Live data from Hacker News

Starting an ISP is hard (2016)

slashgeek.net

21–30 of 36 posts

Re: Starting an ISP is hard (2016)

#21
post #6

Earlier quoted context omitted.

With the power of blockchains by your side, start your own bank today.

Because that's worked so well for a large number of exchanges ... Banks cost a lot to start, because you have regulatory hurdles to leap, and need to spend a lot on security.

Banks don't cost a lot to start. There are nearly 5000 commercial banks in America. That's individual corporations, not branches.

Re: Starting an ISP is hard (2016)

#22

Starting an ISP doesn't have to be that hard at all. If you want to start one in the country where I live you could do so with very little investment. All the issues he talks about are due to the fact that there is no real competition in the US and no infrastructure to support competition. The way it works over here is that there is a (commercially owned) open fiber network which can be used by anyone who wants to, a…

What is this superb country you live in?

Probably UK? I have a choice of about 10 different ISPs where I live in the UK, and none of them would install anything in my house, they all use the same cable that runs to my house, it's just like changing electricity or gas providers.

Re: Starting an ISP is hard (2016)

#23
post #6

Earlier quoted context omitted.

With the power of blockchains by your side, start your own bank today.

Because that's worked so well for a large number of exchanges ... Banks cost a lot to start, because you have regulatory hurdles to leap, and need to spend a lot on security.

That was a joke, though...

Re: Starting an ISP is hard (2016)

#24
post #9

I was involved in starting one of America's biggest ISP's back in the early 90's - I won't name which one, but it was #2 for a while - and I can say that things have radically changed from the way they used to be. But, I don't believe that the fact that "its hard" is a reason not to do it - I think we really need to return to the days of Mom&Pop BBS'es and locally-maintained and operated NOC's in neighborhoods that r…

Your vision is great. The goal should be more ambitious at the social rather than at the business level.

Re: Starting an ISP is hard (2016)

#25
Back when I was 16 and the Dial-up was slowly dying, I took a cable connection and start selling internet to my neighbours.

By the 3rd year the network reached 3 blocks around the neighbourhood and I had ~50 paying clients. I still remember I did that with an old Intel 486 running Slackware Linux and a bunch of ipchains MASQ scripts. I had even a small local website where you were able to see your traffic ( I remember I was paid 10EUR/500MB or something. )

Back then I was not alone. Almost every neighbourhood had those networks, but at one point they started to unite together and the bigger ones were buying the small ones.

Now the biggest problem would be regulation, rather than the hardware requirements.

Re: Starting an ISP is hard (2016)

#26

Earlier quoted context omitted.

What is this superb country you live in?

The Netherlands. Unfortunately, the fiber network doesn't cover the entire country yet. It was bought by the biggest telco a couple of years back and they pretty much halted expansion. Fortunately, a lot of local initiatives have started up with similar conditions.

> "It was bought by the biggest telco a couple of years back and they pretty much halted expansion."

Why did they halt expansion? Would they not make enough profit outside of denser metropolitan areas? Could such a model (whereby the government only allows construction of broadband network for companies that agree to lease out the wires to everyone) while facilitating ISP competition actually stifle deployment of new wire?

Re: Starting an ISP is hard (2016)

#27
I don't know if I agree with this. I cofounded an ISP in the mid-90's. It was in a relatively small town in a rural state in the midwestern US.

I have no idea what the residents options for internet access are now, but at the time the only really choice (unless you were college faculty or a student) was AOL. But there wasn't a local line - you had to call a city 30 miles away that had a long distance toll charge.

That meant we had a relatively captive audience. But, not a lot of people knew what the "internet" was then.

At the time, our largest cost was the 32 phone lines coming in (more expensive than a residential phone line). Initial equipment for the T1 and the modems (sixteen 28.8 ciscos) and setup was only something like $30k. The second largest cost was the T1 (none of us took a salary).

The issue was capitalization. It got to a point where additional significant personal guarantees were going to be required from people who hadn't had a salary for almost 2 years (I was making a little money making static websites). At that time the state telecom called and offered a little money for the customers and to assist in migrating them (I guess they were finally ready to offer rural internet access).

At the end of the day, we might have been a little early to the market, we were under-capitalized, and we didn't realize at the time that if we had offered complementary services around residential internet access (we only had a couple commercial customers, more aggressive web development and hosting services, and a better idea in how to raise capital without more personal guarantees), who knows.

I should add that I never wanted to be in the ISP business, I had an idea about a web based business (that I still think was a really good idea and I'm sorry that running an ISP and burning through all of our collective savings prevented me from doing) and was talked in to the ISP model by the people I went to looking for investment.

TLDR: Technically, starting and running an ISP isn't all that hard. Probably easier today than it was 22 years ago. The business/economic side is the challenge.

Re: Starting an ISP is hard (2016)

#28
post #18
post #9

I was involved in starting one of America's biggest ISP's back in the early 90's - I won't name which one, but it was #2 for a while - and I can say that things have radically changed from the way they used to be. But, I don't believe that the fact that "its hard" is a reason not to do it - I think we really need to return to the days of Mom&Pop BBS'es and locally-maintained and operated NOC's in neighborhoods that r…

I worked at #1! The place all this went bad was when big cable and AT&T got into the game and flushed out all the rest of the competition. Comcast is a reprehensible company from top to bottom--the pricing sucks, cable tv sucks, bundling sucks, and their customer service SUCKS! So I won't do business with them and am relegated to crappy DSL from Frontier which also SUCKS, just not as bad as Comcast as a company. How…

If the Mom & Pop -- real American entrepreneurs -- had grew to national scale, would they have become robber barons? It is just a consequence of size or is more complex than that?

Re: Starting an ISP is hard (2016)

#29
post #19

the original impetus for 'build your own ISP' started from a HN discussion, and frankly i think its where it should end. Detroit is a sterling reminder of the hubris of a generation in the wake of late stage capitalism. Naturally, creating your own internet in the wasteland that was once americas crown jewel of middle class prosperity seems not only natural, but necessary as the economic incentive from private indust…

Why can't people do both prism break stuff and build new ISPs? Those can be complementary. The new ISPs could be built with freedom-respecting FOSS software and hardware.

Also could you elaborate on "limit advertising servers and telimetry servers so as to diminish the unspoken incentive the repeal of net neutrality and operating systems that do not respect privacy." What is the unknown incentive? Aren't there other incentives other than advertising and telemetry?

Re: Starting an ISP is hard (2016)

#30

Earlier quoted context omitted.

The Netherlands. Unfortunately, the fiber network doesn't cover the entire country yet. It was bought by the biggest telco a couple of years back and they pretty much halted expansion. Fortunately, a lot of local initiatives have started up with similar conditions.

> "It was bought by the biggest telco a couple of years back and they pretty much halted expansion." Why did they halt expansion? Would they not make enough profit outside of denser metropolitan areas? Could such a model (whereby the government only allows construction of broadband network for companies that agree to lease out the wires to everyone) while facilitating ISP competition actually stifle deployment of new…

> Why did they halt expansion?

Because it's cheaper to upgrade their existing phone network to VDSL than to install new fiber. Basically, they already own a nation-wide copper phone network, the fiber network was a competitor. They bought the competitor and halted expansion so they could squeeze some more money out of their existing copper network. The official PR bullshit is that by upgrading the copper network they can offer high speed internet to more people in a shorter amount of time than by expanding the fiber network.

> Would they not make enough profit outside of denser metropolitan areas?

The denser metropolitan areas are the main areas where there is no fiber yet. It's the more rural areas that have the best fiber penetration.

Post reply on HN