When you use floats to do the arithmetic, you lose precision in each operation, and particularly when you multiply two numbers with different scales (exponents).
On prediction failure, Hansen's 1988 "A,B,C" forecasts of rising temperature are rapidly diverging from the cooling we are actually experiencing right now, where case C assumed we massively limited CO2 also
In the past ten years, we've seen warming of 0.18 degrees Celsius, which is less than the 0.25 degrees Celsius that was predicted, but it certainly hasn't been cooling. This is why the Arctic ice and Antarctic ice are melting. Yes, stop the presses, the globe is warming!
From recent events, I think both A and B are wrong. When an error is pointed out in research that shows AGW is happening, people use that error as an excuse not to believe any research that AGW is happening, even years after the error is corrected. When an error is pointed out in the IPCC report about a minor effect of climate change, people use that error to doubt all effects of climate change. Correcting the errors or pointing out they don't change the results will not silence the critics. It will only make the critics claim that their opinion is being suppressed even though the science has been indisputably proven to be flawed and therefore cannot be trusted!
Let's think through what would really happen if scientists released their code. The code has bugs, as all code does. People with an ulterior motive would point to the bugs and say "Look here! A bug! The science cannot be trusted!" And millions of sheeple would repeat "Yes! The code has bugs! And therefore I refuse to believe it!" It won't matter whether the bugs are relevant to the science; the fact that there are any bugs at all will cause people who want to disagree to say there's doubt about the results. Meanwhile, they will go about their business using computer systems that are riddled with bugs, but function well enough the vast majority of the time they're not even aware of the bugs.
Opera is massive in Eastern Europe. On Russian, Ukrainian, and Polish sites it makes sense to push Opera. But does it make sense to push Opera to English users, based on Opera usage on mobile devices, when most visitors to the site are using a desktop browser? A study done by a Mozilla employee shows users are more willing to switch to Chrome or Firefox. Few users of non-Opera browsers are willing to switch to Opera.
I don't know whether Opera does or doesn't fully pass the Acid3 test. I can tell you with certainty that I have never seen Opera Software ASA make that claim. If you could point me to where they do make such a claim, I would be truly grateful.
No, that was when they reached a 100% score with correct rendering. Safari did not pass the performance aspect of the test until the post I gave. I still have not seen Opera claim they pass the performance aspect of the test. You can see Opera's acknowledgement of this little detail in the post you cite
There are some remaining issues yet to be fixed, but we hope to have those sorted out shortly.
We're continuing to increase the concentration of carbon dioxide in the atmosphere. That's predicted to cause a few degrees of warming, which is in turn predicted to increase sea level by a meter or more. It is technologically impossible to stop any further rise in sea level, according to the vast majority of climatologists.
I've heard climatologists say that we may have reached a tipping point, meaning that no matter how much we reduce carbon dioxide emissions we could be locked into several degrees Celsius temperature rise and several meters of sea level rise. That will be costly, but it's not going to destroy humanity, or even human civilization.
Who is the Climate-Change-will-destroy-humanity crowd? I haven't heard of anyone saying that climate change will destroy humanity. Climate change could prove disastrous, but disasters happen all the time and don't wipe humanity off the face of the planet.
Arrhenius estimated that halving of CO2 would decrease temperatures by 4 - 5 C (Celsius) and a doubling of CO2 would cause a temperature rise of 5 - 6 C.
You're right. I think we've all taken the wrong approach with huge, bloated standard libraries. Let all developers write all code from scratch. Need to output an integer, just write the code that turns the integer into a stream of characters, then pass that stream of characters into your homebrew I/O functions, which pass them off to your custom built drivers. There's no need for all languages to have this functionality! It just makes developers have to code around the differences and bugs in each runtime!
No one is calling for removing all greenhouse gasses from the atmosphere. That would be suicide. The current plan is to reduce greenhouse emissions dramatically, so that the amount of carbon dioxide in the atmosphere can stabilize.
As for your idea that reduced carbon dioxide emissions have led to cooling, that's nonsense. Carbon dioxide emissions have been rising every year except for the past year or two, and every year the concentration of carbon dioxide in the atmosphere has been increasing. The recent plateau in temperatures is more likely due to a natural cooling phase balancing out the warming due to increased greenhouse gasses.
Yes, there will be some positive effects due to global warming, such as more arable land in the high latitudes. On the other hand, there will also be negative consequences, such as drought and desertification in other areas, and rising sea levels which will cause millions of people to have to relocate. There is general agreement among world leaders that warming over 2 degrees Celsius would result in overall negative consequences, so they're trying to limit the concentration of carbon dioxide in the atmosphere to 450-550 ppm to limit the effects.
Computer modelling has added nothing to this. No computer model had produced better results than simply drawing a straight line through the graph.
What computer modeling has added is that there are not subtle effects that were not in the simpler models that will significantly alter the warming. For example, it has been hypothesized that certain types of clouds will cancel out most of the warming, or that other types of clouds will cause even more warming that previously predicted. The newest models and data show that the previous predictions are pretty accurate.
As far as your assertion that these models are untested, you completely wrong. These models have predicted warming for over a century, and we've been seeing that warming since the 1970s. We have decades of evidence that confirm the models. The fact that people continue to say the models are "untested" is why we need to have more stories about the matter.
Exactly. That's why countries are trying to agree to cut emissions. It needs the cooperation of at least China and the U.S. That's why the climate bill that's before the U.S. Senate is such a big deal. Without a climate bill, the U.S. cannot commit to cutting emissions, and so other countries aren't willing to do so either.
Slashdot Mirror
Why would multiplying numbers cause you to lose precision? I think a more common way to lose precision would be to subtract two numbers that are nearly equal to each other.
In the past ten years, we've seen warming of 0.18 degrees Celsius, which is less than the 0.25 degrees Celsius that was predicted, but it certainly hasn't been cooling. This is why the Arctic ice and Antarctic ice are melting. Yes, stop the presses, the globe is warming!
From recent events, I think both A and B are wrong. When an error is pointed out in research that shows AGW is happening, people use that error as an excuse not to believe any research that AGW is happening, even years after the error is corrected. When an error is pointed out in the IPCC report about a minor effect of climate change, people use that error to doubt all effects of climate change. Correcting the errors or pointing out they don't change the results will not silence the critics. It will only make the critics claim that their opinion is being suppressed even though the science has been indisputably proven to be flawed and therefore cannot be trusted!
Let's think through what would really happen if scientists released their code. The code has bugs, as all code does. People with an ulterior motive would point to the bugs and say "Look here! A bug! The science cannot be trusted!" And millions of sheeple would repeat "Yes! The code has bugs! And therefore I refuse to believe it!" It won't matter whether the bugs are relevant to the science; the fact that there are any bugs at all will cause people who want to disagree to say there's doubt about the results. Meanwhile, they will go about their business using computer systems that are riddled with bugs, but function well enough the vast majority of the time they're not even aware of the bugs.
You can also get the FrontMotion Firefox MSI. Mozilla is also working on their own MSI builds of Firefox.
Opera is massive in Eastern Europe. On Russian, Ukrainian, and Polish sites it makes sense to push Opera. But does it make sense to push Opera to English users, based on Opera usage on mobile devices, when most visitors to the site are using a desktop browser? A study done by a Mozilla employee shows users are more willing to switch to Chrome or Firefox. Few users of non-Opera browsers are willing to switch to Opera.
I don't know whether Opera does or doesn't fully pass the Acid3 test. I can tell you with certainty that I have never seen Opera Software ASA make that claim. If you could point me to where they do make such a claim, I would be truly grateful.
Safari was the first to fully pass Acid3.
Oh, you mean like automatic software updating and automatic crash reporting, for example?
For example, when Opera started supporting GreaseMonkey scripts as user javascript?
I remember that not long ago Opera didn't support client-side XSLT when IE and Firefox had for years.
They'll have to pry my Freedom Bread from my cold, dead body!
We're continuing to increase the concentration of carbon dioxide in the atmosphere. That's predicted to cause a few degrees of warming, which is in turn predicted to increase sea level by a meter or more. It is technologically impossible to stop any further rise in sea level, according to the vast majority of climatologists.
I've heard climatologists say that we may have reached a tipping point, meaning that no matter how much we reduce carbon dioxide emissions we could be locked into several degrees Celsius temperature rise and several meters of sea level rise. That will be costly, but it's not going to destroy humanity, or even human civilization.
Who is the Climate-Change-will-destroy-humanity crowd? I haven't heard of anyone saying that climate change will destroy humanity. Climate change could prove disastrous, but disasters happen all the time and don't wipe humanity off the face of the planet.
Look out! There are llamas!
Google, Google, who can I turn to? 8.67.53.09
You're right. I think we've all taken the wrong approach with huge, bloated standard libraries. Let all developers write all code from scratch. Need to output an integer, just write the code that turns the integer into a stream of characters, then pass that stream of characters into your homebrew I/O functions, which pass them off to your custom built drivers. There's no need for all languages to have this functionality! It just makes developers have to code around the differences and bugs in each runtime!
No one is calling for removing all greenhouse gasses from the atmosphere. That would be suicide. The current plan is to reduce greenhouse emissions dramatically, so that the amount of carbon dioxide in the atmosphere can stabilize.
As for your idea that reduced carbon dioxide emissions have led to cooling, that's nonsense. Carbon dioxide emissions have been rising every year except for the past year or two, and every year the concentration of carbon dioxide in the atmosphere has been increasing. The recent plateau in temperatures is more likely due to a natural cooling phase balancing out the warming due to increased greenhouse gasses.
Yes, there will be some positive effects due to global warming, such as more arable land in the high latitudes. On the other hand, there will also be negative consequences, such as drought and desertification in other areas, and rising sea levels which will cause millions of people to have to relocate. There is general agreement among world leaders that warming over 2 degrees Celsius would result in overall negative consequences, so they're trying to limit the concentration of carbon dioxide in the atmosphere to 450-550 ppm to limit the effects.
Where did I say that anyone predicts a catastrophe? Ironically, I think it is you who are misrepresenting what I said.
What computer modeling has added is that there are not subtle effects that were not in the simpler models that will significantly alter the warming. For example, it has been hypothesized that certain types of clouds will cancel out most of the warming, or that other types of clouds will cause even more warming that previously predicted. The newest models and data show that the previous predictions are pretty accurate.
As far as your assertion that these models are untested, you completely wrong. These models have predicted warming for over a century, and we've been seeing that warming since the 1970s. We have decades of evidence that confirm the models. The fact that people continue to say the models are "untested" is why we need to have more stories about the matter.
Exactly. That's why countries are trying to agree to cut emissions. It needs the cooperation of at least China and the U.S. That's why the climate bill that's before the U.S. Senate is such a big deal. Without a climate bill, the U.S. cannot commit to cutting emissions, and so other countries aren't willing to do so either.