Why doesn't it occur to the technology world that we already are getting a very reasonable adoption per year in those areas that have high speed access for a major percentage of the average citizens entertainment budget! No doubt, decreasing the price would help--that it under the companies total control! But, both low speed and high speed bandwidth is getting more expensive and we have a recession.
The worse thing possible is to subsidize the "socalled" high speed access which has not increased in speed for 3 + years -- still stuck @ ~ 500kb for dsl and 3 mbs for cable. They want a "subsidy" for their current business. If we encourage locking in the current technology, when we can get content beyond what dial up provides fine, i.e. current video content, video conterence and need more speed, the users will be locked into deals at the current speed with the incumbent suppliers and no funds will be available to the post-cable/telco era challengers.
In fact, there is in reality not that much bandwidth available for many apps. A large university with a pipe of say 155 mbps may be
able to support only 5000 users at peak times with only 30kbps, but say only 25 streaming video streams at 6 mbps. And that is all!
When you have an average usage for web surfing of less than 10kbps for even cable users and some apps and downloads of > 1,000 x, an no metering and no knowledge by the typical users when they are moving from light usage to dominating an expensive pipe, you have a disaster waiting to happen. The most innocent user can by sending their home video or having a video conference bring a network to their needs.
All you can do is ration, charge, or expect chaos.
Next years file sharing program could be an entirely different usage, say sharing scientific results, your newest documentary, or the family video album. The bits get used, and everyone else is SOL.
As disk drive capacity keeps going up, there is less and less incentive to compress for usage within a home. i just bought a 160 gb drive for $280 - that holds 500 cds @ 350 mb ave losslessly compressed. If I buy an CD, i want it instantly accessible and at as high quality as i bought it. I don't want some sucky MP3 version unless i have to download into a portable device that doesn't have the storage to store full quality.
Too bad for swappers that broadband bandwidth hasn't sped up for the typical consumer in the last 4 years, that forces people who swap to use MP3 or OGG. But for me, i just want to be able to store music that i buy conveniently and in good quality. For that i need to be able to copy to disk for the foreseeable future. I would prefer not to buy a physical CD with all the limits and costs - I'd be glad to buy the songs from the labels or the band - but i'd like full fidelity. Of course that labels and CD stores only stock bands that manage to get a contract (regardless of their quality) and make me buy my music in lots of a dozen songs at a time, but that's another matter.
but even fewer artists make money from concerts than from the labels. I suspect fewer than 1000 bands get more than a couples months minimum pay from concerts (not gigs at an occasional club)
armed with the time and location the money was last transacted, there is a lot to conclude, even if you don't know the entire trail. if you knew for example where and whom each bill was initially recieved or its destination, you know the community of users. Most of those bills will go directly from the person who withdrew them to a merchant or at most a second person before ending up at another bank or merchant. With a timestamp, you can make some really good correlations of activities. Scary. You certainly would be able to find out who used their fresh bill for coke and then spent it!
It seems that the post office can track not only where mail is sorted, but exactly when it went through a specific sorting machine and what went before and after. This came out in the anthrax investigation where they could say when the lady in CT had her letter pass through a machine and which letters came before and after.
I'd suggest tracking money is even more interesting the tracking mail to many in a position of power. Are you all sure it isn't already happening -- there is little incentive to tell anyone.
Virtually all services we share with others are engineered statistically to provide an acceptable level of service at peak load (or not). Whether we are talking about roads, cell phone capacity, ability to get a dial tone, toilet flushing capacity, or dial up internet usage -- unmetered usage requires an average usage which has a cost that is acceptable to the majority of the users.
If these assumptions change, the costs must change, you must go to variable pricing, or you must cap. This is simply an engineering reality--
whether we are talking about mailing letters to small villages in alaska, or all videoconferencing at dinner time on our broadband connection.
P to P applications, work from home users, and servers use a disappropriate amount of bandwidth. The broadband providers have a finite amount of bandwidth at several layers, the local area, the upstream and downstream connections etc. Some usage patterns are 10 x the typical users -- at what point should the burden go onto the users which actually represent far disaproprate amount of actual capacity costs.
There are only four basic choices:
1. Tragedy of the commons: allow all users to degrade. All users are unhappy, but no increase in cost. This was AOLs initial dial strategy when they instituted "unlimited usage".
2. Control the type of usage: This permits the provider to reduce or eliminate certain kind of heavy bandwidth usage--a hidden bandwidth charge--infinite for some types of bandwidth usage. Makes the unaffected users happy at no extra cost. The affected users are totally disenfranchised.
3. Allow unlimited usage without usage controls: This requires adding infrastructure indefinitely. When videos get common swapped, including the live feeds from all the day's soap operas etc, the costs could go through the roof.
4. Price for usage: This is straightforward in principle, but would be painful for many. There would be no need to control the type of usage. There would probably be a peak set of hours with a higher rate, and certainly a different upstream rate than downstream rate. You might have a initial quota that was "free" or included in the base cost. You could do anything you damn well please with the bandwidth, but you might find P-P apps and downloading movies more expensive than buying the content, at least during peak hours. This is also wretchedly difficult to explain to the typically internet user, much less the newby, and you'd likely want to have a $ meter in the corner of your screen so you could quickly kill a "bad" app or a useless download.
All systems have to be engineered to handle a peak load. There is a cost to that load, and you have cover the true cost someway, either by charging proportionally or overcharging the lighter users!
Does anyone realistically have a solution that makes everyone happy and doesn't result in some sort of limitation to heavy users?
The University of Phoenix online is an accredited degree program including masters degrees. They have more online students than many large universities (>10,000). You might find it interesting that they use means for running their classes: Internet Explorer and a web equivalent interface. Since in fact, they use newsgroups and email along with electronic books, there is nothing that isn't available for the open source users.
Note, that this is run of the few IPOs of the last quarter, and it has gone up (2.5 x!), so they must be doing someing right!
They specifically arrange their classes so that students and teachers do not have to be online at the same time. A friend's experience has been quite good, individual attention to assistments and problems from the teachers, several give out their home phone numbers if you get really stuck.
I can't see that watching a video is better than optimizing for asynchronous conversations and getting professors that are willing to talk to students, not at the lecture, but when the help is needed. This does not seem to be a technlogy issue, but having an administration that expects responsiveness from teachers.
Asynchonous availablity is really important for distant learning, otherwise you have a scheduling nightmare depending on the time zone you happen to be in -- unless you are a full-time student.
the reason udp is used is often due to the fact that the congestion control builtin to tcp would be futile, useless, or counterproductive. in the case of voip, if you can't get the packet there in realtime, there is no congestion control necessary or useful, as the packet is no longer of interest, but the next packet is, if it too can make it in time. a different set of design tradeoffs for this application, guaranteed reception is not one of the issues for sure.
Slashdot Mirror
Why doesn't it occur to the technology world that we already are getting a very reasonable adoption per year in those areas that have high speed access for a major percentage of the average citizens entertainment budget! No doubt, decreasing the price would help--that it under the companies total control! But, both low speed and high speed bandwidth is getting more expensive and we have a recession.
The worse thing possible is to subsidize the "socalled" high speed access which has not increased in speed for 3 + years -- still stuck @ ~ 500kb for dsl and 3 mbs for cable. They want a "subsidy" for their current business. If we encourage locking in the current technology, when we can get content beyond what dial up provides fine, i.e. current video content, video conterence and need more speed, the users will be locked into deals at the current speed with the incumbent suppliers and no funds will be available to the post-cable/telco era challengers.
In fact, there is in reality not that much bandwidth available for many apps. A large university with a pipe of say 155 mbps may be
able to support only 5000 users at peak times with only 30kbps, but say only 25 streaming video streams at 6 mbps. And that is all!
When you have an average usage for web surfing of less than 10kbps for even cable users and some apps and downloads of > 1,000 x, an no metering and no knowledge by the typical users when they are moving from light usage to dominating an expensive pipe, you have a disaster waiting to happen. The most innocent user can by sending their home video or having a video conference bring a network to their needs.
All you can do is ration, charge, or expect chaos.
Next years file sharing program could be an entirely different usage, say sharing scientific results, your newest documentary, or the family video album. The bits get used, and everyone else is SOL.
As disk drive capacity keeps going up, there is less and less incentive to compress for usage within a home. i just bought a 160 gb drive for $280 - that holds 500 cds @ 350 mb ave losslessly compressed. If I buy an CD, i want it instantly accessible and at as high quality as i bought it. I don't want some sucky MP3 version unless i have to download into a portable device that doesn't have the storage to store full quality.
Too bad for swappers that broadband bandwidth hasn't sped up for the typical consumer in the last 4 years, that forces people who swap to use MP3 or OGG. But for me, i just want to be able to store music that i buy conveniently and in good quality. For that i need to be able to copy to disk for the foreseeable future. I would prefer not to buy a physical CD with all the limits and costs - I'd be glad to buy the songs from the labels or the band - but i'd like full fidelity. Of course that labels and CD stores only stock bands that manage to get a contract (regardless of their quality) and make me buy my music in lots of a dozen songs at a time, but that's another matter.
but even fewer artists make money from concerts than from the labels. I suspect fewer than 1000 bands get more than a couples months minimum pay from concerts (not gigs at an occasional club)
armed with the time and location the money was last transacted, there is a lot to conclude, even if you don't know the entire trail. if you knew for example where and whom each bill was initially recieved or its destination, you know the community of users. Most of those bills will go directly from the person who withdrew them to a merchant or at most a second person before ending up at another bank or merchant. With a timestamp, you can make some really good correlations of activities. Scary. You certainly would be able to find out who used their fresh bill for coke and then spent it!
It seems that the post office can track not only where mail is sorted, but exactly when it went through a specific sorting machine and what went before and after. This came out in the anthrax investigation where they could say when the lady in CT had her letter pass through a machine and which letters came before and after.
I'd suggest tracking money is even more interesting the tracking mail to many in a position of power. Are you all sure it isn't already happening -- there is little incentive to tell anyone.
Virtually all services we share with others are engineered statistically to provide an acceptable level of service at peak load (or not). Whether we are talking about roads, cell phone capacity, ability to get a dial tone, toilet flushing capacity, or dial up internet usage -- unmetered usage requires an average usage which has a cost that is acceptable to the majority of the users.
If these assumptions change, the costs must change, you must go to variable pricing, or you must cap. This is simply an engineering reality--
whether we are talking about mailing letters to small villages in alaska, or all videoconferencing at dinner time on our broadband connection.
P to P applications, work from home users, and servers use a disappropriate amount of bandwidth. The broadband providers have a finite amount of bandwidth at several layers, the local area, the upstream and downstream connections etc. Some usage patterns are 10 x the typical users -- at what point should the burden go onto the users which actually represent far disaproprate amount of actual capacity costs.
There are only four basic choices:
1. Tragedy of the commons: allow all users to degrade. All users are unhappy, but no increase in cost. This was AOLs initial dial strategy when they instituted "unlimited usage".
2. Control the type of usage: This permits the provider to reduce or eliminate certain kind of heavy bandwidth usage--a hidden bandwidth charge--infinite for some types of bandwidth usage. Makes the unaffected users happy at no extra cost. The affected users are totally disenfranchised.
3. Allow unlimited usage without usage controls: This requires adding infrastructure indefinitely. When videos get common swapped, including the live feeds from all the day's soap operas etc, the costs could go through the roof.
4. Price for usage: This is straightforward in principle, but would be painful for many. There would be no need to control the type of usage. There would probably be a peak set of hours with a higher rate, and certainly a different upstream rate than downstream rate. You might have a initial quota that was "free" or included in the base cost. You could do anything you damn well please with the bandwidth, but you might find P-P apps and downloading movies more expensive than buying the content, at least during peak hours. This is also wretchedly difficult to explain to the typically internet user, much less the newby, and you'd likely want to have a $ meter in the corner of your screen so you could quickly kill a "bad" app or a useless download.
All systems have to be engineered to handle a peak load. There is a cost to that load, and you have cover the true cost someway, either by charging proportionally or overcharging the lighter users!
Does anyone realistically have a solution that makes everyone happy and doesn't result in some sort of limitation to heavy users?
The University of Phoenix online is an accredited degree program including masters degrees. They have more online students than many large universities (>10,000). You might find it interesting that they use means for running their classes: Internet Explorer and a web equivalent interface. Since in fact, they use newsgroups and email along with electronic books, there is nothing that isn't available for the open source users. Note, that this is run of the few IPOs of the last quarter, and it has gone up (2.5 x!), so they must be doing someing right! They specifically arrange their classes so that students and teachers do not have to be online at the same time. A friend's experience has been quite good, individual attention to assistments and problems from the teachers, several give out their home phone numbers if you get really stuck. I can't see that watching a video is better than optimizing for asynchronous conversations and getting professors that are willing to talk to students, not at the lecture, but when the help is needed. This does not seem to be a technlogy issue, but having an administration that expects responsiveness from teachers. Asynchonous availablity is really important for distant learning, otherwise you have a scheduling nightmare depending on the time zone you happen to be in -- unless you are a full-time student.
the reason udp is used is often due to the fact that the congestion control builtin to tcp would be futile, useless, or counterproductive. in the case of voip, if you can't get the packet there in realtime, there is no congestion control necessary or useful, as the packet is no longer of interest, but the next packet is, if it too can make it in time. a different set of design tradeoffs for this application, guaranteed reception is not one of the issues for sure.