Look buddy, I don't doubt that there are PCIe cards that are useful to professionals. What I doubted was the desire to hook them up to a laptop. This post happily provided one example, so I clearly stand corrected. I still don't buy your premise that Macbook Pros with external boxes for these sorts of things are going to be common.
Just so you know, the second link in your list shows a (non-PCIe card, but rather meant-to-be-external) device available with either a Thunderbolt or USB3.0 interface. There is still a Thunderbolt-only device there too that looks to be the higher-end device, and obviously there's the PCIe card version (which is clearly targeted at non-external use given their external device offerings, btw). But doesn't this demonstrate that USB3.0 *can* solve at least some of these more niche use cases, and thus there *is* overlap with/thread to Thunderbolt?
This is the first legitimate use of Thunderbolt that I've heard of. Thanks for providing the example. I suppose Macs are big in the media industries and Thunderbolt solves this problem nicely.
Why buy a desktop when you can simply plug the PCI-E cards straight into your laptop?
What PCIe cards are you plugging in again? Graphics cards? You still have yet to demonstrate that it is not a novelty. I have never seen a CAD setup like that. Nor have I heard of a gaming rig that uses a laptop CPU but has an external graphics box. Maybe you're right and it will be all the rage in CAD houses.
And if you're a pro with a desktop, and you run out of PCI-E slots
You're kidding, right?
A lot of pros are adopting Thunderbolt because it allows them to use the devices that used to require a desktop quickly and easily with a laptop, and they can reduce their machine count by one.
Is there a real use case for connecting a PCI-E card to a system via an external port? The link you showed was basically an enthusiast/hobbyist novelty. If I actually need that sort of graphics power (gamers or CAD), I'm probably using a gaming rig or a workstation, which both have PCI-E slots in the case. I can't imagine what other sort of PCI-E cards I'd be carrying around with my laptop.
The intrinsic value of stock is the value of all dividends the company will ever pay, discounted using the time value of money (I'm simplifying of course). You would hold the stock in order to collect those future dividends. This of course requires a very long-term approach to investing, and with a penalized secondary market, makes it more difficult to move capital from a bad company (that has very slim prospects of producing a future dividend) to a good one.
Fair enough. Now can we stop arguing about the first part of my original post, and move on to the second part, where I provided a different example that debunks your assertion that anything divided by 0 is 0?
The example I provided didn't cover the case you used by design. You said the answer to dividing by zero in the real world is zero. I provided a different example where it is not.
That was an example to demonstrate the fallacy of your question. You can't use anything other than positive natural numbers (i.e. {1,2,3,4,...}) to enumerate a number of something like a "number of parts". Zero "parts" doesn't make any sense. You haven't shown anything generally wrong with dividing by zero in the real world, you've just shown a problem with your own question.
Nah, you've got a couple problems in your example.
First, your exact real-world problem doesn't make sense with a divisor that is less than one, which is how you get close to zero. Five dollars divided into 0.5 parts? What does that even mean?
Now let's flip around your real world example to something that makes sense. Five dollars divided into how many parts, where each part gets $0.50? 5/0.50 = 10 parts. Five dollars divided into how many parts, where each part gets 0.0000001? 50000000. Five dollars divided into how many parts, where each part gets 0? Infinity.
People generally can't feel temperature differences that fine over their whole bodies. For purposes of weather, Celsius units are closer to what people can actually feel.
I think setting the dynamic range of 0-100 based on human extremes (see below) is fine, but I don't think this is an important point really. Both scales have to have finer granularity than "cold, medium, and hot" and they do.
Only because that's what they're used to. In other measurements, what's "normal" isn't going to be 0-100, and people are used to handling them all the time.
One comment on this before we get to your examples. You can't look at averages, because they aren't meaningful - you have to look at the range, so mean+variance+std deviations. Basically, you look at what the likely numbers that people are going to use.
- What's a fast and a slow driving speed? Is it in units of 0-100?
As a matter of fact, yes. 0 is not moving and numbers near it are slow, 100 mph is "really fast" (using the terminology from the original picture I linked). If you're measuring in km/h, you have the same problem as Celsius where the scale isn't nicely correlated to normal human activity:) I think it's merely a happy coincidence that 0-100 mph works out this way (as opposed to Fahrenheit, see below), but it sure did work out nicely.
- What's an average range for human height? Is it in units of 0-100?
Human height doesn't change all that much, and really only changes for about 20 years out of a lifespan of ~80. (Hey, notice how lifespans are also good with 0-100? 0 is really young, 100 is really old)
I guess I have to concede your point though. The scale can be arbitrary and people will get used to it. That doesn't mean that one scale itself is superior or inferior.
Fahrenheit was just retrofitted to it later to be defined by it based on 32 & 212 degrees to make the original, almost completely arbitrary scale have some logical basis. (It was originally based on chilled brine and human body temperature.)
That's exactly my point (no one figured out why Fahrenheit works out this way until you said it just now). Human body temperature being 100 on the scale sets the scale to something that humans have intrinsic experience with. Obviously not human body temperature directly, but outside weather temperature which is related to human body temperature by virtue of evolution giving us a body temperature that tolerates those outside temperatures. On the low end, Fahrenheit used a freezing point of something lower than water, probably something he considered "really cold". My point being, he set the 0-100 range of his scale to extremes that humanes were reasonably likely to encounter. Just judging by that for human use (ignoring all the other SI benefits), I think this makes it a superior scale to Celsius.
Furthermore, somehow 95% of the world somehow seems to exist just fine without Fahrenheit so the integer granularity advantage you are touting seems to be of dubious value. You also seem to be discounting the benefit of being able to communicate with 95% of the world without using a conversion chart.
If converting units is so trivial (as many, many people on this thread point out), you should have no trouble communicating with anyone using Fahrenheit.
Why is 0-100 a significant number instead of, say, 0-32?
0-100 allows for finer granularity of temperature representation without resorting to fractions or decimals, which, while simple enough, are more cumbersome than integers for the average person to deal with. 0-100 also is a natural measure of "low" and "high", not the least of which is because we use percentages all the time which are based on a 0-100 scale. For people who have experience rebasing their definitions of low and high (like math-oriented people), it doesn't matter. For the average person, it does.
Instead of saying, "I'm 80% of the way to freaking hot today," you can say, "I'm 7/8 of the of way to hot today." Wouldn't that be just as nice?
No. 8ths are harder to deal with than 100ths.
We experience temperature more like a street address that we happen to be on -- it's nice here and maybe a little less nice "further down the block." We don't mathematically weigh a 9 point temperature difference so much as recall from experience what that feels like. For telling how comfortable temperature is, the units don't matter at all so long as they can be related to past experience.
I definitely agree with you there. I'm used to Fahrenheit temperature so I know that 70 degrees F is comfortable and 90 degrees F is hot. If I were used to Celsius then I'd be comfortable with its numbers, sure.
In that respect Fahrenheit has no advantages over Celsius except the familiarity of its defenders with it.
I hope you realize that your same argument applies to Celsius as well.
Now they change the colour of the pill or some such minor change and expect another 20 years.
In this particular instance, you're probably right.
My general point stands though. If the Indian government wants to provide access to modern pharmaceuticals at prices that Indians can afford, time for their government to issue some big-money loans to start an Indian pharmaceutical research group. India doesn't have that kind of money? Sell some debt and pay interest on it, just like the U.S. has been forced to do for years. If all these countries had to do their own original research, maybe the U.S. debt situation wouldn't look so terrible in comparison.
Large, influential industries that wouldn't think a second before sending your job overseas for third world labor
Except that they didn't do that when developing these drugs. They paid first-world salaries for research, development, testing, more testing, still more testing, even more testing, and then regulation compliance. Without those first-world costs, there's no drug that you want to sell for third-world prices.
want the USG to make sure said third world labor pays first-world prices for their drugs.
The world wants the US to foot the bill for their drug research, and then once that hard part is done, sell the drugs for materials and menial labor cost? I don't think so. If the prices are so far out of balance, why don't they start their own drug research institute with third-world salaries, testing, and regulations?
Your tax dollars are being used to subsidize security for the games, not live video coverage of the games. Live video coverage of the games is provided by private industry.
You are looking only at one side of the link. There's two sides.
What will stop Verizon from doing this?
Netflix pays Verizon for 100Gbps upstream links at various peering points in the country for whatever Verizon wants to charge. If Verizon doesn't provide 100Gbps from those links because of a bottleneck on Verizon's own network, Netflix sues them for breach of contract. It's Verizon's job to guarantee Netflix that 100Gbps throughout Verizon's network. Repeat for Netflix on AT&T, Google fiber, TimeWarner/Comcast/whatever cable network.
If Netflix doesn't want to do this, then their Verizon-based subscribers will have shitty service. At that point they will either deal with it, or cancel Netflix, or cancel Verizon and switch to an ISP that Netflix does have a traffic contract with.
How does Net Neutrality play a role? Verizon can't refuse to offer Netflix access to their network, or artificially slow down Netflix's traffic on their own network once there is an agreement between Netflix and Verizon. I'm not sure if net neutrality also specifies that Verizon can't charge arbitrarily high prices for network bandwidth to certain companies. That is a good question.
Except that 64-bit ARM (AArch64) doesn't have Thumb. Source.So in 64-bit mode (which is what these server processors will be running in), x86-64 again has a code density advantage over AArch64.
Fact: The U.S. spent $502 billion subsidizing fossil fuels in 2011. This is the result of directly lowered prices, tax breaks and failing to properly price carbon’s negative externalities.
What pricetag is he attaching to carbon's negative externalities? This article, cited in a comment on an earlier Slashdot story on solar energy, pegs it at $1 trillion, and it seems like they asked Austin Powers what he thought it should be. What is Khosla's pricetag for carbon externalities, and where did he get it?
The parent is referring to cost basis analysis for equities. There are several ways. Tracking individual stocks is one option but tedious. See the wikipedia article for details on the other methods:
http://en.wikipedia.org/wiki/Cost_basis
Pick a "brininess" and energy consumption you want to run at. By definition, the brine produced will be more concentrated than the ocean water flowing into your plant. Run the more-concentrated brine and less-concentrated ocean water through this power system and produce whatever energy you can get from it. It will always be less than the energy that you used to produce the freshwater+brine, but it will always be more than 0 which is what you get if you dump the brine back into the ocean. How you want to slide the bars in terms of brininess and energy consumption is up to you. But either way, you're ahead with this system. Get it?
Nope. Osmosis operates off a relative concentration difference. Regular ocean saltwater is much more dilute than brine and would work just fine as the "freshwater" side of this power generator.
Slashdot Mirror
Look buddy, I don't doubt that there are PCIe cards that are useful to professionals. What I doubted was the desire to hook them up to a laptop. This post happily provided one example, so I clearly stand corrected. I still don't buy your premise that Macbook Pros with external boxes for these sorts of things are going to be common.
Just so you know, the second link in your list shows a (non-PCIe card, but rather meant-to-be-external) device available with either a Thunderbolt or USB3.0 interface. There is still a Thunderbolt-only device there too that looks to be the higher-end device, and obviously there's the PCIe card version (which is clearly targeted at non-external use given their external device offerings, btw). But doesn't this demonstrate that USB3.0 *can* solve at least some of these more niche use cases, and thus there *is* overlap with/thread to Thunderbolt?
This is the first legitimate use of Thunderbolt that I've heard of. Thanks for providing the example. I suppose Macs are big in the media industries and Thunderbolt solves this problem nicely.
Why buy a desktop when you can simply plug the PCI-E cards straight into your laptop?
What PCIe cards are you plugging in again? Graphics cards? You still have yet to demonstrate that it is not a novelty. I have never seen a CAD setup like that. Nor have I heard of a gaming rig that uses a laptop CPU but has an external graphics box. Maybe you're right and it will be all the rage in CAD houses.
And if you're a pro with a desktop, and you run out of PCI-E slots
You're kidding, right?
A lot of pros are adopting Thunderbolt because it allows them to use the devices that used to require a desktop quickly and easily with a laptop, and they can reduce their machine count by one.
What devices are these? Still graphics cards?
USB 3.0 added DMA and async (no-polling) control. CPU usage should be on par with FireWire.
Is there a real use case for connecting a PCI-E card to a system via an external port? The link you showed was basically an enthusiast/hobbyist novelty. If I actually need that sort of graphics power (gamers or CAD), I'm probably using a gaming rig or a workstation, which both have PCI-E slots in the case. I can't imagine what other sort of PCI-E cards I'd be carrying around with my laptop.
The intrinsic value of stock is the value of all dividends the company will ever pay, discounted using the time value of money (I'm simplifying of course). You would hold the stock in order to collect those future dividends. This of course requires a very long-term approach to investing, and with a penalized secondary market, makes it more difficult to move capital from a bad company (that has very slim prospects of producing a future dividend) to a good one.
Fair enough. Now can we stop arguing about the first part of my original post, and move on to the second part, where I provided a different example that debunks your assertion that anything divided by 0 is 0?
The example I provided didn't cover the case you used by design. You said the answer to dividing by zero in the real world is zero. I provided a different example where it is not.
That was an example to demonstrate the fallacy of your question. You can't use anything other than positive natural numbers (i.e. {1,2,3,4,...}) to enumerate a number of something like a "number of parts". Zero "parts" doesn't make any sense. You haven't shown anything generally wrong with dividing by zero in the real world, you've just shown a problem with your own question.
Did you read the rest of the post?
Nah, you've got a couple problems in your example.
First, your exact real-world problem doesn't make sense with a divisor that is less than one, which is how you get close to zero. Five dollars divided into 0.5 parts? What does that even mean?
Now let's flip around your real world example to something that makes sense. Five dollars divided into how many parts, where each part gets $0.50? 5/0.50 = 10 parts. Five dollars divided into how many parts, where each part gets 0.0000001? 50000000. Five dollars divided into how many parts, where each part gets 0? Infinity.
People generally can't feel temperature differences that fine over their whole bodies. For purposes of weather, Celsius units are closer to what people can actually feel.
I think setting the dynamic range of 0-100 based on human extremes (see below) is fine, but I don't think this is an important point really. Both scales have to have finer granularity than "cold, medium, and hot" and they do.
Only because that's what they're used to. In other measurements, what's "normal" isn't going to be 0-100, and people are used to handling them all the time.
One comment on this before we get to your examples. You can't look at averages, because they aren't meaningful - you have to look at the range, so mean+variance+std deviations. Basically, you look at what the likely numbers that people are going to use.
- What's a fast and a slow driving speed? Is it in units of 0-100?
As a matter of fact, yes. 0 is not moving and numbers near it are slow, 100 mph is "really fast" (using the terminology from the original picture I linked). If you're measuring in km/h, you have the same problem as Celsius where the scale isn't nicely correlated to normal human activity :) I think it's merely a happy coincidence that 0-100 mph works out this way (as opposed to Fahrenheit, see below), but it sure did work out nicely.
- What's an average range for human height? Is it in units of 0-100?
Human height doesn't change all that much, and really only changes for about 20 years out of a lifespan of ~80. (Hey, notice how lifespans are also good with 0-100? 0 is really young, 100 is really old)
I guess I have to concede your point though. The scale can be arbitrary and people will get used to it. That doesn't mean that one scale itself is superior or inferior.
Fahrenheit was just retrofitted to it later to be defined by it based on 32 & 212 degrees to make the original, almost completely arbitrary scale have some logical basis. (It was originally based on chilled brine and human body temperature.)
That's exactly my point (no one figured out why Fahrenheit works out this way until you said it just now). Human body temperature being 100 on the scale sets the scale to something that humans have intrinsic experience with. Obviously not human body temperature directly, but outside weather temperature which is related to human body temperature by virtue of evolution giving us a body temperature that tolerates those outside temperatures. On the low end, Fahrenheit used a freezing point of something lower than water, probably something he considered "really cold". My point being, he set the 0-100 range of his scale to extremes that humanes were reasonably likely to encounter. Just judging by that for human use (ignoring all the other SI benefits), I think this makes it a superior scale to Celsius.
Only if you have an integer fetish.
Most people do.
Furthermore, somehow 95% of the world somehow seems to exist just fine without Fahrenheit so the integer granularity advantage you are touting seems to be of dubious value. You also seem to be discounting the benefit of being able to communicate with 95% of the world without using a conversion chart.
If converting units is so trivial (as many, many people on this thread point out), you should have no trouble communicating with anyone using Fahrenheit.
Why is 0-100 a significant number instead of, say, 0-32?
0-100 allows for finer granularity of temperature representation without resorting to fractions or decimals, which, while simple enough, are more cumbersome than integers for the average person to deal with. 0-100 also is a natural measure of "low" and "high", not the least of which is because we use percentages all the time which are based on a 0-100 scale. For people who have experience rebasing their definitions of low and high (like math-oriented people), it doesn't matter. For the average person, it does.
Instead of saying, "I'm 80% of the way to freaking hot today," you can say, "I'm 7/8 of the of way to hot today." Wouldn't that be just as nice?
No. 8ths are harder to deal with than 100ths.
We experience temperature more like a street address that we happen to be on -- it's nice here and maybe a little less nice "further down the block." We don't mathematically weigh a 9 point temperature difference so much as recall from experience what that feels like. For telling how comfortable temperature is, the units don't matter at all so long as they can be related to past experience.
I definitely agree with you there. I'm used to Fahrenheit temperature so I know that 70 degrees F is comfortable and 90 degrees F is hot. If I were used to Celsius then I'd be comfortable with its numbers, sure.
In that respect Fahrenheit has no advantages over Celsius except the familiarity of its defenders with it.
I hope you realize that your same argument applies to Celsius as well.
The 0-100 degreen range of Farenheit better represents the range of temperatures that humans encounter. http://imgur.com/gallery/ucOQh
Now they change the colour of the pill or some such minor change and expect another 20 years.
In this particular instance, you're probably right.
My general point stands though. If the Indian government wants to provide access to modern pharmaceuticals at prices that Indians can afford, time for their government to issue some big-money loans to start an Indian pharmaceutical research group. India doesn't have that kind of money? Sell some debt and pay interest on it, just like the U.S. has been forced to do for years. If all these countries had to do their own original research, maybe the U.S. debt situation wouldn't look so terrible in comparison.
Large, influential industries that wouldn't think a second before sending your job overseas for third world labor
Except that they didn't do that when developing these drugs. They paid first-world salaries for research, development, testing, more testing, still more testing, even more testing, and then regulation compliance. Without those first-world costs, there's no drug that you want to sell for third-world prices.
want the USG to make sure said third world labor pays first-world prices for their drugs.
The world wants the US to foot the bill for their drug research, and then once that hard part is done, sell the drugs for materials and menial labor cost? I don't think so. If the prices are so far out of balance, why don't they start their own drug research institute with third-world salaries, testing, and regulations?
Your tax dollars are being used to subsidize security for the games, not live video coverage of the games. Live video coverage of the games is provided by private industry.
You are looking only at one side of the link. There's two sides.
What will stop Verizon from doing this?
Netflix pays Verizon for 100Gbps upstream links at various peering points in the country for whatever Verizon wants to charge. If Verizon doesn't provide 100Gbps from those links because of a bottleneck on Verizon's own network, Netflix sues them for breach of contract. It's Verizon's job to guarantee Netflix that 100Gbps throughout Verizon's network. Repeat for Netflix on AT&T, Google fiber, TimeWarner/Comcast/whatever cable network.
If Netflix doesn't want to do this, then their Verizon-based subscribers will have shitty service. At that point they will either deal with it, or cancel Netflix, or cancel Verizon and switch to an ISP that Netflix does have a traffic contract with.
How does Net Neutrality play a role? Verizon can't refuse to offer Netflix access to their network, or artificially slow down Netflix's traffic on their own network once there is an agreement between Netflix and Verizon. I'm not sure if net neutrality also specifies that Verizon can't charge arbitrarily high prices for network bandwidth to certain companies. That is a good question.
Except that 64-bit ARM (AArch64) doesn't have Thumb. Source.So in 64-bit mode (which is what these server processors will be running in), x86-64 again has a code density advantage over AArch64.
You need to go to better movie theaters. Alamo Drafthouse is the best and got some good publicity when they did this a couple years ago.
http://drafthouse.com/blog/entry/she_texted._we_kicked_her_out
http://oilprice.com/Latest-Energy-News/World-News/Fossil-Fuel-Industry-Receives-1.3-Trillion-in-Subsidies-Each-Year.html
From the letter:
Fact: The U.S. spent $502 billion subsidizing fossil fuels in 2011. This is the result of directly lowered prices, tax breaks and failing to properly price carbon’s negative externalities.
What pricetag is he attaching to carbon's negative externalities? This article, cited in a comment on an earlier Slashdot story on solar energy, pegs it at $1 trillion, and it seems like they asked Austin Powers what he thought it should be. What is Khosla's pricetag for carbon externalities, and where did he get it?
The parent is referring to cost basis analysis for equities. There are several ways. Tracking individual stocks is one option but tedious. See the wikipedia article for details on the other methods: http://en.wikipedia.org/wiki/Cost_basis
Pick a "brininess" and energy consumption you want to run at. By definition, the brine produced will be more concentrated than the ocean water flowing into your plant. Run the more-concentrated brine and less-concentrated ocean water through this power system and produce whatever energy you can get from it. It will always be less than the energy that you used to produce the freshwater+brine, but it will always be more than 0 which is what you get if you dump the brine back into the ocean. How you want to slide the bars in terms of brininess and energy consumption is up to you. But either way, you're ahead with this system. Get it?
Nope. Osmosis operates off a relative concentration difference. Regular ocean saltwater is much more dilute than brine and would work just fine as the "freshwater" side of this power generator.