I know this has been said again and again, but it's like the music industry doesn't read slashdot;)
You seem to think they can read?
If Jill (Music Industry) gives an encryption key to Jane (a person using a DRM infested player that needs the key to play music) so that Bob the hacker (Who happens to be the same person as Jane) cannot spy on the music being played by Jane and copy it, Jane can still spy on the music and copy it because she is listening to it.
That's equivalent to the analog hole. Music already faces the analog hole big time because things "need" to play on a huge base of analog equipment, from small headphones to big 10,000 watt stereo systems (the real cause of earthquakes and power shortages in California).
At least the motion picture industry got to add a small bit of resistance to the analog hole with the addition of HDCP to block use of analog connections. But this can be circumvented by many people with some electronics knowledge.
And analog does not mean less that perfect copies, either. The analog signal has some clues as to the original digital codes because digital itself is actually imperfect (quantization errors).
There is no way to solve this problem without banning all recording devices and computers that are not running software that is vetted by the government or some corporate entity every time they're turned on. Do you want to live in that world?
Of course not. This is one reason we need to work harder to be sure open source software has a wide user base, and people are accustomed to using it.
...would be to build as just a foundational infrastructure. It would be fiber all the way from each home and business to the various connection points. These would be buildings, not little pedestals. These fibers would then be leased out to any company wanting to provide services over them. There would be 4 actual optical fibers to each home and business (more for special cases), so it would be possible for "light up" providers to offer only one type of service, and customers could get their phone, TV, and internet, from different providers if they so choose. Or people and businesses could lease them directly to have a very high speed point to point service wherever they want.
It's not competing against the telco... it's providing them with a fiber based infrastructure they can use. It's not competing against cable TV... it's providing them with a fiber based infrastructure they can use. It's not competing against broadband services... it's providing them with a fiber based infrastructure they can use.
It's just a road. The city and state generally build roads and let people use them. The directions the telcos and cable TV companies are trying to go is the equivalent to not only them building the road, but also them building all the vehicles and allowing no other vehicles on the road, and them restricting what parts of town people are allowed to even go to.
Cities often provide public transportation. So some basic default services is not out of the question, anyway. But it might get structured so it is not a major competition. For example, it might provide connectivity only within the city itself and not to the world internet. It might carry only over-the-air TV stations, and not all those satellite based national channels.
I'd bet a lot of business would love to jump in and provide services over an infrastructure they don't have to pay all that up front cost to build. Whether it's paid for by leasing the fibers, or by taxes, is something the city would have to decide.
A true bare-metal hypervisor would just run on bare-metal without any assistance of any OS, and would present the appearance of a bare-metal environment (usually one of exactly the same architecture, or architecture class, that it runs on) before any OS is even running on this. There would be no need for special OSes to be run on it. There would be no need for Solaris or Linux to be around. I could run only MS-DOS 1.0 on it, if it were for the x86 architecture.
I agree with this. The law is too vague, and approaches the issue in the wrong way.
Consider someone who wants to print up his own political or commercial speech on sheets of paper and distribute them. So he steals a printing press, paper, and ink, do do this. Does his right to free speech prevail over the property rights of the owner of the press, paper, and ink? What if the owner bought these things to carry out his own right of free speech; the owner's right to free speech have now been trampled on by the thief.
Does one's right to free speech allow them to write their message with paint on the windshields of parked cars?
The law needs to focus on property rights. Free speech does need to be preserved, but it is not free as in free beer. We already have property rights laws in place and they may even be good enough now, if these cases get argued in a way to utilize those laws. And however that might not be the case, new laws that would be needed should focus on exactly how we limit free speech now (e.g. you can't steal a printing press).
E-mail has a special issue. The recipient's property is the major element of delivery for e-mail. The sending server only needs to keep one copy of a bulk e-mailed message. Unlike most traditional mailing lists, most spamware actually does avoid queueing duplicate copies per recipient. And the spamware can readily carry out 1000 SMTP connections per process. The recipient server, due to how some mail may need to be handled, usually has to dedicate a whole process to each incoming message. Only if that message has multiple recipients at that location is any duplication advantage gained. Then the message is "queued" in the recipient mailbox until the mailbox owner picks it up to read it.
Property rights are taken by any email that causes an excess number of processes to be initiated. This can result in delays of mail the recipient really wants to get, and possibly a loss in some cases. This is also an infringement on the right of others wanting to exercise their free speech.
Property rights are taken by any email that fills up the mailbox. This usually results in a non-delivery of email, depriving the mailbox owner of mail that may want. This also infringes the free speech rights of those who want to send mail to that recipient (that the recipient is agreeing to allow his property to be used for).
Striking down an anti-spam law that arbitrarily limits speech is actually the correct ruling. The case itself should have been argued on the issues of the property rights of the recipient, the property rights of other recipients sharing the same server, the property rights of the provider of the services, and the free speech rights infringement of other senders that would be exercising that free speech right in agreement with the recipients.
So by striking down that law, the Virginia Supreme Court did not say we have to allow spam through. It merely said that this particular law is not the right way to deal with it.
... hijack some newspaper web site (easy enough to do) and "publish" his own story about some company going into bankruptcy. Then he or his partner buys up the stock at a discount and waits for the story to be retracted before selling it at the normal price. Someone has to buy stock for the price to be driven back up.
This is just a variation of the old stock scams previously done by email on small penny stocks. But with newspaper web site hijacking, Google news expanding it, and automated trading on Wall Street, this can be done on larger stocks where hundreds of people would be buying in at the bottom, obscuring the scammer in the crowd.
I have a script that fakes a new home directory, and that keeps each new instance of Firefox from passing the URL over to the older process and exiting. This script is run when I launch a new browser, whether with a URL or not. It builds the fake home directory from a template (a tarball), makes the HOME environment variable point at it, and after a few other manipulations, fires up Firefox itself.
In theory it uses more memory. In reality I quit browsers I'm done with and they clean up well when the process exits. So the overall memory usage is actually less. And when one crashes or goes into a loop and needs to be killed (as happened last night on the Comcast web site), I only lose that process. I do not lose all the other browser instances I still may have sitting around and running (typically 5 to 8 started at any given time... one of which is exclusively for Slashdot). It has worked rather well, so far.
... and so I've missed quite many opportunities to refinance my house, to save big money on male enhancement products, and help an ailing Nigerian prince transfer his inheritance to my bank account.
If this fiber-to-the-home growth is based on the FiOS way, then it's really nothing more than an increment above cable TV.
DSL over twisted-pair is a private bandwidth all the way to the central office. There, it is possible to terminate at different providers (even if that termination has to be all or nothing with the same provider).
The cable TV model, however, is shared bandwidth. Although it uses coax with more bandwidth than twisted-pair can deliver, and you can get some significant speeds, you only get it all when others are not using it. When everyone else is using it, you only get a fraction of the promised burst speeds.
FiOS uses the cable TV model. Users share the bandwidth. All the homes in a group are multiplexed into a single bit stream. While this does give an improvement over existing cable TV service (the TV spectrum is on a different optical wavelength than the data), it is still a limiting factor.
And the FiOS model makes it much harder to have competition improve on the type of service provided.
In the future, perhaps as soon as 10 years from now, we'll be wanting TV not from the 500 or so channels FiOS and services like it can offer, but from the millions of "channels" anyone can provide over the internet. But we might not get this with a single provider system that the cable TV model locks us into.
Today we are looking at high definition TV at 1920 x 1080 interlaced at 50 or 60 frames per second. In 10 to 20 years, we'll be seeing ultra definition TV at 5120 x 2160 (cinema aspect, 64:27) progressive at 100 or 120 frames per second. You won't get that over the air. That's over 20 times the needed bandwidth (although it is likely to compress a little better). We'll need at least 200 mbps to deliver such programming.
What I propose is a whole new scheme for data service delivery (DSD) is what I call fiber from provider to customer (FFPTC or just FPC). In this scheme, the fiber (actually a bundle of at least 4 fibers) runs all the way between a home and the central office building. The customer can select which content service provider (CSP) and the fiber terminates into their equipment. Between this provider and this customer over that fiber, they can do all they want. That could be as much as 10 gbps or more.
To do this, there needs to be a clear separation between the delivery and the content. Maybe one company can have both roles, but it really needs to be at minimum separate corporate entities. An alternative would be a government or contracted entity providing this service.
Cost is a huge factor in the improvement of data based services. And most of that cost will initially be in laying the fiber everywhere. This requires very long term returns on investment to really do it right. Businesses that were in it for long term returns would not be doing the risky innovations we still need to provide new kinds of future services. And besides, businesses don't even want to be in the long term returns model. They want fast short term windfall-like growth and profits.
Either we will pay by having a lousy non-competitive system like cable TV and FiOS which cannot sustain the future demands just 10 or 20 years away, or we will over the long term for a system that can serve our needs for the next 100 years or more, includes the content provider competition we need, and drives new economic growth through innovation.
... if using older recording technology that has gaps between tracks and records zeros in raw form. Today's recording involves multi-level coding and scrambling, where even all-zeros will have a big mash-up of flux values, and overlaps the gaps to some degree.
If that 80 GB drive that had been zeroed-out with dd had recorded Osama bin Laden's exact location, you can be sure the data recovery experts at certain nameless US government agencies would scramble to get hold of that drive, regardless. And it would not surprise me if they can recover some data from it. They would not be worried about getting their $60 deposit back, and the drive will likely be destroyed as a hard drive as we know it. The tab for such recovery could be in the millions of dollars, but for that kind of data, it would be worth it.
Is the data on your computer with that to someone?
What middle? It will be on an internal government network that is isolated from the internet. The "man in the middle" will be whatever agency is managing this (CIA? NSA? If I were to tell you I'd have to kill you).
It will be in a private restricted network operated by the government. You can't even send packets on that network without a top secret security clearance. Despite all the screwups of most of government, there are a few things they do right. And you will never come near it.
Comcast management doesn't even get the opportunity to (fail to) understand, because this is something being sold to them by another company. If there are any real geeks working at Comcast, you can be sure they are completely out of any decision/advisory loops.
... delay the traffic of the highest bandwidth (ab)users. By doing this without regard to the content of the traffic, or its TCP port numbers, etc., then they are in a neutral position. How to do this delaying is another matter. They need to avoid focusing on peer-to-peer file sharing just because it happens to be the activity of the biggest users. As long as that is true, focusing on the actual bandwidth hogging will effectively slow down whatever usage is involved.
How to slow down users needs to do something other than forged RST packets. Aside from the legal issues, protocol developers will figure out ways to become RST immune. One simple way is to carry on as of there was no RST and see if a normal packet comes along within a certain time frame (a couple seconds). If not, then the RST is considered real. If there is a normal packet soon enough, then the RST is forged. Comcast is using this technique because it is NOT practical for them to selective drop individual packets in transit; RST forgery is a lower cost injection method. But if they continue this method, geeks will figure out ways around it (plural... there's more than one way to do this).
Ultimately they will have to make it dynamically adjust the bandwidth rate on the customer attachment equipment. If a customer bursts traffic at high rates too much, gradually lower their bandwidth burst rate limit until it reaches the level where continuous traffic solidly for a month equals 250GB.
Actually, it's Flash that ruined the reputation of YouTube. It's not about the video quality. It's about the security. And the browser has a lot to do with this, too, for failing to support video (it supports still pictures and animated GIFs, so why not video and audio, too, as a built-in integrated feature).
As for the encoding, show me a free open source H.264 encoder (even if it is a slow one that takes many times the play time to encode it), or a hardware encoder with free open source drivers that work in BSD and Linux (or at least one of them and can be adapted to the other).
Slashdot Mirror
I know this has been said again and again, but it's like the music industry doesn't read slashdot ;)
You seem to think they can read?
If Jill (Music Industry) gives an encryption key to Jane (a person using a DRM infested player that needs the key to play music) so that Bob the hacker (Who happens to be the same person as Jane) cannot spy on the music being played by Jane and copy it, Jane can still spy on the music and copy it because she is listening to it.
That's equivalent to the analog hole. Music already faces the analog hole big time because things "need" to play on a huge base of analog equipment, from small headphones to big 10,000 watt stereo systems (the real cause of earthquakes and power shortages in California).
At least the motion picture industry got to add a small bit of resistance to the analog hole with the addition of HDCP to block use of analog connections. But this can be circumvented by many people with some electronics knowledge.
And analog does not mean less that perfect copies, either. The analog signal has some clues as to the original digital codes because digital itself is actually imperfect (quantization errors).
There is no way to solve this problem without banning all recording devices and computers that are not running software that is vetted by the government or some corporate entity every time they're turned on. Do you want to live in that world?
Of course not. This is one reason we need to work harder to be sure open source software has a wide user base, and people are accustomed to using it.
... Monticello, Minnesota, then they would just pull out and go away.
...would be to build as just a foundational infrastructure. It would be fiber all the way from each home and business to the various connection points. These would be buildings, not little pedestals. These fibers would then be leased out to any company wanting to provide services over them. There would be 4 actual optical fibers to each home and business (more for special cases), so it would be possible for "light up" providers to offer only one type of service, and customers could get their phone, TV, and internet, from different providers if they so choose. Or people and businesses could lease them directly to have a very high speed point to point service wherever they want.
It's not competing against the telco ... it's providing them with a fiber based infrastructure they can use. It's not competing against cable TV ... it's providing them with a fiber based infrastructure they can use. It's not competing against broadband services ... it's providing them with a fiber based infrastructure they can use.
It's just a road. The city and state generally build roads and let people use them. The directions the telcos and cable TV companies are trying to go is the equivalent to not only them building the road, but also them building all the vehicles and allowing no other vehicles on the road, and them restricting what parts of town people are allowed to even go to.
Cities often provide public transportation. So some basic default services is not out of the question, anyway. But it might get structured so it is not a major competition. For example, it might provide connectivity only within the city itself and not to the world internet. It might carry only over-the-air TV stations, and not all those satellite based national channels.
I'd bet a lot of business would love to jump in and provide services over an infrastructure they don't have to pay all that up front cost to build. Whether it's paid for by leasing the fibers, or by taxes, is something the city would have to decide.
A true bare-metal hypervisor would just run on bare-metal without any assistance of any OS, and would present the appearance of a bare-metal environment (usually one of exactly the same architecture, or architecture class, that it runs on) before any OS is even running on this. There would be no need for special OSes to be run on it. There would be no need for Solaris or Linux to be around. I could run only MS-DOS 1.0 on it, if it were for the x86 architecture.
I agree with this. The law is too vague, and approaches the issue in the wrong way.
Consider someone who wants to print up his own political or commercial speech on sheets of paper and distribute them. So he steals a printing press, paper, and ink, do do this. Does his right to free speech prevail over the property rights of the owner of the press, paper, and ink? What if the owner bought these things to carry out his own right of free speech; the owner's right to free speech have now been trampled on by the thief.
Does one's right to free speech allow them to write their message with paint on the windshields of parked cars?
The law needs to focus on property rights. Free speech does need to be preserved, but it is not free as in free beer. We already have property rights laws in place and they may even be good enough now, if these cases get argued in a way to utilize those laws. And however that might not be the case, new laws that would be needed should focus on exactly how we limit free speech now (e.g. you can't steal a printing press).
E-mail has a special issue. The recipient's property is the major element of delivery for e-mail. The sending server only needs to keep one copy of a bulk e-mailed message. Unlike most traditional mailing lists, most spamware actually does avoid queueing duplicate copies per recipient. And the spamware can readily carry out 1000 SMTP connections per process. The recipient server, due to how some mail may need to be handled, usually has to dedicate a whole process to each incoming message. Only if that message has multiple recipients at that location is any duplication advantage gained. Then the message is "queued" in the recipient mailbox until the mailbox owner picks it up to read it.
Property rights are taken by any email that causes an excess number of processes to be initiated. This can result in delays of mail the recipient really wants to get, and possibly a loss in some cases. This is also an infringement on the right of others wanting to exercise their free speech.
Property rights are taken by any email that fills up the mailbox. This usually results in a non-delivery of email, depriving the mailbox owner of mail that may want. This also infringes the free speech rights of those who want to send mail to that recipient (that the recipient is agreeing to allow his property to be used for).
Striking down an anti-spam law that arbitrarily limits speech is actually the correct ruling. The case itself should have been argued on the issues of the property rights of the recipient, the property rights of other recipients sharing the same server, the property rights of the provider of the services, and the free speech rights infringement of other senders that would be exercising that free speech right in agreement with the recipients.
So by striking down that law, the Virginia Supreme Court did not say we have to allow spam through. It merely said that this particular law is not the right way to deal with it.
I'm using my own hard drive space to backup my own files. That way if my house burns down ... oh wait.
I have more than twice that number of files on my 8 external hard drives.
... hijack some newspaper web site (easy enough to do) and "publish" his own story about some company going into bankruptcy. Then he or his partner buys up the stock at a discount and waits for the story to be retracted before selling it at the normal price. Someone has to buy stock for the price to be driven back up.
This is just a variation of the old stock scams previously done by email on small penny stocks. But with newspaper web site hijacking, Google news expanding it, and automated trading on Wall Street, this can be done on larger stocks where hundreds of people would be buying in at the bottom, obscuring the scammer in the crowd.
... bought UAL stock at just $3? Oh wait.
I have a script that fakes a new home directory, and that keeps each new instance of Firefox from passing the URL over to the older process and exiting. This script is run when I launch a new browser, whether with a URL or not. It builds the fake home directory from a template (a tarball), makes the HOME environment variable point at it, and after a few other manipulations, fires up Firefox itself.
In theory it uses more memory. In reality I quit browsers I'm done with and they clean up well when the process exits. So the overall memory usage is actually less. And when one crashes or goes into a loop and needs to be killed (as happened last night on the Comcast web site), I only lose that process. I do not lose all the other browser instances I still may have sitting around and running (typically 5 to 8 started at any given time ... one of which is exclusively for Slashdot). It has worked rather well, so far.
... and so I've missed quite many opportunities to refinance my house, to save big money on male enhancement products, and help an ailing Nigerian prince transfer his inheritance to my bank account.
If this fiber-to-the-home growth is based on the FiOS way, then it's really nothing more than an increment above cable TV.
DSL over twisted-pair is a private bandwidth all the way to the central office. There, it is possible to terminate at different providers (even if that termination has to be all or nothing with the same provider).
The cable TV model, however, is shared bandwidth. Although it uses coax with more bandwidth than twisted-pair can deliver, and you can get some significant speeds, you only get it all when others are not using it. When everyone else is using it, you only get a fraction of the promised burst speeds.
FiOS uses the cable TV model. Users share the bandwidth. All the homes in a group are multiplexed into a single bit stream. While this does give an improvement over existing cable TV service (the TV spectrum is on a different optical wavelength than the data), it is still a limiting factor.
And the FiOS model makes it much harder to have competition improve on the type of service provided.
In the future, perhaps as soon as 10 years from now, we'll be wanting TV not from the 500 or so channels FiOS and services like it can offer, but from the millions of "channels" anyone can provide over the internet. But we might not get this with a single provider system that the cable TV model locks us into.
Today we are looking at high definition TV at 1920 x 1080 interlaced at 50 or 60 frames per second. In 10 to 20 years, we'll be seeing ultra definition TV at 5120 x 2160 (cinema aspect, 64:27) progressive at 100 or 120 frames per second. You won't get that over the air. That's over 20 times the needed bandwidth (although it is likely to compress a little better). We'll need at least 200 mbps to deliver such programming.
What I propose is a whole new scheme for data service delivery (DSD) is what I call fiber from provider to customer (FFPTC or just FPC). In this scheme, the fiber (actually a bundle of at least 4 fibers) runs all the way between a home and the central office building. The customer can select which content service provider (CSP) and the fiber terminates into their equipment. Between this provider and this customer over that fiber, they can do all they want. That could be as much as 10 gbps or more.
To do this, there needs to be a clear separation between the delivery and the content. Maybe one company can have both roles, but it really needs to be at minimum separate corporate entities. An alternative would be a government or contracted entity providing this service.
Cost is a huge factor in the improvement of data based services. And most of that cost will initially be in laying the fiber everywhere. This requires very long term returns on investment to really do it right. Businesses that were in it for long term returns would not be doing the risky innovations we still need to provide new kinds of future services. And besides, businesses don't even want to be in the long term returns model. They want fast short term windfall-like growth and profits.
Either we will pay by having a lousy non-competitive system like cable TV and FiOS which cannot sustain the future demands just 10 or 20 years away, or we will over the long term for a system that can serve our needs for the next 100 years or more, includes the content provider competition we need, and drives new economic growth through innovation.
... if using older recording technology that has gaps between tracks and records zeros in raw form. Today's recording involves multi-level coding and scrambling, where even all-zeros will have a big mash-up of flux values, and overlaps the gaps to some degree.
If that 80 GB drive that had been zeroed-out with dd had recorded Osama bin Laden's exact location, you can be sure the data recovery experts at certain nameless US government agencies would scramble to get hold of that drive, regardless. And it would not surprise me if they can recover some data from it. They would not be worried about getting their $60 deposit back, and the drive will likely be destroyed as a hard drive as we know it. The tab for such recovery could be in the millions of dollars, but for that kind of data, it would be worth it.
Is the data on your computer with that to someone?
You don't have income tax. And you are fussing about some sales tax?
Except for a-space.gov and a-space.mil.
They could have called it "I-Spy", but I think that name has been taken, already.
They should create a parallel site like Slashdot, called Spydot, where intelligence can be moderated (Score 5: Funny).
What middle? It will be on an internal government network that is isolated from the internet. The "man in the middle" will be whatever agency is managing this (CIA? NSA? If I were to tell you I'd have to kill you).
It will be in a private restricted network operated by the government. You can't even send packets on that network without a top secret security clearance. Despite all the screwups of most of government, there are a few things they do right. And you will never come near it.
Comcast management doesn't even get the opportunity to (fail to) understand, because this is something being sold to them by another company. If there are any real geeks working at Comcast, you can be sure they are completely out of any decision/advisory loops.
... my NYTimes password. Oh wait, I never even registered. Nevermind.
... delay the traffic of the highest bandwidth (ab)users. By doing this without regard to the content of the traffic, or its TCP port numbers, etc., then they are in a neutral position. How to do this delaying is another matter. They need to avoid focusing on peer-to-peer file sharing just because it happens to be the activity of the biggest users. As long as that is true, focusing on the actual bandwidth hogging will effectively slow down whatever usage is involved.
How to slow down users needs to do something other than forged RST packets. Aside from the legal issues, protocol developers will figure out ways to become RST immune. One simple way is to carry on as of there was no RST and see if a normal packet comes along within a certain time frame (a couple seconds). If not, then the RST is considered real. If there is a normal packet soon enough, then the RST is forged. Comcast is using this technique because it is NOT practical for them to selective drop individual packets in transit; RST forgery is a lower cost injection method. But if they continue this method, geeks will figure out ways around it (plural ... there's more than one way to do this).
Ultimately they will have to make it dynamically adjust the bandwidth rate on the customer attachment equipment. If a customer bursts traffic at high rates too much, gradually lower their bandwidth burst rate limit until it reaches the level where continuous traffic solidly for a month equals 250GB.
Actually, it's Flash that ruined the reputation of YouTube. It's not about the video quality. It's about the security. And the browser has a lot to do with this, too, for failing to support video (it supports still pictures and animated GIFs, so why not video and audio, too, as a built-in integrated feature).
As for the encoding, show me a free open source H.264 encoder (even if it is a slow one that takes many times the play time to encode it), or a hardware encoder with free open source drivers that work in BSD and Linux (or at least one of them and can be adapted to the other).
You slashdotted them! No response from their site, now.
What about 1920 x 1080 progressive at 59.94 fps?