Virtualization is also the way to go to save power. Fewer servers.
This is only true if you run relatively few tasks that parallelize poorly. If you need to distribute many tasks that parallelize well across a cluster of machines, virtualization is just unnecessary complexity and cost. Also, the redundancy lost by consolidating servers may represent a higher cost than that saved by the lower energy consumption. But yes, for some, virtualization is the way to go.
The point I was trying to make is that disbelief in the the facts of evolution require a different kind of person than one that disbelieves the reigning Theory of Evolution.
So this is a bit of nitpicking, but since you made your anti-fact claim so strongly, I thought I might respond: Evolution actually has two senses, both fact and theory, and neither should be confused with abiogenesis, which is what creationists frequently mean (attack) when they say evolution. The observed facts of evolution aren't going to be "superseded" by anything, but the theory behind it could be (just like any other scientific theory). Evolution is happening, and if a person doesn't believe that, then no amount of debate or logic is going to convince them otherwise, because they can't even see the data. This goes beyond not being a "science type" and suggests some form of mental impairment, IMO.
Except evolution can be proven: scientists regularly evolve new lines of insects, bacteria, and other rapidly reproducing forms of life. Applying the same concepts to life all over the world is just taking a sufficiently large step back so that you can see the bigger picture. Examining the world and coming up with natural explanations that can be experimentally tested is the very essence of the scientific method, which makes it perfectly appropriate to teach in a science classroom. I would support discussing creationism in a science classroom too, but only to teach children why it's not a scientific theory.
A great point, but bear in mind that by removing patent protection, it allows any business in any other country to in turn profit from this American investment. Is that a better use of NASA research?
(Hmm, Slashdot seems to have eaten my reply to this. Sorry if it appears twice.)
This assumes that I own part of NASA or that NASA pays me something out of its profits. All of which is false.
I disagree that this is false. NASA is "owned" by the US government, and we individually have some say in how it uses its assets. NASA's "profits" are the benefits that NASA provides the people of the US (and the world, if you prefer). If NASA is not providing any sort of benefit, you need to write your legislators and tell them to eliminate it, just like you would any other government program that's not doing anything useful.
So in your world, where NASA would be forbidden from selling these patents, could you explain how NASA could recover the sunk cost of that research? The difference between the money you paid for the research, and the fruits of that research, is gone. It's not coming back. It's the value of that research (i.e. the patents) that matters now. NASA is converting that value from a useless thing (patent) into a useful thing (cash). You and your government have exactly the same value of assets as you had before.
The government spent tax dollars for NASA. NASA did research that resulted in patents. The value of those patents is all that matters at this point. The rest is a sunk cost. An auction swaps two things of equal value: here, a patent license, and cash. NASA will be able to do more with the cash than they will with a patent. The company owner will be able to do more with a patent license than they can with the cash. For taxpayers, it's a zero sum, except now we have products coming to market that we wouldn't have had before.
No, I get to pay more every time I buy products using the patented technology.
The value of the research that happened due to your tax contributions is still in NASA's hands, because NASA traded that research for cash at (presumably) a fair market value, via an auction. It's still your money, if you subscribe to the notion that the government is yours.
The sum of the money you've paid NASA in taxes, and the value of NASA's cash and assets, will not have changed at the instant this deal goes through. The value of the patents will be converted to real cash. If you buy products from the company that wins the auction, you will pay for the company's expenses (the value of the patent), because it will be priced into their products, but the cash originally taxed for NASA is still in NASA's hands (minus the difference in value of the research vs. patent, which is a sunk cost and unrecoverable no matter what you do).
It seems to me that by patenting an invention, it allows the US government to ensure that only US companies (taxpayers) can exploit that patent. Making something the equivalent of public domain would seem to allow companies in other countries to profit from this research. It also seems plausible to me that the WTO might have a few things to say about the US government preferentially licensing patents to US companies.
The patents are being auctioned. This means these corporations are paying your government (thus you, indirectly) cash money. It's not corporate welfare, but it is effectively a one-time tax (you paid taxes to fund the research that went into these patents, and the money resulting from their sale is going to the government, not back to you).
No, you'll still have paid only once, assuming the terms of this sale are equitable. Your original investment (taxes paid) into the research that created these patents is being *sold*, not given, to a private company. NASA is getting something in return for this, which means you are getting something in return for it.
"Classless" subnetting allows the use of 0 and 255. If you have a network 10.0.0.0/23 (not/24 aka class C), that means the network ranges from 10.0.0.0 to 10.0.1.255, with only 0.0 and 1.255 reserved. This means 0.255 and 1.0 are perfectly usable, legitimate addresses. The only time this should be problematic is if you've misconfigured your own network (saw 10.0.0.255 and assumed your subnet was 10.0.0.0/24), or if your network devices are horribly broken.
Your MD5 hashing trick is just about useless at actually hiding anything. All you've done is replace the IP address with something derived directly to the IP address, while chopping off ~12 bits of precision (likely much less since many of those IP addresses won't be valid or active unicast addresses). It's trivial to build up a lookup table from IP address to identifier. MD5 is nice if you can't reverse it, but you don't have to reverse it here. If someone subpoenas your logs, and they know the IP address(es) they're looking for, you've handed them a complete record. If your site were high-traffic, with lots of active IP addresses sharing the same identifier, you might get a bit of anonymity out of that for your users, but I'm not convinced this approach is better than Google's.
Do all web browsers support HTML 5? How many browsers support the Gears plug-in? I know nothing about the HTML5 stuff you're talking about, so I don't know why it would be *removed* in favor of the Gears APIs (why not have both?), but since Gears is available as a plug-in for other browsers, it stands to reason that at least in the short-term, supporting Gears is actually *better* for cross-browser compatibility, at least until someone comes up with a plugin that implements these same features using HTML5, or until all of the browser vendors implement it independently.
I've heard this too. I always assumed it was for the same reason you don't want an air conditioner too large for the space it's cooling: it ends up cycling too quickly/frequently. It sounds perfectly reasonable to have a second a/c unit on standby, however, in case the first fails.
And when all of the cool basement air is heated, then what? It's unlikely that the surface area of a basement the size of a server room will dissipate the heat created by a room full of machines. It might help, marginally, but I doubt the benefit will be significant enough to make someone decide to move machines there.
Now, drill some holes deep into the ground, and run a network of fluid-filled pipes through them, and you might be in business (figuratively speaking). I suspect this will be expensive.
Suck the hot air out and draw cold air from the rest of the building in.
Except that the negative pressure you've created in the rest of the building will draw air in from outside. If it's hot out (or too wet/dry), this air must now be cooled (conditioned). All you've done is move the problem out of the machine room to the rest of the building. And now you have to make sure the building A/C can keep up (it's not just the machines getting uncomfortable now).
This only works if the air outside is the right temperature and humidity, because outside air must enter the building to replace the hot air you're venting out the top of your racks. If it's too hot, or the humidity isn't right, you're still going to need air conditioning. You've just moved the problem: instead of containing it in your server room, it's now spread throughout your building (everywhere the negative pressure allows air to enter). I would NOT want to be the server guy when a heat wave hits and the building A/C can't keep up, because even if your servers aren't cooking yet, everyone else in the building might as well be.
If the government doesn't mandate the labeling requirements who will?
So you have two choices, as you intimate: you can either (a) don't mandate anything, and let civil and criminal law step in when someone puts someone else in danger, or (b) mandate labeling, somehow.
When people suffer real harm that can be traced to a specific source, labeling is always the result. You don't need a government to mandate labeling in these cases, because companies will do it voluntarily. This is why we have warning labels on plastic bags, coffee cups and Happy Meals. If someone gets hurt while using your product, and they think there's a tiny, tiny chance that a jury can assign 1% blame to you, they will sue you (the lawyer will work on contingency), and at some point, the costs of defending those lawsuits exceed the costs of a bunch of extra labeling.
When people don't suffer real harm, though, or they suffer real harm, but can only associate it with exposure to something that they'd have been exposed to in a thousand different places, only then do you need the government to step in and do something. So California decided labeling was the best approach. I would hope that the decision of what to put on The List is based on actual risk. I would think that if a chemical was present in a certain type of building, or a typical instance of a product, and that the risk of getting cancer was increased by some specific percentage, that would be the criteria needed to get it on The List. I have no idea if this is how it actually works. In my idealized implementation of such a law, there would be clear, measurable criteria that a substance would need to meet to trigger mandatory labeling, and it would be clear to business owners when they must and must not put up signs. The idea that business owners can be sued for not having a sign, but still having detectable levels of some carcinogen that they could reasonably have never known was there, is actually counter-productive, because it's cheaper to put up a sign and be done with it than it is to have your business/product re-tested every N months.
The question I have is whether or not labeling does any good. I would be willing to wager far less than 1% of the population have altered their behavior or purchasing habits as a result of the signs. With that data, you can revise your cost/benefit projections for this proposition and decide anew whether or not it's worth keeping it. I rather suspect it's not, but I wish good luck to the politician that decides to take that position in California.
Slashdot Mirror
This is only true if you run relatively few tasks that parallelize poorly. If you need to distribute many tasks that parallelize well across a cluster of machines, virtualization is just unnecessary complexity and cost. Also, the redundancy lost by consolidating servers may represent a higher cost than that saved by the lower energy consumption. But yes, for some, virtualization is the way to go.
The point I was trying to make is that disbelief in the the facts of evolution require a different kind of person than one that disbelieves the reigning Theory of Evolution.
So this is a bit of nitpicking, but since you made your anti-fact claim so strongly, I thought I might respond: Evolution actually has two senses, both fact and theory, and neither should be confused with abiogenesis, which is what creationists frequently mean (attack) when they say evolution. The observed facts of evolution aren't going to be "superseded" by anything, but the theory behind it could be (just like any other scientific theory). Evolution is happening, and if a person doesn't believe that, then no amount of debate or logic is going to convince them otherwise, because they can't even see the data. This goes beyond not being a "science type" and suggests some form of mental impairment, IMO.
Except evolution can be proven: scientists regularly evolve new lines of insects, bacteria, and other rapidly reproducing forms of life. Applying the same concepts to life all over the world is just taking a sufficiently large step back so that you can see the bigger picture. Examining the world and coming up with natural explanations that can be experimentally tested is the very essence of the scientific method, which makes it perfectly appropriate to teach in a science classroom. I would support discussing creationism in a science classroom too, but only to teach children why it's not a scientific theory.
Don't confuse evolution with abiogenesis, and take a read through Evolution as theory and fact.
A great point, but bear in mind that by removing patent protection, it allows any business in any other country to in turn profit from this American investment. Is that a better use of NASA research?
(Hmm, Slashdot seems to have eaten my reply to this. Sorry if it appears twice.)
I disagree that this is false. NASA is "owned" by the US government, and we individually have some say in how it uses its assets. NASA's "profits" are the benefits that NASA provides the people of the US (and the world, if you prefer). If NASA is not providing any sort of benefit, you need to write your legislators and tell them to eliminate it, just like you would any other government program that's not doing anything useful.
So in your world, where NASA would be forbidden from selling these patents, could you explain how NASA could recover the sunk cost of that research? The difference between the money you paid for the research, and the fruits of that research, is gone. It's not coming back. It's the value of that research (i.e. the patents) that matters now. NASA is converting that value from a useless thing (patent) into a useful thing (cash). You and your government have exactly the same value of assets as you had before.
The government spent tax dollars for NASA. NASA did research that resulted in patents. The value of those patents is all that matters at this point. The rest is a sunk cost. An auction swaps two things of equal value: here, a patent license, and cash. NASA will be able to do more with the cash than they will with a patent. The company owner will be able to do more with a patent license than they can with the cash. For taxpayers, it's a zero sum, except now we have products coming to market that we wouldn't have had before.
The value of the research that happened due to your tax contributions is still in NASA's hands, because NASA traded that research for cash at (presumably) a fair market value, via an auction. It's still your money, if you subscribe to the notion that the government is yours.
The sum of the money you've paid NASA in taxes, and the value of NASA's cash and assets, will not have changed at the instant this deal goes through. The value of the patents will be converted to real cash. If you buy products from the company that wins the auction, you will pay for the company's expenses (the value of the patent), because it will be priced into their products, but the cash originally taxed for NASA is still in NASA's hands (minus the difference in value of the research vs. patent, which is a sunk cost and unrecoverable no matter what you do).
How so?
It seems to me that by patenting an invention, it allows the US government to ensure that only US companies (taxpayers) can exploit that patent. Making something the equivalent of public domain would seem to allow companies in other countries to profit from this research. It also seems plausible to me that the WTO might have a few things to say about the US government preferentially licensing patents to US companies.
The patents are being auctioned. This means these corporations are paying your government (thus you, indirectly) cash money. It's not corporate welfare, but it is effectively a one-time tax (you paid taxes to fund the research that went into these patents, and the money resulting from their sale is going to the government, not back to you).
The patents are being auctioned, not given away for "a couple of pennies."
No, you'll still have paid only once, assuming the terms of this sale are equitable. Your original investment (taxes paid) into the research that created these patents is being *sold*, not given, to a private company. NASA is getting something in return for this, which means you are getting something in return for it.
"Classless" subnetting allows the use of 0 and 255. If you have a network 10.0.0.0/23 (not /24 aka class C), that means the network ranges from 10.0.0.0 to 10.0.1.255, with only 0.0 and 1.255 reserved. This means 0.255 and 1.0 are perfectly usable, legitimate addresses. The only time this should be problematic is if you've misconfigured your own network (saw 10.0.0.255 and assumed your subnet was 10.0.0.0/24), or if your network devices are horribly broken.
Your MD5 hashing trick is just about useless at actually hiding anything. All you've done is replace the IP address with something derived directly to the IP address, while chopping off ~12 bits of precision (likely much less since many of those IP addresses won't be valid or active unicast addresses). It's trivial to build up a lookup table from IP address to identifier. MD5 is nice if you can't reverse it, but you don't have to reverse it here. If someone subpoenas your logs, and they know the IP address(es) they're looking for, you've handed them a complete record. If your site were high-traffic, with lots of active IP addresses sharing the same identifier, you might get a bit of anonymity out of that for your users, but I'm not convinced this approach is better than Google's.
Do all web browsers support HTML 5? How many browsers support the Gears plug-in? I know nothing about the HTML5 stuff you're talking about, so I don't know why it would be *removed* in favor of the Gears APIs (why not have both?), but since Gears is available as a plug-in for other browsers, it stands to reason that at least in the short-term, supporting Gears is actually *better* for cross-browser compatibility, at least until someone comes up with a plugin that implements these same features using HTML5, or until all of the browser vendors implement it independently.
Did you file a bug?
I've heard this too. I always assumed it was for the same reason you don't want an air conditioner too large for the space it's cooling: it ends up cycling too quickly/frequently. It sounds perfectly reasonable to have a second a/c unit on standby, however, in case the first fails.
And when all of the cool basement air is heated, then what? It's unlikely that the surface area of a basement the size of a server room will dissipate the heat created by a room full of machines. It might help, marginally, but I doubt the benefit will be significant enough to make someone decide to move machines there.
Now, drill some holes deep into the ground, and run a network of fluid-filled pipes through them, and you might be in business (figuratively speaking). I suspect this will be expensive.
Except that the negative pressure you've created in the rest of the building will draw air in from outside. If it's hot out (or too wet/dry), this air must now be cooled (conditioned). All you've done is move the problem out of the machine room to the rest of the building. And now you have to make sure the building A/C can keep up (it's not just the machines getting uncomfortable now).
This only works if the air outside is the right temperature and humidity, because outside air must enter the building to replace the hot air you're venting out the top of your racks. If it's too hot, or the humidity isn't right, you're still going to need air conditioning. You've just moved the problem: instead of containing it in your server room, it's now spread throughout your building (everywhere the negative pressure allows air to enter). I would NOT want to be the server guy when a heat wave hits and the building A/C can't keep up, because even if your servers aren't cooking yet, everyone else in the building might as well be.
Normally Flash does live video streams using RTMP, which wouldn't normally use HTTP.
So you have two choices, as you intimate: you can either (a) don't mandate anything, and let civil and criminal law step in when someone puts someone else in danger, or (b) mandate labeling, somehow.
When people suffer real harm that can be traced to a specific source, labeling is always the result. You don't need a government to mandate labeling in these cases, because companies will do it voluntarily. This is why we have warning labels on plastic bags, coffee cups and Happy Meals. If someone gets hurt while using your product, and they think there's a tiny, tiny chance that a jury can assign 1% blame to you, they will sue you (the lawyer will work on contingency), and at some point, the costs of defending those lawsuits exceed the costs of a bunch of extra labeling.
When people don't suffer real harm, though, or they suffer real harm, but can only associate it with exposure to something that they'd have been exposed to in a thousand different places, only then do you need the government to step in and do something. So California decided labeling was the best approach. I would hope that the decision of what to put on The List is based on actual risk. I would think that if a chemical was present in a certain type of building, or a typical instance of a product, and that the risk of getting cancer was increased by some specific percentage, that would be the criteria needed to get it on The List. I have no idea if this is how it actually works. In my idealized implementation of such a law, there would be clear, measurable criteria that a substance would need to meet to trigger mandatory labeling, and it would be clear to business owners when they must and must not put up signs. The idea that business owners can be sued for not having a sign, but still having detectable levels of some carcinogen that they could reasonably have never known was there, is actually counter-productive, because it's cheaper to put up a sign and be done with it than it is to have your business/product re-tested every N months.
The question I have is whether or not labeling does any good. I would be willing to wager far less than 1% of the population have altered their behavior or purchasing habits as a result of the signs. With that data, you can revise your cost/benefit projections for this proposition and decide anew whether or not it's worth keeping it. I rather suspect it's not, but I wish good luck to the politician that decides to take that position in California.