It was discussed on Slashdot at the time, but the editors missed what's going on. This comment is a summary of what happened; you can read the story, and the other comments on it, for more details.
Short summary: a fan of Capcom games decides to make a crossover, when Capcom find out, they'd be within their rights to sue them, but instead they decided to publish it, giving it away free as an advert for their game franchises.
As Capcom showed recently, there are often situations where you have the alternatives of engaging in expensive legal battles, or getting a bunch of free marketing and good publicity out of the situation. The second option is rarely taken, but it's nearly always better, and I applaud people for taking it.
I think the point is that it works as a CAPTCHA. The humans are aware that leaving yourself open to fools' mate is a CAPTCHA, and wait a few seconds before playing it to demonstrate that they're human; the bots aren't yet advanced enough and play the mate instantly.
I'd say that in a QvR endgame, you'd expect most of the best players to win, but probably not in the minimum possible number of moves. When you're playing a standard endgame as a human, you typically try to use an algorithm that always works and is easy to remember, rather than taking risks.
No, not usually a general purpose computer. Just as most programs aren't CPU emulators, most silicon chips aren't general purpose computers. Computers are used to replace special-purpose chips more and more nowadays because they're easily mass-produced, but that's far from essential.
No, but a computer is probably the most efficient way to actually test your programs. (Even if they're intended to eventually run on something that isn't a computer, such as wired into hardware.)
You can link statically against LGPL code, so long as you provide the object files you used (allowing end-users to repeat the link with different LGPL files). You're right about GPL, though (except for the special case when you're linking other GPL code in).
If you overclock a monitor, you can expect similar results to if you try to overclock a CPU. (In general, it's going to overheat.) Because it's rather easier to do by mistake than it is for a CPU, monitors tend to check for it nowadays and cut out intentionally.
Windows actually has two root-like permission levels, "administrator", and "SYSTEM" (which is higher and cannot be given to normal accounts). It might be interesting to know which the attack allows escalation to (although I think an attacker could do anything they cared about with only administrator-level permissions, they'd just have to do it a little indirectly).
Also, IIRC the ban on exceptions is because they had a lot of existing code that wasn't exception-safe and didn't want to break it, rather than because they think they're a bad idea in general.
That Gnome 3 complaint is almost certainly a for-technical-reasons one. The problem is that you can look at a window and observe that it's open, and make a good guess about which app it belongs to, but that doesn't necessarily give you enough information to be able to open another copy of the app.
(I've been looking into the technical details behind this in an attempt to improve it for Unity, and I'm reasonably sure there's no 100% reliable way to figure it out. Unity's solution seems to be just to randomly fail when it can't figure it. Gnome 3 is presumably only making it possible when it's sure it'll work.)
I tried using the Ribbon for a while. I found that it was mostly intuitive (although try finding something like the double underline in Microsoft Word if you don't already know where it is; it's awkward in both pre-Ribbon and Ribbon versions, but somewhat harder with the Ribbon), but my main issue is simply that it takes more clicks. With toolbar+menu, if something's used a lot it's on the toolbar in a known place and can be accessed with one click. If something isn't used a lot, I can still access it through the menus. On the other hand, with the Ribbon, the number of clicks it takes to do something depends on whatever you did last, so it's not only more for common tasks than a menu+toolbar is on average, but it's also inconsistent; unless you do a useless click on a Ribbon tab even if it's already open, your muscle memory is going to be screwed up.
(And another observation: why isn't the "save" button on the ribbon in Microsoft Office programs? It got moved to a separate area precisely because it's commonly used and the Ribbon is bad at handling commonly used commands.)
Yeah, in the UK, everyone uses touchtone nowadays, but last we checked the exchanges still understand pulse dial, and will try to interpret the phone as doing that if it doesn't get the tones it's expecting. So cradle dialing may still exist!
Unity has one search that searches everything it knows how to search at once. So if you're searching for local files, by default it sends the information out too.
It's on by default, and has an opt-out that's not hard to find if you look for it but not presented in-your-face either. IMO, this isn't nearly good enough; it should be off by default (and I don't care how prominent the option to turn it back on again is).
The miners put most of the computational effort in for you; typically you pay a small fee to encourage them to put your transaction into the chain. Usually the buyer will pay the transaction fee, and the seller won't count the sale as complete until it goes sufficiently far back in the train. (In other words, while Bitcoin transactions are instant, it takes a nontrivial amount of time to determine for a fact whether a transaction happened or not. And as such, a cautious seller won't release the goods immediately, but wait until there's enough effort in recording that version of events that nobody's unlikely to override it with a new one.)
So if the mining stopped, nobody would be recording transactions any more. As such, double-spending would be trivial; you could repeatedly do a small amount of mining to satisfy one person that they had the coin, then a bit more to change the history's idea of where the coin went and give it to someone else (or a second account you control). As such, people would have to wait for a very long time in order to be satisfied that a transaction were genuine if there were no mining, and this would make transactions basically impossible. A currency is worth nothing if you can't meaningfully trade it.
It's a proof-of-work calculation that records a transaction history. Basically, it's to prevent double-spending; if someone attempts to transfer the same bitcoin to two different people, the person who gets it is the person who had more computational effort go into recording them as having it. If the amount of computational effort in recording transaction histories were low enough, someone could double-spend by recording an alternative history on a powerful computer and having it supersede the transaction history everyone thought they were using.
In order to encourage people to put effort into recording the history, there's a reward for doing so. This lead (perhaps predictably) to "mining", where people race to record the transaction history first in order to get the reward.
So there is a purpose to mining, but it's only to keep the Bitcoin system itself running. If people stopped mining (perhaps because the reward got low enough), Bitcoin would collapse. (It's envisaged that once the mining reward gets low enough, people transferring bitcoins will pay a transfer fee to the miners to encourage them to keep mining, and they'd get their income that way instead.)
More information on what you said: apparently the notices will be in the newspapers on November 16th. The court was a little surprised that, given that the order was made on October 18th, it took that long to get an advert into the newspapers.
Reading the ruling (it's at the last link in the summary) is quite interesting. Basically what happened is that Apple tried to exploit technicalities in the judge's wording, and the judges found a bunch of technicalities in Apple's statement to find it invalid anyway.
(For instance, the judges focused on the facts that the court case had nothing, technically, to do with the iPad, and that patents weren't involved, both of which contradict Apple's original statement.)
Agreed. I'm British, and was watching the debates in 2008 (I didn't in 2012, too busy at work). McCain seemed entirely reasonable before and after the election, but during it, he was insane, presumably in an attempt to win the Republican vote. At the moment, I'm not convinced it's possible to win the Republican primary and also win the election. Perhaps that will change over the next few years.
Hopefully the rules will require them to be no louder both in terms of peak level, and in terms of root-mean-square. That shuts down most of the potential abuses.
The actual reason is so that specific versions of Ubuntu are easy to search for; it's using much the same principle as the googlewhack. Completely invented words would be better for that, but words that are merely rarely used work well enough in combination with the rest of the search term. You can search for version numbers, but it's often confused by other numbers that happen randomly on the page.
Slashdot Mirror
It was discussed on Slashdot at the time, but the editors missed what's going on. This comment is a summary of what happened; you can read the story, and the other comments on it, for more details.
Short summary: a fan of Capcom games decides to make a crossover, when Capcom find out, they'd be within their rights to sue them, but instead they decided to publish it, giving it away free as an advert for their game franchises.
As Capcom showed recently, there are often situations where you have the alternatives of engaging in expensive legal battles, or getting a bunch of free marketing and good publicity out of the situation. The second option is rarely taken, but it's nearly always better, and I applaud people for taking it.
I think the point is that it works as a CAPTCHA. The humans are aware that leaving yourself open to fools' mate is a CAPTCHA, and wait a few seconds before playing it to demonstrate that they're human; the bots aren't yet advanced enough and play the mate instantly.
I'd say that in a QvR endgame, you'd expect most of the best players to win, but probably not in the minimum possible number of moves. When you're playing a standard endgame as a human, you typically try to use an algorithm that always works and is easy to remember, rather than taking risks.
No, not usually a general purpose computer. Just as most programs aren't CPU emulators, most silicon chips aren't general purpose computers. Computers are used to replace special-purpose chips more and more nowadays because they're easily mass-produced, but that's far from essential.
No, but a computer is probably the most efficient way to actually test your programs. (Even if they're intended to eventually run on something that isn't a computer, such as wired into hardware.)
You can link statically against LGPL code, so long as you provide the object files you used (allowing end-users to repeat the link with different LGPL files). You're right about GPL, though (except for the special case when you're linking other GPL code in).
If you overclock a monitor, you can expect similar results to if you try to overclock a CPU. (In general, it's going to overheat.) Because it's rather easier to do by mistake than it is for a CPU, monitors tend to check for it nowadays and cut out intentionally.
Sensible merchants would just use divisible-by-5 prices to avoid issues with rounding.
Windows actually has two root-like permission levels, "administrator", and "SYSTEM" (which is higher and cannot be given to normal accounts). It might be interesting to know which the attack allows escalation to (although I think an attacker could do anything they cared about with only administrator-level permissions, they'd just have to do it a little indirectly).
Also, IIRC the ban on exceptions is because they had a lot of existing code that wasn't exception-safe and didn't want to break it, rather than because they think they're a bad idea in general.
That Gnome 3 complaint is almost certainly a for-technical-reasons one. The problem is that you can look at a window and observe that it's open, and make a good guess about which app it belongs to, but that doesn't necessarily give you enough information to be able to open another copy of the app.
(I've been looking into the technical details behind this in an attempt to improve it for Unity, and I'm reasonably sure there's no 100% reliable way to figure it out. Unity's solution seems to be just to randomly fail when it can't figure it. Gnome 3 is presumably only making it possible when it's sure it'll work.)
I tried using the Ribbon for a while. I found that it was mostly intuitive (although try finding something like the double underline in Microsoft Word if you don't already know where it is; it's awkward in both pre-Ribbon and Ribbon versions, but somewhat harder with the Ribbon), but my main issue is simply that it takes more clicks. With toolbar+menu, if something's used a lot it's on the toolbar in a known place and can be accessed with one click. If something isn't used a lot, I can still access it through the menus. On the other hand, with the Ribbon, the number of clicks it takes to do something depends on whatever you did last, so it's not only more for common tasks than a menu+toolbar is on average, but it's also inconsistent; unless you do a useless click on a Ribbon tab even if it's already open, your muscle memory is going to be screwed up.
(And another observation: why isn't the "save" button on the ribbon in Microsoft Office programs? It got moved to a separate area precisely because it's commonly used and the Ribbon is bad at handling commonly used commands.)
Yeah, in the UK, everyone uses touchtone nowadays, but last we checked the exchanges still understand pulse dial, and will try to interpret the phone as doing that if it doesn't get the tones it's expecting. So cradle dialing may still exist!
Unity has one search that searches everything it knows how to search at once. So if you're searching for local files, by default it sends the information out too.
It's on by default, and has an opt-out that's not hard to find if you look for it but not presented in-your-face either. IMO, this isn't nearly good enough; it should be off by default (and I don't care how prominent the option to turn it back on again is).
The miners put most of the computational effort in for you; typically you pay a small fee to encourage them to put your transaction into the chain. Usually the buyer will pay the transaction fee, and the seller won't count the sale as complete until it goes sufficiently far back in the train. (In other words, while Bitcoin transactions are instant, it takes a nontrivial amount of time to determine for a fact whether a transaction happened or not. And as such, a cautious seller won't release the goods immediately, but wait until there's enough effort in recording that version of events that nobody's unlikely to override it with a new one.)
So if the mining stopped, nobody would be recording transactions any more. As such, double-spending would be trivial; you could repeatedly do a small amount of mining to satisfy one person that they had the coin, then a bit more to change the history's idea of where the coin went and give it to someone else (or a second account you control). As such, people would have to wait for a very long time in order to be satisfied that a transaction were genuine if there were no mining, and this would make transactions basically impossible. A currency is worth nothing if you can't meaningfully trade it.
It's a proof-of-work calculation that records a transaction history. Basically, it's to prevent double-spending; if someone attempts to transfer the same bitcoin to two different people, the person who gets it is the person who had more computational effort go into recording them as having it. If the amount of computational effort in recording transaction histories were low enough, someone could double-spend by recording an alternative history on a powerful computer and having it supersede the transaction history everyone thought they were using.
In order to encourage people to put effort into recording the history, there's a reward for doing so. This lead (perhaps predictably) to "mining", where people race to record the transaction history first in order to get the reward.
So there is a purpose to mining, but it's only to keep the Bitcoin system itself running. If people stopped mining (perhaps because the reward got low enough), Bitcoin would collapse. (It's envisaged that once the mining reward gets low enough, people transferring bitcoins will pay a transfer fee to the miners to encourage them to keep mining, and they'd get their income that way instead.)
Slashdot's community has a habit of picking on the editors whenever they do a sloppy job. It's nice to see them complemented when they do it properly.
More information on what you said: apparently the notices will be in the newspapers on November 16th. The court was a little surprised that, given that the order was made on October 18th, it took that long to get an advert into the newspapers.
Reading the ruling (it's at the last link in the summary) is quite interesting. Basically what happened is that Apple tried to exploit technicalities in the judge's wording, and the judges found a bunch of technicalities in Apple's statement to find it invalid anyway.
(For instance, the judges focused on the facts that the court case had nothing, technically, to do with the iPad, and that patents weren't involved, both of which contradict Apple's original statement.)
Agreed. I'm British, and was watching the debates in 2008 (I didn't in 2012, too busy at work). McCain seemed entirely reasonable before and after the election, but during it, he was insane, presumably in an attempt to win the Republican vote. At the moment, I'm not convinced it's possible to win the Republican primary and also win the election. Perhaps that will change over the next few years.
Hopefully the rules will require them to be no louder both in terms of peak level, and in terms of root-mean-square. That shuts down most of the potential abuses.
No, the problem is that if other people are selling, your optimal way to make money is to sell first but slightly before they do.
The real problem with the markets is that the way you make money on them is to do the same as everyone else, just faster
The actual reason is so that specific versions of Ubuntu are easy to search for; it's using much the same principle as the googlewhack. Completely invented words would be better for that, but words that are merely rarely used work well enough in combination with the rest of the search term. You can search for version numbers, but it's often confused by other numbers that happen randomly on the page.