Sorry to disillusion you, but the Grammar Nazi was right.
effect verb - cause to happen; bring about. affect verb - 1 make a difference to; have an effect on. 2 touch the feelings of. (source: Compact OED, www.askoxford.com)
So 'affect' is the closest verb in meaning to the noun 'effect', which is what 'effective' is derived from. Confusing, but that's English for you.
The European Convention on Human Rights, a document that essentially forms part of the constitution of most european countries, grants a right to not be punished unless you break a law, which would apparently apply here.
Because while Niven's Ringworld is widely considered as an impossible structure due to the extraordinary structural strength requirements it places on his "scrith", something somewhat Halo-like could actually be built by a civilization not incalculably more advanced than our own. Therefore, its details make a more interesting thought exercise than Ringworld's do.
Dyson spheres have similar issues to ringworlds, i.e., if you're using rotation to provide psuedogravity you have stability issues (on a dyson sphere this manifests itself as the habitable area wandering over the surface of the sphere over time). If you're relying on mass, you're going to need a lot of it, much more mass than would be required for a similarly sized ringworld.
I think the answer is to forget the idea of having a star in the middle, which enables you to build substantially smaller. The ringworld from Niven's books is 150 million kilometres in diameter. This scale is unfeasible. The scale of Halo as quoted in the article here is substantially more realistic, but you don't really get to have a proper star in the middle of it. In terms of its SF predecessors, Halo is much more like Iain M. Banks's Orbitals, which rely on a star somewhere outside of their ring to provide light. As well as allowing you to build smaller, this also gives you a more natural day/night cycle as the sun rises over the edge of the orbital in the morning, passes over beyond the far side at noon and falls down behind the orbital in the evening; you don't get the harsh terminators produced by Niven's rotating array of light shields.
Verified by Visa can be broken if a merchant you deal with intercepts your password before sending it on to the payment network, storing it along with all the other details of your card in order to use them fraudulently at a later date.
This scheme doesn't allow that attack scenario, because the passcode would change before that "later date". If they attempted to use it immediately and directly you will have more information about who the attacker is, rather than just the nebulous information that somebody's charged stuff to your credit card that you didn't ask them to.
This scheme does allow an attack scenario that VbV doesn't, however: VbV is not broken by somebody stealing your physical card, whereas this scheme is.
The two are therefore both useful, and for maximum security you should be using both.
"I find this scare-mongering over mercury to be amusing. "
As do I. Why would you HAZMAT a room for 5mg of mercury vapour that will float out the window?
Most of which had already escaped by the time the measurements of mercury levels took place. Detected levels of mercury in the room were apparently 6 times the 300ng per cubic metre level described by the article author as a "safe level", so the room probably only contained a few micrograms of the stuff. The safe level is, of course, bullshit. Background environmental levels are approximately 10ng per litre, which equals 1000ng per cubic metre, or three times over this so-called safe level.
Takes about 2 ng/L Hg in the right environment (anaerobic, sulfur reducing/methylating bacteria - very common btw, especially in landfills) to create a methyl-Hg issue.
Whereas background environmental levels of mercury are approximately 10ng/L.
The specialist found mercury levels in the bedroom in excess of six times the state's "safe" level for mercury contamination of 300 billionths of a gram per cubic meter.
OK, so in the room (probably approximately 10 cubic metres) there were 5 billionths of a gram, or.002mcg, of mercury? Get a life. This is not a level of contamination to worry about. This total amount is only double the level of daily exposure that is estimated for the general population, and 1/100th the amount you absorb if you eat fish (source).
You have a source for that assertion? As I understand the system, SoundExchange are not legally permitted to tie you to a contract: they must license music to you if you meet the requirements of USC Title 17 Section 114 (d) (2). Basically, you provide a list of everything you want to license to SoundExchange after having played them, and if you meet the requirements they are obliged to license them to you for the standardised fee. If you've licensed some of your playlist via alternate means, you merely have to omit those entries from the list you provide to them, and you won't be charged for them.
Of course, licensing per hour of play is also available. This may be what you're refering to, as it works out substantially easier because you don't have to provide a complete listing of everything you've played every month. The options are:
$0.000762 per listener per performance, except that 4% of performances shall bear no royalty. Based on 15 tracks per hour, this would equate to $0.00109728 per listener per hour. Or $0.0117 per listener per hour for predominantly music-based programming. Slightly more expensive, but much easier to manage. Minimum charge $500 p/a (reached in 1 year of 24 hour per day programming with an average of 52 listeners).
A broadcaster that wanted to play a significant proportion of independent tracks would find the first option more appealing, I suspect, as long as they had more than those 52 average listeners. So, yes, what you're saying is effectively right for very small broadcasters, although it has nothing to do with contractual obligations, it is merely a matter of economics. While the broadcaster could not list the track on their submission to SoundExchange and therefore effectively not pay the RIAA for it, the RIAA would get the money via the minimum charge anyway.
But none of this applies in the slightest to what TFA was talking about, which is broadcasters that choose only to deal with independent artists. This is still viable.
So, are request radio shows no longer "broadcast radio" and therefore have to licence their music differently? Or is it only if processing the requests is automated?
The new rules set soundexchange as the collector for all internet radio royalties, even if the artist doesn't want them.
Yes, but that doesn't prevent the artist from licensing their songs directly, independently. It's about giving the broadcaster the choice to license everything they want to play from a single source. The broadcaster doesn't have to use that option if they don't want to.
This is the power play that everyone was expecting. For several years, SoundExchange was all about "collecting for the artists", so that they get their "fair share". Now that new regulations have been set, the true colors are coming out.
Nothing's changed. This is not a power play. The article author misunderstood what SoundExchange is about, is all that's happened. Of course you don't have to pay SoundExchange if you negotiate directly with the artists.
Nobody's penalising you. The article author doesn't understand the law, and doesn't understand what the RIAA are saying on their web site, and it's showing through in what he said.
Here is a post that seems to get it right, and put it in terms we should all be able to understand.
From the article: "SoundExchange will collect Internet radio royalties for your song even if you don't want them to do so"
Are you saying that this isn't that case?
No, that is the case. However "artists can offer to download their music for free, but they cannot offer their songs to Internet radio for free" is completely wrong. Artists can do whatever they want, they just have to license their stuff via SoundExchange in addition to however they want to license it themselves. Which means that the entire premise of the article is rubbish: Webcasters can pay artists directly and not deal with SoundExchange if they prefer to do so and the artists are willing to do so. They just have the option of using SoundExchange instead, if they'd rather pay more and get a simpler single agreement that covers them for everything they want.
Also, from the research note, they only simulated a neuron firing rate of 1Hz. I don't know much about mouse brains, but I know that's nowhere near the firing rate found in mammal brains.
Frankly, I don't think a general purpose computer architecture like BlueGene is appropriate for this kind of research. A cluster of lower-power small nodes with small local memory and a dense interconnect would be much better for the purpose. Their simulation had 4096 nodes simulating approximately 2,000 neurons each. I think 65536 nodes with about 120 neurons on each would be more appropriate for this kind of work.
How can it be half a mouse brain if it has 1/1000 the number of a real half mouse brain? Their simulated neurons also had less synapses than the real thing. So is the 8000 a typo, or am I missing something?
So now E.U. citizens have the explicit right to make private copies for personal non-profit usage
No, that's not what this means. It means it isn't a criminal offence for them to do so; they can still be sued by the copyright holder, they just aren't going to prison for it.
Do it on a commercial scale, and you end up in prison. This is already the case in most EU countries, I think, so I'm not sure what the directive changes. The FIPR objects on the grounds that it extends criminal liability to intentional patent infringement, which certainly sounds bad to me.
They mention, for example, that current processing cores can typically only perform 4 simultaneous operations per-core, and imply that this is some kind of weakness. They completely fail to mention that the vast majority of applications running on those processors don't even use the 4 available scheduling resources in each core. In other words, the number of applications that would benefit from being able to execute more than 4 simultaneous instructions in the same core is vanishingly small.
What the article doesn't point out is that the folks who've designed the processor have done so in parallel with compiler research that has determined that the architecture they've produced is substantially better at allowing parallelism to be exploited. The reason most programs don't use all 4 of the current-gen ALUs at once is because the processor's logic for determining which instructions are ready to be executed isn't particularly good. The 4-instruction limit is a limit of current architectures, not a basic one (although it is a limit fundamental enough to the architecture that you don't really benefit by adding more than 4).
Nice concept, but it doesn't solve many pressing problems in computer architecture, namely the memory wall and parallel programmability.
No, but what it does do is completely eliminate the need for register renaming, which (as I understand it) consumes a significant proportion of the silicon in most modern OOE-capable processors. This saved area can then be diverted into having either more execution units or more cache, whichever best helps with the target problem domain. It also makes thread-level parallelism easier to achieve, because the register renaming cache doesn't have to be duplicated for each executing thread.
Moore's Law doesn't require that speed and power of the CPUs increase every two years. The actual law states that the number of transistors at minimum component cost doubles, not just the number of transistors. http://en.wikipedia.org/Moore's_Law/
And what determines the number of transistors at minimum component cost, other than the size of the transistors? Using current fabrication techniques, AFAIK the cost of a fixed area of die is pretty much constant. Therefore, a decrease in feature size of sqrt(2) will double the number of transistors for the same cost. This is the predominant driving force behind Moore's Law.
Slashdot Mirror
It also keeps stuff like movies from being played if they're on the wrong medium (e.g., BD-R).
Hold on -- you can't record a movie onto a recordable BluRay disc and watch it on a standalone viewer?
Yeah, that'll be popular. They're signing their medium's death warrant right there.
$ `dd if=/dev/urandom of=File.txt bs=1 count=32 && cat File.txt`
09f911029d74e35bd84156c5635688c0
$
"I swear you honor, my computer came up with it randomly"
It's worse than that. Your computer randomly came up with a shell command that produced it. The chances of that are... well... freakily low.
Unfortunately, while very Churchill, this is believed to be a misattribution.
Sorry to disillusion you, but the Grammar Nazi was right.
effect verb - cause to happen; bring about.
affect verb - 1 make a difference to; have an effect on. 2 touch the feelings of.
(source: Compact OED, www.askoxford.com)
So 'affect' is the closest verb in meaning to the noun 'effect', which is what 'effective' is derived from. Confusing, but that's English for you.
The European Convention on Human Rights, a document that essentially forms part of the constitution of most european countries, grants a right to not be punished unless you break a law, which would apparently apply here.
Have to ask: are you the 'szyzyg' I used to know on mono?
Because while Niven's Ringworld is widely considered as an impossible structure due to the extraordinary structural strength requirements it places on his "scrith", something somewhat Halo-like could actually be built by a civilization not incalculably more advanced than our own. Therefore, its details make a more interesting thought exercise than Ringworld's do.
Dyson spheres have similar issues to ringworlds, i.e., if you're using rotation to provide psuedogravity you have stability issues (on a dyson sphere this manifests itself as the habitable area wandering over the surface of the sphere over time). If you're relying on mass, you're going to need a lot of it, much more mass than would be required for a similarly sized ringworld.
I think the answer is to forget the idea of having a star in the middle, which enables you to build substantially smaller. The ringworld from Niven's books is 150 million kilometres in diameter. This scale is unfeasible. The scale of Halo as quoted in the article here is substantially more realistic, but you don't really get to have a proper star in the middle of it. In terms of its SF predecessors, Halo is much more like Iain M. Banks's Orbitals, which rely on a star somewhere outside of their ring to provide light. As well as allowing you to build smaller, this also gives you a more natural day/night cycle as the sun rises over the edge of the orbital in the morning, passes over beyond the far side at noon and falls down behind the orbital in the evening; you don't get the harsh terminators produced by Niven's rotating array of light shields.
Yes, you are. Here's a hint: Cardholder Not Present.
Verified by Visa can be broken if a merchant you deal with intercepts your password before sending it on to the payment network, storing it along with all the other details of your card in order to use them fraudulently at a later date.
This scheme doesn't allow that attack scenario, because the passcode would change before that "later date". If they attempted to use it immediately and directly you will have more information about who the attacker is, rather than just the nebulous information that somebody's charged stuff to your credit card that you didn't ask them to.
This scheme does allow an attack scenario that VbV doesn't, however: VbV is not broken by somebody stealing your physical card, whereas this scheme is.
The two are therefore both useful, and for maximum security you should be using both.
"I find this scare-mongering over mercury to be amusing. "
As do I. Why would you HAZMAT a room for 5mg of mercury vapour that will float out the window?
Most of which had already escaped by the time the measurements of mercury levels took place. Detected levels of mercury in the room were apparently 6 times the 300ng per cubic metre level described by the article author as a "safe level", so the room probably only contained a few micrograms of the stuff. The safe level is, of course, bullshit. Background environmental levels are approximately 10ng per litre, which equals 1000ng per cubic metre, or three times over this so-called safe level.
Takes about 2 ng/L Hg in the right environment (anaerobic, sulfur reducing/methylating bacteria - very common btw, especially in landfills) to create a methyl-Hg issue.
Whereas background environmental levels of mercury are approximately 10ng/L.
Agreed that the article is scaremongering junk.
.002mcg, of mercury? Get a life. This is not a level of contamination to worry about. This total amount is only double the level of daily exposure that is estimated for the general population, and 1/100th the amount you absorb if you eat fish (source).
The specialist found mercury levels in the bedroom in excess of six times the state's "safe" level for mercury contamination of 300 billionths of a gram per cubic meter.
OK, so in the room (probably approximately 10 cubic metres) there were 5 billionths of a gram, or
You have a source for that assertion? As I understand the system, SoundExchange are not legally permitted to tie you to a contract: they must license music to you if you meet the requirements of USC Title 17 Section 114 (d) (2). Basically, you provide a list of everything you want to license to SoundExchange after having played them, and if you meet the requirements they are obliged to license them to you for the standardised fee. If you've licensed some of your playlist via alternate means, you merely have to omit those entries from the list you provide to them, and you won't be charged for them.
Of course, licensing per hour of play is also available. This may be what you're refering to, as it works out substantially easier because you don't have to provide a complete listing of everything you've played every month. The options are:
$0.000762 per listener per performance, except that 4% of performances shall bear no royalty. Based on 15 tracks per hour, this would equate to $0.00109728 per listener per hour.
Or $0.0117 per listener per hour for predominantly music-based programming. Slightly more expensive, but much easier to manage.
Minimum charge $500 p/a (reached in 1 year of 24 hour per day programming with an average of 52 listeners).
A broadcaster that wanted to play a significant proportion of independent tracks would find the first option more appealing, I suspect, as long as they had more than those 52 average listeners. So, yes, what you're saying is effectively right for very small broadcasters, although it has nothing to do with contractual obligations, it is merely a matter of economics. While the broadcaster could not list the track on their submission to SoundExchange and therefore effectively not pay the RIAA for it, the RIAA would get the money via the minimum charge anyway.
But none of this applies in the slightest to what TFA was talking about, which is broadcasters that choose only to deal with independent artists. This is still viable.
So, are request radio shows no longer "broadcast radio" and therefore have to licence their music differently? Or is it only if processing the requests is automated?
The new rules set soundexchange as the collector for all internet radio royalties, even if the artist doesn't want them.
Yes, but that doesn't prevent the artist from licensing their songs directly, independently. It's about giving the broadcaster the choice to license everything they want to play from a single source. The broadcaster doesn't have to use that option if they don't want to.
This is the power play that everyone was expecting. For several years, SoundExchange was all about "collecting for the artists", so that they get their "fair share". Now that new regulations have been set, the true colors are coming out.
Nothing's changed. This is not a power play. The article author misunderstood what SoundExchange is about, is all that's happened. Of course you don't have to pay SoundExchange if you negotiate directly with the artists.
Nobody's penalising you. The article author doesn't understand the law, and doesn't understand what the RIAA are saying on their web site, and it's showing through in what he said.
Here is a post that seems to get it right, and put it in terms we should all be able to understand.
From the article: "SoundExchange will collect Internet radio royalties for your song even if you don't want them to do so"
Are you saying that this isn't that case?
No, that is the case. However "artists can offer to download their music for free, but they cannot offer their songs to Internet radio for free" is completely wrong. Artists can do whatever they want, they just have to license their stuff via SoundExchange in addition to however they want to license it themselves. Which means that the entire premise of the article is rubbish: Webcasters can pay artists directly and not deal with SoundExchange if they prefer to do so and the artists are willing to do so. They just have the option of using SoundExchange instead, if they'd rather pay more and get a simpler single agreement that covers them for everything they want.
Also, from the research note, they only simulated a neuron firing rate of 1Hz. I don't know much about mouse brains, but I know that's nowhere near the firing rate found in mammal brains.
Frankly, I don't think a general purpose computer architecture like BlueGene is appropriate for this kind of research. A cluster of lower-power small nodes with small local memory and a dense interconnect would be much better for the purpose. Their simulation had 4096 nodes simulating approximately 2,000 neurons each. I think 65536 nodes with about 120 neurons on each would be more appropriate for this kind of work.
How can it be half a mouse brain if it has 1/1000 the number of a real half mouse brain? Their simulated neurons also had less synapses than the real thing. So is the 8000 a typo, or am I missing something?
It's a typo. See original research note here.
So now E.U. citizens have the explicit right to make private copies for personal non-profit usage
No, that's not what this means. It means it isn't a criminal offence for them to do so; they can still be sued by the copyright holder, they just aren't going to prison for it.
Do it on a commercial scale, and you end up in prison. This is already the case in most EU countries, I think, so I'm not sure what the directive changes. The FIPR objects on the grounds that it extends criminal liability to intentional patent infringement, which certainly sounds bad to me.
They mention, for example, that current processing cores can typically only perform 4 simultaneous operations per-core, and imply that this is some kind of weakness. They completely fail to mention that the vast majority of applications running on those processors don't even use the 4 available scheduling resources in each core. In other words, the number of applications that would benefit from being able to execute more than 4 simultaneous instructions in the same core is vanishingly small.
What the article doesn't point out is that the folks who've designed the processor have done so in parallel with compiler research that has determined that the architecture they've produced is substantially better at allowing parallelism to be exploited. The reason most programs don't use all 4 of the current-gen ALUs at once is because the processor's logic for determining which instructions are ready to be executed isn't particularly good. The 4-instruction limit is a limit of current architectures, not a basic one (although it is a limit fundamental enough to the architecture that you don't really benefit by adding more than 4).
Nice concept, but it doesn't solve many pressing problems in computer architecture, namely the memory wall and parallel programmability.
No, but what it does do is completely eliminate the need for register renaming, which (as I understand it) consumes a significant proportion of the silicon in most modern OOE-capable processors. This saved area can then be diverted into having either more execution units or more cache, whichever best helps with the target problem domain. It also makes thread-level parallelism easier to achieve, because the register renaming cache doesn't have to be duplicated for each executing thread.
Moore's Law doesn't require that speed and power of the CPUs increase every two years. The actual law states that the number of transistors at minimum component cost doubles, not just the number of transistors. http://en.wikipedia.org/Moore's_Law/
And what determines the number of transistors at minimum component cost, other than the size of the transistors? Using current fabrication techniques, AFAIK the cost of a fixed area of die is pretty much constant. Therefore, a decrease in feature size of sqrt(2) will double the number of transistors for the same cost. This is the predominant driving force behind Moore's Law.