You missed the point: with internet connections rapidly reaching video speeds and the telephone network very much tied to 8khz there is no value in having a 16 khz VoIP. If you're going to up the sampling rate only for VoIP, go straight to 44khz and be done with it. Don't brag because you were dumb enough to select a median value.
we're talking about the full-disclosure model and how that effectively serves malware development
The open source, full-disclosure model improves the pace of ALL software development. All means all, including software development for "bad" purposes.
So what? If you're going to up the sampling rate why not go directly to 44khz stereo (CD quality audio) and be done with it? Jumping from the telephony industry standard 8khz to 16 khz is thoroughly uninspired.
And I was paid roughly the same in my first job after my bachelor's degree ten years ago... Except I was paid on salary, not part time, and I got paid vacation and benefits to too. And it was ten years ago.
As for how well you teach your students, it sounds like you're on track to become a professor. If that's your goal then you're doing the right thing and you shouldn't let me dissuade you. Teaching was not among my goals.
I live in germany. Degree is everything over here.
Fair enough. I live in the US. I've heard the rules in Europe are different.
And if you haven't found out yet that realizing your dream takes dekades and not years, then you haven't picked a career at all.
I realized most of my goals five years past the bachelors' degree. House. Salary. My favorite childhood hobby parlayed in to an enjoyable career. I'd even bootstrapped a moderately successful small business. I had to pick some more ambitious goals.
1. You get to meet and work with people who are pretty clear about what they want.
I do that now. Why do I need graduate school for this?
2. The rest of the world suddenly takes you more seriously.
I just negotiated and won approval for a $600k project. The people I care about already take me seriously.
3. You can use graduate school as an ideal environment for beginning work on a startup.
Or you can spend some time working for startups and parter on the next project with people who have experience and credentials starting a company, not just wild ideas.
4. You can use graduate school as a pivot to change your career.
If it took you that long to figure out you picked the wrong career.
5. You get to pick your choice of work and your work hours.
I do that now.
6. You can get involved in projects that can actually impact the real world.
You can do that in the work force and be well paid for it.
7. You can get involved in projects that have absolutely no impact on the real world. You can work on things simply because they're interesting and fun. You often get paid to do this.
And then you can become a professor/researcher at the school and continue to get paid piddling amounts for someone with your talents. Which might be okay if you had free choice in what you wanted to investigate, but you don't have free choice. You have to write proposals and sell your ideas to various committees and sponsors and fight your way through some vicious office politics on the way. So in the end you don't work on what you want to, but instead settle for what you can get approved.
8. You can do things that you missed out on in your undergraduate school. It's a second chance.
If you need a second chance. But if grad students are folks who needed a second chance to get it right, what does that say about their abilities?
9. If you're good at what you do, you can count on being invited to travel around the world to conferences and seminars.
If you like public speaking. Personally, I'm an introvert.
10. You get to be the TA this time around.
Because I always wanted to be the guy who got paid piddling amounts of money to do a lousy job of teaching students, all of whom clearly understand that I'm doing a lousy job.
1. My focus was comp sci and CS is not an old enough discipline to have a useful postgraduate program. Let me put that another way: Not enough is understand yet in the discipline for there to be more than four years worth of material to teach.
2. The bubble was just kicking in to gear (late '95) and I wanted in on the ground floor. I figured if I was wrong about grad school I could always go back after riding the bubble to its end.
Do I regret it? Absolutely not. I rode the bubble just like I planned, I started a couple of businesses and and spent all but 10 months of the last decade doing work I love to do.
Will I go back to school? At some point I plan to get a law degree. I still don't see any value in a MSCS of a PhD CS. Quite the contrary: the few folks I've encountered with an MSCS (generally from Podunk U) have been blithering idiots.
Actually, I'd be curious to see data too on this too. I haven't really heard anything of note about preforking or single-thread non-blocking IO outside of the C/Unix universe. It strikes me that you only use those methods when you're trying to squeeze out the last ounce of performance from the hardware. Other languages (like Java) generally have different priorities and the programmers using those other languages generally have priorities in line with their language's strengths.
Yeah, I worded it badly. The point I was trying to make is that if you aren't talking about something written in C/C++ then the performance issues associated with threaded/not-threaded are the least of your concerns. Its only when you get to a fairly low-level language like C that they start to make a noticable difference.
Debugging a preforked C program like Apache 1.3 is still much easier than debugging a monolithic process with non-blocking I/O like Squid. Debugging a monolithic C process like Squid is still much easier than debugging multithreaded software like Microsoft Windows.
This isn't likely to change regardless of how many processors you throw in the machine.
The rules change when you move to something like Java where cross-thread contamination of the data structures is relatively difficult. But then if you're talking about Java then why are you seriously considering hardware performance issues?
Do you understand the Secretary's job? I mean really understand it, the official and unofficial parts. Do you understand it enough better than she understands it so that you can build a computer that does all of the things she needs and wants it to do? And don't forget, it needs to do everything her boss decides she needs to do with it.
I'm not -that- smart and I'll bet that you aren't either.
There are places where a closed OS works. Think wireless router or Internet appliance. But the desktop? Not so much.
I'd argue that programmers who won't learn about and accomodate their users' mental models are the ones "unwilling to learn a simple concept" but that argument would fall on deaf ears, wouldn't it?
Let me put it another way: only an ass would insist her name was Zoë, not Zoe. Why would you intentionally program Zoe's computer to behave like an ass? E means e means ë. Mere mortals understand this simple concept. Why does it cause programmers such trouble?
Perhaps you haven't heard, but "the customer is always right."
Solving the problem for Apache doesn't solve it for any of the other applications. Solving it for the filesystem solves the problem for everything except wildcard expansion. Solving it for Apache requires a slow directory search. Solving it for the filesystem takes a short, fast patch.
As for why not FAT, I assume that was a rhetorical question.
Why are folks like you so hostile to Linux having a decent case-insensitive filesystem anyway? Its holding Linux back from desktop adoption. This makes no sense to me.
One of first culture shocks for people moving from Windows to Linux is the case sensitivity of file names.
I wrote a patch that creates an "ext3ci", a case-insensitive ext3 filesystem for Linux. Because it creates a seperate driver, you can have parts of the directory tree set up as classic case-sensitive and parts insensitive. Bash's wildcard expansion still expects case sensitive names but its great for web servers and similar machines where the users are expecting case-insensitive behavior.
I don't know who saved what but I know that the modern generation-4 HP DL385 server (which borrows heavily from both HP and Compaq technologies) kicks the spit out of every comperable machine out there. Whoever came up with the physical design is an effing genius and I'd like to shake his hand.
The map shows open source "vendors" not open source developers. Of course it shows essentially the same distribution as software vendors in general. The surprise would be if it showed anything else.
It also says exactly nothing about the physical distribution of the open source phenomenon.
In a world that is in chaos politically, socially and environmentally, how can the human race sustain another 100 years?
Do nothing. Or rather, keep doing what we're doing. Chaos is the natural state of the world and it has been for longer than the human race has been around. We found ways to solve or live with yesterday's problems and there is no reason to believe we won't do the same with today's.
Could it be that they were intending to monitor domestic calls (and internet traffic) all along
Of course. This was created to satisfy the extension of traditional telephone wiretapping requirements. You remember Carnivore and the related laws, right? No large Internet provider can cost-effectively satisfy a wiretapping subpoena for -only- the data requested. That filtering requires equipment vastly more powerful than the routers they use. I looks to me like AT&T cut a deal: We'll give you access to the total data stream but in return you agree that filtering for the lawfully authorized data is solely and permanantly the Federal government's problem and expense.
From the perspective of fiscal responsibility to the shareholders, its the right choice.
Now of course UTOPIA's long term viability depends on delivering cable television services over their fiber as well.
Yeah, among other things. If they're still solvent in 10 years you win the right to say, "I told you so." I won't hold my breath.
$44/month. How much of that price do you think is being forwarded to UTOPIA?
My guess? $35-$40. That would put them on track to pay the 20-year note for the fiber but they'll still be screwed on the equipment unless they find another source of revenue.
I saw it here in Virginia when Verizon/Bell Atlantic opened up their 1.5 ADSL offering to competitive ISPs. The price to the ISPs was $32 each unless you did volumes in the tens of thousands. Then it fell to $29. The next week, the ads in the paper were for $35-$40. A year later the names were different but the prices were the same. Always some sucker who thinks he can sell at a loss and make it up in volume.
It is worth recognizing that UTOPIA is basically a large, MPLS based Ethernet VLAN.
Which is not even remotely close to the fiber optic mutli-ISP setup Cringley described. If you're willing to battle out oversubscription on gig-e and 10-gig-e links in order to get back to the ISP then you can shorten the fiber links to half a mile as Utopia appears to have done. That costs a lot less than the $15k I claimed for Cringley's design. About $6k in fact over 10 years if Utopia's published numbers are reasonably accurate. That's still 4 times Cringley's "worst case" claim.
Don't get me wrong, I think its great what Utopia is doing and if they offered it here I'd sign up in a heartbeat. I signed up for Vonage too. Vonage is a wonderful service, but I didn't buy stock because I think they're undercharging and I expect them to flame out once their competition gets up to speed.
Wireless ISPs are a big deal in Utah to serve an amazing number of areas
Of course they are. Utah is flat. Try that trick anywhere east of the Appalachians. If your line of sight isn't blocked by hills then its blocked by 4-story trees. The idea that you're going to deliver universal access via wireless in a 30-mile radius here is just plain unrealistic. You'll catch 30% of a particular valley and that's about it.
But all of this is beside the point. Cringley didn't propose covering rural areas with wireless to avoid the fiber expense. He proposed fiber coops for everybody.
And some Republicans are aghast at what the modern robber barons have been allowed to get away with. Individuals will be individuals and you should make the effort to learn the views of the individuals who propose to represent you.
That having been said, there are only a couple dozen technophiles in Congress. The rest tow the party line. Democrats are largely anti-censorship and want to reign in the corporate robber barons. Republicans think network neutrality is communist and would still ban Darwin if they thought they could get away with it.
You can pick one or the other, because with a few exceptions those are the electable choices.
Costs to construct UTOPIA are calculated from the preliminary engineering design. Final engineering will refine the design and, thus, the costs.
Six additional points:
a. There are a few places cheaper to build than Utah, but not many. b. The Utopia figures calculate a cost per household of $1171 versus 100% of the households. Cringley assumes a 40% uptake. That would scale Utopia's numbers to $3000, double Cringley's estimate. c. This is a cost estimate for initial deployment only. Maintenance, backhoes, road moves, etc. cost at least 10% per year. Cringley proposes amortizing over 10 years. That means you have to add 100% to cover maintenance during that period, pushing Utopia's customer cost to at least $6000 or $70/month -- 4 times Cringley's maximum estimate. d. This is NOT for rural Utah. Its almost exclusively for the high-density cities and towns. e. The average fiber run is less than half a mile. This is adequate for Utopia's single-company service-oriented architecture but way to short to offer fiber-ISP choice as Cringly proposes. f. Lets not forget - that $70/mo is for the fiber only. The equipment and service is on top of that. You're going to pay at least another $30/month to light the fiber and put a decent amount of bandwidth on it. Personally, I'd pay $100/mo for a sweet Internet service but will 40% of the households on my block? Not a chance.
So, your data proves you're wrong. How do you like that?
First off, we already own the last mile. That's why its called a "public right of way." Just like a public park, it belongs to all of us. The problem is that our pussy politicians (especially the Republicans) aren't asserting our rights to it. Companies using the right-of-ways are being permitted to tie the associated services with other more expensive and more restrictive services that aren't associated with the right-of-way. Want that to change? Vote Democrat. They're not perfect, but they do believe in actually regulating the companies that consume public resources.
Frankston points out that we build and finance public infrastructure in a public way using public funds with the goal of benefiting economic, social, and cultural development in our communities. So why not do the same with the Internet, which is an information infrastructure?
Because: a) Networking technology continues to undergo rapid change. b) Even the experts don't understand the 50-year requirements very well.
Public infrastructure projects work OK when the technology is stable and well understood. Like roads and bridges. They're a disasterous sinkhole for cash the rest of the time. That's why the money disappeared. 20 years from now when half the politicians are folks who grew up with the Internet and the networking experts can clearly articulate an infrastructure that with reasonable maintenance will remain appropriate and cost-effective for 50 years, then maybe we can look at it as a government infrastructure project.
In the mean time, we should assert our rights to the public right-of-ways. The price of access should be that the companies which use it don't get a unilateral choice in how the resulting products are sold.
The cost per fiber drop, according to Bill's estimate, is $1,000-$1,500 if 40 percent of homes participate.
There have to be some crazy assumptions behind that. Taking 12 strands for a mile with no stops is $15k in ideal circumstances. In downtown DC its $175/foot. If your ISP is not the phone company then there's about a 90% chance that its nearest office is more than 10 miles away. Even for the best case the numbers don't compute... And that's without considering the cost of maintenance and equipment to light the fiber.
Fiber works for the phone company because they multiplex it at about a 16:2 ratio within a few hundred yards of your home and then trunk that cable back to an office that's within about 3 miles. Even then they're banking on your purchase of phone, Internet and TV at $150/month to recoup the cost over the next 10 years.
$1500/customer? That's off by at least an order of magnitude. $1500 might cover the raw cost of the cable itself, but that's about it.
Slashdot Mirror
You missed the point: with internet connections rapidly reaching video speeds and the telephone network very much tied to 8khz there is no value in having a 16 khz VoIP. If you're going to up the sampling rate only for VoIP, go straight to 44khz and be done with it. Don't brag because you were dumb enough to select a median value.
we're talking about the full-disclosure model and how that effectively serves malware development
The open source, full-disclosure model improves the pace of ALL software development. All means all, including software development for "bad" purposes.
So what? If you're going to up the sampling rate why not go directly to 44khz stereo (CD quality audio) and be done with it? Jumping from the telephony industry standard 8khz to 16 khz is thoroughly uninspired.
And I was paid roughly the same in my first job after my bachelor's degree ten years ago... Except I was paid on salary, not part time, and I got paid vacation and benefits to too. And it was ten years ago.
As for how well you teach your students, it sounds like you're on track to become a professor. If that's your goal then you're doing the right thing and you shouldn't let me dissuade you. Teaching was not among my goals.
I live in germany. Degree is everything over here.
Fair enough. I live in the US. I've heard the rules in Europe are different.
And if you haven't found out yet that realizing your dream takes dekades and not years, then you haven't picked a career at all.
I realized most of my goals five years past the bachelors' degree. House. Salary. My favorite childhood hobby parlayed in to an enjoyable career. I'd even bootstrapped a moderately successful small business. I had to pick some more ambitious goals.
1. You get to meet and work with people who are pretty clear about what they want.
I do that now. Why do I need graduate school for this?
2. The rest of the world suddenly takes you more seriously.
I just negotiated and won approval for a $600k project. The people I care about already take me seriously.
3. You can use graduate school as an ideal environment for beginning work on a startup.
Or you can spend some time working for startups and parter on the next project with people who have experience and credentials starting a company, not just wild ideas.
4. You can use graduate school as a pivot to change your career.
If it took you that long to figure out you picked the wrong career.
5. You get to pick your choice of work and your work hours.
I do that now.
6. You can get involved in projects that can actually impact the real world.
You can do that in the work force and be well paid for it.
7. You can get involved in projects that have absolutely no impact on the real world. You can work on things simply because they're interesting and fun. You often get paid to do this.
And then you can become a professor/researcher at the school and continue to get paid piddling amounts for someone with your talents. Which might be okay if you had free choice in what you wanted to investigate, but you don't have free choice. You have to write proposals and sell your ideas to various committees and sponsors and fight your way through some vicious office politics on the way. So in the end you don't work on what you want to, but instead settle for what you can get approved.
8. You can do things that you missed out on in your undergraduate school. It's a second chance.
If you need a second chance. But if grad students are folks who needed a second chance to get it right, what does that say about their abilities?
9. If you're good at what you do, you can count on being invited to travel around the world to conferences and seminars.
If you like public speaking. Personally, I'm an introvert.
10. You get to be the TA this time around.
Because I always wanted to be the guy who got paid piddling amounts of money to do a lousy job of teaching students, all of whom clearly understand that I'm doing a lousy job.
I skipped grad school because:
1. My focus was comp sci and CS is not an old enough discipline to have a useful postgraduate program. Let me put that another way: Not enough is understand yet in the discipline for there to be more than four years worth of material to teach.
2. The bubble was just kicking in to gear (late '95) and I wanted in on the ground floor. I figured if I was wrong about grad school I could always go back after riding the bubble to its end.
Do I regret it? Absolutely not. I rode the bubble just like I planned, I started a couple of businesses and and spent all but 10 months of the last decade doing work I love to do.
Will I go back to school? At some point I plan to get a law degree. I still don't see any value in a MSCS of a PhD CS. Quite the contrary: the few folks I've encountered with an MSCS (generally from Podunk U) have been blithering idiots.
Actually, I'd be curious to see data too on this too. I haven't really heard anything of note about preforking or single-thread non-blocking IO outside of the C/Unix universe. It strikes me that you only use those methods when you're trying to squeeze out the last ounce of performance from the hardware. Other languages (like Java) generally have different priorities and the programmers using those other languages generally have priorities in line with their language's strengths.
Yeah, I worded it badly. The point I was trying to make is that if you aren't talking about something written in C/C++ then the performance issues associated with threaded/not-threaded are the least of your concerns. Its only when you get to a fairly low-level language like C that they start to make a noticable difference.
Debugging a preforked C program like Apache 1.3 is still much easier than debugging a monolithic process with non-blocking I/O like Squid.
Debugging a monolithic C process like Squid is still much easier than debugging multithreaded software like Microsoft Windows.
This isn't likely to change regardless of how many processors you throw in the machine.
The rules change when you move to something like Java where cross-thread contamination of the data structures is relatively difficult. But then if you're talking about Java then why are you seriously considering hardware performance issues?
Do you understand the Secretary's job? I mean really understand it, the official and unofficial parts. Do you understand it enough better than she understands it so that you can build a computer that does all of the things she needs and wants it to do? And don't forget, it needs to do everything her boss decides she needs to do with it.
I'm not -that- smart and I'll bet that you aren't either.
There are places where a closed OS works. Think wireless router or Internet appliance. But the desktop? Not so much.
You're French? That explains a lot.
I'd argue that programmers who won't learn about and accomodate their users' mental models are the ones "unwilling to learn a simple concept" but that argument would fall on deaf ears, wouldn't it?
Let me put it another way: only an ass would insist her name was Zoë, not Zoe. Why would you intentionally program Zoe's computer to behave like an ass? E means e means ë. Mere mortals understand this simple concept. Why does it cause programmers such trouble?
Perhaps you haven't heard, but "the customer is always right."
Solving the problem for Apache doesn't solve it for any of the other applications. Solving it for the filesystem solves the problem for everything except wildcard expansion. Solving it for Apache requires a slow directory search. Solving it for the filesystem takes a short, fast patch.
As for why not FAT, I assume that was a rhetorical question.
Why are folks like you so hostile to Linux having a decent case-insensitive filesystem anyway? Its holding Linux back from desktop adoption. This makes no sense to me.
One of first culture shocks for people moving from Windows to Linux is the case sensitivity of file names.
I wrote a patch that creates an "ext3ci", a case-insensitive ext3 filesystem for Linux. Because it creates a seperate driver, you can have parts of the directory tree set up as classic case-sensitive and parts insensitive. Bash's wildcard expansion still expects case sensitive names but its great for web servers and similar machines where the users are expecting case-insensitive behavior.
http://bill.herrin.us/freebies/
You mean the battery for the on-board smartarray? It slides in between the edge of the case and the left-most drives. That's why the cable is so long.
I don't know who saved what but I know that the modern generation-4 HP DL385 server (which borrows heavily from both HP and Compaq technologies) kicks the spit out of every comperable machine out there. Whoever came up with the physical design is an effing genius and I'd like to shake his hand.
The map shows open source "vendors" not open source developers. Of course it shows essentially the same distribution as software vendors in general. The surprise would be if it showed anything else.
It also says exactly nothing about the physical distribution of the open source phenomenon.
In a world that is in chaos politically, socially and environmentally, how can the human race sustain another 100 years?
Do nothing. Or rather, keep doing what we're doing. Chaos is the natural state of the world and it has been for longer than the human race has been around. We found ways to solve or live with yesterday's problems and there is no reason to believe we won't do the same with today's.
Received in a tech support email:
Internet is connected to phone line #1. It should be connected to #2. How do I change this?
Could it be that they were intending to monitor domestic calls (and internet traffic) all along
Of course. This was created to satisfy the extension of traditional telephone wiretapping requirements. You remember Carnivore and the related laws, right? No large Internet provider can cost-effectively satisfy a wiretapping subpoena for -only- the data requested. That filtering requires equipment vastly more powerful than the routers they use. I looks to me like AT&T cut a deal: We'll give you access to the total data stream but in return you agree that filtering for the lawfully authorized data is solely and permanantly the Federal government's problem and expense.
From the perspective of fiscal responsibility to the shareholders, its the right choice.
Now of course UTOPIA's long term viability depends on delivering cable television services over their fiber as well.
Yeah, among other things. If they're still solvent in 10 years you win the right to say, "I told you so." I won't hold my breath.
$44/month. How much of that price do you think is being forwarded to UTOPIA?
My guess? $35-$40. That would put them on track to pay the 20-year note for the fiber but they'll still be screwed on the equipment unless they find another source of revenue.
I saw it here in Virginia when Verizon/Bell Atlantic opened up their 1.5 ADSL offering to competitive ISPs. The price to the ISPs was $32 each unless you did volumes in the tens of thousands. Then it fell to $29. The next week, the ads in the paper were for $35-$40. A year later the names were different but the prices were the same. Always some sucker who thinks he can sell at a loss and make it up in volume.
It is worth recognizing that UTOPIA is basically a large, MPLS based Ethernet VLAN.
Which is not even remotely close to the fiber optic mutli-ISP setup Cringley described. If you're willing to battle out oversubscription on gig-e and 10-gig-e links in order to get back to the ISP then you can shorten the fiber links to half a mile as Utopia appears to have done. That costs a lot less than the $15k I claimed for Cringley's design. About $6k in fact over 10 years if Utopia's published numbers are reasonably accurate. That's still 4 times Cringley's "worst case" claim.
Don't get me wrong, I think its great what Utopia is doing and if they offered it here I'd sign up in a heartbeat. I signed up for Vonage too. Vonage is a wonderful service, but I didn't buy stock because I think they're undercharging and I expect them to flame out once their competition gets up to speed.
Wireless ISPs are a big deal in Utah to serve an amazing number of areas
Of course they are. Utah is flat. Try that trick anywhere east of the Appalachians. If your line of sight isn't blocked by hills then its blocked by 4-story trees. The idea that you're going to deliver universal access via wireless in a 30-mile radius here is just plain unrealistic. You'll catch 30% of a particular valley and that's about it.
But all of this is beside the point. Cringley didn't propose covering rural areas with wireless to avoid the fiber expense. He proposed fiber coops for everybody.
And some Republicans are aghast at what the modern robber barons have been allowed to get away with. Individuals will be individuals and you should make the effort to learn the views of the individuals who propose to represent you.
That having been said, there are only a couple dozen technophiles in Congress. The rest tow the party line. Democrats are largely anti-censorship and want to reign in the corporate robber barons. Republicans think network neutrality is communist and would still ban Darwin if they thought they could get away with it.
You can pick one or the other, because with a few exceptions those are the electable choices.
From the link, emphasis mine:
Costs to construct UTOPIA are calculated from the preliminary engineering design. Final engineering will refine the design and, thus, the costs.
Six additional points:
a. There are a few places cheaper to build than Utah, but not many.
b. The Utopia figures calculate a cost per household of $1171 versus 100% of the households. Cringley assumes a 40% uptake. That would scale Utopia's numbers to $3000, double Cringley's estimate.
c. This is a cost estimate for initial deployment only. Maintenance, backhoes, road moves, etc. cost at least 10% per year. Cringley proposes amortizing over 10 years. That means you have to add 100% to cover maintenance during that period, pushing Utopia's customer cost to at least $6000 or $70/month -- 4 times Cringley's maximum estimate.
d. This is NOT for rural Utah. Its almost exclusively for the high-density cities and towns.
e. The average fiber run is less than half a mile. This is adequate for Utopia's single-company service-oriented architecture but way to short to offer fiber-ISP choice as Cringly proposes.
f. Lets not forget - that $70/mo is for the fiber only. The equipment and service is on top of that. You're going to pay at least another $30/month to light the fiber and put a decent amount of bandwidth on it. Personally, I'd pay $100/mo for a sweet Internet service but will 40% of the households on my block? Not a chance.
So, your data proves you're wrong. How do you like that?
First off, we already own the last mile. That's why its called a "public right of way." Just like a public park, it belongs to all of us. The problem is that our pussy politicians (especially the Republicans) aren't asserting our rights to it. Companies using the right-of-ways are being permitted to tie the associated services with other more expensive and more restrictive services that aren't associated with the right-of-way. Want that to change? Vote Democrat. They're not perfect, but they do believe in actually regulating the companies that consume public resources.
Frankston points out that we build and finance public infrastructure in a public way using public funds with the goal of benefiting economic, social, and cultural development in our communities. So why not do the same with the Internet, which is an information infrastructure?
Because:
a) Networking technology continues to undergo rapid change.
b) Even the experts don't understand the 50-year requirements very well.
Public infrastructure projects work OK when the technology is stable and well understood. Like roads and bridges. They're a disasterous sinkhole for cash the rest of the time. That's why the money disappeared. 20 years from now when half the politicians are folks who grew up with the Internet and the networking experts can clearly articulate an infrastructure that with reasonable maintenance will remain appropriate and cost-effective for 50 years, then maybe we can look at it as a government infrastructure project.
In the mean time, we should assert our rights to the public right-of-ways. The price of access should be that the companies which use it don't get a unilateral choice in how the resulting products are sold.
The cost per fiber drop, according to Bill's estimate, is $1,000-$1,500 if 40 percent of homes participate.
There have to be some crazy assumptions behind that. Taking 12 strands for a mile with no stops is $15k in ideal circumstances. In downtown DC its $175/foot. If your ISP is not the phone company then there's about a 90% chance that its nearest office is more than 10 miles away. Even for the best case the numbers don't compute... And that's without considering the cost of maintenance and equipment to light the fiber.
Fiber works for the phone company because they multiplex it at about a 16:2 ratio within a few hundred yards of your home and then trunk that cable back to an office that's within about 3 miles. Even then they're banking on your purchase of phone, Internet and TV at $150/month to recoup the cost over the next 10 years.
$1500/customer? That's off by at least an order of magnitude. $1500 might cover the raw cost of the cable itself, but that's about it.