I'm in currently in a rural village in Spain where there's a node sponsored by city hall. When you read the article it sounds really ideal, but it isn't. What started in this vivillage with the intention of connecting older people with basic internet (Wikipedia, newspapers, Google...), has turned into a unusable connection that the local electrician charges more than 3x what it costs to install at your home. It is an absolute nightmare for configuring it, rendering unintelligible for older people. Even if they somehow got to the Internet, the local node has been hijacked by all the people that come for the summer and are lured by the "free" connection which is then gobble, mostly by torrenting pirated movies, leaving basically no bandwidth for the rest of the town.
I now don't recommend Guifi for anyone in this area, as it has become a costly, unusable, sham, "vaporware"-network here.
that would be against the terms of use. from the article:
You have the freedom to use the network for any purpose as long as you don’t harm the operation of the network itself, the rights of other users, or the principles of neutrality that allow contents and services to flow without deliberate interference.
You have the right to understand the network and its components, and to share knowledge of its mechanisms and principles.
You have the right to offer services and content to the network on your own terms.
You have the right to join the network, and the obligation to extend this set of rights to anyone according to these same terms.
If any of the other users are part owners (shareholders) of any copyrighted material being downloaded illegally, then those downloading it are accessing the network against the ToS.
Surely by "rights of other users" they mean to include the right to not "have my property stolen."
"The rights of other users" most probably refer to the rights mentioned in this chart. Not copyright.
It is unlikely anyone in the network has the rights to the copyrighted material being exchanged. There may be some exceptions, but you'd have to dig for them, in an extremely invasive and non-neutral way.
Also, copyright infringement is nothing like property theft, mostly because copy & paste doesn't have the same effects as cut & paste. Stop using this broken analogy, it only hurts the debate.
Am i the only one who gets irked by having to read half an article of "how you should feel about this" crap to get an answer to the "WTF is this and how it works" question, which should have taken one line at the start?
Anyway, i find the title odd, because at the end of things it's still plugged into the Telefonica/Comcast/Whatever. So, what's the point?
This basically describes a community/small-business ISP, nothing special.
Obviously the article is not targeted to a technical audience. A quick summary:
- guifi.net (you can visit the website for more info) is a non-profit foundation to support the development of an open, user-owned network.
- Users, alone or by groups buy networking gear (nodes) that connects them to other nodes.
- The result is a mesh network between all those nodes. Guifi.net develops and mantains (free-software based) software to manage the network. They have a website that centralizes information, manages the private IP space, and provides firmwares for the supported gear.
- The network may be used for any purpose. If you want to build a VPN between two physical locations that both have guifi.net connevtivity you don't need to go to public internet. Aside from maintaining the networking gear, there are no fees either.
- Getting internet connectivity is possible. You "just" need someone giving you an "exit gateway. This can be a friend with guifi.net connectivity AND a traditional telefonica/comcast connection. You can also buy internet access directly from the carriers at the CatNIX peering center (the foundation has laid out fiber connecting to that exchange). I don't know the fees.
It is a cool project, but I can see how users can get unhappy if they pay for a node under the premise of "free connectivity" only to find themselves on a (quite large) private mesh network.
This seems similar to starry (https://starry.com/internet). They are rolling out their wifi routers now and looking forward to their internet service that uses millimeter waves.
This is by the same guy who did the Aero startup. The large broadcasting companies were completely threatened by it.
Interested to see what happens with Guifi.net and Starry.
I've spent most of the last 15 years building and maintaining wireless ISP networks like this one. I've known several people (in the US) who have started an ISP like this in their garage and built it up to a few hundred costumers - enough to pay themselves a decent living salary. I'm also working on a startup in this space. Happy to answer questions.
Does this line of service target solely the lack of connectivity, i.e. ISPs will not service you, or other issues like a slow/congested networks?
My main point is figuring out where the other end of the line is, at the original ISP, or beyond that. And please forgive my lack of vocabulary, I have limited networks knowledge.
If I understand your question correctly I think you're asking what network connectivity is 'upstream' from the wireless network, and how/if that compares to purchasing a connection from the telco?
In my experience the way this works is that we buy a leased fiber connection - say 1GBPS, usually terminated either in a regular datacenter or in a business park somewhere. Sometimes you're buying that line from the same entity that is also selling residential Internet access in nearby neighborhoods (IE CenturyLink) but the business group you're dealing with is almost entirely distinct from the residential Internet service group.
1GBPS will go further than you think - you can put 700-800 heavy usage customers (cord cutters, college students, etc) on a 1GBPS link and never (or very rarely) max it out. Selling higher speed plans doesn't really increase the max throughput on the network either, except in fairly predictable 'bands'. That seems unintuitive, but most people don't really increase their Internet usage depending on the speed of the connection (other than at a few thresholds, IE going from <1mbps to something that allows cable-cutting) so selling a faster connection just gets them their data faster and frees up your capacity for the next request.
What does a user mean in this case - a single person or a family ? And how does this work in prime time, when 50%+ of the people are watching netflix etc, at 4 Mbps average(1080p + 720p), with other video services even less compressed ?
a single family (or a single connection to the network, anyway. One customer from the ISP perspective.)
You would think that at least 50% of people would be watching Netflix at the same time each night, but as of yet that doesn't seem to be the case. (At least not on any of the networks that I've been familiar with, which are in pretty tech-savvy areas like I mentioned, with heavy users.) It also helps that you can get one of Netflix's caching servers (can't remember what they're called) once you hit a certain threshold of traffic to their network.
The key though is that the more customers you aggregate the less traffic you can plan for per customer. So for example if you have 100 customers you might need to budget your network at 4-5mbps/customer to make it through peak times without slowing every one down too bad, while if you have 1000 customer you can get by with ~1.5 mbps per customer.
I'm not currently building a WISP, I'm building an auxiliary business for WISPs (Email in my profile if you're curious what it is - not really trying to promote here just yet) but since you asked I'll answer from past experience:
1) I've used Cambium and Ubiquiti and several others. Ubiquiti is hard to avoid because it's so damn cheap and it does work reasonably well, although Cambium is noticeably more robust. The latest stuff coming out is fixed 4G LTE, several companies pushing that gear in to the WISP space and I suspect that's where things will start to take off.
2) 30-50 for Cambium and Ubiquiti APs. Fewer if possible. On a large network the APs tend to be the bottle neck.
3) Fiber to primary POPs and then high capacity licensed wireless backhauls 2 to maybe 3 towers deep in the network. Depending on the local weather it's not unrealistic to get a licensed link that can support 800mbps to maybe 1.2 gbps over 3 to maybe 5 miles.
Spent a Saturday a while back with Detroit's digital stewards. They are true mesh network wizards and have gone all over the world helping communities set up their own mesh networks. I am in awe of what these people have accomplished with literally no money.
Something like 70% of Detroiters lack net access and this group is slowly trying to change that. They could be better organized about asking for outside help. If you've got an old router or want to just send them a buck it goes towards putting more families on the Internet. They're also looking for developers to write programs for specific community needs, most of them are hardware people.
On the bright side, it illustrates the profit margins and excess bandwidth that ISPs have. They are making plenty, and for a few thousand people to "slice some off the top" for sharing isn't going to crush them.
But this may prove to not be scalable. It's almost certainly against most terms of service for those that actually pay to connect to the internet. Of course, I'd say "if everyone connected this way, where would the internet come from?" but then they'd all be connected anyway...
Guifi has a connection to the local Internet Exchange in barcelona (catnix) which is managed by exo.cat (an association of people paying a bi-yearly fee to support the expansion of the guifi network) as well as other connections around barcelona and other cities / towns.
A lot of people shared their internet bandwidth which here in Spain was not illegal or blocked by the ISPs if you where smart. There are also several small ISPs that have been created that sells internet and telephone services over the guifi network.
> On the bright side, it illustrates the profit margins and excess bandwidth that ISPs have.
How? Though I totally agree with the general idea that wireless is the way to go. You can dramatically reduce build-out costs by just building fiber to the node and point-to-point wireless for the last hop.
Remember when you could connect to the internet via an ISP some guy ran down the street where he had a rack in his closet in the mid 90s? Those were some kick ass times. The media companies just bought everything up in the dot-com boom.
The sad thing is this isn't a NEW approach. Community approach is how the internet spread before it got commercialized by the big companies. Spread via universities and small mom and pop ISPs
Wilson, NC got tired of waiting for telecom companies to wire the town with high speed affordable Internet — so it did it itself. Then the telecom companies succesfully lobbied the state legislature to pass a law preventing any other municipalities from doing the same.
Its a good question... in the article it states that Guifi was able to get a "peering" connection.
Internet service providers have to connect to each other somehow and this is done at IXP (internet exchange points: https://en.wikipedia.org/wiki/Internet_exchange_point )-- basically a huge building where separate providers can come together and exchange traffic.
I imagine that a peering connection in such a facility requires a heroic amount of maneuvering around 800 lb gorilla companies, but it certainly is possible if one has enough traffic and money.
The Guifi network has a connection to the local internet exchange CATNIX. That connection is paid for by the association exo.cat, which is supported by members paying a bi-yearly fee.
Peering doesn't get you to the internet at large though, it just gets you to the people connected to the particular exchange (and they're not usually willing to relay your traffic to other networks)
Now I am really mixed up. How else would an independent ISP get to "the internet at large"? Isn't that what a peering agreement is all about at an IXP?
They'd have to buy transit from a provider. Peering is connectivity between two ISPs. Transit is connectivity between you and the rest of the internet.
As usual, the devil is in the "just". Typical US $50/month broadband service includes about $1/month worth of wholesale bandwidth; the rest goes into maintaining the last ~10 miles and profit. There may be room for considerable efficiency gains.
I'd sure be interested in what those efficiency gains could be. The ISP business isn't exactly a gold mine and margins are pretty thin. It's only at (massive) scale that things start to look better.
I'm sure it'd be much easier if people just moved into their closest DC. That last mile really is such a drag.
No way. ISPs that own infrastructure are capital intensive business that print money over long time horizons.
Their operating margins are like 97% profit, and depending on the maturity of the network, the costs on debt service are something like 25% to 90% of costs. It's a business that generates consistent long term profit and cash flow.
In the last 10mi of course. Don't dig cable, don't push tv, have members buy and deploy the mesh network. Also don't pay lobbyists, don't pay for advertising, installers Cust support, CEO salaries, etc.
> The ISP business isn't exactly a gold mine and margins are pretty thin.
Whoa, who told you that? Lets think about this in simplified terms, if you could run a coax cable to houses around you without caring about it being buried or hidden and start earning $50-$100 a month would you do it? I would be doing nothing else until I collapsed. I would go up and down the street with cable spools of cable putting posts in the ground to split it to people's houses. In one long day you could be earning what you would get from a minimum wage job. In a week you would be middle class. In a month you would live like a lawyer, in two you would live like a doctor, and you would never have to work again to make the same amount of money. Gold mines don't even do this well.
Building networks is, to no one's surprise, expensive. Financial analysts last year estimated that Google had to spend $84 million to build a fiber network that passed 149,000 homes in Kansas City, with the cost per home at $500 to $674. That figure did not include additional costs for actually connecting each home that requests service. A national Google Fiber build out passing 15 percent of US homes would cost $11 billion a year for five years, Wall Street analysts have estimated.
Montgomery discussed the challenge of finding what some people call "patient capital," or investors who are willing to put money into companies that might take five or 10 years to make a big profit.
"Costs are fixed. You're going pay your staff and maintenance people whether you have one customer or 10,000. All your costs are basically going to be the same," he said. "For a fiber network, you need to reach about 30 percent of the local market within a couple of years, or you're going to go out of business."
---
"One big reason we lack Internet competition: Starting an ISP is really hard
Creating an ISP? You'll need millions of dollars, patience, and lots of lawyers."
"Patient capital" is desperate for moderate returns over time. Ask any pension or endowment investor -- they are screwed now because there is no way for their funds to earn yield.
The reason AT&T was able to gobble up the entire country in the 20th century was in part their ability to sell convertible securities that delivered premium returns with low/moderate risk.
I thought it was the billions of dollars AT&T took from the federal government, as recently as two years ago (although traditional financing played a part):
24/7 operations team at the NOC. Billing. Collections. Technical support. Physical support. Lawyers. Local easements. Sales and Advertising. Cost of capital. Goes on and on.
If you're a cable company who already built all of that, it's marginally very profitable to layer on internet service.
If you think that you can go run a coax cable to someone's house and start collecting a monthly check from them, and collect routinely, without a) billing issues, b) collections issues, c) service outages, etc. etc. -- I mean, maybe for 3 neighbors it would work. But not at scale.
Btw 'at scale' all the overhead you mentioned decreases. I doubt it costs $70 a month per customer. You see issues constantly, but how big is the customer base? 100k? That's 7 million a month. With bandwidth costs of around $1 per person I'm going to guess there is a lot of room for profit.
> The first node went live in the early summer of 2004, when Roca turned on a router with a directional antenna he’d installed at the top of a tall building near the local government headquarters. That office was also the only place in town with Internet access, a DSL line Telefonica had run to municipal governments throughout the region. The antenna was aimed, line of sight, toward Roca’s home about six kilometers away.
One person can pay to have an expensive connection and shuffle all the internet through that. As more and more people join and bring on their own hardwire connections it becomes much more tolerant to disruption.
Given the challenges I've had with Comcast this week, I'm wondering what it would take to get a neighborhood mesh up and running (each home is still segregated but with shared egress) for improved uptime and probably better bandwidth.
I've considered trying to get fiber for gigabit+ wired in my subdivision here in the Bay Area (with HOA approval). We have fiber coming up to the main road and have been wired for ATT/Comcast for a long time.
Since we already have ATT/Comcast I think they will be pain to deal with when trying to run new fiber.
There are a few providers doing fixed wireless, but even that can only reach a fraction of everyone in the community.
ATT leases space since they built out more recently. Comcast I have no idea, it was probably built out long ago and they have some agreement. I did find an interesting case of a HOA that is doing their own fiber via Paxio in Milpitas. I will certainly contact them.
What's different between them and other companies? Didn't really get it. There are many small internet service providers who pretty much rent the infrastructure from the big guys. You still have the same problems though.
The internet consists of the backbone and the last mile. The backbone is actually pretty diversified and competitive. Installing one cable that carries millions of connections is a pretty good business model. There are uplinks all over the country where you can connect to a data center and get huge connections for a few hundred dollars and up. For example, Hurricane Electric runs an uplink serving the northern Bay Area in Fremont, I believe.
The last mile is much more of a pain in the ass. You have to bring connections to everyone's house and you are then installing one cable that carries one connection. This is why there has been little competition and people are generally unhappy with their ISPs.
Guifi is addressing the last mile market with a network built by a cooperative non profit. They are in no way "renting infrastructure from the big guys". They are reselling bandwidth in the way that a supermarket resells produce. If the big guys had put last mile infrastructure into Catalonia, Guifi would not have started.
As much as I want to love Monkeybrains.. 20mbps for $35 is way more expensive than Comcast, and their weird coverage map is unclear where you can actually get it. I tried to contact their customer support and didn't manage to get a reply. At least comcast tells you you're getting screwed, and doesn't ghost you.
Sorry you didn't get a reply. We respond to every inquiry, but sometimes the responses end up in one's spam folder. We run our own mail servers and despite our best efforts (DKIM, SPF, never being on any blacklists) some mail providers bin us.
If you give us a call (415-974-1313) we can look up your address and give you a better idea of what to expect, though we tend to under-promise and over-deliver.
The website is a bit... dated, but we're working on a new one with a new coverage map[0] and more details about our service in buildings that are already on-net.
I'm in Duboce Triangle and get the same 65Mbps up and down. There have only been two short outages since I got the service two months ago. Given I'm paying a third of Comcast's price, I can stand 30 minutes of no internet.
