Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Rico]

between 16,000 and 500,000/mi depending on location and what exactly you include in the bid, last I checked.

[Shinra]

Compared to the implementation costs of copper it's basically nothing, and the price continues to sink daily. Fiber is the reason that ISPs in the US can offer 20-100 meg connections now, why broadband penetration is so high, and why i find it increasingly hard to take the concept of a bandwidth cap seriously.

[Rico]

This is especially true if fiber or at least conduit is installed during government subsidized road work, like for instance the giant recovery act from 2009 which encourages that road work.

[Brentai]

Not a network expert but I'm guessing that the bottleneck is in the routing servers, not the tubes.

The internet may not be a truck but it still needs drivers.

[Classic]

What Brentai said was actually my concern, but it occurred to me that "laying fiber" might be an industry term that already included routing infrastructure in its price. I don't doubt Shinra when he says that ISPs are a long way from being unprofitable, but he may have forgotten some of the costs to build networks and was giving readers a false impression of just how cheap they are.

As Thad can no doubt tell you, it takes serious effort to keep a network.

[Saturn]

The problem I have with bandwidth capping is it's just encouraging already bad behavior. It's not an exaggeration to say that the refusal of telecom companies to upgrade or replace existing infrastructure has cost consumers billions in increased costs and service interruptions, and this is just one more reason AT&T can put on the list of reasons they'll never upgrade their existing hardware. The central offices they're running all this fiber out of these days are pretty much built around decaying bell telephone company equipment, much of it jury-rigged over the years to support software and hardware many, many hardware generations ahead of it. And the process to get new circuits installed? I can't describe it in words. It takes months. You have to go through provisioning teams, implementation teams, if you have to operate on a clec or a rival rboc's turf it's even worse. If AT&T would spend the money to upgrade their infrastructure - from the ground up - they could streamline the circuit install process, eliminate a whole lot of redundancy and upgrade the available bandwidth for an area in days rather than months. On the money they'd save alone from all the extra manhours of work, they'd pay back the cost of the infrastructure upgrades.

haven't AT&T and the other telecoms taken Federal money to UPGRADE THEIR SHIT already?

[Classic]

Yes, and notably failed to deliver as promised. But last I heard about that was in '08 or '09. So maybe they've actually delivered now?

I can't find anything that doesn't look like foaming angry rants on the subject though.

EDIT:
Hmm, it looks like I haven't heard about it since... 2006? Hrm. I didn't think I'd lost that much time.

[Norondor]

as though stealing money from the government was a new thing for corporations to do to boost profits (see also charter schools)

[Classic]

On a completely <threadsplit> note, I've been told I need to listen to NPR's American Life "Crybaby" episode. From the summaries I hear, it's about people in financial industries and how they're not at all contrite or happy with their bailout.

[Shinra]

What Brentai said was actually my concern, but it occurred to me that "laying fiber" might be an industry term that already included routing infrastructure in its price. I don't doubt Shinra when he says that ISPs are a long way from being unprofitable, but he may have forgotten some of the costs to build networks and was giving readers a false impression of just how cheap they are.

As Thad can no doubt tell you, it takes serious effort to keep a network.

They have to constantly upgrade the routing hardware anyway to make room for new customers. No matter what their bandwidth consumption is, you hook too many seperate connections into insufficient hardware, you're going to have issues with service interruptions. And with how much money they're making off of recurring subscriptions, the one-time expense of upgrading hardware for a given set of users is paid off in very little time, anyway.

[Shinra]

Hi, I'm resurrecting this thread to point out the following:

It costs your average big telco company 3 cents to transfer 1gb of data, including all overhead costs.

You could transfer 1.5 terabytes of data and your ISP would still be profiting if you were paying 50 dollars a month for your internet service at that rate.

  sounds like a challenge to get my money's worth to me.

[sei]

(1.5 terabytes) / (1 month) = 612.459935 kBps

Good luck sustaining that.

[Thad]

Now, I'm opposed to per-byte pricing in PRACTICE simply because there is absolutely no free-market mechanism keeping prices reasonable (hi, Internet connection that just went down for 20 minutes for thousandth time since last November!  I'd quit paying for you, but the only other broadband provider in the area offers 1/12 the connection speed!), but I can see hypothetical situations in which it would be okay.

To expand with an example I should have thought of earlier: I DO pay per-byte prices...for Usenet.  I pay Astraweb $25 for every 180GB.  Given my bandwidth use, this comes out to considerably less than their monthly rates ($10 and up).

So yeah, there is such a thing as reasonable per-byte pricing.  I just don't expect it from Cox or Qwest.