Hacker News Reader

Comment by mulmen | Hacker News Reader

If we really care about net neutrality is it viable to create our own ISPs? Are net neutrality rules different between the backbone networks?

View on HN

If we really care about net neutrality, we should focus on privacy. If we build the internet so that it hides the destination data from the carrier, net neutrality will automatically happen as a byproduct, and we don't need the FCC or have to care about different backbone networks.

We should stop arguing that AT&T should be legally barred from doing something with the data it has access to, and start making it so that AT&T doesn't have access to that data.

We should stop arguing that AT&T should be legally barred from doing something with the data it has access to

Why should we stop arguing that? The postal service operates under the same constraints.

I thought I explained that in the remainder of my sentence, but I'll go ahead and clarify that I do think AT&T should be legally barred, but I also think it's a waste of time to argue about it. It's a question of effectiveness and of time & energy and how to best spend it. Why spend time asking lawmakers for net neutrality regulation, when instead we could do what we do best as engineers and build the network such that net neutrality and privacy is built in. Then we don't have to sit around hoping that AT&T and others are not tempted to seek ways to make money. Why spend time writing articles that complain about the Republican pro-industry FCC, when we have the power amongst ourselves to build a better internet and take the question of net neutrality off the table entirely?

The USPS analogy is a bit apples to oranges; the USPS is a government organization and AT&T is one of many private for-profit entities that carry our digital communication. Furthermore, AT&T already was legally barred from listening to conversations going through their network and they did it anyway. I don't exactly trust them, even if there are laws.

I don't think we should expect companies to not try to make money. Their function is to try to make money. If they see ways to try to make money, by asking consumers to pay extra for faster connections to Netflix, they should be allowed to offer premium services.

A big problem with the question of net neutrality is that it's subjective. It's fair to say they're slowing people down if they don't pay more. It's also fair to say they're just offering premium service for a premium fee. These is no winning this argument, both sides have valid arguments. Focusing on privacy instead will answer a question that has a clearer "right" answer, and it will automatically yield a network that has net neutrality built in.

I think this is a great point. Expecting companies not to act in their own best interest is foolish. It's also foolish to expect our elected officials to act in the citizens' best interest rather than for their donors. Instead, let's make net Neutrality the only option so we can focus on technical challenges.

You can certainly do that, but only in countries that have LLU (local loop unbundling) where you can create an ISP without having to own a ton of wire. The US does not have this market-promoting measure.

The difficulty, is that where this matters is in the "last mile" leg of the connection.

Which, to do at scale, is notoriously difficult given the required capital expenditures, as well as navigating the regulation minefield, that the incumbents do everything in their power to keep difficult.

Is it really last _mile_?

If the missing leg of the connection is truly that short, could there be an opportunity for inexpensive cooperative wireless mesh routing in neighborhoods with several outlet nodes onto more regulation-sparse fiber?

Or is 'last mile' just a euphemism for distances that in actuality are much larger?

Last mile is a euphemism for:

* Interconnects

* Customer Support

* Pole Access

* Physical Plant

* Customer Premise Equipment (CPE)

Perhaps if the FCC would designate more long-range friendly frequencies as unlicensed, perhaps with slightly higher power limits, you could eliminate a few of those items?

Don't get me wrong -- there are many cooperatives very successfully running stable and competitive wireless mesh networks, but that model doesn't work very well in major metro areas, with lots of buildings, and very high-density users.

This was tried in the mid to late 2000s and many of the mesh networks, at least the ones run by cities, were shut down or are much smaller now. The problem with doing things through the legal system or through government is that mesh routing can be seen as competitive to the existing oligopoly (which has more $$ to lobby government to increase barriers to entry).

Sharing one beefy connection with 10-20 people in a few nearby houses would be a good first step because it's under the radar of the big telecomm companies and you can't be legally barred from sharing your connection?

I love cooperative mesh networks, but they can't really compete on speed and latency with regular internet connections.

Can you get around last mile monopolies?

Without laying fiber or networks to the millions of homes around the entire united states (kind of like google tried and found not economically viable), no.

That is, ironically, a problem largely created by public interest minded people. Most places in the U.S., it is illegal for an ISP startup to be like Apple: targeting the lucrative high end of the market and working down. You have to build out to every neighborhood, even in nehgborhoods where the cost of build out will be more than what you can hope to make from subscribers there.

That means even behemoths find it difficult to enter each others' markets. Baltimore, for example, wants fiber. Verizon was ready to build fiber. It would have been easy--Baltimore's suburbs are already wired. But the city wouldn't give Verizon a television license (fundamental to the economics of broadband) without universal build out. But it just didn't make sense to do that in a city where a third of the population is below the poverty line.

In fact, nobody has been willing to build fiber in Baltimore on those terms, despite the city railing against Comcast and trying to get a competitor.

Not quite, that is an issue, but not remotely the biggest one. The biggest problem (and google themselves admit this!) is planning and zoning happens at the city and even county / neighborhood level. If you want to move into a city like say Chicago you don't just need to pass city muster. You need to get the alderman to agree for each small neighborhood. What if you need to lay fiber through neighborhood B to reach neighborhood C from neighborhood A and the zoning committee in B denies you? Well that just cut off a large swath of potential clients. This is fundamentally a problem with all infrastructure improvements in the US. The local communities have too much control over what happens. This is also why things like infra happen to move so quickly in a place like say China where the federal govt can trample over all local regulations. For things like infrastructure, this is a great thing.

Sometimes you aren't allowed to - the local regulations can prohibit you from building your own last mile even if you wanted to pay the huge cost to needlessly duplicate it.

It would be great if I can buy my physical line (to the interconnect) from my current ISP.

Then I'm responsible for getting someone to service the quality of that connection between my house and this connection.

I really hope the new wireless attempts to the last mile are successful. I think both Google and Facebook are seriously trying.

Reply on news.ycombinator.com