I know anti-aliased text "looks" better than aliased -- more like a photo of the printed page, rather than a pixelated approximation. But I'm not so sure anti-aliasing really helps me read text more easily. I'm getting more used to it now, but in general, I think the extra "fuzziness" makes it harder for (my) eyes to make out the edges of letter shapes and quickly identify them. That is, I have generally found it a lot quicker and less-fatiguing to read well-hinted, high-contrast, sharp-edged, aliased text, rather than photo-like anti-aliased text.
I think this is similar to the difference between reading pure black text (e.g., from a typewriter or laser printer) vs. reading text printed through a half-tone screen (e.g., in coarsely-screened photos or illustrations).
Has anyone else had this experience? Does anyone have a more complete explanation for it?
I get frustrated, too, with the people who double-click everything, who are too timid to look in the control panel, who don't even know they have a directory browser on their computer (or what a directory/folder is). But I don't think any of your analogies apply all that well.
Operating a computer is far more complicated than driving -- there are lots more essential controls to use, many of them are hidden, and computer icons are far less understandable than road signs. Also, people don't have the experience of "riding around" with someone else operating a computer for 16-odd years before they have to use one, the way they would with a car.
Computers today might be a little like cars were at the beginning of the last century -- no one had grown up with them, there were no clear standards, no licenses, no drivers ed courses -- early adopters just used 'em and figured it out for themselves. But maybe, like cars computers will become more standardized and familiar to the general public, and this problem will clear up.
Jumping analogies, I think people relate to computers about the same way they would respond to being suddenly made chef at a restaurant, without ever having been in a kitchen before. They'd ask around for some pointers and "recipes" -- "OK, so I get the eggs from this big metal box (will they always be there? where do they come from?), I crack them into this metal pan-thing, I put it in just the right place on this other big metal box and turn this knob. Then the pan gets hot, and eventually they get cooked" -- then they'd follow the script religiously. If their restaurant served exactly the same dishes every night, this might be as far as they'd get. After a while, they might start to notice patterns in how things work, and a few people might start buying cookbooks and improvising. But most people, if they had to churn out the same dishes day in and day out, would just stick to the routine they know, and be glad when they get to go home.
I'm not sure where you're finding a copy of OS X 10.2 to upgrade with. But I had good results upgrading my OS 9 PowerBook (Lombard/bronze) to OS X 10.0 a few months ago. I assume your experience should be similar for later versions of OS X.
I just followed the instructions in the installation package. I think this first required updating my OS 9 to the latest version, so I'd have a compatible installation to use in the Classic environment. Then I booted from the OS X CD and installed OS X on the same HD. I know lots of people recommend setting up a separate partition, but that was more hassle than I was ready for. Installing OS X on the same partition it worked fine for me.
As a result, I retained my original OS 9 setup and added a new OS X setup. So I could boot directly into my original OS 9, or I could use it unaltered as my Classic environment in OS X. I think the installer also copied the original system settings (e.g., network preferences) from OS 9 to OS X, so it mirrored my original setup as well as possible. After this, both operating systems cohabited quite peacefully. I could continue using all my old OS 9 apps, and move on to OS X as well.
Rage Pro acceleration is an enormous improvement, at least on my iBook. For reference, try scrolling quickly through a big Slashdot page in IE or OmniWeb with 10.1.4 -- the scroll "thumb" lags about half a second behind the mouse cursor. Then try it with 10.1.5 -- the "thumb" stays right there with the mouse cursor (maybe 0.1 second behind). Window resizing is also much improved. This makes the computer feel a lot snappier.
Maybe not that much, but you're right, I forgot to account for the daily variability. That would probably add another factor of 5 or 6, pushing it up to 35-45 square meters. Still pretty manageable, but expensive at today's prices.
The sun only delivers 2400 watts / m^2 maximum -- do some calculations, you'd need a lot of cells to supply the average household, let alone business!
Hmm, let's see. 1000 W/m^2 * 15% efficiency / 1000 W per average household = 7 m^2 per household. Gee, my roof is only 160 square meters. I don't know where I'll ever find 7 square meters for solar panels.
Also, you don't recall correctly about the composition of solar cells. The most efficient lab cells a little while ago used gallium arsenide, and they may have moved on to something more obscure since then. But the vast majority of commercial solar cells use silicon, which doesn't present much of a disposal problem. Solvents used in manufacturing are also reused within the factory, so solar cell manufacture is not especially dirty.
Wind power: the best solution until it shows up in your backyard.
Wind power goes mostly on open land where there is lots of wind. The farmers and ranchers who own this land love it, because they keep on farming and ranching below the turbines, and they get extra income from their land. There is no reason to put turbines in cities (except maybe this).
> Anyway, if you want to know anything specifically, post here...
Mostly, I want to know if the user interface feels more responsive. On my late-2001 iBook with 10.1, whenever you drag a window corner to resize it, the screen redraw lags behind the mouse by a second or so, so if you move quickly, you end up with the window corner running an inch or two behind the mouse, and eventually catching up when you stop moving the mouse. (Interestingly, if you drag quickly and stop suddenly, the corner never catches up with where the mouse is -- the corner sits where the mouse was a couple of steps ago, until you give it another mouse event by nudging the mouse or releasing the button; then it redraws in the right position.) This lag happens most in window resizing, especially in Internet Explorer or OmniWeb with complex windows. But it also happens to a lesser degree when resizing Finder windows or scrolling any window.
When the system doesn't respond immediately to mouse movements, it makes it feel sluggish, and it kind of grates on me. This lag even happens on a newish G4 PowerMac I use at school, although the response on that machine is good enough that you hardly notice it in normal use.
I have a late-2001 iBook too. How does Jaguar work on it? Are scrolling and window-resizing faster, even using "plain 'old' Quartz"? That's what I want to see.
All I want to know is, will 10.2 finally give me instantaneous scrolling and window resizing on my iBook in Finder, OmniWeb, Explorer, etc.? It is so frustrating to have the window lag a second or two behind my mouse during scroll and resize operations -- it feels like I'm working through molasses, and makes the system "feel" kludgy. This is my biggest gripe about OS X.
I get immediate window-drawing on a 300 MHz Pentium II running Windows 98, so why don't I on my 600 MHz iBook? (especially since my iBook smokes the Windows machine on other graphics-intensive tasks, like QuickTime playback)
I know the Quartz window resize is more sophisticated and complex than the Windows window resize, but it still seems like Apple should be able to make this work. Does anyone know if it will on 10.2?
Re:doesnt seem economical
on
Lunar Power
·
· Score: 1
Solar cells on earth? We have clouds. We have day and night. The moon (thanks to an astronomical quirk) has permanent day and night.
Which part of the moon has permanent day and night? Certainly one side of the moon always faces the Earth, but that doesn't make it always light. Nor does the moon really have a "dark" side. As far as I know, any spot on the moon will get light half the time, just like the Earth. (That's why it's full sometimes and not others.)
Sure, we don't get as much solar energy per square meter at the Earth's surface as you would get on the moon, but we have plenty of square meters. Last time I checked, we would need about 0.02% of the Earth's surface to meet our energy needs with solar panels. So you can make up for reduced insolation simply by adding more panels.
The only issue is cost, and (just like with computers) the cost of solar panels drops as you make more of them. Once we get solar power above a one or two percent of U.S. energy demand, it will be cheaper to buy new panels than to dig up coal or oil and burn it. At that point, solar power becomes self-financing.
Intermittency won't be a real problem until we're supplying about 15-30% of our power from solar, and by then we may have worked out some of the storage issues. I'm pretty sure this will be cheaper than trying to send massive gizmos into space (or build them there).
why this will never happen, and why that's OK
on
Cringely: OS X on Intel
·
· Score: 2, Insightful
First, Cringely's article is 90% about what this port would do for consumers and Microsoft. There's no compelling case here for why Apple would want to do it, which means there's no story about why it would happen.
Steve Jobs already tried this game with NextStep and it didn't work out. I bet he did that under duress, and since it didn't work out, there's no way he'll be convinced to try it again.
Cringely also writes, "So Apple has to make at least a "good faith" effort with this OS X port, reflecting the realities of Intel hardware." This points to a fundamental reason this will never happen. With Apple hardware, Apple can sell a product that "just works." With Intel hardware, Apple is stuck with a massive and unprofitable effort to develop and test drivers for all the cheap Intel-compatible devices out there, or they're stuck with a bunch of customers screaming about how their machines don't work with OS X. Either way, Apple loses.
Finally, I'm amazed by this whole business about Apple hardware being "too expensive." Look, obviously some people are buying it at this price, so it must not be "too expensive" for them -- i.e., it offers something they're willing to pay for. It's just "too expensive" for you, and that's why you're griping. And I think that's outlandish for coders, because we can pretty much universally afford to pay the $400 extra to get a really good box. Some of you just seem to feel entitled to perfect hardware at bottom-of-market prices, which I don't understand. I've worked in carpentry, and I know that you have to pay extra for good tools, but they make the work experience so much better. If you're making your living off the hardware, a few extra dollars is nothing. Think about all the other things you throw money away on, yet you balk at investing in decent tools for your work?
The questions they're answering are more philosophical than practical. A lot of people would like to know where we and our world came from and where we're going. This is a stab at answering part of that.
Knowing that our planet will just barely escape the sun might also reinforce the sense that it's special somehow.
How to cut the cost of new technology
on
Super Hard Steel
·
· Score: 1
The technology doesn't actually have to be super-competitive at first in order to have long-run potential. In fact, if it uses a lot of off-the-shelf processes, that may actually limit its potential for reducing costs later. Here's why:
a) Most fabulous, new technologies are more expensive than the competing technologies when they are first introduced. But after a few generations of production, costs come down into the competitive range. Formally speaking, this is because most new technologies are governed by a "learning curve" or "experience curve." Every time you double the amount of the new thing that you've produced, you cut costs by a fixed percentage (e.g., 20%). New technologies can overtake old technologies in price, because output of a brand-new technology can grow exponentially, producing linear drops in price, while the established industries are stuck at more-or-less linear growth levels, yielding only logarithmic(?) drops in price.
b) The learning curve effect is most pronounced when you don't use off-the-shelf technology in your innovation. You can get the cost of the first generation down quite a bit by using a lot of off-the-shelf components, but then there's nowhere to go from there -- the prices are already set for all your components and they're not falling. On the other hand, if your new technology uses a lot of customized processes or materials, you can get rapid drops in price as you boost output -- every successive generation is a chance to improve your customized approaches.
Computers are a good example of the learning curve effect (and I bet you'll find that Moore's law only holds true as long as you have exponential growth in output). Wind and solar energy technology also show huge learning effects, while coal and gas plants don't, so there's hope for us yet.
The one caveat here is that output of a new technology may grow quite slowly at first if it isn't competitive for many applications when it is first introduced (although every new technology is competive for some applications). So progress down the learning curve may be slow early on, if you start with very high costs. But once your prices come down a little and you start competing in more applications, your costs can start falling really fast if there are learning curves at work. On the other hand, if your technology is based entirely on off-the-shelf components, you'll be stuck where you started.
This suggestion may be either dorky or rash:
Could someone engineer an "anti-virus" to patch all the unpatched IIS servers in the world? It could spread itself like the Code Red virus, but then it could unload the Code Red virus and/or install a patch on the affected server to close the hole. This should be possible in theory, because the security hole allows full access to the computer.
(1) Natural gas produces less CO2 than coal, because more of the energy comes from the hydrogen component of the fuel, and less from the carbon component (Natural gas is CH4, and coal is roughly C135H96O9NS). Natural gas isn't great, but it produces half as much CO2 as coal per unit of energy.
(2) Yes, China has bad smog, and yes that is a big part of the reason they want to switch away from coal, and yes, smog doesn't have much to do with global warming. But China is still doing more than the U.S., which is doing NOTHING.
(3) Exponential growth of any sort is dangerous for the environment. But you're letting yourself off too easy by pointing to China's population growth. Their population may be growing by 3% a year (or whatever), but we are doing at least as much harm if our per capita consumption is growing by 3% per year.
You've launched a rather incomplete assault on renewable energy sources here. Yes, large-scale hydro is no longer an option -- it causes too much harm to existing ecosystems, and there aren't many places left to put big dams anyway. And yes, wave- or tidal-power collectors or ocean thermal taps cannot produce large amounts of power without huge pieces of machinery. But you've skipped right over the renewable technologies that are actually being installed today -- wind and solar systems.
Wind: "Good wind areas, which cover 6% of the contiguous U.S. land area, have the potential to supply more than one and a half times the current electricity consumption of the United States." And the wind turbines being built today can capture this energy at about the same cost as operating a natural gas turbine or a coal plant, and less than half of the cost of nuclear power (wind has no fuel cost, so all of the cost of power comes from paying off the original investment in the equipment).
Solar: Enough sunlight strikes the U.S. to meet our electricity needs 700 times over. Solar cells are currently quite expensive to manufacture, so their power costs more than coal or gas. Unfortunately solar cells are now caught in a catch-22. They could become the cheapest power source in the world, but not until they are manufactured in really high volume, and that's not happening because they're so expensive. (We all know what happens when you start manufacturing silicon chips in high volume...) Although they're not there yet, prices are falling, and solar cells will become the preferred energy source eventually.
The great thing about photovoltaic cells is that they let us get our electricity from a huge fusion reactor that is conveniently located 100,000,000 miles from any population center.
P.S.: Solar cells have an energy payback time of 1-4 years, so they produce 8-30 times more electricity in their lifetime than they require to manufacture.
I think a few clarifications need to be made here and for some of the messages further down:
(1) CFCs have almost nothing to do with global warming. They are the main cause of ozone depletion, which is a different problem. Ozone depletion allows more ultraviolet light to reach the Earth's surface. The most commonly cited threat from ozone depletion is a rise in skin cancer.
Global warming (a different problem) is caused by a buildup of greenhouse gases in the atmosphere, which trap solar heat reflected from the earth, and send it back down again, raising the planet's average temperature. The most commonly cited threat from global warming is a rise in sea level. But there are other scary problems. The one that I worry about is the fact that global warming could shift lots of climate bands hundreds of miles toward the poles. Existing ecosystems are built around the current temperature regimes -- forests and other ecosystems simply can't move as fast as the climate bands will shift, and they could be gradually weakened or killed off.
(2) This is not "bogus science." The Intergovernmental Panel on Climate Change is the single most authoritative source on this topic. They were established with the blessing of most world governments and with the mandate to come up with a genuine scientific consensus on the issue of climate change. They started in the early '90s with meek statements about the possibility of human-induced climate change, and have gradually become more decisive as more evidence has accumulated. The most important thing about this report is not the exact numbers that they are estimating, but the fact that most of the world's scientists who know anything about the topic (including a number of former skeptics) now believe that the evidence shows beyond a reasonable doubt that humans are causing global warming which is clearly distinct from the Earth's natural variation. You can always find some scientists (and more often pseudoscientists) who will disagree, but they are now a slim minority.
Slashdot Mirror
I know anti-aliased text "looks" better than aliased -- more like a photo of the printed page, rather than a pixelated approximation. But I'm not so sure anti-aliasing really helps me read text more easily. I'm getting more used to it now, but in general, I think the extra "fuzziness" makes it harder for (my) eyes to make out the edges of letter shapes and quickly identify them. That is, I have generally found it a lot quicker and less-fatiguing to read well-hinted, high-contrast, sharp-edged, aliased text, rather than photo-like anti-aliased text.
I think this is similar to the difference between reading pure black text (e.g., from a typewriter or laser printer) vs. reading text printed through a half-tone screen (e.g., in coarsely-screened photos or illustrations).
Has anyone else had this experience? Does anyone have a more complete explanation for it?
I get frustrated, too, with the people who double-click everything, who are too timid to look in the control panel, who don't even know they have a directory browser on their computer (or what a directory/folder is). But I don't think any of your analogies apply all that well.
Operating a computer is far more complicated than driving -- there are lots more essential controls to use, many of them are hidden, and computer icons are far less understandable than road signs. Also, people don't have the experience of "riding around" with someone else operating a computer for 16-odd years before they have to use one, the way they would with a car.
Computers today might be a little like cars were at the beginning of the last century -- no one had grown up with them, there were no clear standards, no licenses, no drivers ed courses -- early adopters just used 'em and figured it out for themselves. But maybe, like cars computers will become more standardized and familiar to the general public, and this problem will clear up.
Jumping analogies, I think people relate to computers about the same way they would respond to being suddenly made chef at a restaurant, without ever having been in a kitchen before. They'd ask around for some pointers and "recipes" -- "OK, so I get the eggs from this big metal box (will they always be there? where do they come from?), I crack them into this metal pan-thing, I put it in just the right place on this other big metal box and turn this knob. Then the pan gets hot, and eventually they get cooked" -- then they'd follow the script religiously. If their restaurant served exactly the same dishes every night, this might be as far as they'd get. After a while, they might start to notice patterns in how things work, and a few people might start buying cookbooks and improvising. But most people, if they had to churn out the same dishes day in and day out, would just stick to the routine they know, and be glad when they get to go home.
I'm not sure where you're finding a copy of OS X 10.2 to upgrade with. But I had good results upgrading my OS 9 PowerBook (Lombard/bronze) to OS X 10.0 a few months ago. I assume your experience should be similar for later versions of OS X.
I just followed the instructions in the installation package. I think this first required updating my OS 9 to the latest version, so I'd have a compatible installation to use in the Classic environment. Then I booted from the OS X CD and installed OS X on the same HD. I know lots of people recommend setting up a separate partition, but that was more hassle than I was ready for. Installing OS X on the same partition it worked fine for me.
As a result, I retained my original OS 9 setup and added a new OS X setup. So I could boot directly into my original OS 9, or I could use it unaltered as my Classic environment in OS X. I think the installer also copied the original system settings (e.g., network preferences) from OS 9 to OS X, so it mirrored my original setup as well as possible. After this, both operating systems cohabited quite peacefully. I could continue using all my old OS 9 apps, and move on to OS X as well.
Rage Pro acceleration is an enormous improvement, at least on my iBook. For reference, try scrolling quickly through a big Slashdot page in IE or OmniWeb with 10.1.4 -- the scroll "thumb" lags about half a second behind the mouse cursor. Then try it with 10.1.5 -- the "thumb" stays right there with the mouse cursor (maybe 0.1 second behind). Window resizing is also much improved. This makes the computer feel a lot snappier.
Maybe not that much, but you're right, I forgot to account for the daily variability. That would probably add another factor of 5 or 6, pushing it up to 35-45 square meters. Still pretty manageable, but expensive at today's prices.
The sun only delivers 2400 watts / m^2 maximum -- do some calculations, you'd need a lot of cells to supply the average household, let alone business!
Hmm, let's see. 1000 W/m^2 * 15% efficiency / 1000 W per average household = 7 m^2 per household. Gee, my roof is only 160 square meters. I don't know where I'll ever find 7 square meters for solar panels.
Also, you don't recall correctly about the composition of solar cells. The most efficient lab cells a little while ago used gallium arsenide, and they may have moved on to something more obscure since then. But the vast majority of commercial solar cells use silicon, which doesn't present much of a disposal problem. Solvents used in manufacturing are also reused within the factory, so solar cell manufacture is not especially dirty.
Wind power: the best solution until it shows up in your backyard.
Wind power goes mostly on open land where there is lots of wind. The farmers and ranchers who own this land love it, because they keep on farming and ranching below the turbines, and they get extra income from their land. There is no reason to put turbines in cities (except maybe this).
> Anyway, if you want to know anything specifically, post here...
Mostly, I want to know if the user interface feels more responsive. On my late-2001 iBook with 10.1, whenever you drag a window corner to resize it, the screen redraw lags behind the mouse by a second or so, so if you move quickly, you end up with the window corner running an inch or two behind the mouse, and eventually catching up when you stop moving the mouse. (Interestingly, if you drag quickly and stop suddenly, the corner never catches up with where the mouse is -- the corner sits where the mouse was a couple of steps ago, until you give it another mouse event by nudging the mouse or releasing the button; then it redraws in the right position.) This lag happens most in window resizing, especially in Internet Explorer or OmniWeb with complex windows. But it also happens to a lesser degree when resizing Finder windows or scrolling any window.
When the system doesn't respond immediately to mouse movements, it makes it feel sluggish, and it kind of grates on me. This lag even happens on a newish G4 PowerMac I use at school, although the response on that machine is good enough that you hardly notice it in normal use.
Is this any better in 10.2?
I have a late-2001 iBook too. How does Jaguar work on it? Are scrolling and window-resizing faster, even using "plain 'old' Quartz"? That's what I want to see.
All I want to know is, will 10.2 finally give me instantaneous scrolling and window resizing on my iBook in Finder, OmniWeb, Explorer, etc.? It is so frustrating to have the window lag a second or two behind my mouse during scroll and resize operations -- it feels like I'm working through molasses, and makes the system "feel" kludgy. This is my biggest gripe about OS X.
I get immediate window-drawing on a 300 MHz Pentium II running Windows 98, so why don't I on my 600 MHz iBook? (especially since my iBook smokes the Windows machine on other graphics-intensive tasks, like QuickTime playback)
I know the Quartz window resize is more sophisticated and complex than the Windows window resize, but it still seems like Apple should be able to make this work. Does anyone know if it will on 10.2?
Solar cells on earth? We have clouds. We have day and night. The moon (thanks to an astronomical quirk) has permanent day and night.
Which part of the moon has permanent day and night? Certainly one side of the moon always faces the Earth, but that doesn't make it always light. Nor does the moon really have a "dark" side. As far as I know, any spot on the moon will get light half the time, just like the Earth. (That's why it's full sometimes and not others.)
Sure, we don't get as much solar energy per square meter at the Earth's surface as you would get on the moon, but we have plenty of square meters. Last time I checked, we would need about 0.02% of the Earth's surface to meet our energy needs with solar panels. So you can make up for reduced insolation simply by adding more panels.
The only issue is cost, and (just like with computers) the cost of solar panels drops as you make more of them. Once we get solar power above a one or two percent of U.S. energy demand, it will be cheaper to buy new panels than to dig up coal or oil and burn it. At that point, solar power becomes self-financing.
Intermittency won't be a real problem until we're supplying about 15-30% of our power from solar, and by then we may have worked out some of the storage issues. I'm pretty sure this will be cheaper than trying to send massive gizmos into space (or build them there).
First, Cringely's article is 90% about what this port would do for consumers and Microsoft. There's no compelling case here for why Apple would want to do it, which means there's no story about why it would happen.
Steve Jobs already tried this game with NextStep and it didn't work out. I bet he did that under duress, and since it didn't work out, there's no way he'll be convinced to try it again.
Cringely also writes, "So Apple has to make at least a "good faith" effort with this OS X port, reflecting the realities of Intel hardware." This points to a fundamental reason this will never happen. With Apple hardware, Apple can sell a product that "just works." With Intel hardware, Apple is stuck with a massive and unprofitable effort to develop and test drivers for all the cheap Intel-compatible devices out there, or they're stuck with a bunch of customers screaming about how their machines don't work with OS X. Either way, Apple loses.
Finally, I'm amazed by this whole business about Apple hardware being "too expensive." Look, obviously some people are buying it at this price, so it must not be "too expensive" for them -- i.e., it offers something they're willing to pay for. It's just "too expensive" for you, and that's why you're griping. And I think that's outlandish for coders, because we can pretty much universally afford to pay the $400 extra to get a really good box. Some of you just seem to feel entitled to perfect hardware at bottom-of-market prices, which I don't understand. I've worked in carpentry, and I know that you have to pay extra for good tools, but they make the work experience so much better. If you're making your living off the hardware, a few extra dollars is nothing. Think about all the other things you throw money away on, yet you balk at investing in decent tools for your work?
This matters why?
The questions they're answering are more philosophical than practical. A lot of people would like to know where we and our world came from and where we're going. This is a stab at answering part of that.
Knowing that our planet will just barely escape the sun might also reinforce the sense that it's special somehow.
The technology doesn't actually have to be super-competitive at first in order to have long-run potential. In fact, if it uses a lot of off-the-shelf processes, that may actually limit its potential for reducing costs later. Here's why:
a) Most fabulous, new technologies are more expensive than the competing technologies when they are first introduced. But after a few generations of production, costs come down into the competitive range. Formally speaking, this is because most new technologies are governed by a "learning curve" or "experience curve." Every time you double the amount of the new thing that you've produced, you cut costs by a fixed percentage (e.g., 20%). New technologies can overtake old technologies in price, because output of a brand-new technology can grow exponentially, producing linear drops in price, while the established industries are stuck at more-or-less linear growth levels, yielding only logarithmic(?) drops in price.
b) The learning curve effect is most pronounced when you don't use off-the-shelf technology in your innovation. You can get the cost of the first generation down quite a bit by using a lot of off-the-shelf components, but then there's nowhere to go from there -- the prices are already set for all your components and they're not falling. On the other hand, if your new technology uses a lot of customized processes or materials, you can get rapid drops in price as you boost output -- every successive generation is a chance to improve your customized approaches.
Computers are a good example of the learning curve effect (and I bet you'll find that Moore's law only holds true as long as you have exponential growth in output). Wind and solar energy technology also show huge learning effects, while coal and gas plants don't, so there's hope for us yet.
The one caveat here is that output of a new technology may grow quite slowly at first if it isn't competitive for many applications when it is first introduced (although every new technology is competive for some applications). So progress down the learning curve may be slow early on, if you start with very high costs. But once your prices come down a little and you start competing in more applications, your costs can start falling really fast if there are learning curves at work. On the other hand, if your technology is based entirely on off-the-shelf components, you'll be stuck where you started.
This suggestion may be either dorky or rash: Could someone engineer an "anti-virus" to patch all the unpatched IIS servers in the world? It could spread itself like the Code Red virus, but then it could unload the Code Red virus and/or install a patch on the affected server to close the hole. This should be possible in theory, because the security hole allows full access to the computer.
(1) Natural gas produces less CO2 than coal, because more of the energy comes from the hydrogen component of the fuel, and less from the carbon component (Natural gas is CH4, and coal is roughly C135H96O9NS). Natural gas isn't great, but it produces half as much CO2 as coal per unit of energy.
(2) Yes, China has bad smog, and yes that is a big part of the reason they want to switch away from coal, and yes, smog doesn't have much to do with global warming. But China is still doing more than the U.S., which is doing NOTHING.
(3) Exponential growth of any sort is dangerous for the environment. But you're letting yourself off too easy by pointing to China's population growth. Their population may be growing by 3% a year (or whatever), but we are doing at least as much harm if our per capita consumption is growing by 3% per year.
You've launched a rather incomplete assault on renewable energy sources here. Yes, large-scale hydro is no longer an option -- it causes too much harm to existing ecosystems, and there aren't many places left to put big dams anyway. And yes, wave- or tidal-power collectors or ocean thermal taps cannot produce large amounts of power without huge pieces of machinery. But you've skipped right over the renewable technologies that are actually being installed today -- wind and solar systems.
Wind: "Good wind areas, which cover 6% of the contiguous U.S. land area, have the potential to supply more than one and a half times the current electricity consumption of the United States." And the wind turbines being built today can capture this energy at about the same cost as operating a natural gas turbine or a coal plant, and less than half of the cost of nuclear power (wind has no fuel cost, so all of the cost of power comes from paying off the original investment in the equipment).
Solar: Enough sunlight strikes the U.S. to meet our electricity needs 700 times over. Solar cells are currently quite expensive to manufacture, so their power costs more than coal or gas. Unfortunately solar cells are now caught in a catch-22. They could become the cheapest power source in the world, but not until they are manufactured in really high volume, and that's not happening because they're so expensive. (We all know what happens when you start manufacturing silicon chips in high volume...) Although they're not there yet, prices are falling, and solar cells will become the preferred energy source eventually.
The great thing about photovoltaic cells is that they let us get our electricity from a huge fusion reactor that is conveniently located 100,000,000 miles from any population center.
P.S.: Solar cells have an energy payback time of 1-4 years, so they produce 8-30 times more electricity in their lifetime than they require to manufacture.
I think a few clarifications need to be made here and for some of the messages further down:
(1) CFCs have almost nothing to do with global warming. They are the main cause of ozone depletion, which is a different problem. Ozone depletion allows more ultraviolet light to reach the Earth's surface. The most commonly cited threat from ozone depletion is a rise in skin cancer.
Global warming (a different problem) is caused by a buildup of greenhouse gases in the atmosphere, which trap solar heat reflected from the earth, and send it back down again, raising the planet's average temperature. The most commonly cited threat from global warming is a rise in sea level. But there are other scary problems. The one that I worry about is the fact that global warming could shift lots of climate bands hundreds of miles toward the poles. Existing ecosystems are built around the current temperature regimes -- forests and other ecosystems simply can't move as fast as the climate bands will shift, and they could be gradually weakened or killed off.
(2) This is not "bogus science." The Intergovernmental Panel on Climate Change is the single most authoritative source on this topic. They were established with the blessing of most world governments and with the mandate to come up with a genuine scientific consensus on the issue of climate change. They started in the early '90s with meek statements about the possibility of human-induced climate change, and have gradually become more decisive as more evidence has accumulated. The most important thing about this report is not the exact numbers that they are estimating, but the fact that most of the world's scientists who know anything about the topic (including a number of former skeptics) now believe that the evidence shows beyond a reasonable doubt that humans are causing global warming which is clearly distinct from the Earth's natural variation. You can always find some scientists (and more often pseudoscientists) who will disagree, but they are now a slim minority.