This is a very good point indeed! I actually tried to do exactly this at work. But it didn't work out very well. Basically spreadsheets are a "visual" way to do vectorized operations, at least most trivial operations. So people type in one column, type in the other next one and then type in the third cell something like: =D12+E12. Then they take the mouse and drag the operation through every input value. How would you do it in R? You would have to create two vectors with the data:
A = 1:10
B = c(1.1, 3.3,...)
The nice thing about R (or octave/scilab) is obtaining the results:
C = A+B
It is readable, it is simple, self documented but there are a few catches:
(1) Switching "context" - editing mode/calculation
(2) Editing mode is somewhat difficult. Even if you use data.entry or edit.data.frame (this functions open up a very simple spreadsheet window where you can edit the data.
Most people just learn a "simple" tool (aka spreadsheet) and do with it. I know smart people who are stuck with excel for more than 10 years. And basically what they know today is what they new three weeks after using their first spreadsheet program.
After failing miserably trying to convince a very few people to play around with R I'm convinced that if you want get rid of spreadsheets you will need something that looks and feels the same, at least on basic operations. I thought about implementing a spreadsheet->R converter environment. It would look and feel like a spreadsheet but you could open up an editor and write complex functions. I was too lazy and these guys made something very similar running on.NET so I guess you can write things in any language that targets.NET. Once I have time I might try to start developping somthing like that on Linux.
Instead of making more money, people could work less. Instead of buying all sorts of shit, people could do much cleaner things such as talking, writing, riding a bike, going to a brothel, taking a walk, singing, playing a part on a play, painting, fighting (if not pushed too far it is not necessarily bad for some people...).
It is way too simplistic to say that there is a law of nature that says we will end up using every resource available. We are supposed to be rational beings even if we often do stupid things. One of the things of being rational (or partly rational) is that we can choose what we do. We don't simply answer any call of the wild (even if there is such a thing).
We are changing from a production society to a consumer one. We are becoming a bunch of morons that just sit and receive stuff. Not very different from the Eloi in H. G. Wells' The Time Machine. I don't think just consuming is satisfying enough. It is much easier (and faster) just to watch a movie than it is to tell a story and we end up watching 10 movies. Maybe a little boredom is good for creativity. It certainly is much cleaner than riding a car 100 km to do anything "new".
Well, one could say that it is always a political problem...
You are actually right. It is a political problem. But on the other hand, political motivation has a direct relation to public opinion and anyone on/. knows how knowledgeable is public opinion on scientific matters. Most news outlets just say that only dumb people or loony green/anarchist/etc question the use of GM crops.
For whatever reasons, many media outlets automatically side with whatever any large corporation say. Monsanto, a couple of years ago, was running a large PR campaign in Brazil. This campaign used pretty girls to distribute leaflets on university campuses and inter-cities bus stations (that's two places a saw them). I don't see any problem with that except that there were some outright lies: they said that the main motivation of GM crops (soybean) was to reduce use of herbicides. Why would one want to plant a herbicide resistant crop if the objective is to reduce the use of herbicides (I know, things are not that simple, there are nonlinearities involved...).
You can be sure that any judge reviewing a related process will be affected by this sort of campaign. IMHO, democracy is not possible without free flow of information from every side involved. Monsanto's load of money is really affecting government policies, media opinion and scientific research. It is stiffening debate on a very important issue: food security on a world were pests and diseases are global and the weather has gone crazy (whatever the causes).
Paulo
From the faq http://softwarecommunity.intel.com/tbbWiki/FAQ/591 .htm :
FAQ
What compilers, operating systems and processors are supported?
Version 2, Changed by RANDY SMITH on 7/23/2007
Created by: JAMES REINDERS
The project is dedicated to supporting all compilers, all OSes and all processors as a cornerstone objective of the project. Up to date information on status is available on the web site.
You said: You mean food that doesn't need to be sprayed with (as much) pesticide because it's biologically resistant to insects?.
But the most common GM crop in the world is monsanto's soybean whose name is Roundup Ready (RR) which is resistant to Roundup which is a pesticide manufactured by monsanto. Why would someone use a pesticide resistant plant if not for using more pesticide?
But, is it possible to GM a plant to be resistant to a pest? Yes but i would only work with that pest. Pests change from place to place and from time to time. A crop resistant to one disease would be useful in one place at a period of time. You know, it is not exactly cheap to engineer a resistant plant.
Again: is it possible? YES. Will it be done? Probably not. Monsanto will stick with pesticide resistant genetic modifications.
Let me just tell you about some problems in Brazil. There is a state, Parana that was trying to stay GM free. Federal justice said that a state can not rule on that and they had to open up to GM soy bean. Non GM grain has a higher value (europe doesn't buy GM grain). Taxes are paid when the grain arrives on the port. If you declare that you are producing GM grain, you will pay 2% taxes. If you don't declare that, the government will test your production for GM grain. If the government finds more than 1% of GM grain, your crop is considered GM and the taxes rise to 3% and you have to pay for the tests, which is quite expensive (I don't know how much). 1% GM limit is easily reached through cross-pollination. They still have to pay royalties to Monsanto.
Since these costs are fairly high, many producers are choosing to declare their crops GM. Overall effects: basically, you can not choose to grow non GM crops.
Not saying that I agree with those numbers but you don't know much about statistics or uncertainties. You don't even know much about significant figures! Every measurement has an uncertainty. Sometimes this uncertainty is small and sometimes it is large. If you go to the butcher and ask for 1kg of meat, The guy will charge you for 1kg but you won't get 1kg, probably something like 1.002 (probably worse). Governments regulate this sort of thing and that's why there are large and costly institutions like NIST (in the USA).
I did not read the paper that was mentioned but if it was published in a scientific journal it will probably mention somewhere the uncertainty value. So the actual number is something like 150.000 +/- 50.000 with a probability of 90% (there are other ways to declare the uncertainty). Now these hypothetical numbers mean that the actual number of deaths could be larger than 200.000 or lower than 100.000 but the probability of the number being outside the given range is close to 10%. 150.000 is just the "nice" number closest most likely number.
If you want to question the numbers, the way to go about it is to question the sampling techniques. That's how polls and research is usually manipulated. They certainly didn't go around the world determining what cause each death. The sample used to get these numbers could be biased (counting only the deaths near some sort of disaster for instance). Good sampling is *really* hard if not impossible. To get a good sample it takes a lot of experience, deep knowledge of the subject under study and a lot of honesty.
BTW, didn't your high school physics teacher tell you to drop all those digits of the calculator display?
Actually, the definition of statistics is to torture data until it confesses what you!
Re:Danger of 'GPLv3 or later'
on
GPLv2 Vs. GPLv3
·
· Score: 2, Informative
That's the reason for the word 'or'. The software is double licensed. If you don't like the provisions of GPL version n just use the other side of 'or'. Basically this mean that you can not, in practice, restrict the code anymore than the most permissive side of the 'or' clause.
I don't think you know very well how a scientific model is developed. "Trained" models exist but they usually don't get very far (far as in getting published in a respected journal). Before model predictions are published, the model must be analyzed thoroughly: contributions should be specified and things left out should be justified. By the way, what new contributions do you read every week? In the mainstream the contributions are hardly ever mentioned, except some major one that the article is trying to refute or support. What is more controversial is how the parameters affect and are handled by the model.
How these contributions affect the problem is another issue that is both difficult and very important. In complex models, each contribution might be the object of one or several papers. They should reflect knowledge of other models, experimental results and observations.
In this sense the model is trained. It should reproduce historic observations. If it can not do that what is its use? Every contribution and term must be justified and tested. That's why there is peer review. Every scientific funding agency has a rank of publications and top publications are usually very thorough, reviewers comments are though and it might take years to publish a study.
What will the models look like in 30 years, who knows. That's why uncertainty analysis is so important and any decent model provides that. Future models will certainly be much more accurate but well done uncertainty analysis usually provides good error bounds to predictions. You should not mistake uncertainty with a chess program bug.
I know they are in thew bottom 10 of the list but anyway the oppression in Cuba does not come close to those countries where people are sometimes chopped to death. Foreign journalists can go basically anywhere they want in Cuba and they will not get killed or tortured. Hell, they will probably not even get arrested (foreign journalists). Is freedom limited in Cuba? Of course it is! But to compare it to Uzbequistan or whatever does not help. There is a reason why the cuban regime is so stable even after almost 50 years. And it is not because of brutality. So, if you believe that cubans should have more freedom stop repeating what the US government says (or what US funded NGOs say). This only helps the hard cores of the cuban regime.
One additional thing that I didn't mention on my previous post. RMS wasn't talking about Cuba. He was talking about things happening in an US base. So it is an american criticizing the US government. I don't see anything wrong with that.
In my opinion, talking about free software is very helpful to get people thinking about freedom of speech. You might not realize it but most people just don't care about "political stuff" such as "freedom of speech". I know many people who simply did not ever think about freedom. Once they started to use free software that changed radically (two of these guys are Opus Dei).
Instead of the US always trying to bomb their way to a solution, maybe not trying to block internet and other developments from Cuba could actually be a good way to promote democracy. On the other hand this would make it difficult to substitute the Castro Regime by a friendly one led by terrorists such as Posada Carriles.
Not saying that Cuba is paradise or even a good place but come on! Second worst? You are quoting from Reporters without Borders, a NGO that receive a lot of cash from US State Department and other departments. That's why their main enemies lately are Cuba and Venezuela.
It is difficult to believe that Cuba is worse than Saudi Arabia, a country that does not allow non business visits by any non-muslim. Uzbequistan is certainly much better than Cuba. The list of US client states that are a "paradise" compared to Cuba is long.
Iraq is also much better! If you don't get killed by the different kind of wackos you will be popped by some trigger happy "coalition" soldier. China is another fine example of free press.
With so many countries with free press, no wonder Cuba and Venezuela are at the bottom.
Trains are used primarily for freight, at least in most places. It is *much* cheaper.
In the case of passenger travel, there are situations where trains are better. Even in Europe planes are often much cheaper than trains. But if two cities are something like 800 km apart and there are many passengers, high speed trains are much better. 3 ours between city centers. No need to go to the airport and hardly any delay.
Weight is not unlimited, it just not as important as in airplanes, that's what I meant. The rails and trains have to be designed for the weight. Heavy compositions take a while to accelerate and going uphill can make the engines sweat.
You make good points, but to be more exact the drag of airplanes and high speed trains do depend on length.
Why?
Because they are streamlined and very little flow separation occurs at the back and most of the drag is the result surface shear and this depends on length (kind of like pressure drop in long pipes).
On the other hand, planes fly at high altitude and air density is considerably lower (drag has a linear relation to density).
But I think the most important aspect is the weight. In trains weight is basically "unlimited" and your assertion of drag/passenger ratio is central in this discussion. But for planes, weight requires more lift which, even with ow drag/lift ratios, results in larger lifting surfaces and higher drag.
I believe that most of the Solaris nice features that could benefit Linux is not very easy to port. On the other hand, Linux does have one very nice thing that could really benefit Solaris without much trouble: tons of drivers. Probably they would have to develop a few interfaces but this would be good for solaris.
I guess it is hard to say who would benefit the most and who would lose given a license compatibility.
1) Nobody said anything about Windows. Why are Linux users so unable to let Linux stand on its own? You never see Mac users constantly comparing everything about OS X to Windows, instead they judge OS X on its own merits and criticize it for its own failings.
Are you insane? I have never heard of a Mac User admitting a failing of OS X!
Violent weather only over oceans or large bodies of water. To be able to sustain a strong storm, a continuous flux of energy wich is provided by the sun and high water surface temperature. Without this energy input storms die out quickly.
Any way I think that the best and most accurate way to determine variations in the sun is by direct measurement which is not difficult to make and if anything were going on it would have been discovered already (probably). I think it is a very long shot saying that climate change is a result of variations in the sun because of perceived variations in the weather of Mars of which we know next to nothing.
Don't worry. If there is any credibility to this, it will be used to exhaustion.
There is one thing called super-saturated mixtures. In the atmosphere it is common to have relative humidities of 700% without drop formation.
An important aspect in cloud formation is the presence of particles from where the drops can form. Sometimes silver nitrate (if I remember correctly) was employed as seeding to develop the drops (or larger drops).
Cities are a large source of dust particles that contribute to drop formation.
Now, how all these factors work together to produce rain or control temperature is a much, much more difficult problem.
Slashdot Mirror
Actually, the first successful ICBM test was realized in the USSR. The wikipedia article on ICBM's has some of the history.
This is a very good point indeed! I actually tried to do exactly this at work. But it didn't work out very well. Basically spreadsheets are a "visual" way to do vectorized operations, at least most trivial operations. So people type in one column, type in the other next one and then type in the third cell something like: =D12+E12. Then they take the mouse and drag the operation through every input value. How would you do it in R? You would have to create two vectors with the data:
...)
.NET so I guess you can write things in any language that targets .NET. Once I have time I might try to start developping somthing like that on Linux.
A = 1:10
B = c(1.1, 3.3,
The nice thing about R (or octave/scilab) is obtaining the results: C = A+B
It is readable, it is simple, self documented but there are a few catches:
(1) Switching "context" - editing mode/calculation
(2) Editing mode is somewhat difficult. Even if you use data.entry or edit.data.frame (this functions open up a very simple spreadsheet window where you can edit the data.
Most people just learn a "simple" tool (aka spreadsheet) and do with it. I know smart people who are stuck with excel for more than 10 years. And basically what they know today is what they new three weeks after using their first spreadsheet program.
After failing miserably trying to convince a very few people to play around with R I'm convinced that if you want get rid of spreadsheets you will need something that looks and feels the same, at least on basic operations. I thought about implementing a spreadsheet->R converter environment. It would look and feel like a spreadsheet but you could open up an editor and write complex functions. I was too lazy and these guys made something very similar running on
Let's just hope that these new solar cells will help us build a warp drive so we can conquer the alpha quadrant!
These are *VERY* wise words!
Instead of making more money, people could work less. Instead of buying all sorts of shit, people could do much cleaner things such as talking, writing, riding a bike, going to a brothel, taking a walk, singing, playing a part on a play, painting, fighting (if not pushed too far it is not necessarily bad for some people...).
It is way too simplistic to say that there is a law of nature that says we will end up using every resource available. We are supposed to be rational beings even if we often do stupid things. One of the things of being rational (or partly rational) is that we can choose what we do. We don't simply answer any call of the wild (even if there is such a thing).
We are changing from a production society to a consumer one. We are becoming a bunch of morons that just sit and receive stuff. Not very different from the Eloi in H. G. Wells' The Time Machine. I don't think just consuming is satisfying enough. It is much easier (and faster) just to watch a movie than it is to tell a story and we end up watching 10 movies. Maybe a little boredom is good for creativity. It certainly is much cleaner than riding a car 100 km to do anything "new".
Well, one could say that it is always a political problem...
/. knows how knowledgeable is public opinion on scientific matters. Most news outlets just say that only dumb people or loony green/anarchist/etc question the use of GM crops.
You are actually right. It is a political problem. But on the other hand, political motivation has a direct relation to public opinion and anyone on
For whatever reasons, many media outlets automatically side with whatever any large corporation say. Monsanto, a couple of years ago, was running a large PR campaign in Brazil. This campaign used pretty girls to distribute leaflets on university campuses and inter-cities bus stations (that's two places a saw them). I don't see any problem with that except that there were some outright lies: they said that the main motivation of GM crops (soybean) was to reduce use of herbicides. Why would one want to plant a herbicide resistant crop if the objective is to reduce the use of herbicides (I know, things are not that simple, there are nonlinearities involved...).
You can be sure that any judge reviewing a related process will be affected by this sort of campaign. IMHO, democracy is not possible without free flow of information from every side involved. Monsanto's load of money is really affecting government policies, media opinion and scientific research. It is stiffening debate on a very important issue: food security on a world were pests and diseases are global and the weather has gone crazy (whatever the causes). Paulo
From the faq http://softwarecommunity.intel.com/tbbWiki/FAQ/591 .htm :
FAQ
What compilers, operating systems and processors are supported?
Version 2, Changed by RANDY SMITH on 7/23/2007
Created by: JAMES REINDERS
The project is dedicated to supporting all compilers, all OSes and all processors as a cornerstone objective of the project. Up to date information on status is available on the web site.
You are right. But I guess a better term would be agrotoxic.
You said: You mean food that doesn't need to be sprayed with (as much) pesticide because it's biologically resistant to insects?.
But the most common GM crop in the world is monsanto's soybean whose name is Roundup Ready (RR) which is resistant to Roundup which is a pesticide manufactured by monsanto. Why would someone use a pesticide resistant plant if not for using more pesticide?
But, is it possible to GM a plant to be resistant to a pest? Yes but i would only work with that pest. Pests change from place to place and from time to time. A crop resistant to one disease would be useful in one place at a period of time. You know, it is not exactly cheap to engineer a resistant plant.
Again: is it possible? YES. Will it be done? Probably not. Monsanto will stick with pesticide resistant genetic modifications.
Let me just tell you about some problems in Brazil. There is a state, Parana that was trying to stay GM free. Federal justice said that a state can not rule on that and they had to open up to GM soy bean. Non GM grain has a higher value (europe doesn't buy GM grain). Taxes are paid when the grain arrives on the port. If you declare that you are producing GM grain, you will pay 2% taxes. If you don't declare that, the government will test your production for GM grain. If the government finds more than 1% of GM grain, your crop is considered GM and the taxes rise to 3% and you have to pay for the tests, which is quite expensive (I don't know how much). 1% GM limit is easily reached through cross-pollination. They still have to pay royalties to Monsanto.
Since these costs are fairly high, many producers are choosing to declare their crops GM. Overall effects: basically, you can not choose to grow non GM crops.
I think that your turnkey solution will become more common as virtualization becomes popular. This could be a solution to several problems.
Not saying that I agree with those numbers but you don't know much about statistics or uncertainties. You don't even know much about significant figures! Every measurement has an uncertainty. Sometimes this uncertainty is small and sometimes it is large. If you go to the butcher and ask for 1kg of meat, The guy will charge you for 1kg but you won't get 1kg, probably something like 1.002 (probably worse). Governments regulate this sort of thing and that's why there are large and costly institutions like NIST (in the USA). I did not read the paper that was mentioned but if it was published in a scientific journal it will probably mention somewhere the uncertainty value. So the actual number is something like 150.000 +/- 50.000 with a probability of 90% (there are other ways to declare the uncertainty). Now these hypothetical numbers mean that the actual number of deaths could be larger than 200.000 or lower than 100.000 but the probability of the number being outside the given range is close to 10%. 150.000 is just the "nice" number closest most likely number. If you want to question the numbers, the way to go about it is to question the sampling techniques. That's how polls and research is usually manipulated. They certainly didn't go around the world determining what cause each death. The sample used to get these numbers could be biased (counting only the deaths near some sort of disaster for instance). Good sampling is *really* hard if not impossible. To get a good sample it takes a lot of experience, deep knowledge of the subject under study and a lot of honesty. BTW, didn't your high school physics teacher tell you to drop all those digits of the calculator display?
And remember, copying a CD is much worse offense than tax fraud!!!
You do realize that your idea is way too intelligent for these people? The morons believe that they humans used to ride vegetarian T-REXs!
Well, according to Bakunin (God and State) Eve was the first revolutionary. She became free when she ate the apple.
Actually, the definition of statistics is to torture data until it confesses what you!
That's the reason for the word 'or'. The software is double licensed. If you don't like the provisions of GPL version n just use the other side of 'or'. Basically this mean that you can not, in practice, restrict the code anymore than the most permissive side of the 'or' clause.
I don't think you know very well how a scientific model is developed. "Trained" models exist but they usually don't get very far (far as in getting published in a respected journal). Before model predictions are published, the model must be analyzed thoroughly: contributions should be specified and things left out should be justified. By the way, what new contributions do you read every week? In the mainstream the contributions are hardly ever mentioned, except some major one that the article is trying to refute or support. What is more controversial is how the parameters affect and are handled by the model.
How these contributions affect the problem is another issue that is both difficult and very important. In complex models, each contribution might be the object of one or several papers. They should reflect knowledge of other models, experimental results and observations.
In this sense the model is trained. It should reproduce historic observations. If it can not do that what is its use? Every contribution and term must be justified and tested. That's why there is peer review. Every scientific funding agency has a rank of publications and top publications are usually very thorough, reviewers comments are though and it might take years to publish a study.
What will the models look like in 30 years, who knows. That's why uncertainty analysis is so important and any decent model provides that. Future models will certainly be much more accurate but well done uncertainty analysis usually provides good error bounds to predictions. You should not mistake uncertainty with a chess program bug.
I know they are in thew bottom 10 of the list but anyway the oppression in Cuba does not come close to those countries where people are sometimes chopped to death. Foreign journalists can go basically anywhere they want in Cuba and they will not get killed or tortured. Hell, they will probably not even get arrested (foreign journalists). Is freedom limited in Cuba? Of course it is! But to compare it to Uzbequistan or whatever does not help. There is a reason why the cuban regime is so stable even after almost 50 years. And it is not because of brutality. So, if you believe that cubans should have more freedom stop repeating what the US government says (or what US funded NGOs say). This only helps the hard cores of the cuban regime.
One additional thing that I didn't mention on my previous post. RMS wasn't talking about Cuba. He was talking about things happening in an US base. So it is an american criticizing the US government. I don't see anything wrong with that.
In my opinion, talking about free software is very helpful to get people thinking about freedom of speech. You might not realize it but most people just don't care about "political stuff" such as "freedom of speech". I know many people who simply did not ever think about freedom. Once they started to use free software that changed radically (two of these guys are Opus Dei).
Instead of the US always trying to bomb their way to a solution, maybe not trying to block internet and other developments from Cuba could actually be a good way to promote democracy. On the other hand this would make it difficult to substitute the Castro Regime by a friendly one led by terrorists such as Posada Carriles.
Paulo
Not saying that Cuba is paradise or even a good place but come on! Second worst? You are quoting from Reporters without Borders, a NGO that receive a lot of cash from US State Department and other departments. That's why their main enemies lately are Cuba and Venezuela.
It is difficult to believe that Cuba is worse than Saudi Arabia, a country that does not allow non business visits by any non-muslim. Uzbequistan is certainly much better than Cuba. The list of US client states that are a "paradise" compared to Cuba is long.
Iraq is also much better! If you don't get killed by the different kind of wackos you will be popped by some trigger happy "coalition" soldier. China is another fine example of free press.
With so many countries with free press, no wonder Cuba and Venezuela are at the bottom.
Trains are used primarily for freight, at least in most places. It is *much* cheaper.
In the case of passenger travel, there are situations where trains are better. Even in Europe planes are often much cheaper than trains. But if two cities are something like 800 km apart and there are many passengers, high speed trains are much better. 3 ours between city centers. No need to go to the airport and hardly any delay.
Weight is not unlimited, it just not as important as in airplanes, that's what I meant. The rails and trains have to be designed for the weight. Heavy compositions take a while to accelerate and going uphill can make the engines sweat.
You make good points, but to be more exact the drag of airplanes and high speed trains do depend on length.
Why?
Because they are streamlined and very little flow separation occurs at the back and most of the drag is the result surface shear and this depends on length (kind of like pressure drop in long pipes).
On the other hand, planes fly at high altitude and air density is considerably lower (drag has a linear relation to density).
But I think the most important aspect is the weight. In trains weight is basically "unlimited" and your assertion of drag/passenger ratio is central in this discussion. But for planes, weight requires more lift which, even with ow drag/lift ratios, results in larger lifting surfaces and higher drag.
I believe that most of the Solaris nice features that could benefit Linux is not very easy to port. On the other hand, Linux does have one very nice thing that could really benefit Solaris without much trouble: tons of drivers. Probably they would have to develop a few interfaces but this would be good for solaris.
I guess it is hard to say who would benefit the most and who would lose given a license compatibility.
1) Nobody said anything about Windows. Why are Linux users so unable to let Linux stand on its own? You never see Mac users constantly comparing everything about OS X to Windows, instead they judge OS X on its own merits and criticize it for its own failings.
Are you insane? I have never heard of a Mac User admitting a failing of OS X!
Violent weather only over oceans or large bodies of water. To be able to sustain a strong storm, a continuous flux of energy wich is provided by the sun and high water surface temperature. Without this energy input storms die out quickly.
Any way I think that the best and most accurate way to determine variations in the sun is by direct measurement which is not difficult to make and if anything were going on it would have been discovered already (probably). I think it is a very long shot saying that climate change is a result of variations in the sun because of perceived variations in the weather of Mars of which we know next to nothing.
Don't worry. If there is any credibility to this, it will be used to exhaustion.
There is one thing called super-saturated mixtures. In the atmosphere it is common to have relative humidities of 700% without drop formation.
An important aspect in cloud formation is the presence of particles from where the drops can form. Sometimes silver nitrate (if I remember correctly) was employed as seeding to develop the drops (or larger drops).
Cities are a large source of dust particles that contribute to drop formation.
Now, how all these factors work together to produce rain or control temperature is a much, much more difficult problem.