How is this supposed fairness issue of drive to work versus drive to the bar different from being taxed on the gas you consume like you already are? Oh, right, because you won't know what jurisdiction to credit the taxes to. As if that can't be solved just as accurately by pooling the taxes in a region and then counting cars.
Honestly, the notion that the government needs to track where I drive in order to tax me fairly is truly bizarre.
So tax based milage times a formula derived from your home and work addresses and require corporate (but not personal) cars to carry a GPS logger and tax it differently. Every tax-based argument you can throw at me for GPS trackers in personal vehicles is trivially worked around.
Mileage based user fees? Couldn't they simply read the odometer during the annual or biannual safety and emissions inspection? Why do they need to track you via GPS to assess a road use fee?
in order to manage a queue, you have to have a queue to manage. For most of the Internet connections, their performance is unacceptable long before you have a queue large enough to make management of the queue noticeably affect the performance.
Yes, that's exactly what I was trying to say. You just said it more eloquently.
A traffic prioitization service already exists. It's Akamai's whole business model: They buy pipes to strategic locations with many service providers, cache servers near the customer and route requests to the best-choice server. You buy space on their servers and your data gets to the customer faster.
What Mr Smith wants to do is, well, asinine. He wants to allow the data pipes on his network to fill to 100% and then prioritize the traffic based on who pays. This suggests such a flawed understanding of the technology that as the chief technology officer, he should be fired.
See, here's the problem: For a router to make a priority-based switching decision between packets, it has to have more than one packet cached in memory waiting for free space in the outgoing pipe. But, if you havn't started transmitting the first packet by the time the second packet finishes arriving then you've already lost the speed game. Fast service means that you don't hold on to the packets. You send them out the next link as soon as you get them. Any other architecture would result in transmission speeds that are two to three times slower, even for the highest priority packets! Duh!
So if you don't want your network to suck rocks, you still have to keep the utilization below 80%, and if you keep the utilization down then except for rare bursts of traffic the prioritization function will never be used.
As a search engine, why on earth would I buy priority on your network knowing that either A) it almost never gets used or B) your network is piss slow either way? Answer: I wouldn't.
Fire Mr. Smith. He doesn't understand the technology he's charged with overseeing.
1300 gigs = 1,300,000,000,000 bytes = 10400000000000 bits = 10400000 megabits. At 100 mbps, that's at least 104000 seconds or 29 hours to copy the whole thing.
Yeah, backups are going to be a problem unless of course that machine IS the backup server where the backups are being stored.
I'm looking at the picture and it looks like a solid front-plate. I don't care how many fans you use, you can't pull air through a solid piece of plastic. Meager ventilation equals burnout city. Poor design. It may be a month, it may be a year but those drives will fail before the box is removed for obsolescence.
Actually, if you have a coupler (they're $5-$15) you can join two SC cables together the same way you join two RJ45 cables together with a coupler. Works just like an extension cord, no expertise needed. You get some signal loss, but unless you couple too many cables together in a row it will still work fine.
And with kits like Amphenol's Lightcrimp Plus, you don't need tremendous expertise to crimp fiber optic cables either. Its more expensive than copper (about $1200 for the kit and about $10 per jack) but its about the same level of skill as crimping an RJ45 cable with very good results.
INVERSELY proportional. Ohm's Law: P=EI which can be rewritten as I=P/E. As power consumption (P) by the PC remains constant, the necessary amperage (I) varies as the inverse of the voltage (E). If the voltage drops by a multiple of.4 (120->48) then the amperage jumps by a multiple of 2.5 (1/0.4=2.5) and vice versa.
Yes, I'm oversimplifying since AC and DC volts aren't equivalent, but that certainly doesn't change Ohm's Law to mean its opposite!
If current doubles, power quadruples. If voltage doubles, current also doubles, and power quadruples.
Dude, you need to retake basic electricity. P=EI. If only I doubles, P doubles. If I and E both double then P quadruples, but nothing in the equations requires E and I to double at the same time.
Not sure how you figure it saves power since you still use the bulk of the power supply logic to convert the voltage from the 48V supply down to the 12V, 5V and 3.3V that the server uses... And still have to convert the incoming mains power from AC to DC somewhere in the building.
Delivering power at 48V instead of 120V means you're dealing with 150% more amperage... Transmission loss generally follows the amperage. That's why they boost the voltage on the high-power lines. And its the amperage not the voltage that kills you. That's why many folks survive high-voltage lightning strikes.
You'll also pay three times as much for the power supplies since they're not made in the same quantity as normal PC power supplies.
There are only really two major benefits to DC power systems:
1) AC power induces currents in any nearby parallel wires, such as your network cables. DC power doesn't. 2) Battery systems are trivial to implement in a DC power system: simply connect the batteries in parallel with the main supply with some simple cutoff circuitry for if the voltage drops too low.
The general user doesn't interface with IP addresses at all. He sees www.google.com, not 72.14.207.99.
Barring a killer app which just won't work without IPv6 (and that's very unlikely) the general user won't drive v6 adoption. The network and systems administrators will. They care very much about having to buy and maintain two sets of addresses indefinately instead of just one.
No, you just misunderstood my complaint. I'll paraphrase your example: I want my PS2 to play my PS1 games without also buying a PS1.
A better metaphor is this: I don't want Windows to dual-boot to DOS so I can run my DOS programs, I want it to run my DOS programs in a Window.
IPv6 can't use my IPv4 addresses. The sockets API can recognize a mapped IPv4 address and establish IPv4 connections for me, but it doesn't do IPv6 connections with my existing IPv4 addresses. If I want to actually use IPv6 I have to get new addresses. THAT is a lack of backwards compatibility.
While the actual packets on the network will be IPv4, the logical connection will be presented as an IPv6 one to the application.
That ain't it. It allows compatibility at the sockets API (like I said) but it doesn't actually allow you to establish IPv6 connections using a valid, routable address mapped from the your existing IPv4 registration.
Call me when that host attempts an IPv6 connection to the mapped IPv4 address using its mapped IPv4 address as the source. And when the router understands that mask to mean "refer to the IPv4 BGP table to route this IPv6 packet." THEN I'll be excited about deploying v6.
Hrmph. V6 is compatible with V4 the way Netbeui is compatible with IPX. Which is to say its not.
Internet Protocol Version 6 is a backwards-compatible replacement for the current Internet protocol
Is this true? I was under the impression that the compatibility more or less ended at the socket API. Is the v4 address space actually mapped in to the v6 address space now so that hosts with v4 addresses are automatically capable of talking v6 if there is a v6 path?
No? That's what I thought. No, you have to go buy (cha-ching) seperate v6 space a number all your servers and routers with two seperate addresses, one v4 and one v6, manage new DNS for your v6 hosts, etc. etc. v1 had more compatibility with v4 than v4 has with v6. At least with the move to v4 the existing registrations mapped in to the new address space.
Feel free to point me at the documentation that says I'm wrong about this. No, really, I would like to be wrong about this. But last I heard they wanted to start the registration process over from scratch with this move and that means you don't have backwards compatibility.
Yes, it should. Why? Because if it doesn't, Red Hat and Novell (one or the other) will make one and the other will follow suite. And the one they make will end up being mediocre with ease of forward porting given priority over effectiveness.
Novell has bet the farm on Linux and they need this to keep the hardware vendors happy and convince them to make the move from supporting Netware to supporting SuSE. If we refuse it from the main kernel tree then it'll be yet another incompatibility in the vendor kernels and yet another roadblock to using a generic kernel on those distros.
Netflix has already established what the market will bear. Its a little under $2 per movie (12 movies a month for around $22). Yes, I know that's for rentals, not purchases (wink, nod). Sony is welcome to try for $8 but they're in for a painful learning experience if they do.
Excellent points, but you need some pretty tall ceilings to make convection work. Otherwise the cabinets that pull more air through than is coming out of the floor create a circulation loop above the cabinet so that the top machines are sucking on their own hot air. With low ceilings, injecting the air at the top diffuses any circulating hot air.
In fairness, there is also a big difference between designing a 10,000 foot data center and a 1,000 foot computer room. In a large data center the added underfloor pathway has an intrinsic value of its own as you endeavor to cope with the massive amounts of cabling and piping and ducting that you need. In the small comptuer room its superfluous; you have at most two air conditioners anyway and the flow pattern is trivial. The raised floor with the additional building code requirements and the space consumed by ramps and so forth is a waste of money in the small computer room.
Raised floor cooling was designed back when the computer room held mainframe and telephone switch equipment with vertical boards in 5-7 foot tall cabinets. The tile was holed or removed directly under each cabinet, so cool air flowed up, past the boards and out through the top of the cabinet. It then wandered its way across the ceiling to the air conditioners' intakes and the cycle repeated.
Telecom switching equipment still uses vertically mounted boards for the most part and still expects to intake air from the bottom and exhaust it out the top. Have any AT&T/Lucent/Avaya equipment in your computer room? Go look.
Now look at your rack mount computer case. Doesn't matter which one. Does it suck air in at the bottom and exhaust it out at the top? No. No, it doesn't. Most suck air in the front and exhaust it out the back. Some suck it in one side and exhaust it out the other. The bottom is a solid slab of metal which obstructs 100% of any airflow directed at it.
Gee, how's that going to work?
Well, the answer is: with some hacks. Now the holed tiles are in front of the cabinet instead of under it. But wait, that basically defeats the purpose of using the raised floor to move air in the first place. Worse, that mild draft of cold air competes with the rampaging hot air blown out of the next row of cabinets. So, for the most part your machines get to suck someone elses hot air!
So what's the solution? A hot aisle / cold aisle approach. Duct cold air overhead to the even-numbered aisles. Have the front of the machines face that cold aisle in the cabinets to either side. Duct the hot air back from the odd-numbered aisles to the air conditioners. Doesn't matter that the hot aisles are 10-15 degrees hotter than the cold aisles because air from the hot aisles doesn't enter the machines.
As I recall, the most fun game sessions I was involved in back in the day were where the DM basically threw the rules to the winds. They existed as a framework so that we all more or less had the same picture of what was going on, but the DM decided who got hit and how badly generally without stopping to record the score. In fact, the DMs who spent 10 minutes scoring a 10-second round of combat were major killjoys.
Perhaps that's why the D&D-branded computer RPG's have not been the top performers. As a backdrop for the imaginatino it was fantastic but if you actually followed the rules it was no fun.
I don't disagree with you, but Granny won't comprehend your argument on an emotional level and she votes too. Convince her that competition would bring back the live operator and she may help you but when you complain about the price of Internet service when she pays ten bucks for Net Zero if she uses it at all, well, its not a compelling argument -- particularly if the oposition can credibly claim that your facts are bogus.
The article is a bunch of bunk. With the exception of Canada, no country on the list has large rural areas. Korea, for example, you could drive between any two points in about the same time it takes more than a few in the US to commute to work each day. The huge swaths of nothing that are common in the US make broadband for those areas more expensive.
US Urban areas just about all have decent broadband now too. Canada has rural broadband because the government subsidizes a big chunk of the cost -- not something we should be asking of our federal government in a time of record deficits.
The author goes on to lament that 3mbps is too slow for high definition video without mentioning that the average bit rate on a DVD movie is 3 mbps while the fixed bit rate on a directv channel is 2 mbps. With newer technologies like mpeg4, 1.5 mbps is more than enough to support television quality video. Right now. Today. Before you even consider the technologies in the pipeline.
And lets talk about one of the technologies in the pipeline. Verizon, awful hated monopoly Verizon, is deploying FiOS at a rate that exceeds 2,000,000 homes per year. That's fiber optic to the home. Rates start at $50 for 15mbps. Add Vonage for $25 like I have and what the heck are you complaining about?
Don't get me wrong. Some of his comments about the FCC are dead on target. Bush's FCC has seriously screwed the pooch on competition. But the data the author uses to support his case is weak at best, leaning towards farcical.
The customers aren't demanding it. I'd still do it for the heck of it, but when the customers aren't demanding it I can't justify creating a multi-thousand dollar obligation to ARIN and filling out a mass of paperwork to get an allocation that's actually usable on the Internet. And of course the folks I buy bandwidth from aren't using it either for largely the same reason, so it does me no good.
If either Level3 or Cogent was buying a "default" service from a third party, their customers wouldn't have a problem. The moment the peering connection was cut the lower-priority BGP routes from the third party would have taken over and their traffic would have gone through the third-party link.
The reason these two jokers are having this problem is that they made a business decision to only move traffic with reciprocal peering and then failed to keep that peering alive. That's because they're both cheap-ass bastards; peering costs a heck of a lot less than buying transit.
Go buy from someone else who who isn't a cheap-ass. Someone who buys transit for anything they can't peer. You won't have a problem.
The only lesson here is that most time honored of lessons: you get what you pay for.
They don't like the fact that the U.S. has complete physical control over [.com,.net and.org].
Tough beans. We got here first, we created it and those three domains are now ours. While I can see why other countries may want to cash in at this late date, I can't imagine a reason why the US bend over and take it.
They aren't monopolies. There are quite literally hundreds of top level domains to choose from. If you don't like the US rules for.com, don't register in.com. If you register in.com, understand that you're buying a US product.
a www user's client will nearly always resolve the address by first hitting on a U.S. controlled root server
This is controlled at the ISP level. The ISP's servers are typically configured to start at the US root servers for non-authorative domains.
This is trivially changable for any country that (A) wants to and (B) has at least some competence in DNS. They simply set up their own root servers (starting with the 62kb anonymous-ftpable root zone from the ICANN) and then madate that all ISPs in the country use their root servers instead. Truthfully, the ISP'd get better performance anyway running traffic to a nearby server.
Slashdot Mirror
How is this supposed fairness issue of drive to work versus drive to the bar different from being taxed on the gas you consume like you already are? Oh, right, because you won't know what jurisdiction to credit the taxes to. As if that can't be solved just as accurately by pooling the taxes in a region and then counting cars.
Honestly, the notion that the government needs to track where I drive in order to tax me fairly is truly bizarre.
So tax based milage times a formula derived from your home and work addresses and require corporate (but not personal) cars to carry a GPS logger and tax it differently. Every tax-based argument you can throw at me for GPS trackers in personal vehicles is trivially worked around.
Mileage based user fees? Couldn't they simply read the odometer during the annual or biannual safety and emissions inspection? Why do they need to track you via GPS to assess a road use fee?
in order to manage a queue, you have to have a queue to manage. For most of the Internet connections, their performance is unacceptable long before you have a queue large enough to make management of the queue noticeably affect the performance.
Yes, that's exactly what I was trying to say. You just said it more eloquently.
A traffic prioitization service already exists. It's Akamai's whole business model: They buy pipes to strategic locations with many service providers, cache servers near the customer and route requests to the best-choice server. You buy space on their servers and your data gets to the customer faster.
What Mr Smith wants to do is, well, asinine. He wants to allow the data pipes on his network to fill to 100% and then prioritize the traffic based on who pays. This suggests such a flawed understanding of the technology that as the chief technology officer, he should be fired.
See, here's the problem: For a router to make a priority-based switching decision between packets, it has to have more than one packet cached in memory waiting for free space in the outgoing pipe. But, if you havn't started transmitting the first packet by the time the second packet finishes arriving then you've already lost the speed game. Fast service means that you don't hold on to the packets. You send them out the next link as soon as you get them. Any other architecture would result in transmission speeds that are two to three times slower, even for the highest priority packets! Duh!
So if you don't want your network to suck rocks, you still have to keep the utilization below 80%, and if you keep the utilization down then except for rare bursts of traffic the prioritization function will never be used.
As a search engine, why on earth would I buy priority on your network knowing that either A) it almost never gets used or B) your network is piss slow either way? Answer: I wouldn't.
Fire Mr. Smith. He doesn't understand the technology he's charged with overseeing.
Hmm, lets see...
1300 gigs = 1,300,000,000,000 bytes = 10400000000000 bits = 10400000 megabits. At 100 mbps, that's at least 104000 seconds or 29 hours to copy the whole thing.
Yeah, backups are going to be a problem unless of course that machine IS the backup server where the backups are being stored.
I'm looking at the picture and it looks like a solid front-plate. I don't care how many fans you use, you can't pull air through a solid piece of plastic. Meager ventilation equals burnout city. Poor design. It may be a month, it may be a year but those drives will fail before the box is removed for obsolescence.
Try this link: http://www.ampnetconnect.com/systems/fiber/FiberSy s.asp?SysCom=0&partOrSysID=16&systemname=LightCrim p%20Plus%20System
Actually, if you have a coupler (they're $5-$15) you can join two SC cables together the same way you join two RJ45 cables together with a coupler. Works just like an extension cord, no expertise needed. You get some signal loss, but unless you couple too many cables together in a row it will still work fine.
And with kits like Amphenol's Lightcrimp Plus, you don't need tremendous expertise to crimp fiber optic cables either. Its more expensive than copper (about $1200 for the kit and about $10 per jack) but its about the same level of skill as crimping an RJ45 cable with very good results.
Amps and volts are proportional (Ohm's Law).
.4 (120->48) then the amperage jumps by a multiple of 2.5 (1/0.4=2.5) and vice versa.
INVERSELY proportional. Ohm's Law: P=EI which can be rewritten as I=P/E. As power consumption (P) by the PC remains constant, the necessary amperage (I) varies as the inverse of the voltage (E). If the voltage drops by a multiple of
Yes, I'm oversimplifying since AC and DC volts aren't equivalent, but that certainly doesn't change Ohm's Law to mean its opposite!
If current doubles, power quadruples. If voltage doubles, current also doubles, and power quadruples.
Dude, you need to retake basic electricity. P=EI. If only I doubles, P doubles. If I and E both double then P quadruples, but nothing in the equations requires E and I to double at the same time.
Not sure how you figure it saves power since you still use the bulk of the power supply logic to convert the voltage from the 48V supply down to the 12V, 5V and 3.3V that the server uses... And still have to convert the incoming mains power from AC to DC somewhere in the building.
Delivering power at 48V instead of 120V means you're dealing with 150% more amperage... Transmission loss generally follows the amperage. That's why they boost the voltage on the high-power lines. And its the amperage not the voltage that kills you. That's why many folks survive high-voltage lightning strikes.
You'll also pay three times as much for the power supplies since they're not made in the same quantity as normal PC power supplies.
There are only really two major benefits to DC power systems:
1) AC power induces currents in any nearby parallel wires, such as your network cables. DC power doesn't.
2) Battery systems are trivial to implement in a DC power system: simply connect the batteries in parallel with the main supply with some simple cutoff circuitry for if the voltage drops too low.
The general user doesn't interface with IP addresses at all. He sees www.google.com, not 72.14.207.99.
Barring a killer app which just won't work without IPv6 (and that's very unlikely) the general user won't drive v6 adoption. The network and systems administrators will. They care very much about having to buy and maintain two sets of addresses indefinately instead of just one.
No, you just misunderstood my complaint. I'll paraphrase your example: I want my PS2 to play my PS1 games without also buying a PS1.
A better metaphor is this: I don't want Windows to dual-boot to DOS so I can run my DOS programs, I want it to run my DOS programs in a Window.
IPv6 can't use my IPv4 addresses. The sockets API can recognize a mapped IPv4 address and establish IPv4 connections for me, but it doesn't do IPv6 connections with my existing IPv4 addresses. If I want to actually use IPv6 I have to get new addresses. THAT is a lack of backwards compatibility.
From the link:
While the actual packets on the network will be IPv4, the logical connection will be presented as an IPv6 one to the application.
That ain't it. It allows compatibility at the sockets API (like I said) but it doesn't actually allow you to establish IPv6 connections using a valid, routable address mapped from the your existing IPv4 registration.
Call me when that host attempts an IPv6 connection to the mapped IPv4 address using its mapped IPv4 address as the source. And when the router understands that mask to mean "refer to the IPv4 BGP table to route this IPv6 packet." THEN I'll be excited about deploying v6.
Hrmph. V6 is compatible with V4 the way Netbeui is compatible with IPX. Which is to say its not.
Internet Protocol Version 6 is a backwards-compatible replacement for the current Internet protocol
Is this true? I was under the impression that the compatibility more or less ended at the socket API. Is the v4 address space actually mapped in to the v6 address space now so that hosts with v4 addresses are automatically capable of talking v6 if there is a v6 path?
No? That's what I thought. No, you have to go buy (cha-ching) seperate v6 space a number all your servers and routers with two seperate addresses, one v4 and one v6, manage new DNS for your v6 hosts, etc. etc. v1 had more compatibility with v4 than v4 has with v6. At least with the move to v4 the existing registrations mapped in to the new address space.
Feel free to point me at the documentation that says I'm wrong about this. No, really, I would like to be wrong about this. But last I heard they wanted to start the registration process over from scratch with this move and that means you don't have backwards compatibility.
Yes, it should. Why? Because if it doesn't, Red Hat and Novell (one or the other) will make one and the other will follow suite. And the one they make will end up being mediocre with ease of forward porting given priority over effectiveness.
Novell has bet the farm on Linux and they need this to keep the hardware vendors happy and convince them to make the move from supporting Netware to supporting SuSE. If we refuse it from the main kernel tree then it'll be yet another incompatibility in the vendor kernels and yet another roadblock to using a generic kernel on those distros.
Netflix has already established what the market will bear. Its a little under $2 per movie (12 movies a month for around $22). Yes, I know that's for rentals, not purchases (wink, nod). Sony is welcome to try for $8 but they're in for a painful learning experience if they do.
Excellent points, but you need some pretty tall ceilings to make convection work. Otherwise the cabinets that pull more air through than is coming out of the floor create a circulation loop above the cabinet so that the top machines are sucking on their own hot air. With low ceilings, injecting the air at the top diffuses any circulating hot air.
In fairness, there is also a big difference between designing a 10,000 foot data center and a 1,000 foot computer room. In a large data center the added underfloor pathway has an intrinsic value of its own as you endeavor to cope with the massive amounts of cabling and piping and ducting that you need. In the small comptuer room its superfluous; you have at most two air conditioners anyway and the flow pattern is trivial. The raised floor with the additional building code requirements and the space consumed by ramps and so forth is a waste of money in the small computer room.
Raised floor cooling was designed back when the computer room held mainframe and telephone switch equipment with vertical boards in 5-7 foot tall cabinets. The tile was holed or removed directly under each cabinet, so cool air flowed up, past the boards and out through the top of the cabinet. It then wandered its way across the ceiling to the air conditioners' intakes and the cycle repeated.
Telecom switching equipment still uses vertically mounted boards for the most part and still expects to intake air from the bottom and exhaust it out the top. Have any AT&T/Lucent/Avaya equipment in your computer room? Go look.
Now look at your rack mount computer case. Doesn't matter which one. Does it suck air in at the bottom and exhaust it out at the top? No. No, it doesn't. Most suck air in the front and exhaust it out the back. Some suck it in one side and exhaust it out the other. The bottom is a solid slab of metal which obstructs 100% of any airflow directed at it.
Gee, how's that going to work?
Well, the answer is: with some hacks. Now the holed tiles are in front of the cabinet instead of under it. But wait, that basically defeats the purpose of using the raised floor to move air in the first place. Worse, that mild draft of cold air competes with the rampaging hot air blown out of the next row of cabinets. So, for the most part your machines get to suck someone elses hot air!
So what's the solution? A hot aisle / cold aisle approach. Duct cold air overhead to the even-numbered aisles. Have the front of the machines face that cold aisle in the cabinets to either side. Duct the hot air back from the odd-numbered aisles to the air conditioners. Doesn't matter that the hot aisles are 10-15 degrees hotter than the cold aisles because air from the hot aisles doesn't enter the machines.
As I recall, the most fun game sessions I was involved in back in the day were where the DM basically threw the rules to the winds. They existed as a framework so that we all more or less had the same picture of what was going on, but the DM decided who got hit and how badly generally without stopping to record the score. In fact, the DMs who spent 10 minutes scoring a 10-second round of combat were major killjoys.
Perhaps that's why the D&D-branded computer RPG's have not been the top performers. As a backdrop for the imaginatino it was fantastic but if you actually followed the rules it was no fun.
I don't disagree with you, but Granny won't comprehend your argument on an emotional level and she votes too. Convince her that competition would bring back the live operator and she may help you but when you complain about the price of Internet service when she pays ten bucks for Net Zero if she uses it at all, well, its not a compelling argument -- particularly if the oposition can credibly claim that your facts are bogus.
The article is a bunch of bunk. With the exception of Canada, no country on the list has large rural areas. Korea, for example, you could drive between any two points in about the same time it takes more than a few in the US to commute to work each day. The huge swaths of nothing that are common in the US make broadband for those areas more expensive.
US Urban areas just about all have decent broadband now too. Canada has rural broadband because the government subsidizes a big chunk of the cost -- not something we should be asking of our federal government in a time of record deficits.
The author goes on to lament that 3mbps is too slow for high definition video without mentioning that the average bit rate on a DVD movie is 3 mbps while the fixed bit rate on a directv channel is 2 mbps. With newer technologies like mpeg4, 1.5 mbps is more than enough to support television quality video. Right now. Today. Before you even consider the technologies in the pipeline.
And lets talk about one of the technologies in the pipeline. Verizon, awful hated monopoly Verizon, is deploying FiOS at a rate that exceeds 2,000,000 homes per year. That's fiber optic to the home. Rates start at $50 for 15mbps. Add Vonage for $25 like I have and what the heck are you complaining about?
Don't get me wrong. Some of his comments about the FCC are dead on target. Bush's FCC has seriously screwed the pooch on competition. But the data the author uses to support his case is weak at best, leaning towards farcical.
The customers aren't demanding it. I'd still do it for the heck of it, but when the customers aren't demanding it I can't justify creating a multi-thousand dollar obligation to ARIN and filling out a mass of paperwork to get an allocation that's actually usable on the Internet. And of course the folks I buy bandwidth from aren't using it either for largely the same reason, so it does me no good.
If either Level3 or Cogent was buying a "default" service from a third party, their customers wouldn't have a problem. The moment the peering connection was cut the lower-priority BGP routes from the third party would have taken over and their traffic would have gone through the third-party link.
The reason these two jokers are having this problem is that they made a business decision to only move traffic with reciprocal peering and then failed to keep that peering alive. That's because they're both cheap-ass bastards; peering costs a heck of a lot less than buying transit.
Go buy from someone else who who isn't a cheap-ass. Someone who buys transit for anything they can't peer. You won't have a problem.
The only lesson here is that most time honored of lessons: you get what you pay for.
They don't like the fact that the U.S. has complete physical control over [.com, .net and .org].
.com, don't register in .com. If you register in .com, understand that you're buying a US product.
Tough beans. We got here first, we created it and those three domains are now ours. While I can see why other countries may want to cash in at this late date, I can't imagine a reason why the US bend over and take it.
They aren't monopolies. There are quite literally hundreds of top level domains to choose from. If you don't like the US rules for
a www user's client will nearly always resolve the address by first hitting on a U.S. controlled root server
This is controlled at the ISP level. The ISP's servers are typically configured to start at the US root servers for non-authorative domains.
This is trivially changable for any country that (A) wants to and (B) has at least some competence in DNS. They simply set up their own root servers (starting with the 62kb anonymous-ftpable root zone from the ICANN) and then madate that all ISPs in the country use their root servers instead. Truthfully, the ISP'd get better performance anyway running traffic to a nearby server.