (a competition is possible and anti-monopoly agencies watch out for collusions, etc.) - things get fixed just like that. A contender comes, says "hey folks, we have simply decent service with no BS" and people sign up in droves.
Please don't get me wrong, though. I totally understand the present-day US-specific issues. Net Neutrality seems to be a necessity, but it must be treated as a temporary measure at best, always mentioned with a large footnote to it. It is not a solution - it's a band-aid until the wound is allowed to heal (if it is allowed, and NN by itself _doesn't_ heal anything), so such solutions should really focus on ensuring that they're not hindering any grassroot competition but rather welcome them. And specifically in the US (which is way less dense than most places) some push towards community fiber projects may be a necessity.
Need just one of those for every place, and relaxed zoning laws between such IXes (where ISPs come to play). That assuming that those singleton last-mile networks are ran by the people for the people, in a low corruption environment (which I believe the US is, at least on a smaller-scale levels). If democracy doesn’t fix unfairness I’m not sure what can do.
so, yeah, i'd like to see this notion quashed.
How many of those do you think are going to be profitable?
This fiber should be collectively owned by the property owners, with legally required peering and network policy voting rights for whoever lives on the properties (aka ensuring that landlord can't lock their tenant(s) down to any particular ISPs or impose a crappy network policy on them).
Then careful planning and relaxed permit processes for running lines between IXes, so ISPs can come and build the actual backbones, offer their own services and so on.