Utilities meter because they have to produce what gets used. Every gallon of water you use has to be filtered and pumped to you. Every watt of power you use has to be generated. If there's no demand, it isn't produced.
Internet bandwidth isn't like that. A DS3 is 45mbps - period. Packets or idle pattern, it is always and forever 45mbps. Or put another way, if everyone downloads 10GB this month and 100GB next month, TWC's bill(s) to other ISPs are unchanged. Their bill is based entirely on rate, completely independant of volume. Yet, they want to charge us for both rate and volume.
almost every broadband ISP has overbuilt their network
Nope. They've all overSOLD their network. The whole problem is that they haven't been building their network to keep up. TW's SEC filings clearly (for inventive definitions) show their profits increasing (by ~18%) while operating costs are decreasing (by ~11%.) Over the last decade, they've made on the order of $20 BILLION in profit with operating costs totally approx. $1.5 BILLION. TWC's CEO has been quoted saying the broadband business is highly profitable and their operating costs are constantly falling. Yet, the greedy f***ers want even MORE money, to be even more insanely profitable, while doing almost nothing to improve their actual network. They've even had the balls to state DOCSIS 3.0 rates (when they roll them out) of 50mbps and 100mbps -- which means you get to use it for 1 hour per month before exceeding your cap.
TWC is doing this for exactly two reasons... (1) C. A. S. H. and (2) to kill broadband streaming as a means of proping up the traditional cable TV business.
All this while Time Warner -- the media company and parent company of Time Warner Cable -- plans to offer their media library for streaming to any traditional subscriber. If you have cable or sat, you can stream the same content over the internet as you get on your TV. Gee, I wonder who's content is not going to count towards your cap. *cough*MONOPOLY*cough*
Much! Floppies are highly unstable magnetic storage systems. They suffer what has been termed "bit rot"... the magnetic information weakens over time. USB keys use flash memory that doesn't use magnetic fields for storage. It should, in theory, last hundreds of years vs. a floppy being dead after just a few years.
Copy to a USB HD drive.
Yes, because those things last decades. Don't waste your money on external USB HD's. The constant power cycling and rough handling (plus the changing operating enviroments) usually kills them after just a few years (read: if it lasts 3 years, you're lucky.) The warantee on it should be very telling as to how soon it's going to die.
Burn to CD/DVD. not much better than floppy again:(
Sadly, yes. CD/DVD as an archival medium is a gamble. Many cheap discs will fail after a few years. And all of them degrade when explosed to light -- sunlight will destroy one in few hours.
Tape is still good longterm, a used tape drive perhaps(DDS4+controller=$100)
That's been true for decades. It can be an expensive option, but tape is a proven technology -- NASA has tapes half a century old that can still be read; I have tapes 20 years old that are still perfectly readable and I've done nothing special to preserve them.
Paper and a file cabinet. At least paper is still readable in 10 years.
Sometimes. It depends on the paper, inks, and how it was (mis)handled. Go talk to your closest archivist society if you want to know the particulars.
Paper has it's problems as well. Ever seen a file room get flooded. I have. *grin* A few words on building a proper file room... never put it in the basement or other low area where water would naturally accumulate. Install barriers in the ceiling and walls to prevent water infiltration in the event of leaks. Never put the file cabnets directly on the floor -- elevate them about a foot so there's room for standing water before any documents are sunk. Install water sensors and alarms. And lastly, install some type of air handler(s) to maintain a set temp and humidity -- your papers will last a lot longer that way.
(I've seen flooded data centers as well. But with raised floors, those are kind of unremarkable.)
No you wont. This is not 62.5/50micron multi-mode fiber. A coupler in a single-mode fiber causes a great deal of signal loss. I have never seen anyone terminate SM fiber anywhere but a termination point (i.e. at the equipment, repeater, or patch pannel inside a building.) "Just install a repeater" is laughable... those things are not free and require power that isn't found in most ditches.
Today, we have very good equipment for making fusion splices -- to the point it's almost automatic. The real time consuming process is getting to the fiber to fix it in the first place. Followed closely behind by the tediousness of getting each strand connected to the correct other half.
I switched to CFL when I moved into my townhouse 2 years ago. I've replaced the light in the kitchen 4 times -- I'm on bulb 5! None of the remaining incandecent lights have been replaced, making them more than 2 years old. (some of them much more than 2.) Contrast that with the 3ft full FL lights in my former apartment... they hadn't been changed since the building was built in '95 -- I lived there 11 years and never changed them.
No it's not. I routinely netboot systems for repairs, upgrades, reimaging, etc. And with USB booting available on almost everything these days, it's just a matter of walking up to it...
Hard drives cheaper... depends on how you look at it, but I'll give you that one. But they most f'ing certainly DO NOT hold their data longer. I have boxes of new drives that failed in less than 2 years. I have drives that sat on a shelf for 6 months and were no longer reliablly readable -- being IDE, they are now trash. Expensive enterprise (read: SCSI and FC) drives may last for 5+ years, but they have small capacities and huge costs. (and even they don't fair well sitting on a shelf -- but they can be low-level formated and put back into production, unlike today's cheap shit IDE, SATA, and SAS drives that put their firmware on the platters instead of a $0.10 EEPROM.)
Compatibility? Over the span of decades, everything grows old and obsolete. It won't be long before PCs are made without IDE (PATA) interfaces at all. Not too long ago, every Mac used SCSI -- now they're all IDE/SATA. Sun workstations used to be all SCSI (and even FC), but they went to IDE a long time ago -- and I stopped buying them because of it. As another commenter pointed out, yes, I have an MFM drive from decades long past -- even have controller cards for it, but I don't have any operational machines with ISA ports to run it (and the data is almost certainly gone.) I used to have an ESDI system; good luck finding that hardware today. Anyone remember EISA or MCA? (I have an alpha with EISA slots. And an RS6000 with micro-channel slots. Neither have been on for years.)
And for a real world example, I have a 25-ish year old Exabyte 8mm "8205" tape, last written to 15 years ago. At a consulting gig about 2 years ago, I popped it into their Eliant 820 drive and read every mm of that tape with zero issues. Let's see you read back the data on a hard drive that's been banging around in a drawer in your kitchen for a decade. (Note: my warehouse only has sony AIT drives.)
Input Output ----- ------ Protocol Packet Count Packet Count Byte Count Byte Count 30sec Bit Rate (bps) 30sec Bit Rate (bps) 30sec Max Bit Rate (bps) 30sec Max Bit Rate (bps)
Total 1714020 1017809 2113060473 198288797 4000 3000 5887000 586000
That's 2.1G in and 198M out in about a day. And no, I don't waste my time with bittorrent at home -- the connection's too slow.
(Note: NBAR is hideously inaccurate at protocol matching as it only looks at port numbers. But the totals are correct and are 64bit counters so they don't roll over easily.)
Since when does TWC have anything to do with networking in Oz? This is entire arguement is about the evolving practices of a US based ISP doing business in parts of the USA. Bottom line... they, like almost all US based ISPs, have never sold "unlimited" consumer accounts with any usage caps. But now, they've "seen the light" and want to tap into that massive pile of cash in the form of "overage fees" that have made cellphone providers so insanely rich.
What rubs everyone the wrong way is how much they out-right lie about their motivations. Money. Money is their entire motivation for everything. They push up the advertised rates ("up to X Mbps") to remain competitive -- yet they're bumping into the limits of what they can realistically provide. But then cap your "unlimited" connection to laughablly low levels to generate more revenue and otherwise discourage you from getting media anywhere else but from them (read: cable tv, dvd, and cd.) (i.e. stiffling innovation.) And to futher the insult, when confronted about it, they dismiss everyone with "it's only a trial" and "we have no plans to roll this out everywhere" (100% bullshit) They've even been quoted as saying people with contracts will see no change; one of their HUGE ad claims over their competition (assuming there is any) is the entire lack of any contracts of any kind.
The speed at their various interconnects aren't really at issue. The network, as is, is working just fine. Even with all these evil bittorrent users cloging the tubes, it still works just fine. However without upgrading their (decade old) technology, they cannot keep pushing the rates higher and higher as a response to DSL and FiOS -- who are upgrading their technology.
Correct. However, there are other problems... the biggest is the move away from traditional pay TV service (read: cable) to internet streaming, which means a (huge) drop in revenue. The other big issue is the amount of bandwidth necessary to provide DOCSIS 3.0 speeds. Many operators have extreme congestion in the cable network (due entirely to all this BS on-demand and iControl crap) so doubling or tripling the number of channels used by data services is going to be a hard sell. And moving to high frequencies means upgrading a lot of (expensive) equipment. (most cable networks only go to 850Mhz, which is as high as they needed to go in the days of analog cable.)
The problem with T1s are the regulations. Tariffs dictate the price -- which is hundreds of times higher than it actually costs to deliver today. We pay ~500$ per month for a T1 (voice and data), but the "last mile" is a 300ft run -- all inside the building -- from an optical shelf to our phone room. (the same shelf feeds us a DS3, 'tho I don't see the bills for that one.)
Just to carify... they can change the rate on your modem pretty much instantly. It's a number in the modem's config file. My speed has been increased 3 times over the last few years, and all it takes is a modem reset (so it loads the new config.) DOCSIS 3.0 requires new hardware to do channel bonding.
No. No they cannot. Look at all the municple wifi projects, and city run fiber networks that have been sued relentlessly by both cable and telco companies. While it may be cheap on paper, the red tape and legal hurdles are daunting.
Actually, ISDN proved to be a bit of a problem for the aging infrastructure. When ISDN was installed to my apartment in '97, Bellsouth had to run 4 miles of trunk cable first (seems noone told them about 800 new apartment being built:-)), then it took "Bob" 3 hours to find a pair that would work. After that, it worked flawlessly (short of a CO crash and a lightning strike) for 11 years. I'd still have it if the world could be seen through a 64k/128k straw.
Plus, ISDN is way too complicated for the average house wife. You cannot simply plug the phone in and go. I've seen a lot of gear attempt "auto spid", but none have ever worked properly.
You do realize gmail uses significantly more bandwidth than traditional email applications. POP clients download the message once. Good IMAP clients retrieve a message only once. (Note: PINE is not one of them.) Web email systems, of which gmail is one, render and send you the message (wrapped in a bunch of HTML) every time you click on it. Unless you delete most of your email without reading it, Eudora (etc.) is much less "expensive" than gmail.
Internet innovations have always come at the cost of increased bandwidth. Putting a cap on a large portion of the internet will kill those innovations. People will have to be aware of not only how long it takes to load a flash app, but the actual money cost of those bytes. And then there's the waste of everybody's time, and now money, of ads stuck everywhere on every web page out there.
Step away from this constant "everybody is downloading full rate HD broadcast TV shows" crap. Yes, some people do that, but the numbers show most aren't -- and the networks certainly won't be able to support everyone doing it, ever. The issue is everyday, "normal", network usage... downloading email, browsing the web, IM crap (twitter, etc.), voip, skype, voice/video conferencing, windows patches, downloading the new version of itunes,... Just browsing the web -- slashdot for instance -- uses far more bandwidth than people realize. And the trend is upwards, and has been for a long time.
Exactly. They've upped the advertised rate (the link rate to the customer) to boost business and retain customers, but have not invested in the necessary infrastructure to actually support it. It has always been assumed "90% will use 10% of their bandwidth and 10% will use 90%". That's still (mostly) true today. However, 10% of 5M is 500k, not the 50-100k typical a few years ago.
The issue is that they've done business without caps for decades. But now want to add caps, and worse fees for going over some stupid low limit -- with no warning. They created this mess by continually increasing the available speed to customers without investing in adequate supporting systems. They oversold their network by a factor of thousands, and now it's bitten their ass off. That highly profitable business is about to tank -- upgrades are expensive and users aren't going to take higher bills when they have other options. It's not as if they've been blind-sided, either. Their traffic engineering data has been showing the rise in usage for many years. And in all seriousness, I see no need, ZERO, for any of this crap... even with the massive oversubscription, it still works just fine. No, everyone cannot get full speed all the time. But everyone doesn't need to or, as the numbers show, want to.
And honestly, what's wrong with selling cheaper packages with slower rates instead. Or throttling a connection instead of $1/GB overage fees. The answer is, of course, they want more money and $1/GB extortion is how they plan to get it.
No, it's not. Hulu, Youtube, netflix and blockbuster streaming, the ever increasing application of Flash... The "web" is much higher bandwidth place today than it was even a year ago. Or do you want to go back to what the "web" looked like in 1995... almost entirely ASCII text with no formating, very few graphics, no flashy interactive content, just text and links to more text. Today, the protocol overhead that you never see is larger than the entire web page of the pre-broadband days. But that's not innovation, that's just a bunch of people sharing TV shows.
Without the innovations that gave us broadband in the first place, we wouldn't have much of a world-wide-web. There'd be no rich text, formated email (which personally, I wish didn't exist.) People wouldn't be getting emailed pictures of their grandkids. Streaming video would still be tiny, grainy, low frame rate, crap. There would be no java, javascript, flash, or any of the richness they've brought to the world.
Limit the amount of bandwidth people get to use, and you will end up holding back internet innovation. If you only get to download 40GB per month, why does anyone need a 100Mbps DOCSIS3.0 modem?
bandwidth is a shared resource. Internet pipes aren't cheap
True and true. However, ISPs have always oversold their capacity by rather large (and ever increasing) margins. In the dialup days, 10:1 to 12:1 was a "good ISP". Today cable providers oversell by over 100:1. (I once worked for an ISP that sold 1.5M DSL connections off a DSLAM with only a 768k connection.) Their entire business model is "screw the customer". They want to advertise 10M service for $19.95 per month, charge you $50 per month, and not even provide 1M of service. "We'll sell you a 100M connection, but only let you use it for 3.9s per month."
Throttling connections improves throughput for others -- or it can where the network is significantly overloaded. However, it means no increase in revenue. How many times do I have to say it? It's all about their lust for money. They cannot jack the prices up because it will drive away business. DSL is cheaper (and faster) in almost every market where RR is sold. FiOS has forced them to offer speeds their network cannot support (for more than a half dozen people.) They have begrudgingly increased speeds only after competing technologies best them -- 3M cable when DSL was 1.5M (a decade ago); when DSL was commonly available @ 5M, cable matched it; when DSL was commonly available @ 6-7M, cable matched it again; where FiOS has popped up, they'll sell 10/1 (and better) but rare see it. Thing is, in 1995 -- pre-DOCSIS anything -- the technology was capable of 30/10 (max.) -- that's what an "uncapped" modem would give you, and yes, it's very disruptive to everyone else. Today, almost 15 years later, they're still using the same technology. To get any faster -- and I'm talking way faster -- requires channel bonding, which is part of DOCSIS 3.0. But that's new, expensive equipment, and consumes precious cable TV space, and means people will use even more bandwidth.
BTW, RoadRunner already has 4 speed based teirs. They've advertised them for several years. (because DSL always has... gotta be like the Jones's!)
Here in Raleigh, NC, when I last looked, TWC used 6 QAMs for data services. That's an average of ~74TB per month per node (headend), in total bits. That means, with their 40GB tier, ~2000 customers can use 100% of their cap. I don't know that they put that many people on a single node, but I wouldn't be surprised.
The network is handling everything just fine right now. But the trend is for higher usage. And their network won't be able to handle that for long. So, they want to stop that trend -- hold back innovation, if you will -- and at the same time, make more money. (infrastructure is expensive.) I don't know how to make it any clearer... they do not want to lose customers; they want to increase revenue by charging people more money. (The longer they can delay spending for upgrades, means more money they get to put in their pockets. And for the record, I think they're out of rope on that one.)
Actually, my (bellsouth) DSL is faster than the cablemodem... 7Mbps vs. 5Mbps. Plus, it's cheaper, and it includes a static IP. (There's no FiOS here, so TW has no reason to push higher speeds.)
Slashdot Mirror
Utilities meter because they have to produce what gets used. Every gallon of water you use has to be filtered and pumped to you. Every watt of power you use has to be generated. If there's no demand, it isn't produced.
Internet bandwidth isn't like that. A DS3 is 45mbps - period. Packets or idle pattern, it is always and forever 45mbps. Or put another way, if everyone downloads 10GB this month and 100GB next month, TWC's bill(s) to other ISPs are unchanged. Their bill is based entirely on rate, completely independant of volume. Yet, they want to charge us for both rate and volume.
Nope. They've all overSOLD their network. The whole problem is that they haven't been building their network to keep up. TW's SEC filings clearly (for inventive definitions) show their profits increasing (by ~18%) while operating costs are decreasing (by ~11%.) Over the last decade, they've made on the order of $20 BILLION in profit with operating costs totally approx. $1.5 BILLION. TWC's CEO has been quoted saying the broadband business is highly profitable and their operating costs are constantly falling. Yet, the greedy f***ers want even MORE money, to be even more insanely profitable, while doing almost nothing to improve their actual network. They've even had the balls to state DOCSIS 3.0 rates (when they roll them out) of 50mbps and 100mbps -- which means you get to use it for 1 hour per month before exceeding your cap.
TWC is doing this for exactly two reasons... (1) C. A. S. H. and (2) to kill broadband streaming as a means of proping up the traditional cable TV business.
All this while Time Warner -- the media company and parent company of Time Warner Cable -- plans to offer their media library for streaming to any traditional subscriber. If you have cable or sat, you can stream the same content over the internet as you get on your TV. Gee, I wonder who's content is not going to count towards your cap. *cough*MONOPOLY*cough*
Much! Floppies are highly unstable magnetic storage systems. They suffer what has been termed "bit rot"... the magnetic information weakens over time. USB keys use flash memory that doesn't use magnetic fields for storage. It should, in theory, last hundreds of years vs. a floppy being dead after just a few years.
Yes, because those things last decades. Don't waste your money on external USB HD's. The constant power cycling and rough handling (plus the changing operating enviroments) usually kills them after just a few years (read: if it lasts 3 years, you're lucky.) The warantee on it should be very telling as to how soon it's going to die.
Sadly, yes. CD/DVD as an archival medium is a gamble. Many cheap discs will fail after a few years. And all of them degrade when explosed to light -- sunlight will destroy one in few hours.
That's been true for decades. It can be an expensive option, but tape is a proven technology -- NASA has tapes half a century old that can still be read; I have tapes 20 years old that are still perfectly readable and I've done nothing special to preserve them.
Sometimes. It depends on the paper, inks, and how it was (mis)handled. Go talk to your closest archivist society if you want to know the particulars.
Paper has it's problems as well. Ever seen a file room get flooded. I have. *grin* A few words on building a proper file room... never put it in the basement or other low area where water would naturally accumulate. Install barriers in the ceiling and walls to prevent water infiltration in the event of leaks. Never put the file cabnets directly on the floor -- elevate them about a foot so there's room for standing water before any documents are sunk. Install water sensors and alarms. And lastly, install some type of air handler(s) to maintain a set temp and humidity -- your papers will last a lot longer that way.
(I've seen flooded data centers as well. But with raised floors, those are kind of unremarkable.)
No you wont. This is not 62.5/50micron multi-mode fiber. A coupler in a single-mode fiber causes a great deal of signal loss. I have never seen anyone terminate SM fiber anywhere but a termination point (i.e. at the equipment, repeater, or patch pannel inside a building.) "Just install a repeater" is laughable... those things are not free and require power that isn't found in most ditches.
Today, we have very good equipment for making fusion splices -- to the point it's almost automatic. The real time consuming process is getting to the fiber to fix it in the first place. Followed closely behind by the tediousness of getting each strand connected to the correct other half.
I switched to CFL when I moved into my townhouse 2 years ago. I've replaced the light in the kitchen 4 times -- I'm on bulb 5! None of the remaining incandecent lights have been replaced, making them more than 2 years old. (some of them much more than 2.) Contrast that with the 3ft full FL lights in my former apartment... they hadn't been changed since the building was built in '95 -- I lived there 11 years and never changed them.
No it's not. I routinely netboot systems for repairs, upgrades, reimaging, etc. And with USB booting available on almost everything these days, it's just a matter of walking up to it...
Hard drives cheaper... depends on how you look at it, but I'll give you that one. But they most f'ing certainly DO NOT hold their data longer. I have boxes of new drives that failed in less than 2 years. I have drives that sat on a shelf for 6 months and were no longer reliablly readable -- being IDE, they are now trash. Expensive enterprise (read: SCSI and FC) drives may last for 5+ years, but they have small capacities and huge costs. (and even they don't fair well sitting on a shelf -- but they can be low-level formated and put back into production, unlike today's cheap shit IDE, SATA, and SAS drives that put their firmware on the platters instead of a $0.10 EEPROM.)
Compatibility? Over the span of decades, everything grows old and obsolete. It won't be long before PCs are made without IDE (PATA) interfaces at all. Not too long ago, every Mac used SCSI -- now they're all IDE/SATA. Sun workstations used to be all SCSI (and even FC), but they went to IDE a long time ago -- and I stopped buying them because of it. As another commenter pointed out, yes, I have an MFM drive from decades long past -- even have controller cards for it, but I don't have any operational machines with ISA ports to run it (and the data is almost certainly gone.) I used to have an ESDI system; good luck finding that hardware today. Anyone remember EISA or MCA? (I have an alpha with EISA slots. And an RS6000 with micro-channel slots. Neither have been on for years.)
And for a real world example, I have a 25-ish year old Exabyte 8mm "8205" tape, last written to 15 years ago. At a consulting gig about 2 years ago, I popped it into their Eliant 820 drive and read every mm of that tape with zero issues. Let's see you read back the data on a hard drive that's been banging around in a drawer in your kitchen for a decade. (Note: my warehouse only has sony AIT drives.)
It's not one full day, but we'll assume it is...
gw#show ip nbar protocol-discovery interface e0/0
Input Output
----- ------
Protocol Packet Count Packet Count
Byte Count Byte Count
30sec Bit Rate (bps) 30sec Bit Rate (bps)
30sec Max Bit Rate (bps) 30sec Max Bit Rate (bps)
Total 1714020 1017809
2113060473 198288797
4000 3000
5887000 586000
That's 2.1G in and 198M out in about a day. And no, I don't waste my time with bittorrent at home -- the connection's too slow.
(Note: NBAR is hideously inaccurate at protocol matching as it only looks at port numbers. But the totals are correct and are 64bit counters so they don't roll over easily.)
Since when does TWC have anything to do with networking in Oz? This is entire arguement is about the evolving practices of a US based ISP doing business in parts of the USA. Bottom line... they, like almost all US based ISPs, have never sold "unlimited" consumer accounts with any usage caps. But now, they've "seen the light" and want to tap into that massive pile of cash in the form of "overage fees" that have made cellphone providers so insanely rich.
What rubs everyone the wrong way is how much they out-right lie about their motivations. Money. Money is their entire motivation for everything. They push up the advertised rates ("up to X Mbps") to remain competitive -- yet they're bumping into the limits of what they can realistically provide. But then cap your "unlimited" connection to laughablly low levels to generate more revenue and otherwise discourage you from getting media anywhere else but from them (read: cable tv, dvd, and cd.) (i.e. stiffling innovation.) And to futher the insult, when confronted about it, they dismiss everyone with "it's only a trial" and "we have no plans to roll this out everywhere" (100% bullshit) They've even been quoted as saying people with contracts will see no change; one of their HUGE ad claims over their competition (assuming there is any) is the entire lack of any contracts of any kind.
The speed at their various interconnects aren't really at issue. The network, as is, is working just fine. Even with all these evil bittorrent users cloging the tubes, it still works just fine. However without upgrading their (decade old) technology, they cannot keep pushing the rates higher and higher as a response to DSL and FiOS -- who are upgrading their technology.
Well, I had the advantage of working for an ISP :-)
A lot more than they want to speed to support all the 5M modems right now.
Correct. However, there are other problems... the biggest is the move away from traditional pay TV service (read: cable) to internet streaming, which means a (huge) drop in revenue. The other big issue is the amount of bandwidth necessary to provide DOCSIS 3.0 speeds. Many operators have extreme congestion in the cable network (due entirely to all this BS on-demand and iControl crap) so doubling or tripling the number of channels used by data services is going to be a hard sell. And moving to high frequencies means upgrading a lot of (expensive) equipment. (most cable networks only go to 850Mhz, which is as high as they needed to go in the days of analog cable.)
The problem with T1s are the regulations. Tariffs dictate the price -- which is hundreds of times higher than it actually costs to deliver today. We pay ~500$ per month for a T1 (voice and data), but the "last mile" is a 300ft run -- all inside the building -- from an optical shelf to our phone room. (the same shelf feeds us a DS3, 'tho I don't see the bills for that one.)
(BTW, it's VDSL for that 300ft.)
Just to carify... they can change the rate on your modem pretty much instantly. It's a number in the modem's config file. My speed has been increased 3 times over the last few years, and all it takes is a modem reset (so it loads the new config.) DOCSIS 3.0 requires new hardware to do channel bonding.
No. No they cannot. Look at all the municple wifi projects, and city run fiber networks that have been sued relentlessly by both cable and telco companies. While it may be cheap on paper, the red tape and legal hurdles are daunting.
Yes, but they still have the "unwritten" 250GB cap; which will get you booted if you exceed repeatedly.
Actually, ISDN proved to be a bit of a problem for the aging infrastructure. When ISDN was installed to my apartment in '97, Bellsouth had to run 4 miles of trunk cable first (seems noone told them about 800 new apartment being built :-)), then it took "Bob" 3 hours to find a pair that would work. After that, it worked flawlessly (short of a CO crash and a lightning strike) for 11 years. I'd still have it if the world could be seen through a 64k/128k straw.
Plus, ISDN is way too complicated for the average house wife. You cannot simply plug the phone in and go. I've seen a lot of gear attempt "auto spid", but none have ever worked properly.
You do realize gmail uses significantly more bandwidth than traditional email applications. POP clients download the message once. Good IMAP clients retrieve a message only once. (Note: PINE is not one of them.) Web email systems, of which gmail is one, render and send you the message (wrapped in a bunch of HTML) every time you click on it. Unless you delete most of your email without reading it, Eudora (etc.) is much less "expensive" than gmail.
Internet innovations have always come at the cost of increased bandwidth. Putting a cap on a large portion of the internet will kill those innovations. People will have to be aware of not only how long it takes to load a flash app, but the actual money cost of those bytes. And then there's the waste of everybody's time, and now money, of ads stuck everywhere on every web page out there.
Step away from this constant "everybody is downloading full rate HD broadcast TV shows" crap. Yes, some people do that, but the numbers show most aren't -- and the networks certainly won't be able to support everyone doing it, ever. The issue is everyday, "normal", network usage... downloading email, browsing the web, IM crap (twitter, etc.), voip, skype, voice/video conferencing, windows patches, downloading the new version of itunes, ... Just browsing the web -- slashdot for instance -- uses far more bandwidth than people realize. And the trend is upwards, and has been for a long time.
Exactly. They've upped the advertised rate (the link rate to the customer) to boost business and retain customers, but have not invested in the necessary infrastructure to actually support it. It has always been assumed "90% will use 10% of their bandwidth and 10% will use 90%". That's still (mostly) true today. However, 10% of 5M is 500k, not the 50-100k typical a few years ago.
The issue is that they've done business without caps for decades. But now want to add caps, and worse fees for going over some stupid low limit -- with no warning. They created this mess by continually increasing the available speed to customers without investing in adequate supporting systems. They oversold their network by a factor of thousands, and now it's bitten their ass off. That highly profitable business is about to tank -- upgrades are expensive and users aren't going to take higher bills when they have other options. It's not as if they've been blind-sided, either. Their traffic engineering data has been showing the rise in usage for many years. And in all seriousness, I see no need, ZERO, for any of this crap... even with the massive oversubscription, it still works just fine. No, everyone cannot get full speed all the time. But everyone doesn't need to or, as the numbers show, want to.
And honestly, what's wrong with selling cheaper packages with slower rates instead. Or throttling a connection instead of $1/GB overage fees. The answer is, of course, they want more money and $1/GB extortion is how they plan to get it.
No, it's not. Hulu, Youtube, netflix and blockbuster streaming, the ever increasing application of Flash... The "web" is much higher bandwidth place today than it was even a year ago. Or do you want to go back to what the "web" looked like in 1995... almost entirely ASCII text with no formating, very few graphics, no flashy interactive content, just text and links to more text. Today, the protocol overhead that you never see is larger than the entire web page of the pre-broadband days. But that's not innovation, that's just a bunch of people sharing TV shows.
Without the innovations that gave us broadband in the first place, we wouldn't have much of a world-wide-web. There'd be no rich text, formated email (which personally, I wish didn't exist.) People wouldn't be getting emailed pictures of their grandkids. Streaming video would still be tiny, grainy, low frame rate, crap. There would be no java, javascript, flash, or any of the richness they've brought to the world.
Limit the amount of bandwidth people get to use, and you will end up holding back internet innovation. If you only get to download 40GB per month, why does anyone need a 100Mbps DOCSIS3.0 modem?
True and true. However, ISPs have always oversold their capacity by rather large (and ever increasing) margins. In the dialup days, 10:1 to 12:1 was a "good ISP". Today cable providers oversell by over 100:1. (I once worked for an ISP that sold 1.5M DSL connections off a DSLAM with only a 768k connection.) Their entire business model is "screw the customer". They want to advertise 10M service for $19.95 per month, charge you $50 per month, and not even provide 1M of service. "We'll sell you a 100M connection, but only let you use it for 3.9s per month."
Throttling connections improves throughput for others -- or it can where the network is significantly overloaded. However, it means no increase in revenue. How many times do I have to say it? It's all about their lust for money. They cannot jack the prices up because it will drive away business. DSL is cheaper (and faster) in almost every market where RR is sold. FiOS has forced them to offer speeds their network cannot support (for more than a half dozen people.) They have begrudgingly increased speeds only after competing technologies best them -- 3M cable when DSL was 1.5M (a decade ago); when DSL was commonly available @ 5M, cable matched it; when DSL was commonly available @ 6-7M, cable matched it again; where FiOS has popped up, they'll sell 10/1 (and better) but rare see it. Thing is, in 1995 -- pre-DOCSIS anything -- the technology was capable of 30/10 (max.) -- that's what an "uncapped" modem would give you, and yes, it's very disruptive to everyone else. Today, almost 15 years later, they're still using the same technology. To get any faster -- and I'm talking way faster -- requires channel bonding, which is part of DOCSIS 3.0. But that's new, expensive equipment, and consumes precious cable TV space, and means people will use even more bandwidth.
BTW, RoadRunner already has 4 speed based teirs. They've advertised them for several years. (because DSL always has... gotta be like the Jones's!)
Here in Raleigh, NC, when I last looked, TWC used 6 QAMs for data services. That's an average of ~74TB per month per node (headend), in total bits. That means, with their 40GB tier, ~2000 customers can use 100% of their cap. I don't know that they put that many people on a single node, but I wouldn't be surprised.
The network is handling everything just fine right now. But the trend is for higher usage. And their network won't be able to handle that for long. So, they want to stop that trend -- hold back innovation, if you will -- and at the same time, make more money. (infrastructure is expensive.) I don't know how to make it any clearer... they do not want to lose customers; they want to increase revenue by charging people more money. (The longer they can delay spending for upgrades, means more money they get to put in their pockets. And for the record, I think they're out of rope on that one.)
Ummm, what contract? One of their biggest selling points in every ad campaign I've seen in the last few years is the whole "no contracts" line.
Actually, my (bellsouth) DSL is faster than the cablemodem... 7Mbps vs. 5Mbps. Plus, it's cheaper, and it includes a static IP. (There's no FiOS here, so TW has no reason to push higher speeds.)