You've written a very indepth and informative post, but you are still wrong on a couple of issues. ADSL access in the uk market is not sold as a commodity. In the LLU market things are different but the vast majority of ISPs in this country are still on Datastream / IPstream products. So it is not as fanciful as you claim for somebody to be paying per packet. The BT pricing structure is actually based on the size of the outgoing pipe (not the customer links) and is sold by capacity. The charging of the ISP by BT is based on the usage of this pipe - it is NOT a product of how many customers they are connected to. The pricing model is here.
Apart from the fact that we are obviously talking about different markets (a lot of what you've written appears to be the US market for broadband) I agree with you about most of your points. But I still have to disagree about the costs that a cable company incurs. Yes - they don't have enough upstream bandwidth. Yes - it's a tree network where the packets have to reach the head. BUT, if their costs rise because they've developed their network the wrong way then it is still a capital / maintenance cost over the entire network. That is not the same as a cost directly incurred by transiting traffic to another network. In particular their network costs are very "steppy" and as long as their internal pipes are big enough they can handle small increases in internal traffic before they incur a cost to make an upgrade.
Any ISP that is not big enough to peer (most of them) is charged for upstream according to traffic. So ANY increase will cause a direct increase in their costs. This is not an apples to apples comparison as you claim. Partly this is because we are describing two different markets, but here the situation is very different between the cable providers and the DSL companies. At the moment telewest (one of the larger cable companies here) is offering a bi-directional 10Mb link for a reasonable price. They don't care about internal bandwidth because they have capacity coming out of their ass. They don't care about traffic because they are large enough to peer at Linx and they have a mixture of consumer and server business to get a profitable peering arrangement.
You're making the same mistake as the OP. "Free" versus nonfree traffic is not a question of capital costs. ISPs have to pay per packet that moves upstream and off their network (depending on their peering status). The question of building infrastructure is irrelevant. It is a cost that the ISP has to invest to be in business, and can be written off against whichever of their activities seems most appropriate. It is not a DIRECT cost of the traffic. But every packet that transits to a backbone has a DIRECT cost for them. Every packet that remains within their network incurs no cost.
It sounds nothing like multicasting. One is limiting peers based on their address ranges, and the other is broadcasting to multiple peers at once. Now, if multicasting did work properly it would revolutionise p2p as it would break the limit (total_upload=total_download) across the swarm.
Who gets stung with a deal like that? The UK is considered to be an example of what can go wrong with non-unbundled services, but even here ISPs paid a flat rental fee for access to the line, and then had to deal with BT for upstream access. Most the of the current bandwidth limits here are a direct result of how BT charges for that upstream bandwidth. Something like this system would be a great boon for them.
You've completely missed the point. Regardless of how the cable infrastructure is arranged within the ISPs own network - they get chared at the transit point. Traffic between their own customers does not transit to another network at any point and so it is free for them.
Although you are theoretically correct what you have said is wrong in practice. If I want to run a classical algorithm on a larger piece of data then I can just simulate a larger computer (to the extent that I page things in or out of memory, implement 64bit operations in 32bit etc). For a quantum computer I need a bigger computer. I can't simulate a 8-qubit machine on a 4-qubit machine with just a polynomial slowdown.
I can't read the actual article at home so I don't know how large their machine is. Shor's algorithm has actually been run on a 4-qubit machine before so the summary is incorrect. I believe that the number they factored was 15. The point being that I need a quantum machine large enough to factor the RSA number. As building a 8-qubit machine is not as simple as slapping two 4-qubit machines together (because of problems with quantum coherence) there will always be a state-of-the-art for how large a Quantum Computer can be, and public crypto with keys significantly larger than that will be safe until a larger machine is developed. Sort of a faster version of the battle between cryptographers and cryptoanalysts that we see at the moment.
You'll notice that nobody made the same claims when EPFL sieved a 1024-bit number recently - instead everyone said use larger keys. The situation is likely to be the same as Quantum Computers increase in size. Lastly, not all public key crypto is shafted, only things that rely on factorisation as a problem. ECC will be quite safe until (if?) somebody develops a quantum algorithm for discrete logs.
Disclaimer: I don't do research in quantum - I work in cryptography, but the quantum guys have an office down the corridor and occasionally I understand what they are talking about. Ashley, don't beat me around the head for getting the details wrong:)
You know, if you don't get the literary reference that the GP is alluding to then maybe you should look it up instead of posting a sniping response. Given the way it was modded I think you are in the minority. Here's a hint: look up quotes by Arthur C. Clarke.
Magic, as the term is commonly used (especially by hackers) is anything that you don't understand. It doesn't imply a supernatural explanation in this context. The empirical approach that the GP described is exactly how we turn magic into science.
You may want to investigate the relevation that different people have different opinions, and that getting a +5 insightful mod doesn't make you authoritative.
If I write to a block that is already in use (say 2) then I have two blocks in flight, my new block N, and the contents of block 2 which either have to go somewhere else or get cached in memory. So *moving* the contents of a block so that I can write into that location takes two writes instead of one. How is this improving the duty cycle?
OK, all three replies came up with the same points but I'll reply to you since you've put them most succinctly. Your point about reservation makes sense, and it does change the overall picture. Spare blocks on the disk would increase the ability to do load leveling. So lets consider the case where we have a disk with 10 blocks, and lets say 4 are free (these can be a mixture of really free, and reserved blocks). If I repeatedly write to a used block (updating / overwriting) then I have five choices of where to put the data - either the original block, or mark that block as free and pick one of the other 4 blocks.
So I claim that the load can be leveled over 50% of the disk at most - this depends entirely (as you've pointed out) on my second assumption. If I decide that I'm going to pick one of the other 5 used slots to write to, then I have to keep that data. So that block needs to be copied to one of the original choice of 5 places. Unless I choose to keep the dirty data in memory, and hope that I don't run out of storage / lose the power then I can indeed level over all 10 blocks in the disk. But is it reasonable to assume that I can cache the dirty values this way to level over the whole disk rather than just the free space?
You're assuming that the 2GB a day could be spread evenly over the disk. This would vary depending on how much free space you have on the device. If your drive is 1% full then you can distribute your writes over the other 99%. But most people don't keep their storage mainly empty. In fact people tend to run just under the limit - hence the saying that crap always expands to fill the available space. If your drive was 99% full then you can't distribute the writes over the parts with data (as it would have to be moved somewhere else negating the benefit), and then you run into the problem with the limited duty cycle.
Having said all of that, I don't think my throughput is anything like 2Gb, and most of it would be swap (hasn't happened much this past couple of years) and/tmp. Given that/tmp would be better suited to a RAM disk anyway I don't think that either would pose a problem, and the lifespan of these flash disks is probably comparable to a magnetic platter. As another reply pointed out, when the duty cycle is exceeded you can't alter the sector anymore. On a magnetic disk when a sector dies you're SOFL. Once the price comes down to an afforable level these drives will be beautiful...
The author doesn't mention the copyright-transfer form (if any) that he signed. If he refused to sign the copyright transfer form then it is hard to understand how the journal could publish it in the first place. But if he did then he has a fair point as they don't have the right to sell the published form of the paper.
It's nice to hear somebody raising this as a problem because the current system of copyright transfers is a bitch. However, he wants to distribute the paper to his own students - why does he need the printed from the journal? He wrote the original paper, and presumably has the original source for the paper, so why not just print that off instead. He sounds like somebody doing this to try and make a noise, rather than because it is actually hindering him.
Your fourth point is quite interesting, as if they are systematically charging for CC material then they seem to have screwed up.
It's nice to read one of these rare insightful posts on slashdot. In fact I may close my browser before scrolling any further:)
I think you've hit the nail on the head that registry is the key problem. This reminded me of Rothemund's work although their pattern structures are slightly finer than the DNA scaffolding that he created.
If either technique paid off (ie to the extent that components could be attached to arbitrary points in the pattern) then it would revolutionise chip design. But, that is quite a big if. It should be interesting to see if this work pans out enough to come to market.
OK, you raise a good point about China and that definitely counts against Google. But evens out? Have you heard of a little conflict called the Crusades?
Given that a large chunk of that money is currently being paid into funds for use in future sex-abuse cases I wouldn't worry about the Pope declaring them evil. When the head of an organisation that has overtly avoid routing out pedophiles in their ranks goes on to call you evil - the comparison actually works in your favour. Number of people directly abused or killed by Google - 0. Catholic church - can any of us count that high? And lets just say that the Godwin clock on this thread is starting unusually low....
Why does he ask? Let me tidy up his submission a litte:
Dear Slashdot, We haven't had a really good flamefest for ages. As all flames end up in political arguments, and all political arguments end up being about Libertarians. Can we just cut out the middle man and get to the good stuff?
Yours expectantly, A troll who got a story through firehose
I heard there was this one *LAWYER* who just totally flipped out and started sueing people. There was guys that weren't even involved and he sued them before they could even get ready. He even sued pirates...
Remember: FACT - laywers are mammals FACT - laywers totally flip out all the time
So what you're saying is that it can be defeated by building some custom hardware and installing it in the physical location.... oh, much like a keylogger I suppose. So try again, why doesn't this defeat shoulder surfing - the casual swiping of passwords by people who just happen to be in the area.
I'm not really interested in converting anyone to a "cause". I use linux because it is more productive for me, and it has all of the software that I want in an easier form than windows. This won't be the case for everyone, and infact I'd even go as far as to say that I'm an outlier there myself.
The only time that I have to reboot to window is to play a game that isn't supported in Cedega. I actually find it quite amusing that windows users are going to be in the same boat soon. A game that doesn't have any technical reason not to work on their operating system is going to require an upgrade to play. I don't buy the technical arguements, yes I've heard the same reasons for why the driver model wouldn't support DX10 and they are bullshit. Pure and simple - Microsoft could release DX10 on XP without any technical hitches. But why would they want to? All of the cool features (and yes I did like some of the vaporware that has disappeared from vista) are gone so they need something to drive adoption.
The funniest part will be when third-party solutions appear to hack DX10 support into XP. Because then Microsoft will be caught with their pants down. The whole "vista experience" has been odd for me. For years the progression in the linux desktop has been "catching up" with where microsoft is perceived to be. But I've had all of the new "vista" eye candy on my linux desktop for over a year. It was fun for a while, but it's been mostly turned off now as it does get irritating. The real transparancy is useful sometimes.
Odd thing though - it didn't require a major upgrade to activate it all. So if microsoft are telling the truth about their driver woes then maybe linux just has a better architecture design...
So you want to disclaim all the FUD on here by giving us your experience? Are you suggesting that you are a "normal", or perhaps "average" user so that your own anecdotal experience has any value to a wider audience at all?
Lets see, you regularly use 6 machines! (Check, that's definitely average) And one of those machines is a quad-core with a DX10 graphics card (Check, completely average)
So obviously your experience translates well for everyone. Gosh, we should all stop bitching and listen up. So apart from your moaning about those FSF boys not rolling over and spreading for Microsoft when asked, what of value do you have to say? Have you touched on the issue of the discussion - Microsoft bullshitting that they have technical reasons for holding back DX10 from XP when everyone and their mother knows that it's a decision made to drive sales of a failed operating system that nobody wants.
Ooh, that's right you forgot to mention that subject.
Slashdot Mirror
You've written a very indepth and informative post, but you are still wrong on a couple of issues. ADSL access in the uk market is not sold as a commodity. In the LLU market things are different but the vast majority of ISPs in this country are still on Datastream / IPstream products. So it is not as fanciful as you claim for somebody to be paying per packet. The BT pricing structure is actually based on the size of the outgoing pipe (not the customer links) and is sold by capacity. The charging of the ISP by BT is based on the usage of this pipe - it is NOT a product of how many customers they are connected to. The pricing model is here.
Apart from the fact that we are obviously talking about different markets (a lot of what you've written appears to be the US market for broadband) I agree with you about most of your points. But I still have to disagree about the costs that a cable company incurs. Yes - they don't have enough upstream bandwidth. Yes - it's a tree network where the packets have to reach the head. BUT, if their costs rise because they've developed their network the wrong way then it is still a capital / maintenance cost over the entire network. That is not the same as a cost directly incurred by transiting traffic to another network. In particular their network costs are very "steppy" and as long as their internal pipes are big enough they can handle small increases in internal traffic before they incur a cost to make an upgrade.
Any ISP that is not big enough to peer (most of them) is charged for upstream according to traffic. So ANY increase will cause a direct increase in their costs. This is not an apples to apples comparison as you claim. Partly this is because we are describing two different markets, but here the situation is very different between the cable providers and the DSL companies. At the moment telewest (one of the larger cable companies here) is offering a bi-directional 10Mb link for a reasonable price. They don't care about internal bandwidth because they have capacity coming out of their ass. They don't care about traffic because they are large enough to peer at Linx and they have a mixture of consumer and server business to get a profitable peering arrangement.
You're making the same mistake as the OP. "Free" versus nonfree traffic is not a question of capital costs. ISPs have to pay per packet that moves upstream and off their network (depending on their peering status). The question of building infrastructure is irrelevant. It is a cost that the ISP has to invest to be in business, and can be written off against whichever of their activities seems most appropriate. It is not a DIRECT cost of the traffic. But every packet that transits to a backbone has a DIRECT cost for them. Every packet that remains within their network incurs no cost.
It sounds nothing like multicasting. One is limiting peers based on their address ranges, and the other is broadcasting to multiple peers at once. Now, if multicasting did work properly it would revolutionise p2p as it would break the limit (total_upload=total_download) across the swarm.
Who gets stung with a deal like that? The UK is considered to be an example of what can go wrong with non-unbundled services, but even here ISPs paid a flat rental fee for access to the line, and then had to deal with BT for upstream access. Most the of the current bandwidth limits here are a direct result of how BT charges for that upstream bandwidth. Something like this system would be a great boon for them.
You've completely missed the point. Regardless of how the cable infrastructure is arranged within the ISPs own network - they get chared at the transit point. Traffic between their own customers does not transit to another network at any point and so it is free for them.
Although you are theoretically correct what you have said is wrong in practice. If I want to run a classical algorithm on a larger piece of data then I can just simulate a larger computer (to the extent that I page things in or out of memory, implement 64bit operations in 32bit etc). For a quantum computer I need a bigger computer. I can't simulate a 8-qubit machine on a 4-qubit machine with just a polynomial slowdown.
:)
I can't read the actual article at home so I don't know how large their machine is. Shor's algorithm has actually been run on a 4-qubit machine before so the summary is incorrect. I believe that the number they factored was 15. The point being that I need a quantum machine large enough to factor the RSA number. As building a 8-qubit machine is not as simple as slapping two 4-qubit machines together (because of problems with quantum coherence) there will always be a state-of-the-art for how large a Quantum Computer can be, and public crypto with keys significantly larger than that will be safe until a larger machine is developed. Sort of a faster version of the battle between cryptographers and cryptoanalysts that we see at the moment.
You'll notice that nobody made the same claims when EPFL sieved a 1024-bit number recently - instead everyone said use larger keys. The situation is likely to be the same as Quantum Computers increase in size. Lastly, not all public key crypto is shafted, only things that rely on factorisation as a problem. ECC will be quite safe until (if?) somebody develops a quantum algorithm for discrete logs.
Disclaimer: I don't do research in quantum - I work in cryptography, but the quantum guys have an office down the corridor and occasionally I understand what they are talking about. Ashley, don't beat me around the head for getting the details wrong
You know, if you don't get the literary reference that the GP is alluding to then maybe you should look it up instead of posting a sniping response. Given the way it was modded I think you are in the minority. Here's a hint: look up quotes by Arthur C. Clarke.
Magic, as the term is commonly used (especially by hackers) is anything that you don't understand. It doesn't imply a supernatural explanation in this context. The empirical approach that the GP described is exactly how we turn magic into science.
You may want to investigate the relevation that different people have different opinions, and that getting a +5 insightful mod doesn't make you authoritative.
Thanks for the explanation - it makes a lot more sense now. Yes, both my original assumptions were complete tosh then
If I write to a block that is already in use (say 2) then I have two blocks in flight, my new block N, and the contents of block 2 which either have to go somewhere else or get cached in memory. So *moving* the contents of a block so that I can write into that location takes two writes instead of one. How is this improving the duty cycle?
OK, all three replies came up with the same points but I'll reply to you since you've put them most succinctly. Your point about reservation makes sense, and it does change the overall picture. Spare blocks on the disk would increase the ability to do load leveling. So lets consider the case where we have a disk with 10 blocks, and lets say 4 are free (these can be a mixture of really free, and reserved blocks). If I repeatedly write to a used block (updating / overwriting) then I have five choices of where to put the data - either the original block, or mark that block as free and pick one of the other 4 blocks.
So I claim that the load can be leveled over 50% of the disk at most - this depends entirely (as you've pointed out) on my second assumption. If I decide that I'm going to pick one of the other 5 used slots to write to, then I have to keep that data. So that block needs to be copied to one of the original choice of 5 places. Unless I choose to keep the dirty data in memory, and hope that I don't run out of storage / lose the power then I can indeed level over all 10 blocks in the disk. But is it reasonable to assume that I can cache the dirty values this way to level over the whole disk rather than just the free space?
You're assuming that the 2GB a day could be spread evenly over the disk. This would vary depending on how much free space you have on the device. If your drive is 1% full then you can distribute your writes over the other 99%. But most people don't keep their storage mainly empty. In fact people tend to run just under the limit - hence the saying that crap always expands to fill the available space. If your drive was 99% full then you can't distribute the writes over the parts with data (as it would have to be moved somewhere else negating the benefit), and then you run into the problem with the limited duty cycle.
/tmp. Given that /tmp would be better suited to a RAM disk anyway I don't think that either would pose a problem, and the lifespan of these flash disks is probably comparable to a magnetic platter. As another reply pointed out, when the duty cycle is exceeded you can't alter the sector anymore. On a magnetic disk when a sector dies you're SOFL. Once the price comes down to an afforable level these drives will be beautiful...
Having said all of that, I don't think my throughput is anything like 2Gb, and most of it would be swap (hasn't happened much this past couple of years) and
You couldn't be more wrong unless you are somehow counting research performed by the US military as some kind of market force...
The author doesn't mention the copyright-transfer form (if any) that he signed. If he refused to sign the copyright transfer form then it is hard to understand how the journal could publish it in the first place. But if he did then he has a fair point as they don't have the right to sell the published form of the paper.
It's nice to hear somebody raising this as a problem because the current system of copyright transfers is a bitch. However, he wants to distribute the paper to his own students - why does he need the printed from the journal? He wrote the original paper, and presumably has the original source for the paper, so why not just print that off instead. He sounds like somebody doing this to try and make a noise, rather than because it is actually hindering him.
Your fourth point is quite interesting, as if they are systematically charging for CC material then they seem to have screwed up.
It's nice to read one of these rare insightful posts on slashdot. In fact I may close my browser before scrolling any further :)
I think you've hit the nail on the head that registry is the key problem. This reminded me of Rothemund's work although their pattern structures are slightly finer than the DNA scaffolding that he created.
If either technique paid off (ie to the extent that components could be attached to arbitrary points in the pattern) then it would revolutionise chip design. But, that is quite a big if. It should be interesting to see if this work pans out enough to come to market.
Must be a sufficiently advanced technology then...
Well enough to know that the Catholics slaughtered hundreds of thousands. Are you disputing that or did you have another point?
OK, you raise a good point about China and that definitely counts against Google. But evens out? Have you heard of a little conflict called the Crusades?
Given that a large chunk of that money is currently being paid into funds for use in future sex-abuse cases I wouldn't worry about the Pope declaring them evil. When the head of an organisation that has overtly avoid routing out pedophiles in their ranks goes on to call you evil - the comparison actually works in your favour. Number of people directly abused or killed by Google - 0. Catholic church - can any of us count that high? And lets just say that the Godwin clock on this thread is starting unusually low....
I heard there was this one *LAWYER* who just totally flipped out and started sueing people. There was guys that weren't even involved and he sued them before they could even get ready. He even sued pirates...
Remember:
FACT - laywers are mammals
FACT - laywers totally flip out all the time
So what you're saying is that it can be defeated by building some custom hardware and installing it in the physical location.... oh, much like a keylogger I suppose. So try again, why doesn't this defeat shoulder surfing - the casual swiping of passwords by people who just happen to be in the area.
Why is that the obvious solution? Are you mistaking what is required for privacy and what is required for authentication?
I'm not really interested in converting anyone to a "cause". I use linux because it is more productive for me, and it has all of the software that I want in an easier form than windows. This won't be the case for everyone, and infact I'd even go as far as to say that I'm an outlier there myself.
The only time that I have to reboot to window is to play a game that isn't supported in Cedega. I actually find it quite amusing that windows users are going to be in the same boat soon. A game that doesn't have any technical reason not to work on their operating system is going to require an upgrade to play. I don't buy the technical arguements, yes I've heard the same reasons for why the driver model wouldn't support DX10 and they are bullshit. Pure and simple - Microsoft could release DX10 on XP without any technical hitches. But why would they want to? All of the cool features (and yes I did like some of the vaporware that has disappeared from vista) are gone so they need something to drive adoption.
The funniest part will be when third-party solutions appear to hack DX10 support into XP. Because then Microsoft will be caught with their pants down. The whole "vista experience" has been odd for me. For years the progression in the linux desktop has been "catching up" with where microsoft is perceived to be. But I've had all of the new "vista" eye candy on my linux desktop for over a year. It was fun for a while, but it's been mostly turned off now as it does get irritating. The real transparancy is useful sometimes.
Odd thing though - it didn't require a major upgrade to activate it all. So if microsoft are telling the truth about their driver woes then maybe linux just has a better architecture design...
So you want to disclaim all the FUD on here by giving us your experience? Are you suggesting that you are a "normal", or perhaps "average" user so that your own anecdotal experience has any value to a wider audience at all?
Lets see, you regularly use 6 machines! (Check, that's definitely average)
And one of those machines is a quad-core with a DX10 graphics card (Check, completely average)
So obviously your experience translates well for everyone. Gosh, we should all stop bitching and listen up. So apart from your moaning about those FSF boys not rolling over and spreading for Microsoft when asked, what of value do you have to say? Have you touched on the issue of the discussion - Microsoft bullshitting that they have technical reasons for holding back DX10 from XP when everyone and their mother knows that it's a decision made to drive sales of a failed operating system that nobody wants.
Ooh, that's right you forgot to mention that subject.