Don't know about reading the article, but submitting that post might have infringed. Notice how Slashdot analyzed your post, found some text that represented a HTML link, and inserted instructions that tell your browser it's one, and that if you right-click said link you'll get a selection of actions. That's very close indeed to claim 1 of the patent; the only catch (though it's a big one) is that the process is split between Slashdot's servers and the user's browser.
A dick head who's full of crap, too. 5,946,647 at the very least is infringed upon by pretty much all decent phones. (It covers detecting things like phone numbers in text, offering up a choice of actions to carry out on said items, and acting on user instructions to carry them out.) Not just smartphones either - I just checked and my old Samsung dumbphone seems to meet all the required elements of Claim 1. Android might actually be better off than the dumbphones here: they have copy-and-paste, which means if necessary they can just make users select a phone number before choosing actions to take based on it, whereas the dumb- and featurephones don't have that option.
Not sure about 6,343,263 because most phone manufacturers don't release enough details to know if they infringe, but it's an incredibly broad patent and I suspect pretty much everyone is screwed on that one.
It's funny you should mention cars. Apparently car uptake in the US was delayed by a similarly broad and useless patent that just basically patented combining an internal combustion engine with some wheels to make a car, without even bothering to specify how to overcome the technical difficulties. One of the reasons the Henry Ford motor company was so successful is because they blatently ignored it and eventually manged to get it overturned.
By the way, you do know the coins weren't just stolen from some random rich account holder like Mt Gox originally claimed, right? They were essentially created from thin air by a hacker that'd gained write access to the DB. If they hadn't shut things down promptly it's likely they'd have ended up not actually having enough real bitcoins to cover everyone's bitcoin balances; that's also why they insisted on rolling back the trades.
This was obvious for a while because half a million bitcoins were sold and based on examining the public record in the blockchain they didn't seem to actually have that many, but they did admit it in the end. It's fairly clear they must've known all along too; there's no way they could honestly have claimed that it was just a single account's bitcoins that were stolen and that they had enough bitcoins to cover their deposits.
Assets that hold no intrinsic value outside of what some people are willing to pay for with money, which also has no intrinsic value
That's fairly obviously not true in general. For example, a bicycle factory has intrinsic value because you can use it to build bicycles, and the bicycles have intrinsic value because you can use them to get places. Likewise, a design for a microprocessor has value because you can build microprocessors from it which can be used in all sorts of applications that need one.
They pay people dividends of play money called USD/Euro/etc, which has no intrinsic value
Kind of. The money is obviously only worth something if other people will accept it, but in a sense it's just a convenient way of distributing the actual value the company created by taking some assets and making use of them in a productive way to shareholders.
First off, BitCoin isn't a stock.
For some reason, Bitcoin supporters keep on comparing it to one in order to justify its legitimacy, though. Probably because it really doesn't act like a currency at all.
Requires Mac OS X 10.4.x NOT compatible with Mac OS X 10.6 Leopard.
Of course, it and all of Unsanity's other Haxies works by binary-patching programs and libraries to intercept function calls and modify program behaviour, so they're pretty much always broken by major OS releases anyway. They're a great big ugly hack.
Of course, this does mean that someone has to do the hard work of actually merging the two... and it probably is going to be very hard work. Even IBM's attempts at actual upstream contributions to OO.org seem to have been more or less unmergable for some reason
The Radius Pivot apparently used a mercury tilt switch. At the time the price of accelerometers was too high to use them for applications like this. Since then they've come down in price significantly to the point that they're often used in places where previously you'd have found a mercury tilt switch; using one for this is more than a little obvious.
Which is an entirely absurd distinction - there's no technical challenge whatsoever in taking something like this that's already been done with non-touch screens and applying it to touch screens, nor is there any kind of conceptual leap required to realize that you can do. It's the kind of absurd non-distinction that the patent office seems to like though.
OK. If I'm reading this timing report correctly it looks like Xilinx's tools can fit a 150 MHash/sec bitcoin miner onto the smallest Kintex-7 FPGA (the XC7K70T) without even really trying. That roughly corresponds to a slightly overclocked Radeon 5750, which isn't great, though it's probably possible to achieve better performance.
They're kinda... interesting. As I understand it, part of the reason CCP are charging so much for stuff like monocles is that there's far too much PLEX in circulation as a result of their officially sanctioned way of buying in-world ISK for real money. They allowed users to buy PLEX for real money, trade them with other players for in-world ISK, and then the players purchasing the PLEX could use them to extend their Eve subscription in lieu of spending the real money themselves. So it was effectively RMT that no-one could make real money from other than CCP. The trouble is that there's too much demand for RMT purchases of ISK, so PLEXs have been entering the market far faster than they can be consumed by players wanting to pay their subscriptions using ISK.
In a way, you could argue the $70 monocles are CCP's way of indirectly making money from people that buy in-game ISK using real money.
Ah. I remember now: bitmarket.eu was one of the quirky sites. Rather than just executing trades for you semi-automatically, it requires you to pick out a particular set of outstanding orders up to the amount you want to trade, then either manually send them the money or wait for them to manually send you the money. At that point, the party selling the Bitcoins is meant to release them and hope that the half of the transaction paying them doesn't later get reversed (for example, because the user paid using a stolen Paypal account or card). The person attempting to get money for their bitcoins takes on all the risk in the transaction.
Inflation entices sellers to sell later (but buyers to buy now).
Only if the sellers can't find something to invest their money in that'd give a better return than simply sitting on their goods. That's incredibly unusual, especially as many goods actually depreciate in value if you hold onto them. If you think about it, there's basically no-one that only sells stuff - businesses have capital and raw material and wage expenses, they're looking to expand into new markets and come up with more efficient ways of doing things, etc...
It doesn't matter how many instructions per second... Different instructions will take different numbers of CPU/GPU cycles to perform.
All arithmetic instructions used by this code execute in a single cycle on modern AMD GPUs. (In fact, basically all their arithmetic instructions do.) That's not an entirely useful metric, because AMD GPUs are a VLIW architecture and it depends how many execution units you can saturate, but IIRC Bitcoin mining wins here too.
As it stands, I don't know which one uses more GPU cycles in bit rotating, nor do I know what frequency the GPUs are running at.
NVidia GPUs run at a higher frequency, but they're a lot less area-efficient so in the end that doesn't help them.
Except that's a totally irrelevant metric - you can (and probably should) implement all the logic required for a Bitcoin miner on just the FPGA logic elements without even touching the special DSP slices. A very rough ballpark figure is around 75,000 LEs for a design achieving 1 hash per clock cycle, with actual usage and clock speed depending on the FPGA. I haven't synthesised a miner for Xilinx 7-series FPGAs so I can't give more exact estimates.
Second Life's Linden dollars have things you can spend them on that justify the price - virtual things, sure, but still things - and most purchasers of L$ are buying them to spend on stuff. There's really not much you can spend Bitcoins on. If you look at the trading forum on the official site, most of the activity is buying and selling speculation-related items like mining rigs, and that's after discounting all the Ponzi schemes and scams!
Several days? Bah - if you read the forums, people have been waiting for weeks for money from Bitcoins to transfer into their USD or EUR current account. Fees for converting between Bitcoins and USD on the exchange with 90% of the volume (which is the one you probably want to use) are also usually 1.3% - half of which is payed by each party, but it still works out as costing 1.3% in total. Then there's the fees for getting your USD in and out: the cheapest option is Dwolla at a $0.25 flat fee. Dwolla are basically a US-only online payment provider like Paypal, so if both parties are using them why not just cut out the middleman? The other alternatives are Liberty Reserve (2% fee, kinda shady), and for EU residents direct transfer to a bank account in the SEPA region (2% fee, assuming they manage to get this up and running again - it's not currently available).
So basically that works out as a 3.3% fee normally for anyone not living in the US. The exchange has temporarily halved its exchange fees to discourage all its customers from running away after a serious security incident, but even that only reduces it to 2.7%. There are one or two much smaller alternative exchanges that don't have such high fees, but they also don't have enough volume to convert more than a small amount of Bitcoins on them.
That's not counting any fees your buyer had to pay to get the Bitcoins in the first place! (At some point in the future you're going to have to start paying fees to transfer Bitcoins from one person to another too.)
I can just go to a website and say "Give me [exchange rate] for X coins" and it becomes actual folding dollars?
Assuming someone's willing to buy them off you, sort of. You can convert them to virtual dollars that exist as a number in Mt Gox's database (which recently got hacked by a combination of a trivally-exploited SQL injection and the fact that the original site owner had a DB account with full access under the same username and easily-cracked password as his account on the site - after which they lied to everyone repeatedly until well after the point their story was obviously false).
There's also persistent rumours that Mt Gox doesn't actually have enough US dollars to back its clients' USD deposits as a result of criminals transferring in stolen money from other people's bank accounts and converting it to Bitcoins, and the transactions getting reversed later. These rumours predated Mt Gox admitting that this was in fact happening.
Slashdot Mirror
Don't know about reading the article, but submitting that post might have infringed. Notice how Slashdot analyzed your post, found some text that represented a HTML link, and inserted instructions that tell your browser it's one, and that if you right-click said link you'll get a selection of actions. That's very close indeed to claim 1 of the patent; the only catch (though it's a big one) is that the process is split between Slashdot's servers and the user's browser.
A dick head who's full of crap, too. 5,946,647 at the very least is infringed upon by pretty much all decent phones. (It covers detecting things like phone numbers in text, offering up a choice of actions to carry out on said items, and acting on user instructions to carry them out.) Not just smartphones either - I just checked and my old Samsung dumbphone seems to meet all the required elements of Claim 1. Android might actually be better off than the dumbphones here: they have copy-and-paste, which means if necessary they can just make users select a phone number before choosing actions to take based on it, whereas the dumb- and featurephones don't have that option.
Not sure about 6,343,263 because most phone manufacturers don't release enough details to know if they infringe, but it's an incredibly broad patent and I suspect pretty much everyone is screwed on that one.
IIRC, TSMC don't actually share the same process with anyone these days, and even if they did it doesn't exactly work like that...
It's funny you should mention cars. Apparently car uptake in the US was delayed by a similarly broad and useless patent that just basically patented combining an internal combustion engine with some wheels to make a car, without even bothering to specify how to overcome the technical difficulties. One of the reasons the Henry Ford motor company was so successful is because they blatently ignored it and eventually manged to get it overturned.
Of course, this particular Wikileaks leak did have an impact that was felt pretty much throughout the world...
By the way, you do know the coins weren't just stolen from some random rich account holder like Mt Gox originally claimed, right? They were essentially created from thin air by a hacker that'd gained write access to the DB. If they hadn't shut things down promptly it's likely they'd have ended up not actually having enough real bitcoins to cover everyone's bitcoin balances; that's also why they insisted on rolling back the trades.
This was obvious for a while because half a million bitcoins were sold and based on examining the public record in the blockchain they didn't seem to actually have that many, but they did admit it in the end. It's fairly clear they must've known all along too; there's no way they could honestly have claimed that it was just a single account's bitcoins that were stolen and that they had enough bitcoins to cover their deposits.
Assets that hold no intrinsic value outside of what some people are willing to pay for with money, which also has no intrinsic value
That's fairly obviously not true in general. For example, a bicycle factory has intrinsic value because you can use it to build bicycles, and the bicycles have intrinsic value because you can use them to get places. Likewise, a design for a microprocessor has value because you can build microprocessors from it which can be used in all sorts of applications that need one.
They pay people dividends of play money called USD/Euro/etc, which has no intrinsic value
Kind of. The money is obviously only worth something if other people will accept it, but in a sense it's just a convenient way of distributing the actual value the company created by taking some assets and making use of them in a productive way to shareholders.
First off, BitCoin isn't a stock.
For some reason, Bitcoin supporters keep on comparing it to one in order to justify its legitimacy, though. Probably because it really doesn't act like a currency at all.
Requires Mac OS X 10.4.x
NOT compatible with Mac OS X 10.6 Leopard.
Of course, it and all of Unsanity's other Haxies works by binary-patching programs and libraries to intercept function calls and modify program behaviour, so they're pretty much always broken by major OS releases anyway. They're a great big ugly hack.
Of course, this does mean that someone has to do the hard work of actually merging the two... and it probably is going to be very hard work. Even IBM's attempts at actual upstream contributions to OO.org seem to have been more or less unmergable for some reason
Have a look at some of the patents they're using against HTC. Yes, some of them are broad enough that pretty much all modern OSes probably infringe...
The Radius Pivot apparently used a mercury tilt switch. At the time the price of accelerometers was too high to use them for applications like this. Since then they've come down in price significantly to the point that they're often used in places where previously you'd have found a mercury tilt switch; using one for this is more than a little obvious.
Which is an entirely absurd distinction - there's no technical challenge whatsoever in taking something like this that's already been done with non-touch screens and applying it to touch screens, nor is there any kind of conceptual leap required to realize that you can do. It's the kind of absurd non-distinction that the patent office seems to like though.
OK. If I'm reading this timing report correctly it looks like Xilinx's tools can fit a 150 MHash/sec bitcoin miner onto the smallest Kintex-7 FPGA (the XC7K70T) without even really trying. That roughly corresponds to a slightly overclocked Radeon 5750, which isn't great, though it's probably possible to achieve better performance.
They're kinda... interesting. As I understand it, part of the reason CCP are charging so much for stuff like monocles is that there's far too much PLEX in circulation as a result of their officially sanctioned way of buying in-world ISK for real money. They allowed users to buy PLEX for real money, trade them with other players for in-world ISK, and then the players purchasing the PLEX could use them to extend their Eve subscription in lieu of spending the real money themselves. So it was effectively RMT that no-one could make real money from other than CCP. The trouble is that there's too much demand for RMT purchases of ISK, so PLEXs have been entering the market far faster than they can be consumed by players wanting to pay their subscriptions using ISK.
In a way, you could argue the $70 monocles are CCP's way of indirectly making money from people that buy in-game ISK using real money.
Ah. I remember now: bitmarket.eu was one of the quirky sites. Rather than just executing trades for you semi-automatically, it requires you to pick out a particular set of outstanding orders up to the amount you want to trade, then either manually send them the money or wait for them to manually send you the money. At that point, the party selling the Bitcoins is meant to release them and hope that the half of the transaction paying them doesn't later get reversed (for example, because the user paid using a stolen Paypal account or card). The person attempting to get money for their bitcoins takes on all the risk in the transaction.
Inflation entices sellers to sell later (but buyers to buy now).
Only if the sellers can't find something to invest their money in that'd give a better return than simply sitting on their goods. That's incredibly unusual, especially as many goods actually depreciate in value if you hold onto them. If you think about it, there's basically no-one that only sells stuff - businesses have capital and raw material and wage expenses, they're looking to expand into new markets and come up with more efficient ways of doing things, etc...
It doesn't matter how many instructions per second... Different instructions will take different numbers of CPU/GPU cycles to perform.
All arithmetic instructions used by this code execute in a single cycle on modern AMD GPUs. (In fact, basically all their arithmetic instructions do.) That's not an entirely useful metric, because AMD GPUs are a VLIW architecture and it depends how many execution units you can saturate, but IIRC Bitcoin mining wins here too.
As it stands, I don't know which one uses more GPU cycles in bit rotating, nor do I know what frequency the GPUs are running at.
NVidia GPUs run at a higher frequency, but they're a lot less area-efficient so in the end that doesn't help them.
Except that's a totally irrelevant metric - you can (and probably should) implement all the logic required for a Bitcoin miner on just the FPGA logic elements without even touching the special DSP slices. A very rough ballpark figure is around 75,000 LEs for a design achieving 1 hash per clock cycle, with actual usage and clock speed depending on the FPGA. I haven't synthesised a miner for Xilinx 7-series FPGAs so I can't give more exact estimates.
Second Life's Linden dollars have things you can spend them on that justify the price - virtual things, sure, but still things - and most purchasers of L$ are buying them to spend on stuff. There's really not much you can spend Bitcoins on. If you look at the trading forum on the official site, most of the activity is buying and selling speculation-related items like mining rigs, and that's after discounting all the Ponzi schemes and scams!
Several days? Bah - if you read the forums, people have been waiting for weeks for money from Bitcoins to transfer into their USD or EUR current account. Fees for converting between Bitcoins and USD on the exchange with 90% of the volume (which is the one you probably want to use) are also usually 1.3% - half of which is payed by each party, but it still works out as costing 1.3% in total. Then there's the fees for getting your USD in and out: the cheapest option is Dwolla at a $0.25 flat fee. Dwolla are basically a US-only online payment provider like Paypal, so if both parties are using them why not just cut out the middleman? The other alternatives are Liberty Reserve (2% fee, kinda shady), and for EU residents direct transfer to a bank account in the SEPA region (2% fee, assuming they manage to get this up and running again - it's not currently available).
So basically that works out as a 3.3% fee normally for anyone not living in the US. The exchange has temporarily halved its exchange fees to discourage all its customers from running away after a serious security incident, but even that only reduces it to 2.7%. There are one or two much smaller alternative exchanges that don't have such high fees, but they also don't have enough volume to convert more than a small amount of Bitcoins on them.
That's not counting any fees your buyer had to pay to get the Bitcoins in the first place! (At some point in the future you're going to have to start paying fees to transfer Bitcoins from one person to another too.)
I can just go to a website and say "Give me [exchange rate] for X coins" and it becomes actual folding dollars?
Assuming someone's willing to buy them off you, sort of. You can convert them to virtual dollars that exist as a number in Mt Gox's database (which recently got hacked by a combination of a trivally-exploited SQL injection and the fact that the original site owner had a DB account with full access under the same username and easily-cracked password as his account on the site - after which they lied to everyone repeatedly until well after the point their story was obviously false).
What you'll have trouble doing is converting those virtual dollars into actual dollars. People have found their attempts to cash out repeatedly delayed no matter which method they use, and anyone in Europe is almost totally screwed - due to a combination of factors, including black hats stealing money from people's bank accounts by phishing and using Mt Gox to launder it through bitcoins their European bank account has been closed, and all the other exchanges don't have enough volume to sell large amounts of bitcoins.
There's also persistent rumours that Mt Gox doesn't actually have enough US dollars to back its clients' USD deposits as a result of criminals transferring in stolen money from other people's bank accounts and converting it to Bitcoins, and the transactions getting reversed later. These rumours predated Mt Gox admitting that this was in fact happening.
Nah, all the FPGA dev kits big enough to mine Bitcoins at a decent rate are far too expensive to make this viable as far as I can tell.
I think AMD cards can pull more instructions/second on this kind of workload too; NVidia users lose out both ways.
Alternatively, the Ubertooth One costs a lot less than $10,000 and can do this, at least in theory.
I wouldn't be surprised if someone already has, to be honest.