No, I'm not wrong. From BP, in its 2007 report, estimated at 2006 end that there were 909,064 million tons of proven coal reserves worldwide, or 147 years reserves-to-production ratio. This figure only includes reserves classified as "proven"; exploration drilling programs by mining companies, particularly in under-explored areas, are continually providing new reserves. In many cases, companies are aware of coal deposits that have not been sufficiently drilled to qualify as "proven".
Going from 143yrs proven reserves to the 200+ years I stated isn't a big stretch, especially as supplied get tighter, conservation becomes more prevalent, and the price and technologies develop to extract the yet unproven or unknown reserves.
The bottom line is that fossil fuels are running out, and at most we've got 250 years, but since we don't yet have a viable replacement for oil, we've really got maybe 50 years to make the transition to largely renewable sources.
Japan is one of the few places that could possibly be powered completely by geothermal. There isn't nearly enough wave energy to supply the planet, nor is there sufficient wave energy near Japan to supply Japan. With a combination of geothermal, wind, hydro, and possibly some solar or wave, Japan might be able to go completely renewable. Most industrialized countries don't have access the the abundant geothermal resources Japan has (due to their location on the edge of the "ring of fire").
Of renewable sources, solar and wind are the ones that can supply enough power for the world, but both are intermittent sources that are not well suited to supplying either base-load or peak-load power without a significant amount of on-demand energy storage added to the grid. On demand energy storage can be in the form of batteries, super capacitors, gravity reservoirs (e.g. pump water uphill to a reservoir during periods of excess generation, release it through turbines when needed), etc. However, solar requires huge amounts of land. Solar and wind each need more than 4x average demand installed (even with on-demand storage, more still without on-demand storage) because they only average ~25% of installed capacity. Neither solar or wind is viable in all areas, and with it's intermittent nature, the grid must have significantly more capacity to route from locations with excess to locations with a shortage.
Bottom line, for most of the world, nuclear and/or fossil fuels are the only currently viable means to meet the difference between renewable capacity and peak demand. Fossil fuels will be exhausted in 50-250 years (~50 yrs oil, slightly longer for natural gas, 200+ years coal). Since plants have a 40-80yr life span, fossil fuel plant built today, could run out of fuel before the plant is used up. Nuclear is the only long term solution that is viable today, and even that needs to move to a thorium fuel model with breeder reactors and fuel reprocessing in order to last more than a few hundred years.
...not lengthened. Shorten it to 40 years (which is twice the 20 yr life of a patent). Copyright holders can then do what everyone else in the world has to do, save for retirement, and/or keep working. And, yes I have created numerous copyrighted works, I speak from experience. As much as the idea of owning copyright for my entire life sounds great (and I'm not opposed to making it 40 years or natural lifetime of a human creator, whichever is greater), I don't see that it's any benefit to the copyright holder or to society to have longer term copyrights. I'm willing to consider a longer term if someone can make a strong case for it, but I haven't yet seen a convincing argument.
Corporations (non human entities) should only hold a copyright for at most 40 years as there is no "natural lifetime" of a corporation. Frankly, if a corporation hasn't recouped their investment after 20-30 years, they're almost certainly not going to, so 20-30 years may be a better maximum for a corporation, but that's another issue. After those years of a corporation holding a copyright, it should either revert to the individual authors/performers, or become public domain.
As for this law, it doesn't extend the copyright on the original lyrics/music, so it doesn't help the songwriters. Who it helps are the entities (e.g. record labels) who hold the copyrights to specific recordings. The performers will see some benefit, but anyone who has looked at recording contracts knows that the performers typically get 10% or less (usually much less after the labels perform their biased math).
Perhaps you missed the significance of a +/- 5 minute variance of a 9 day orbit. That's 5/(9*1440) = 5/12960 = 1/2592. That's nearly 0.04% variance in the orbital period of a planet. To achieve that much change in orbital period, the velocity change needs to be at least that great, and likely at least 2x that great (since speed changes won't be instantaneous, it will have to slow down 2x as much to average 0.04% slower orbital rate).
Given gravity is inversely proportional to the square of the distance between the centers of mass, how much more massive must the other body be to produce that much change in velocity? Even if the other body is in a 4:3 resonance, making it "relatively close", it would still need to be vastly more massive than the observed planet. If it's vastly more massive and in orbit around the same star in a 4:3 resonance, we would definitely be able to detect it by it's doppler shift of the star unless it's orbital plane is almost perfectly perpendicular to our line of sight, and then we would still likely observe an effect on the star.
The other possibility is that there is a third massive body in the system that playing tug-o-war with the planet, pulling it away from the star, then the star pulls it closer, constantly changing it's orbit. That way the velocity of the planet doesn't have to change as much, however, how such a system could produce such an oscillation in the planet's orbit, sometimes increasing it and sometimes decreasing it resulting in such large changes in orbital period without also causing measurable effects on the star is mystery.
We may eventually figure it out, but for now, that much variation in orbital rate is really bizarre.
650 LY, Not 650,000 LY. 1.3M years ago doesn't come into the picture anywhere. Keep your units straight.
We're seeing the system as it was 650 years ago (circa 1360 AD in our reference), therefore, the gravitational effects we're observing mean that there was something there 650 years ago. If it was 650LY away 650 years ago. I started from my explicitly stated premise that the fastest any massive object has been observed to travel is less then 0.001C (actually ~ 0.0004C, then allowed doubled that as a closure rate, then rounded up to 0.001C to make math simpler), then it could be no closer than 650 - (650*0.001C) LY = 649.350 LY from earth.
And since I've now stated it 3 times, I'm not sure how you keep missing the fact that you have your units wrong and you're ignoring my explicitly stated premise. Now, stop being an asshole.
No matter how you count it, if they were 650 LY away 650 years ago (i.e. we're seeing the effects in starlight now), and they can travel at the outrageous speed of 1/1000 the speed of light, then they're still over 649 LY away.
I hope they do jump on this one. Then we'll see how they try to explain how a planet 650 LY away is any threat to humanity. Even the fastest objects we've detected travel only ~ 250k mph (400k k/h). Let's be very generous and say approximately 1.08M k/h ~ 300 k/s closure rate or about 1/1000 (0.001x) the speed of light. Therefore, it will take at least 650,000 years for it to reach Earth. Oh no, we're doomed!
+/- 5 minutes in a 9 day orbit is a huge variation. This almost has to be a binary planet system, or planet with a massive moon, or something similar. Enough gravitational force to slow or speed up a planet large enough that we can detect it by transit dimming of it's star 650 LY from Earth, that's either a really light planet, or it's got a massive companion orbiting it. The other possibility is that there is a dark star (white/brown dwarf) orbiting the same star, but we should be able to detect that wobble via doppler shift, so the companion moon/planet seems more likely.
I'm familiar with DSDP, ODP, IODP, and Kola. And yes, drilling into rock at 180c-300c requires some engineering. But we have cutting tools that handle much higher temperatures than that. The bottom line is that we've drilled only a little deeper (about 900m) than the Challenger Deep in the Mariani Trench, the lowest known place on earth. Kola only made it about 1/3 of the estimated distance through the crust.
So, which part of "it's about the overall design, not just rounded corners" doesn't get through to you? Nothing anyone has posted other than the Galaxy Tab looks like an iPad or iPhone. There is more to that design patent than a rounded corner rectangle. There are at least 10 other tablet manufacturers who Apple hasn't sued for infringing upon this design patent, because those others don't look strikingly similar to an iPad and don't have all of the specific features covered in the design. The Galaxy tab is the only one anyone has shown that is close enough to violate the design patent.
That's only valid if the designers only used events and data available prior to that event in their design and training of the model. Using knowledge of later events in the design of the model means it's already being "informed by" the future, so even if you retroactively give it only data prior to a past event, the model itself is influenced by future events. Since it's not really possible to keep well known events that occur prior to the design of the model from influencing it's design, and it might not be possible to keep events that occur after design but during implementation from influencing the implementation, you can't use it's accuracy at predicting any events prior to it's completion to test it's ability to predict.
So, you're correct that I understood, and you're completely incorrect about drawing any conclusions from it's ability to "predict" past events based upon feeding it only info prior to the data of the event. Such systems can only be tested after they're completed, and they need to then demonstrate that they're good at predicting events that occur well after their completion. Until they do that, it's very difficult to eliminate design and implementation bias in any results produced while it was being developed, so those results must be considered biased and therefore, are not a reliable indicator of it's ability to predict the future events from current events.
Of course it's good at "predicting" previous events, that's what it was designed to to (and/or was trained by analyzing previous events). If it wasn't good at it, it would be a terrible model. That's not prediction, that's analysis and training. I predict the NYSE will crash on a Monday in Oct 1929, and again on a Monday in Oct 1987. I predict Germany will be on the losing side in both world wars. I predict MS Windows will take 90% of the desktop market despite the Apple Macintosh having nearly a 2 year lead in the market for GUIs and a commanding lead selling Apple IIs in the education market. Wow, I'm better than Nostradamus.
You want to impress me, design and train a system using only knowledge of events prior to 1990, then show me how effective it is at predicting events from 1990 to now using only information available prior to the event it's predicting (you can change 1990 to 2000 or 2005 if that helps). How long before an event does it predict that something is likely? What's the time window of the prediction? How specific are the predictions? What's the accuracy rate?
Weather forecasting is pretty specific, Mostly sunny with a 30% chance of rain in xx area, high temperature of 89F, low of 62F. They're only useful out to about 10 days, but they're pretty accurate and specific. Storm forecasts are (amount of rain, snow, wind speed, hail/size, etc) aren't quite as accurate or specific as the temperature and cloud predictions, and their still pretty reliable, but on a shorter time scale. We've got good (but definitely not perfect) models for near term weather forecasting.
You mean this LG Prada that looks much more like a Palm or WinCE based organizer? The LG Prada that hit the market 2 years after Apple's design patent? The same LG Prada that LG claimed Apple copied the design for the iPhone from, but doesn't look like an iPhone, has more than the all important "single button" on the front, and that LG never sued over (they just claimed the iPhone copied it). That LG Prada?
Wow, the second link sounds a whole lot like the the iPad that was released nearly a year after the article. However, that's only relevant if you ignore the fact that the iPhone and iPod Touch came out about a year before that article, and have essentially the same features and design, only in a smaller form factor, so Apple still has the first to market advantage any way you look at it.
Slashdot Mirror
Seems I messed up including the wikipedia link for Coal in my post above.
No, I'm not wrong. From BP, in its 2007 report, estimated at 2006 end that there were 909,064 million tons of proven coal reserves worldwide, or 147 years reserves-to-production ratio. This figure only includes reserves classified as "proven"; exploration drilling programs by mining companies, particularly in under-explored areas, are continually providing new reserves. In many cases, companies are aware of coal deposits that have not been sufficiently drilled to qualify as "proven".
Going from 143yrs proven reserves to the 200+ years I stated isn't a big stretch, especially as supplied get tighter, conservation becomes more prevalent, and the price and technologies develop to extract the yet unproven or unknown reserves.
The bottom line is that fossil fuels are running out, and at most we've got 250 years, but since we don't yet have a viable replacement for oil, we've really got maybe 50 years to make the transition to largely renewable sources.
Japan is one of the few places that could possibly be powered completely by geothermal. There isn't nearly enough wave energy to supply the planet, nor is there sufficient wave energy near Japan to supply Japan. With a combination of geothermal, wind, hydro, and possibly some solar or wave, Japan might be able to go completely renewable. Most industrialized countries don't have access the the abundant geothermal resources Japan has (due to their location on the edge of the "ring of fire").
Of renewable sources, solar and wind are the ones that can supply enough power for the world, but both are intermittent sources that are not well suited to supplying either base-load or peak-load power without a significant amount of on-demand energy storage added to the grid. On demand energy storage can be in the form of batteries, super capacitors, gravity reservoirs (e.g. pump water uphill to a reservoir during periods of excess generation, release it through turbines when needed), etc. However, solar requires huge amounts of land. Solar and wind each need more than 4x average demand installed (even with on-demand storage, more still without on-demand storage) because they only average ~25% of installed capacity. Neither solar or wind is viable in all areas, and with it's intermittent nature, the grid must have significantly more capacity to route from locations with excess to locations with a shortage.
Bottom line, for most of the world, nuclear and/or fossil fuels are the only currently viable means to meet the difference between renewable capacity and peak demand. Fossil fuels will be exhausted in 50-250 years (~50 yrs oil, slightly longer for natural gas, 200+ years coal). Since plants have a 40-80yr life span, fossil fuel plant built today, could run out of fuel before the plant is used up. Nuclear is the only long term solution that is viable today, and even that needs to move to a thorium fuel model with breeder reactors and fuel reprocessing in order to last more than a few hundred years.
...not lengthened. Shorten it to 40 years (which is twice the 20 yr life of a patent). Copyright holders can then do what everyone else in the world has to do, save for retirement, and/or keep working. And, yes I have created numerous copyrighted works, I speak from experience. As much as the idea of owning copyright for my entire life sounds great (and I'm not opposed to making it 40 years or natural lifetime of a human creator, whichever is greater), I don't see that it's any benefit to the copyright holder or to society to have longer term copyrights. I'm willing to consider a longer term if someone can make a strong case for it, but I haven't yet seen a convincing argument.
Corporations (non human entities) should only hold a copyright for at most 40 years as there is no "natural lifetime" of a corporation. Frankly, if a corporation hasn't recouped their investment after 20-30 years, they're almost certainly not going to, so 20-30 years may be a better maximum for a corporation, but that's another issue. After those years of a corporation holding a copyright, it should either revert to the individual authors/performers, or become public domain.
As for this law, it doesn't extend the copyright on the original lyrics/music, so it doesn't help the songwriters. Who it helps are the entities (e.g. record labels) who hold the copyrights to specific recordings. The performers will see some benefit, but anyone who has looked at recording contracts knows that the performers typically get 10% or less (usually much less after the labels perform their biased math).
Perhaps you missed the significance of a +/- 5 minute variance of a 9 day orbit. That's 5/(9*1440) = 5/12960 = 1/2592. That's nearly 0.04% variance in the orbital period of a planet. To achieve that much change in orbital period, the velocity change needs to be at least that great, and likely at least 2x that great (since speed changes won't be instantaneous, it will have to slow down 2x as much to average 0.04% slower orbital rate).
Given gravity is inversely proportional to the square of the distance between the centers of mass, how much more massive must the other body be to produce that much change in velocity? Even if the other body is in a 4:3 resonance, making it "relatively close", it would still need to be vastly more massive than the observed planet. If it's vastly more massive and in orbit around the same star in a 4:3 resonance, we would definitely be able to detect it by it's doppler shift of the star unless it's orbital plane is almost perfectly perpendicular to our line of sight, and then we would still likely observe an effect on the star.
The other possibility is that there is a third massive body in the system that playing tug-o-war with the planet, pulling it away from the star, then the star pulls it closer, constantly changing it's orbit. That way the velocity of the planet doesn't have to change as much, however, how such a system could produce such an oscillation in the planet's orbit, sometimes increasing it and sometimes decreasing it resulting in such large changes in orbital period without also causing measurable effects on the star is mystery.
We may eventually figure it out, but for now, that much variation in orbital rate is really bizarre.
650 LY, Not 650,000 LY. 1.3M years ago doesn't come into the picture anywhere. Keep your units straight.
We're seeing the system as it was 650 years ago (circa 1360 AD in our reference), therefore, the gravitational effects we're observing mean that there was something there 650 years ago. If it was 650LY away 650 years ago. I started from my explicitly stated premise that the fastest any massive object has been observed to travel is less then 0.001C (actually ~ 0.0004C, then allowed doubled that as a closure rate, then rounded up to 0.001C to make math simpler), then it could be no closer than 650 - (650*0.001C) LY = 649.350 LY from earth.
And since I've now stated it 3 times, I'm not sure how you keep missing the fact that you have your units wrong and you're ignoring my explicitly stated premise. Now, stop being an asshole.
No matter how you count it, if they were 650 LY away 650 years ago (i.e. we're seeing the effects in starlight now), and they can travel at the outrageous speed of 1/1000 the speed of light, then they're still over 649 LY away.
Except, we're still seeing the effect on a planet that is only 650 LY away, so they were 650 LY away 650 years ago.
I hope they do jump on this one. Then we'll see how they try to explain how a planet 650 LY away is any threat to humanity. Even the fastest objects we've detected travel only ~ 250k mph (400k k/h). Let's be very generous and say approximately 1.08M k/h ~ 300 k/s closure rate or about 1/1000 (0.001x) the speed of light. Therefore, it will take at least 650,000 years for it to reach Earth. Oh no, we're doomed!
You can start here and read all the problems you want to.
You think wrong. Don't try to put words in my mouth, I stated exactly what I meant.
+/- 5 minutes in a 9 day orbit is a huge variation. This almost has to be a binary planet system, or planet with a massive moon, or something similar. Enough gravitational force to slow or speed up a planet large enough that we can detect it by transit dimming of it's star 650 LY from Earth, that's either a really light planet, or it's got a massive companion orbiting it. The other possibility is that there is a dark star (white/brown dwarf) orbiting the same star, but we should be able to detect that wobble via doppler shift, so the companion moon/planet seems more likely.
I'm familiar with DSDP, ODP, IODP, and Kola. And yes, drilling into rock at 180c-300c requires some engineering. But we have cutting tools that handle much higher temperatures than that. The bottom line is that we've drilled only a little deeper (about 900m) than the Challenger Deep in the Mariani Trench, the lowest known place on earth. Kola only made it about 1/3 of the estimated distance through the crust.
And then we need a really big straw.
Browser prefetching?
...may still cost you everything that makes it worth living.
So, which part of "it's about the overall design, not just rounded corners" doesn't get through to you? Nothing anyone has posted other than the Galaxy Tab looks like an iPad or iPhone. There is more to that design patent than a rounded corner rectangle. There are at least 10 other tablet manufacturers who Apple hasn't sued for infringing upon this design patent, because those others don't look strikingly similar to an iPad and don't have all of the specific features covered in the design. The Galaxy tab is the only one anyone has shown that is close enough to violate the design patent.
At least 10 other companies make tablets that don't look like iPads and they aren't being sued by Apple.
That's only valid if the designers only used events and data available prior to that event in their design and training of the model. Using knowledge of later events in the design of the model means it's already being "informed by" the future, so even if you retroactively give it only data prior to a past event, the model itself is influenced by future events. Since it's not really possible to keep well known events that occur prior to the design of the model from influencing it's design, and it might not be possible to keep events that occur after design but during implementation from influencing the implementation, you can't use it's accuracy at predicting any events prior to it's completion to test it's ability to predict.
So, you're correct that I understood, and you're completely incorrect about drawing any conclusions from it's ability to "predict" past events based upon feeding it only info prior to the data of the event. Such systems can only be tested after they're completed, and they need to then demonstrate that they're good at predicting events that occur well after their completion. Until they do that, it's very difficult to eliminate design and implementation bias in any results produced while it was being developed, so those results must be considered biased and therefore, are not a reliable indicator of it's ability to predict the future events from current events.
Of course it's good at "predicting" previous events, that's what it was designed to to (and/or was trained by analyzing previous events). If it wasn't good at it, it would be a terrible model. That's not prediction, that's analysis and training. I predict the NYSE will crash on a Monday in Oct 1929, and again on a Monday in Oct 1987. I predict Germany will be on the losing side in both world wars. I predict MS Windows will take 90% of the desktop market despite the Apple Macintosh having nearly a 2 year lead in the market for GUIs and a commanding lead selling Apple IIs in the education market. Wow, I'm better than Nostradamus.
You want to impress me, design and train a system using only knowledge of events prior to 1990, then show me how effective it is at predicting events from 1990 to now using only information available prior to the event it's predicting (you can change 1990 to 2000 or 2005 if that helps). How long before an event does it predict that something is likely? What's the time window of the prediction? How specific are the predictions? What's the accuracy rate?
Weather forecasting is pretty specific, Mostly sunny with a 30% chance of rain in xx area, high temperature of 89F, low of 62F. They're only useful out to about 10 days, but they're pretty accurate and specific. Storm forecasts are (amount of rain, snow, wind speed, hail/size, etc) aren't quite as accurate or specific as the temperature and cloud predictions, and their still pretty reliable, but on a shorter time scale. We've got good (but definitely not perfect) models for near term weather forecasting.
You mean this LG Prada that looks much more like a Palm or WinCE based organizer? The LG Prada that hit the market 2 years after Apple's design patent? The same LG Prada that LG claimed Apple copied the design for the iPhone from, but doesn't look like an iPhone, has more than the all important "single button" on the front, and that LG never sued over (they just claimed the iPhone copied it). That LG Prada?
It's very similar to and closely related to the Streisand effect.
The formula to calculating the perimeter of a rounded corner rectangle is known as Apple Pi.
Well, you got at least one thing right. Anything is better than Vista.
Yes, but the releases of Snow Leopard and especially Lion have already demonstrated the bug FU to PPC owners AND developers. So this is nothing new.
Wow, the second link sounds a whole lot like the the iPad that was released nearly a year after the article. However, that's only relevant if you ignore the fact that the iPhone and iPod Touch came out about a year before that article, and have essentially the same features and design, only in a smaller form factor, so Apple still has the first to market advantage any way you look at it.