There are a number of good realtime black lists (RBLs) that do just that. I use them at my mail server and it blocks a huge percentage of spam. Furthermore, I have my mail server (postfix) set up to tarpit those senders. In other words, it holds onto the connection for 20-30 seconds before sending any sort of reply, effectively slowing down the spammer a bit and consuming their resources.
My home spam filter does not seem to be affected much. I run dspam which has a feature in that over time it will forget words if they are not used in spam. Since the text is usually different or random, it does not have any significant effect on generating false positives. In the years I have been running dspam with tens of thousands of emails, I have only gotten 3-4 false positives.
By having a baysian filter forget over time, it also helps shrink down the database and helps it adapt as the contents of spam change over time.
Of course I also use other spam blocking techniques, like using realtime black lists (RBLs) and blocking a number of Chinese subnets... I should add tpnet.pl and Verizon as well.
This will help. It will be interesting to see how it compares to gasoline in terms of pollutants. Particulates are still a big problem though. Even with a 30% reduction with biodiesel, it is still far more than gasoline or ethanol, both of which are negligible. The other drawback right now at least around where I live is that diesel is significantly more expensive than gasoline, in part due to the low sulfur requirement.
I might also add that I am not favoring the oil companies, far from it. I just do not see any silver bullet that can magically solve the problems. It's a trade-off. Trade gasoline for biodiesel and you get more air pollution. Trade gasoline for ethanol and you get more water pollution, that is if we can even meet the demand for ethanol, which is not possible with corn, especially given how much water corn requires and the fact that the water supply in the big corn growing regions is a limited resource that is being depleted.
So far the closest thing I have seen that could make a big dent without significantly increasing pollution is reverse polymerization, the output of which is oil which can be further refined. This process can turn just about any form of carbon-based waste into a decent grade of oil and is something like 80% efficient. Of course, increasing fuel efficiency in our current vehicles also would help.
Now if they can solve the particulate problem and bring NOx to levels as low as a clean burning gasoline engine then it can be an alternative. It also remains to be seen if enough can be grown to replace diesel. Diesel has a lot of advantages over gasoline, though currently pollution is not one of them.
California tried to get an exemption from requiring ethanol, or for that matter any oxygenate from its fuel when it proved that using certain blends of gasoline reduced pollution further. The oxygenates only help for really old cars, of which there are not many left on the roads. For newer cars, it increases pollution and lowers the fuel economy. Not to mention that ethanol is a huge subsidy to ADM.
Biodiesel also produces MORE NOx than regular diesel (look it up). It may reduce particulate emissions, but still puts out far more than gasoline. Now it might make sense to run existing diesel vehicles on biodiesel, but it still does not make sense to convert gasoline vehicles to biodiesel until these problems are solved.
The NOx problem with any form of diesel is caused by the high compression ratio required.
Now it may cut down on other pollutants, but NOx is still a significant one that causes that orange haze over polluted cities.
Many people believe that Ethanol and Biodiesel are the answer to the shrinking oil supplies and global warming. There are some major obstacles to overcome, though.
In the case of biodiesel, it actually produces more NOx than diesel. While other pollutants are often reduced, this is the major one that forms that orange cloud over heavily polluted cities. NOx is a major pollutant of diesel engines due to the high compression ratios and still has not been effectively solved. While particulate matter is less than diesel, it is still significant, far more than gasoline. While technologies exist to reduce NOx, they are sensitive to sulfur, and while sulfur is virtually eliminated from biodiesel, it would mean that an engine designed to run on it could not use regular diesel without destroying the NOx smog equipment.
For ethanol there are other problems. First of all is the amount of energy required to harvest the corn that is currently used in this country. The best efficiencies are around 20%. In other words, for each gallon of ethanol produced, 0.8 gallons are used for growing and harvesting the corn used. Additionally, a lot of fertilizer and fresh water is required for growing the corn. I am assuming no pesticides, though pesticides like glyphosphate (RoundUp) are also often used as well, even with GM corn. The fertilizer often ends up in the waterways, causing significant pollution, like the dead zone in the Gulf of Mexico. Other sources of ethanol are better, such as sugar beets or sugar cane and possibly other sources. New techniques are being developed to produce ethanol from cellulose, meaning that agricultural waste and other crops like switchgrass or even hemp could be used to produce ethanol.
One other potential energy source is to use reverse polymerization. This process can convert almost any form of organic waste into oil and is fairly efficient and not very complex. This could easily supplement much of our current demand for oil. It can also use agricultural waste like ethanol.
A PC platform will not scale like a decent router will. The memory latency becomes the bottleneck once you start thrashing your CPU cache when you have a lot of routes or ACLs. For small setups, it might work, but it will not compete with dedicated hardware solutions once the complexity grows.
I have implemented routers, and the biggest bottleneck is typically memory latency. Once the routing tables grow beyond what will fit in the cache, the latency kills you. Dedicated hardware routers are designed with this in mind, often with multiple banks of low-latency memory or CAMS.
Try to handle 1M packets per second with 100K routes, MPLS VPNs, ACLs, policy based routing, QoS, policing, marking, reverse path forwarding checks, etc. and it will fall apart. Plus it has to keep statistics on everything. All of that will quickly exceed the memory bandwidth available on any PC platform.
A decent router can do all of this and run at gigabit speeds or more. Add to that all the various interfaces that are available and the much higher port densities.
My experience is that around 60-75% of the spam I receive comes from China. On my home mail server I finally broke down and started blocking the worst offending subnets and the amount of spam I received dropped dramatically. There is a RBL for China, cn.blackhole.us, or a combination of China and Korea (cn-kr.blackhole.us), though these are no longer listed and will likely disappear soon.
I also use several other RBLs which have helped a lot.
I also decided to add the worst offending subnets in China as rules for my firewall to block. The worst offending subnet is 221.208.208.x where my firewall reports an almost constant barrage of IM spam, and from what I've read, this subnet has been a problem for years.
For your own blocking, the following script will get all the subnets used by China (or any other country you're interested in, just change $ctry):
At work, where I cannot do this, most of my spam is also received from China.
Out of the rest of the spam I receive, the US is actually pretty far down on the list of sources, though still much higher than places like the UK, Germany or France. The rest seems to come from places like Poland, Romania and Estonia.
I can easily exceed the rate of whatever memory I put in my Nikon D70s when I shoot pictures back to back. Then again, I often shoot in raw mode to allow me to perform better post processing of the images. Each image is 5-6MB in size, so at 3 pictures per second the flash will not keep up. The faster flash definitely makes a difference if I am shooting a lot of pictures, since the raw buffer is only 4 pictures.
Granted, the higher end cameras have larger buffers which helps mitigate the problem, but faster flash is definitely an advantage in these cases. Now with JPEG, it would be a different story, especially with normal compression.
As far as I know there still is no effective solution to the amount of NOx smog produced by a diesel engine. A lot of this is due to the high compression ratios used, not necessarily the type of fuel.
If this article is accurate, biodiesel can actually produce more NOx and ozone pollution. It may significantly cut down on the soot, which is considered highly toxic. As I type this, there is another spare-the-air day where I live and all public transit is free. Most of this is due to NOx and ozone.
Diesel-electric is perfect for trains, since they need a huge amount of torque at zero RPM and electric motors solve this nicely, basically eliminating any requirements for a complex transmission.
The online update support for SuSE 10.1 is horrible compared to past versions. It is *extremely* slow. Adding a new repository took well over 30 minutes to process it, the CPU remaining busy the entire time. Granted, the machine is only a 1.5GHz P4, but it should take nowhere near this amount of time.
Bringing up the software install tool takes 150MB of RAM. This is excessive.
Then OpenSuse keeps moving repositories around, or deleting them. They removed the KDE 3.5.3 repository recently, for example.
I'm almost ready to switch to another distribution, maybe Kubuntu or some other up-to-date KDE based distro.
Well, some of us can only see a few dozen stars when we go outside due to all the heavy light pollution. I wish they would just shut off most of the street lights at times. I think the only way the local observatories are able to work is that they have sodium filters to filter out most of the street lights.
Not all of us live in areas with low population densities.
I can *never* see the Milky Way when looking up at night unless I travel quite a ways.
Now just allow the user to substitute Austin Powers and things will be set. Maybe even offer it as a bonus after the game has been completed once or twice.
Hmmm, my Prius uses NiMH batteries, as do all the other hybrids I know of. NiCd don't last nearly as long and have a lower energy density. NiMH do not contain cadmium and are capable of also delivering high amounts of current. In fact, all of the rechargable AA batteries I see in the stores for cameras and whatnot are now NiMH instead of NiCD. I think NiMH might be a bit trickier to charge, but they suffer much less from the memory problem NiCDs have.
I doubt that neither ATI nor nVidia could release open source drivers even if they wanted to. As others have said, the drivers probably contain a lot of 3rd party proprietary code that cannot be released.
I'll tell you right now that it will kill Linux in the embedded world if manufacturers were forced to release all their modules. A lot of hardware is proprietary and remains a black box on purpose, to keep competitors out or to hide trade secrets, and yes, a lot of hardware does that.
On the VxWorks product I work on we had one vendor that tried to send us binary only object files (which were not compatible with out linker) then moved to obfuscated code. They finally moved to release non-obfuscated code, but it is still highly proprietary and cost us a small fortune, many tens of thousands of dollars. They also release Linux drivers since so much of their market has moved to Linux. At least with this, I can say that they don't muck about at all in the kernel, but just use the standard driver interfaces.
Why charge so much and make the software proprietary? Because the hardware is very high performance and not very high volume. They'd have a tough time recovering their engineering expenses if they gave away the software when a lot of their customers may buy only a few hundred chips a year.
I think nVidia has worked with kernel developers and people at Xorg and it is known what interfaces they use. I think all of the code that deals directly with the kernel is released, since it is basically a wrapper around their proprietary code.
There was actually a good article on this in the December Scientific American. There would be no shortage of fuel if we were to build breeder reactors which can convert U238, where we have a large supply, to plutonium and other fissionable fuels. The article goes on to say that the amount of dangerous radioactive waste from this type of reactor is a small fraction of other types of reactors (including pebble bed). The article can be found at Scientific American: Smarter Use of Nuclear Waste [ ENERGY ]
Unfortunately, it requires registration to read.
Actually, the acceleration of the current generation of Prius is much better than that... 0 to 60 in 10.3 seconds. Granted, it's not 7-8 seconds like some of the more powerful V-6 cars, but it's decent enough for most driving conditions. The luggage capacity of the Prius is significantly more than both the Accord and Civic, and the passenger area is nearly the same, the Accord having more shoulder and hip room and slightly more front leg room, but the Prius having more head room and rear leg room. The Prius gets roughly double the milage in city driving, however and significantly more on the highway, even ignoring the unrealistic EPA estimates (which are often wrong for most vehicles).
I had this exact same problem... I finally tracked it down to one of my computers and put a filter on the power line that fixed the problem... unfortunately the rest of X10 seems to suck as well. Some devices refuse to be controlled by the RF controller (which itself is rather flakey), but the same devices will work if controlled by another hard-wired controller. I had my bedroom light randomly turn on until I added the filter. The system just seems to degrade over time.
Not to mention that I went through the X10 outlets usually in 6 months because some cheap plastic part in them would break.
In the mean time we can thank the Europeans for reliability of electronics going to hell. It is especially problematic for the more expensive and complex equipment which tends to have a lot of BGA chips with very fine ball spacings. The tin whiskers will cause leads to short out. A good document I found is at http://nepp.nasa.gov/whisker/background/index.htm
To date it looks like you can have either lead free solder or reliability, but rarely both.
That sounds like my experience with the Hertz "N^Heverlost" system. At the time I was in Austin, Tx and we needed to go back to the airport, so we went through the menus and chose "airport" as the destination. It directed us to the old airport which had been closed for several years. We finally figured out we had to tell the system we wanted to go back home and it directed us correctly.
Online is one thing, how about encouraging your students to print it out before class and take the notes with them to the lecture? Or better yet, sell them in the book store?
I had a physics class like that and it was a completely different environment. It was the first time I could actually sit back and listen to the lecture, thinking about it, since the notes I bought at the book store were identical to what he showed in the lecture. I didn't waste all my time scribbling down what was written on his overhead slides, since I already had it. It was one of the best classes I ever had. A laptop would not be of much use for taking notes, since there were a lot of diagrams and equations. Only a pencil was needed to write additional notes from time to time (rarely needed) in the margins. Even better would be to make both printed and electronic available, but require that the students have at least the printed version and bring it to class.
Now this is probably a lot harder for the professor, since they need to know at least several weeks ahead of time exactly what they'll cover in a lecture. Maybe a professor shouldn't do it the first time they teach a class, but the second time and later they certainly could. That's probably why the professor used hand-drawn overhead slides instead of the whiteboard. He probably drew them in a previous quarter and just kept them all. By selling the notes at the bookstore I would think the professor could also make a bit of extra money as well.
One of my best classes in college was a physics class where the professor made all his notes available ahead of time in the book store. He used an overhead projector for his notes, but the beauty was we could just listen to his lecture and scribble additional notes (rarely needed). We didn't spend all of our time transcribing his notes, but instead just listening to his lecture. As I recall, I didn't take that many notes in the class, but I learned far more in that class than my other physics classes.
This was in the early 1990s. I doubt things have changed much, though, except maybe professors using Powerpoint.
Towards the end of my college education the company I worked for during the summer let me keep the laptop I was working with. I brought it in to class and used it to take notes. That was a godsend. I could easily search all of my notes and reference them much faster than by pencil and paper. Not to mention, I could type a hell of a lot faster than write in a notebook. Other students would ask for copies of my notes as well. If the notes were not available ahead of time, this was the next best thing. Of course, back then laptops were common in college.
For doing batch processing of photos I have been using Bibble Pro for Linux lately. I like it since it has good support for the raw format of my SLR and has a lot of batch processing features. For example, it's easy to select a group of 50 photos and adjust the white balance, or use the one click lens distortion fix on all my photos. Best of all, it runs under Linux. It gives me the best of both worlds. It gives me batch processing as well as the ability to individually make changes to each picture. I.e. I can bring out the shadows in a group of pictures, then straighten a couple of them if the camera was crooked and crop them as needed. It also does everything at 16 bits per color.
Now, granted, it does not run on the command line, but it easily lets me select a source and target directory to batch process as well as letting me select individual pictures. I can't really compare it with ImageMagick since I haven't used it directly.
Slashdot Mirror
There are a number of good realtime black lists (RBLs) that do just that. I use them at my mail server and it blocks a huge percentage of spam. Furthermore, I have my mail server (postfix) set up to tarpit those senders. In other words, it holds onto the connection for 20-30 seconds before sending any sort of reply, effectively slowing down the spammer a bit and consuming their resources.
I like the sbl-xbl blacklist at spamhaus.org, which combines several of them together.
My home spam filter does not seem to be affected much. I run dspam which has a feature in that over time it will forget words if they are not used in spam. Since the text is usually different or random, it does not have any significant effect on generating false positives. In the years I have been running dspam with tens of thousands of emails, I have only gotten 3-4 false positives.
By having a baysian filter forget over time, it also helps shrink down the database and helps it adapt as the contents of spam change over time.
Of course I also use other spam blocking techniques, like using realtime black lists (RBLs) and blocking a number of Chinese subnets... I should add tpnet.pl and Verizon as well.
This will help. It will be interesting to see how it compares to gasoline in terms of pollutants. Particulates are still a big problem though. Even with a 30% reduction with biodiesel, it is still far more than gasoline or ethanol, both of which are negligible. The other drawback right now at least around where I live is that diesel is significantly more expensive than gasoline, in part due to the low sulfur requirement.
I might also add that I am not favoring the oil companies, far from it. I just do not see any silver bullet that can magically solve the problems. It's a trade-off. Trade gasoline for biodiesel and you get more air pollution. Trade gasoline for ethanol and you get more water pollution, that is if we can even meet the demand for ethanol, which is not possible with corn, especially given how much water corn requires and the fact that the water supply in the big corn growing regions is a limited resource that is being depleted.
So far the closest thing I have seen that could make a big dent without significantly increasing pollution is reverse polymerization, the output of which is oil which can be further refined. This process can turn just about any form of carbon-based waste into a decent grade of oil and is something like 80% efficient. Of course, increasing fuel efficiency in our current vehicles also would help.
Now if they can solve the particulate problem and bring NOx to levels as low as a clean burning gasoline engine then it can be an alternative. It also remains to be seen if enough can be grown to replace diesel. Diesel has a lot of advantages over gasoline, though currently pollution is not one of them.
California tried to get an exemption from requiring ethanol, or for that matter any oxygenate from its fuel when it proved that using certain blends of gasoline reduced pollution further. The oxygenates only help for really old cars, of which there are not many left on the roads. For newer cars, it increases pollution and lowers the fuel economy. Not to mention that ethanol is a huge subsidy to ADM.
Biodiesel also produces MORE NOx than regular diesel (look it up). It may reduce particulate emissions, but still puts out far more than gasoline. Now it might make sense to run existing diesel vehicles on biodiesel, but it still does not make sense to convert gasoline vehicles to biodiesel until these problems are solved.
The NOx problem with any form of diesel is caused by the high compression ratio required.
Now it may cut down on other pollutants, but NOx is still a significant one that causes that orange haze over polluted cities.
Many people believe that Ethanol and Biodiesel are the answer to the shrinking oil supplies and global warming. There are some major obstacles to overcome, though.
In the case of biodiesel, it actually produces more NOx than diesel. While other pollutants are often reduced, this is the major one that forms that orange cloud over heavily polluted cities. NOx is a major pollutant of diesel engines due to the high compression ratios and still has not been effectively solved. While particulate matter is less than diesel, it is still significant, far more than gasoline. While technologies exist to reduce NOx, they are sensitive to sulfur, and while sulfur is virtually eliminated from biodiesel, it would mean that an engine designed to run on it could not use regular diesel without destroying the NOx smog equipment.
For ethanol there are other problems. First of all is the amount of energy required to harvest the corn that is currently used in this country. The best efficiencies are around 20%. In other words, for each gallon of ethanol produced, 0.8 gallons are used for growing and harvesting the corn used. Additionally, a lot of fertilizer and fresh water is required for growing the corn. I am assuming no pesticides, though pesticides like glyphosphate (RoundUp) are also often used as well, even with GM corn. The fertilizer often ends up in the waterways, causing significant pollution, like the dead zone in the Gulf of Mexico. Other sources of ethanol are better, such as sugar beets or sugar cane and possibly other sources. New techniques are being developed to produce ethanol from cellulose, meaning that agricultural waste and other crops like switchgrass or even hemp could be used to produce ethanol.
One other potential energy source is to use reverse polymerization. This process can convert almost any form of organic waste into oil and is fairly efficient and not very complex. This could easily supplement much of our current demand for oil. It can also use agricultural waste like ethanol.
A PC platform will not scale like a decent router will. The memory latency becomes the bottleneck once you start thrashing your CPU cache when you have a lot of routes or ACLs. For small setups, it might work, but it will not compete with dedicated hardware solutions once the complexity grows.
I have implemented routers, and the biggest bottleneck is typically memory latency. Once the routing tables grow beyond what will fit in the cache, the latency kills you. Dedicated hardware routers are designed with this in mind, often with multiple banks of low-latency memory or CAMS.
Try to handle 1M packets per second with 100K routes, MPLS VPNs, ACLs, policy based routing, QoS, policing, marking, reverse path forwarding checks, etc. and it will fall apart. Plus it has to keep statistics on everything. All of that will quickly exceed the memory bandwidth available on any PC platform.
A decent router can do all of this and run at gigabit speeds or more. Add to that all the various interfaces that are available and the much higher port densities.
-Aaron
My experience is that around 60-75% of the spam I receive comes from China. On my home mail server I finally broke down and started blocking the worst offending subnets and the amount of spam I received dropped dramatically. There is a RBL for China, cn.blackhole.us, or a combination of China and Korea (cn-kr.blackhole.us), though these are no longer listed and will likely disappear soon.
? country=$ctry`;
print join "\n", /([0-9\.]+\/[0-9]+)/g;
I also use several other RBLs which have helped a lot.
I also decided to add the worst offending subnets in China as rules for my firewall to block. The worst offending subnet is 221.208.208.x where my firewall reports an almost constant barrage of IM spam, and from what I've read, this subnet has been a problem for years.
For your own blocking, the following script will get all the subnets used by China (or any other country you're interested in, just change $ctry):
#!/usr/bin/perl $ctry = shift || 'cn'; $_ = `wget -O - http://www.apnic.net/apnic-bin/ipv4-by-country.pl
At work, where I cannot do this, most of my spam is also received from China.
Out of the rest of the spam I receive, the US is actually pretty far down on the list of sources, though still much higher than places like the UK, Germany or France. The rest seems to come from places like Poland, Romania and Estonia.
I can easily exceed the rate of whatever memory I put in my Nikon D70s when I shoot pictures back to back. Then again, I often shoot in raw mode to allow me to perform better post processing of the images. Each image is 5-6MB in size, so at 3 pictures per second the flash will not keep up. The faster flash definitely makes a difference if I am shooting a lot of pictures, since the raw buffer is only 4 pictures.
Granted, the higher end cameras have larger buffers which helps mitigate the problem, but faster flash is definitely an advantage in these cases. Now with JPEG, it would be a different story, especially with normal compression.
As far as I know there still is no effective solution to the amount of NOx smog produced by a diesel engine. A lot of this is due to the high compression ratios used, not necessarily the type of fuel.
If this article is accurate, biodiesel can actually produce more NOx and ozone pollution. It may significantly cut down on the soot, which is considered highly toxic. As I type this, there is another spare-the-air day where I live and all public transit is free. Most of this is due to NOx and ozone.
Diesel-electric is perfect for trains, since they need a huge amount of torque at zero RPM and electric motors solve this nicely, basically eliminating any requirements for a complex transmission.
-Aaron
Even better: just import into kword. All the redacted text magically appears right in your word processor. No copy/paste necessary.
KWord can import PDF files directly into a document.
-Aaron
The online update support for SuSE 10.1 is horrible compared to past versions. It is *extremely* slow. Adding a new repository took well over 30 minutes to process it, the CPU remaining busy the entire time. Granted, the machine is only a 1.5GHz P4, but it should take nowhere near this amount of time.
Bringing up the software install tool takes 150MB of RAM. This is excessive.
Then OpenSuse keeps moving repositories around, or deleting them. They removed the KDE 3.5.3 repository recently, for example.
I'm almost ready to switch to another distribution, maybe Kubuntu or some other up-to-date KDE based distro.
I was using blackholes.us as a RBL to block countries like China and Russia,
? country=$ctry`; /([0-9\.]+\/[0-9]+)/g;
though sadly it looks like they shut down.
I did find a nice script to grab the IP blocks of various countries, though:
#!/usr/bin/perl
$ctry = shift || 'tw';
$_ = `wget -O - http://www.apnic.net/apnic-bin/ipv4-by-country.pl
print join "\n",
This will grab all the IP blocks of Taiwan. Just use 'cn' or whatever.
Well, some of us can only see a few dozen stars when we go outside due to all the heavy light pollution. I wish they would just shut off most of the street lights at times. I think the only way the local observatories are able to work is that they have sodium filters to filter out most of the street lights.
Not all of us live in areas with low population densities.
I can *never* see the Milky Way when looking up at night unless I travel quite a ways.
Now just allow the user to substitute Austin Powers and things will be set. Maybe even offer it as a bonus after the game has been completed once or twice.
Hmmm, my Prius uses NiMH batteries, as do all the other hybrids I know of. NiCd don't last nearly as long and have a lower energy density. NiMH do not contain cadmium and are capable of also delivering high amounts of current. In fact, all of the rechargable AA batteries I see in the stores for cameras and whatnot are now NiMH instead of NiCD. I think NiMH might be a bit trickier to charge, but they suffer much less from the memory problem NiCDs have.
I doubt that neither ATI nor nVidia could release open source drivers even if they wanted to. As others have said, the drivers probably contain a lot of 3rd party proprietary code that cannot be released.
I'll tell you right now that it will kill Linux in the embedded world if manufacturers were forced to release all their modules. A lot of hardware is proprietary and remains a black box on purpose, to keep competitors out or to hide trade secrets, and yes, a lot of hardware does that.
On the VxWorks product I work on we had one vendor that tried to send us binary only object files (which were not compatible with out linker) then moved to obfuscated code. They finally moved to release non-obfuscated code, but it is still highly proprietary and cost us a small fortune, many tens of thousands of dollars. They also release Linux drivers since so much of their market has moved to Linux. At least with this, I can say that they don't muck about at all in the kernel, but just use the standard driver interfaces.
Why charge so much and make the software proprietary? Because the hardware is very high performance and not very high volume. They'd have a tough time recovering their engineering expenses if they gave away the software when a lot of their customers may buy only a few hundred chips a year.
I think nVidia has worked with kernel developers and people at Xorg and it is known what interfaces they use. I think all of the code that deals directly with the kernel is released, since it is basically a wrapper around their proprietary code.
There was actually a good article on this in the December Scientific American. There would be no shortage of fuel if we were to build breeder reactors which can convert U238, where we have a large supply, to plutonium and other fissionable fuels. The article goes on to say that the amount of dangerous radioactive waste from this type of reactor is a small fraction of other types of reactors (including pebble bed). The article can be found at Scientific American: Smarter Use of Nuclear Waste [ ENERGY ]
Unfortunately, it requires registration to read.
Actually, the acceleration of the current generation of Prius is much better than that... 0 to 60 in 10.3 seconds. Granted, it's not 7-8 seconds like some of the more powerful V-6 cars, but it's decent enough for most driving conditions. The luggage capacity of the Prius is significantly more than both the Accord and Civic, and the passenger area is nearly the same, the Accord having more shoulder and hip room and slightly more front leg room, but the Prius having more head room and rear leg room. The Prius gets roughly double the milage in city driving, however and significantly more on the highway, even ignoring the unrealistic EPA estimates (which are often wrong for most vehicles).
I had this exact same problem... I finally tracked it down to one of my computers and put a filter on the power line that fixed the problem... unfortunately the rest of X10 seems to suck as well. Some devices refuse to be controlled by the RF controller (which itself is rather flakey), but the same devices will work if controlled by another hard-wired controller. I had my bedroom light randomly turn on until I added the filter. The system just seems to degrade over time.
Not to mention that I went through the X10 outlets usually in 6 months because some cheap plastic part in them would break.
In the mean time we can thank the Europeans for reliability of electronics going to hell. It is especially problematic for the more expensive and complex equipment which tends to have a lot of BGA chips with very fine ball spacings. The tin whiskers will cause leads to short out. A good document I found is at http://nepp.nasa.gov/whisker/background/index.htm
To date it looks like you can have either lead free solder or reliability, but rarely both.
That sounds like my experience with the Hertz "N^Heverlost" system. At the time I was in Austin, Tx and we needed to go back to the airport, so we went through the menus and chose "airport" as the destination. It directed us to the old airport which had been closed for several years. We finally figured out we had to tell the system we wanted to go back home and it directed us correctly.
Fortunately we gave ourselves plenty of time.
Online is one thing, how about encouraging your students to print it out before class and take the notes with them to the lecture? Or better yet, sell them in the book store?
I had a physics class like that and it was a completely different environment. It was the first time I could actually sit back and listen to the lecture, thinking about it, since the notes I bought at the book store were identical to what he showed in the lecture. I didn't waste all my time scribbling down what was written on his overhead slides, since I already had it. It was one of the best classes I ever had. A laptop would not be of much use for taking notes, since there were a lot of diagrams and equations. Only a pencil was needed to write additional notes from time to time (rarely needed) in the margins. Even better would be to make both printed and electronic available, but require that the students have at least the printed version and bring it to class.
Now this is probably a lot harder for the professor, since they need to know at least several weeks ahead of time exactly what they'll cover in a lecture. Maybe a professor shouldn't do it the first time they teach a class, but the second time and later they certainly could. That's probably why the professor used hand-drawn overhead slides instead of the whiteboard. He probably drew them in a previous quarter and just kept them all. By selling the notes at the bookstore I would think the professor could also make a bit of extra money as well.
One of my best classes in college was a physics class where the professor made all his notes available ahead of time in the book store. He used an overhead projector for his notes, but the beauty was we could just listen to his lecture and scribble additional notes (rarely needed). We didn't spend all of our time transcribing his notes, but instead just listening to his lecture. As I recall, I didn't take that many notes in the class, but I learned far more in that class than my other physics classes.
This was in the early 1990s. I doubt things have changed much, though, except maybe professors using Powerpoint.
Towards the end of my college education the company I worked for during the summer let me keep the laptop I was working with. I brought it in to class and used it to take notes. That was a godsend. I could easily search all of my notes and reference them much faster than by pencil and paper. Not to mention, I could type a hell of a lot faster than write in a notebook. Other students would ask for copies of my notes as well. If the notes were not available ahead of time, this was the next best thing. Of course, back then laptops were common in college.
-Aaron
For doing batch processing of photos I have been using Bibble Pro for Linux lately. I like it since it has good support for the raw format of my SLR and has a lot of batch processing features. For example, it's easy to select a group of 50 photos and adjust the white balance, or use the one click lens distortion fix on all my photos. Best of all, it runs under Linux. It gives me the best of both worlds. It gives me batch processing as well as the ability to individually make changes to each picture. I.e. I can bring out the shadows in a group of pictures, then straighten a couple of them if the camera was crooked and crop them as needed. It also does everything at 16 bits per color.
Now, granted, it does not run on the command line, but it easily lets me select a source and target directory to batch process as well as letting me select individual pictures. I can't really compare it with ImageMagick since I haven't used it directly.
-Aaron