1. It allows ISPs to take bandwidth at a $40/Mbps bulk/wholesale rate and parcel it to consumers for $40/6 Mbps (roughly $7/Mbps). If broadband consumers wanted to buy committed bandwidth in the 5-10 Mbps range, they could expect to pay in the range of $70/Mbps (not much bulk). Looks like a 10:1 oversubscription ratio is very fair in comparison to any traditional business model of a supplier buying bulk and marking up.
2. Oversubscription allows Joe Average User to help finance Joe Power User's connection (a fact that/.ers shouldn't be griping about!) You advocates of pay-per-packet would soon find that your BitTorrent usage would drive your monthly bill to double what it used to be, while granny up the street gets a cost break.
3. No data to back this, but consumers seem to prefer unlimited over metered services. Consider the increasing popularity of "unlimited long distance" plans, mega-rollover minutes (wireless marketing spin for flat-rate). Consider the failure of metered consumer services like ISDN! Sheesh, I wish I could get my energy and water bills flat-rated! My guess is that consumers like the freedom of using a communication service as much as they want, and the predictability of knowing what the bill will be each month.
(Despite what the parent says about paying fixed rates, all tier 1 ISPs offer burst-billed services as well, but ISPs tend to like the same thing as their consumers: predictability of cost...makes the budget easier to figure.)
4. The original justification for packet-switched networks vs. circuit-switched networks was to reduce costs and allow scalability. In short, to more efficiently use available network resources. A dedicated conduit to each subscriber is simply inefficient when you want to scale. This same principle is what makes the economics of VoIP disruptive in comparison to traditional telecom.
5. I hope anyone pushing for a committed bandwidth consumer broadband products isn't simultaneously throwing their weight behind Net Neutrality, because it would essentially make it illegal for telcos and cablecos to provide such a service.
---- Re: honesty in marketing, consumer broadband should be marketed as "up to" a given speed. And that's the way the fine print likely reads. But is this any worse than equipment manufacturers that sold you a 56k modems that will never reach 56k? HDD manufacturers that sold you a 1 GB HDD that was 1000 MB instead of 1024? How about the "up to 20 hour battery life" Sony promised you? Those products will NEVER reach their advertised capacities, and at least your broadband may burst to it's maximum on occasion.
And finally, if anyone thinks the economics of running an ISP are so great and broadband providers are all making off like bandits, think again. It's a business of tight margins and heavy price pressure (enough to sink 100s if not 1000s of small ISPs). ---
my home town of Readlyn, Iowa (population ~900 (yes, nine hundred)
Isn't Readlyn the town with a sign that says
Population: 899 smiling faces and 1 grouchy old bastard
Betcha the grouchy old bastard is the business man that dreamed up the goldmine of a captive audience that will pay for high speed Internet access because there sure ain't much else to do in Readlyn.:)
Re: Cedar Falls Utilities' high speed Internet service, the reviews that I've heard basically lean towards 'CFU is finding out that apparently high speed Internet is slightly more complicated than electricity'...IOW, relatively dismal.
However, I'm with you all the way on inventive attitudes in rural small towns. Iowa Network Services (mentioned in this article) was my first dialup provider in my hometown, in conjunction with the rural telephone cooperative (that's their gimmick). But they're about 6 years too late to bring digital cable and high speed 'net to my hometown. When I moved away in '96, 7 Mbps cable and digital cable had been available for a few months. This is in a town of ~5000 people.
The trick is, if you have a small enough town, everybody's 3100 ft from the CO.:-P Or at least maybe a reasonable distance for typical SDSL. Eat your hearts out, suburbia.
Now, if they could just do something about the unforgiving subzero winters and annoyingly hot and humid summers...
...is to avoid doing business with an Incumbent
Local Exchange Provider (ILEC) like USWest!
Even if you order DSL through a CLEC, you're still
using a Baby Bell telephone system for which DSL is a huge hack (and a limited one at that). As
wonderful as I've found DSL to be, many people are
out of range, subject to artifacts of the legacy
phone network, or in some way unable to get DSL.
Cable modems use a backbone designed for broadcast
services, and limit your choice of ISP to your local cable company (If they can't get TV right, what makes us think they can do 'net access?)
A well planned wireless deployment bears an
infrastructure designed from the ground up to
distribute data. Reinforced with redundancy, it should also overcome availability issues.
Most importantly, with a wireless connection,
you're not as likely to have 3-5 companies making
money off just the last few miles of your connection, which should make a significant
difference in price once infrastructure is established.
I guess I'm of the school of thought that the ISPs for mobile devices ought to filter the content and rewrite it to suit mobile phones.
Huh?! Last time I checked, my ISP was responsible
for providing me bandwidth and...well, bandwidth.
Try to think of one reason why many destinations (gobs of ISPs) should try to filter the immense variety of web content, instead of each "content provider," who has easy access to the data, creating output viewable on numerous devices (once, mind you, not once/transfer), before it is in transit.
The idea of rewriting code (ie, http://www.amazon.com/phone/) for different access devices has always struck me as somewhat foolish.
Yeah, different devices, like say, IE and Netscape? It may be inconvenient for web developers to do so, but hopefully standards like this will decrease the likelihood of a dominant proprietary system.
Disclaimer: I work for a DSL provider.
Delays in DSL installation are nearly inevitable,
and these are the causes I see most often:
ILEC claims lack of facilities.
Bridge taps in the line.
Load coils in the line.
Pre-qualification distance is far less than actual distance.
These all stem from the fact that, until a circuit is built, no one knows it's length, or any other potential problem.
It's an inherent problem in retrofitting new technology, DSL, to operate on top of a legacy phone system which is *extremely* varied (this also causes 56K woes). In reality, every DSL provider will have to deal with these issues unless they trench their own cable from their network to yours. (ha!)
DSL is great: cheap, reliable, secure...if you live pretty near a phone CO. Otherwise, expect to
see wireless become a popular solution in you neck of the woods.
Slashdot Mirror
Oversubscription is good for power users!
/.ers shouldn't be griping about!) You advocates of pay-per-packet would soon find that your BitTorrent usage would drive your monthly bill to double what it used to be, while granny up the street gets a cost break.
1. It allows ISPs to take bandwidth at a $40/Mbps bulk/wholesale rate and parcel it to consumers for $40/6 Mbps (roughly $7/Mbps). If broadband consumers wanted to buy committed bandwidth in the 5-10 Mbps range, they could expect to pay in the range of $70/Mbps (not much bulk). Looks like a 10:1 oversubscription ratio is very fair in comparison to any traditional business model of a supplier buying bulk and marking up.
2. Oversubscription allows Joe Average User to help finance Joe Power User's connection (a fact that
3. No data to back this, but consumers seem to prefer unlimited over metered services. Consider the increasing popularity of "unlimited long distance" plans, mega-rollover minutes (wireless marketing spin for flat-rate). Consider the failure of metered consumer services like ISDN! Sheesh, I wish I could get my energy and water bills flat-rated! My guess is that consumers like the freedom of using a communication service as much as they want, and the predictability of knowing what the bill will be each month.
(Despite what the parent says about paying fixed rates, all tier 1 ISPs offer burst-billed services as well, but ISPs tend to like the same thing as their consumers: predictability of cost...makes the budget easier to figure.)
4. The original justification for packet-switched networks vs. circuit-switched networks was to reduce costs and allow scalability. In short, to more efficiently use available network resources. A dedicated conduit to each subscriber is simply inefficient when you want to scale. This same principle is what makes the economics of VoIP disruptive in comparison to traditional telecom.
5. I hope anyone pushing for a committed bandwidth consumer broadband products isn't simultaneously throwing their weight behind Net Neutrality, because it would essentially make it illegal for telcos and cablecos to provide such a service.
----
Re: honesty in marketing, consumer broadband should be marketed as "up to" a given speed. And that's the way the fine print likely reads. But is this any worse than equipment manufacturers that sold you a 56k modems that will never reach 56k? HDD manufacturers that sold you a 1 GB HDD that was 1000 MB instead of 1024? How about the "up to 20 hour battery life" Sony promised you? Those products will NEVER reach their advertised capacities, and at least your broadband may burst to it's maximum on occasion.
And finally, if anyone thinks the economics of running an ISP are so great and broadband providers are all making off like bandits, think again. It's a business of tight margins and heavy price pressure (enough to sink 100s if not 1000s of small ISPs).
---
my home town of Readlyn, Iowa (population ~900 (yes, nine hundred)
:)
:-P Or at least maybe a reasonable distance for typical SDSL. Eat your hearts out, suburbia.
Isn't Readlyn the town with a sign that says
Population: 899 smiling faces and 1 grouchy old bastard
Betcha the grouchy old bastard is the business man that dreamed up the goldmine of a captive audience that will pay for high speed Internet access because there sure ain't much else to do in Readlyn.
Re: Cedar Falls Utilities' high speed Internet service, the reviews that I've heard basically lean towards 'CFU is finding out that apparently high speed Internet is slightly more complicated than electricity'...IOW, relatively dismal.
However, I'm with you all the way on inventive attitudes in rural small towns. Iowa Network Services (mentioned in this article) was my first dialup provider in my hometown, in conjunction with the rural telephone cooperative (that's their gimmick). But they're about 6 years too late to bring digital cable and high speed 'net to my hometown. When I moved away in '96, 7 Mbps cable and digital cable had been available for a few months. This is in a town of ~5000 people.
The trick is, if you have a small enough town, everybody's 3100 ft from the CO.
Now, if they could just do something about the unforgiving subzero winters and annoyingly hot and humid summers...
Even if you order DSL through a CLEC, you're still using a Baby Bell telephone system for which DSL is a huge hack (and a limited one at that). As wonderful as I've found DSL to be, many people are out of range, subject to artifacts of the legacy phone network, or in some way unable to get DSL.
Cable modems use a backbone designed for broadcast services, and limit your choice of ISP to your local cable company (If they can't get TV right, what makes us think they can do 'net access?)
A well planned wireless deployment bears an infrastructure designed from the ground up to distribute data. Reinforced with redundancy, it should also overcome availability issues.
Most importantly, with a wireless connection, you're not as likely to have 3-5 companies making money off just the last few miles of your connection, which should make a significant difference in price once infrastructure is established.
Huh?! Last time I checked, my ISP was responsible for providing me bandwidth and...well, bandwidth.
Try to think of one reason why many destinations (gobs of ISPs) should try to filter the immense variety of web content, instead of each "content provider," who has easy access to the data, creating output viewable on numerous devices (once, mind you, not once/transfer), before it is in transit.
The idea of rewriting code (ie, http://www.amazon.com/phone/) for different access devices has always struck me as somewhat foolish.
Yeah, different devices, like say, IE and Netscape? It may be inconvenient for web developers to do so, but hopefully standards like this will decrease the likelihood of a dominant proprietary system.
Tom
If you're planning to serve SSL pages only, it might be better to compare Apache-SSL and statically compiled mod_ssl, and see which performs the best.
Delays in DSL installation are nearly inevitable, and these are the causes I see most often:
- ILEC claims lack of facilities.
- Bridge taps in the line.
- Load coils in the line.
- Pre-qualification distance is far less than actual distance.
These all stem from the fact that, until a circuit is built, no one knows it's length, or any other potential problem. It's an inherent problem in retrofitting new technology, DSL, to operate on top of a legacy phone system which is *extremely* varied (this also causes 56K woes). In reality, every DSL provider will have to deal with these issues unless they trench their own cable from their network to yours. (ha!)DSL is great: cheap, reliable, secure...if you live pretty near a phone CO. Otherwise, expect to see wireless become a popular solution in you neck of the woods.