I've always considered bloated to be a vague criticism that can't be refuted. After all, I can't demonstrate that Firefox isn't bloated because it has no specific meaning. Do you have an actual, specific criticism of Firefox, or are you just dissing it for no particular reason? It's fine if you don't like it, but I'd rather that people admit there's no particular reason for their dislike instead of coming up with a meaningless excuse.
Yes, that is a valid point. But that doesn't mean you can't find a control group for clouds, which is what his assertion is. Finding a control group for clouds is exactly the same as in every other experiment -- random assignment.
Test after test shows that Firefox uses less memory than other browsers. What memory issues are you referring to? If you can explain how to reproduce them, someone can file a bug report for you.
I've been using Firefox for years. If anything, it's only become even better. The other browsers have been catching up, but still aren't as good. I've tried Chrome, but it has quite a few problems. I do have minor problems with Firefox, but worse problems with other browsers. I have two extensions installed, including AdBlock Plus. It works great!
Yes, you can do controlled experiments, but you have to do them in a controlled environment.
No, you do not need a controlled environment. All you need to be able to do is to alter the independent variable and measure the dependent variable. The sampling method makes sure that the differences from the individual trials cancels out by making the probability distribution of the control group and the experimental group close to each other. If you measure a significant difference in the dependent variable between the two groups, you've demonstrated the likelihood of a causal relationship between the indpendent and dependent variable. Try reading the other posts for details explanations of why this is true. Or read any description of the scientific method.
Why would they be losing Google's funding, and if they do, why wouldn't they be able to get funding elsewhere? If Ubuntu switches to a different browser, Firefox will lose only a small fraction of its users. I don't think they've ever lead the market in features; they've led the market in quality. You may have a point on H.264, but they're making an ideological stand to support only freely available technologies. If they need to support H.264, they'll do it.
True, getting rid of waste would be the least of our problems. When anyone could replicate any amount of drugs, poisons, bombs, or nuclear weapons that they could get their hands on, we'd have more pressing issues than environmental concerns.
In any case, it never ceases to amaze me how on Slashdot so many readers can see only the benefits of a new policy or technology. Some put the whatcouldpossiblygowrong tag on articles about any new technology, but it's troubling to think that so many think that getting rid of patents and copyrights could cause only desirable effects. Are they really that naive, or are they just not taking the time to thinking through all the repercussions?
If not being paid removes the incentive to create new products, then how do you explain Linux, or any other Free Software?
I think most development on most open source products is done by paid full-time programmers. In the case of Firefox, the work is mostly done by Mozilla developers who get money from Google because Google is the default search engine. If goods are free, Google can't advertise goods, Mozilla's money dries up, and Firefox development slows way down.
Something breaks, you won't need to get it serviced because you would just get a new one.
I wonder what the repercussions of that would be. Where would we put all the broken possessions? In any case, you would need someone to haul away the old thing and possibly install the new thing. Services, man, services.
Let's just think through the repercussions of replicator technology a bit. Just because material goods are free does not mean services will be free. To get medical treatment, you would need money. To get money, you would need to do something that would enable you to get money, such as provide some kind of service to others. I don't think that just because our basic material needs would be met that we would have 16 hours a day to do what we want. Our jobs would simply change from producing goods to providing services (which is something that is already happening anyway).
Is the "economy collapsing" a good thing or a bad thing? A good thing because everyone has all they want for free? Or a bad thing because now that there's no incentive to pay for products (information, entertainment, ideas) that there's no incentive to create new products (information, entertainment, ideas)?
Yes, I understand what you're saying. It can be difficult to test which lure is better. But it is not impossible. If I and others consistently catch far more fish with lure A than with lure B, we have demonstrated that lure A is better than lure B. If, as you explain, the other variables in the experiment matter so much more than the lures, then in the worst case we would get inconclusive results. In other words, there would be no statistical difference in the number of fish caught by lure A and lure B. I certainly do get your point, but it doesn't mean we cannot do controlled experiments on lures.
Academically the clouds and atmospheric conditions have to be and I emphasize have to be identical in the control group to be considered valid.
Why? And, if so, how can you apply the scientific method to a clinical drug trial? According to your reasoning, you cannot perform clinical drug trials, because you cannot possibly give a placebo and the drug to identical people with identical conditions.
Yes, you are correct that you could be using lure B in the wrong places, if you use lure A in one place and use lure B in another. That's why one would not perform the experiment in this manner. Which is the variable causing the difference in catches? Is it the lure, or the location? If you use both lures in the same location, surely you will be able to tell if one is much more effective than the other. That's how science has been done for hundreds of years. It's the scientific method.
I don't need to be an expert on meteorology to know how to perform controlled experiments on cloud seeding, just as I do not need to know about physiology to know how to perform a controlled experiment on drugs. The scientific method is far easier than you're making it out to be. Talk to your professors about it on Monday.
I agree that fish can behave differently on different days. This is why we repeat experiments instead of taking the results of one experiment as conclusive. Are you seriously saying that if when I use lure A I consistently catch twice as many fish as lure B, I still cannot say that lure A is more effective?
How about if when I give Tylenol to test subjects with headaches, 50% report improvement, vs. 20% for a placebo? Can we not conclude anything about the effectiveness of Tylenol, just because each of the subjects is different and headaches just go away naturally?
There is certainly potential for the experiment to not produce significant results. But there's no reason to believe that we would somehow not be able to make a control group for clouds. To take your example for rocks, you could test the effectiveness of different hammers on breaking rocks.
Let's say we want to test how well a steel hammer compares with a rubber mallet for breaking rocks. We have a pile of 200 assorted igneous rocks, all different. We randomly split these 200 rocks into two piles of 100 rocks each. Even though the rocks may be complicated and all different from each other, the distribution of the properties of the two groups of rocks will be similar. If strike the rocks in one pile with a steel hammer and we strike the rocks in the other pile with a rubber mallet, and the steel hammer breaks 50 rocks and the rubber mallet breaks 25 rocks, why wouldn't we be able to conclude that the steel hammer is better able to break igneous rocks?
Why would the fact that the rocks are all different mean that we would not be able to make a control group?
Yes, I agree that the 200 clouds will all be different. Nonetheless, if you randomly assign 100 clouds to one group and the other 100 to another group, then add up the precipitation totals of the two groups, the totals should be approximately equal. Even though the individual clouds are different, the law of large numbers says that the larger the sample of clouds, the lower the standard deviation of the sample will be. This is basic statistics.
If you then seed one group of clouds and not the other group, and you observe that the seeded group produces much more precipitation than the control group, you would then conclude that cloud seeding works. That the clouds are all different makes no difference, just as it makes no difference that all the people in a clinical drug trial are all different. You may understand meteorology, but you don't understand statistics and controlled experiments.
Yes, exactly. You could perform the test on cumulus
clouds, on stratus clouds, on cirrus clouds, and on
nimbus clouds. Just come up with a criterion for which clouds to try your experiment on, then randomly assign some clouds to a control group and the rest to the experimental group. Just like every other control experiment ever performed. What is it about the scientific method that's so difficult for so many to understand?
If I were going to test a fishing lure, say a worm, I would test it against some other lure, say a small fish. If I caught twice as many trout with the worm than with the small fish, could I not say that the worm is the more effective lure?
If I picked 200 clouds to fly through, and randomly sprinkled iodide crystals in 100 of them and randomly sprinkled salt crystals in the other 100, and the ones sprinkled with iodide crystals produced twice as much rain, could I not say that the iodide crystals are effective at seeding clouds? Why is it that you say it's "impossible to pick a control group"? It's just using a random generator!
It's not a matter of whether I think I'm smart. Understanding how to create controlled experiments is just not that hard.
Slashdot Mirror
I've always considered bloated to be a vague criticism that can't be refuted. After all, I can't demonstrate that Firefox isn't bloated because it has no specific meaning. Do you have an actual, specific criticism of Firefox, or are you just dissing it for no particular reason? It's fine if you don't like it, but I'd rather that people admit there's no particular reason for their dislike instead of coming up with a meaningless excuse.
Yes, that is a valid problem. It would be expensive to perform controlled experiments on clouds. But not impossible.
Yes, that is a valid point. But that doesn't mean you can't find a control group for clouds, which is what his assertion is. Finding a control group for clouds is exactly the same as in every other experiment -- random assignment.
Test after test shows that Firefox uses less memory than other browsers. What memory issues are you referring to? If you can explain how to reproduce them, someone can file a bug report for you.
I've been using Firefox for years. If anything, it's only become even better. The other browsers have been catching up, but still aren't as good. I've tried Chrome, but it has quite a few problems. I do have minor problems with Firefox, but worse problems with other browsers. I have two extensions installed, including AdBlock Plus. It works great!
No, you do not need a controlled environment. All you need to be able to do is to alter the independent variable and measure the dependent variable. The sampling method makes sure that the differences from the individual trials cancels out by making the probability distribution of the control group and the experimental group close to each other. If you measure a significant difference in the dependent variable between the two groups, you've demonstrated the likelihood of a causal relationship between the indpendent and dependent variable. Try reading the other posts for details explanations of why this is true. Or read any description of the scientific method.
Why would they be losing Google's funding, and if they do, why wouldn't they be able to get funding elsewhere? If Ubuntu switches to a different browser, Firefox will lose only a small fraction of its users. I don't think they've ever lead the market in features; they've led the market in quality. You may have a point on H.264, but they're making an ideological stand to support only freely available technologies. If they need to support H.264, they'll do it.
But, but, Google is evil!
All they have to say is that Google steals their data with Chrome and Flock disables those features, and Flock will instantly be very popular.
True, getting rid of waste would be the least of our problems. When anyone could replicate any amount of drugs, poisons, bombs, or nuclear weapons that they could get their hands on, we'd have more pressing issues than environmental concerns.
In any case, it never ceases to amaze me how on Slashdot so many readers can see only the benefits of a new policy or technology. Some put the whatcouldpossiblygowrong tag on articles about any new technology, but it's troubling to think that so many think that getting rid of patents and copyrights could cause only desirable effects. Are they really that naive, or are they just not taking the time to thinking through all the repercussions?
I think most development on most open source products is done by paid full-time programmers. In the case of Firefox, the work is mostly done by Mozilla developers who get money from Google because Google is the default search engine. If goods are free, Google can't advertise goods, Mozilla's money dries up, and Firefox development slows way down.
I wonder what the repercussions of that would be. Where would we put all the broken possessions? In any case, you would need someone to haul away the old thing and possibly install the new thing. Services, man, services.
Let's just think through the repercussions of replicator technology a bit. Just because material goods are free does not mean services will be free. To get medical treatment, you would need money. To get money, you would need to do something that would enable you to get money, such as provide some kind of service to others. I don't think that just because our basic material needs would be met that we would have 16 hours a day to do what we want. Our jobs would simply change from producing goods to providing services (which is something that is already happening anyway).
Is the "economy collapsing" a good thing or a bad thing? A good thing because everyone has all they want for free? Or a bad thing because now that there's no incentive to pay for products (information, entertainment, ideas) that there's no incentive to create new products (information, entertainment, ideas)?
IBM has already developed a high-fidelity 3-D copier. They scrapped the project when they realized they would likely sell only two units.
Yes, I understand what you're saying. It can be difficult to test which lure is better. But it is not impossible. If I and others consistently catch far more fish with lure A than with lure B, we have demonstrated that lure A is better than lure B. If, as you explain, the other variables in the experiment matter so much more than the lures, then in the worst case we would get inconclusive results. In other words, there would be no statistical difference in the number of fish caught by lure A and lure B. I certainly do get your point, but it doesn't mean we cannot do controlled experiments on lures.
Why? And, if so, how can you apply the scientific method to a clinical drug trial? According to your reasoning, you cannot perform clinical drug trials, because you cannot possibly give a placebo and the drug to identical people with identical conditions.
He's even chosen me as a Slashdot foe (he's listed as one of my freaks). I'm honored!
Yes, you are correct that you could be using lure B in the wrong places, if you use lure A in one place and use lure B in another. That's why one would not perform the experiment in this manner. Which is the variable causing the difference in catches? Is it the lure, or the location? If you use both lures in the same location, surely you will be able to tell if one is much more effective than the other. That's how science has been done for hundreds of years. It's the scientific method.
I don't need to be an expert on meteorology to know how to perform controlled experiments on cloud seeding, just as I do not need to know about physiology to know how to perform a controlled experiment on drugs. The scientific method is far easier than you're making it out to be. Talk to your professors about it on Monday.
I agree that fish can behave differently on different days. This is why we repeat experiments instead of taking the results of one experiment as conclusive. Are you seriously saying that if when I use lure A I consistently catch twice as many fish as lure B, I still cannot say that lure A is more effective?
How about if when I give Tylenol to test subjects with headaches, 50% report improvement, vs. 20% for a placebo? Can we not conclude anything about the effectiveness of Tylenol, just because each of the subjects is different and headaches just go away naturally?
There is certainly potential for the experiment to not produce significant results. But there's no reason to believe that we would somehow not be able to make a control group for clouds. To take your example for rocks, you could test the effectiveness of different hammers on breaking rocks.
Let's say we want to test how well a steel hammer compares with a rubber mallet for breaking rocks. We have a pile of 200 assorted igneous rocks, all different. We randomly split these 200 rocks into two piles of 100 rocks each. Even though the rocks may be complicated and all different from each other, the distribution of the properties of the two groups of rocks will be similar. If strike the rocks in one pile with a steel hammer and we strike the rocks in the other pile with a rubber mallet, and the steel hammer breaks 50 rocks and the rubber mallet breaks 25 rocks, why wouldn't we be able to conclude that the steel hammer is better able to break igneous rocks?
Why would the fact that the rocks are all different mean that we would not be able to make a control group?
Yes, I agree that the 200 clouds will all be different. Nonetheless, if you randomly assign 100 clouds to one group and the other 100 to another group, then add up the precipitation totals of the two groups, the totals should be approximately equal. Even though the individual clouds are different, the law of large numbers says that the larger the sample of clouds, the lower the standard deviation of the sample will be. This is basic statistics.
If you then seed one group of clouds and not the other group, and you observe that the seeded group produces much more precipitation than the control group, you would then conclude that cloud seeding works. That the clouds are all different makes no difference, just as it makes no difference that all the people in a clinical drug trial are all different. You may understand meteorology, but you don't understand statistics and controlled experiments.
Yes, exactly. You could perform the test on cumulus clouds, on stratus clouds, on cirrus clouds, and on nimbus clouds. Just come up with a criterion for which clouds to try your experiment on, then randomly assign some clouds to a control group and the rest to the experimental group. Just like every other control experiment ever performed. What is it about the scientific method that's so difficult for so many to understand?
If I were going to test a fishing lure, say a worm, I would test it against some other lure, say a small fish. If I caught twice as many trout with the worm than with the small fish, could I not say that the worm is the more effective lure?
If I picked 200 clouds to fly through, and randomly sprinkled iodide crystals in 100 of them and randomly sprinkled salt crystals in the other 100, and the ones sprinkled with iodide crystals produced twice as much rain, could I not say that the iodide crystals are effective at seeding clouds? Why is it that you say it's "impossible to pick a control group"? It's just using a random generator!
It's not a matter of whether I think I'm smart. Understanding how to create controlled experiments is just not that hard.