A small number of samples is enough to estimate the traffic state using their doppler shift. For example, if there is even one cellphone at 60MPH out of hundreds of vehicles on a given section of road it is very unlikely that traffic is jammed. While it is possible that some vehicles will drive much slower than others if you get just two or three 8MPH doppler signatures and no higher ones you probably have slow traffic.
One of the serious problems with massively parallel supercumputers is heat dissipation. I'll leave the rest of the calculation as an exercise for the reader.
This idea has been discussed to death over the last years. There are a few devices like this already shipping but they are very expensive. They used for special application like the healthcare industry (yes, they replace the doctor's clipboard).
The problem is not coming up with ideas, any slashdotter can do that. The problem is implementing them in a way that works and at the right price.
Of course they did nothing wrong. You don't get to be a monopoly by doing anything wrong. You only get to be a monopoly if you do everything right. If Microsoft behaved any differently than it did it should have been sued by its stockholders, and rightly so.
But that is not the point.
Antitrust law is not about "you have been a bad boy, now you will be punished". It's not about right and wrong. It's about what works, and monopolys don't work for consumers. The free market works remarkably well most of the time but when it doesn't antittrust laws kick in.
One way to avoind storing card numbers is to have the client reenter them at each purchase. Unfortunately, people don't like that. They like the convenience of one-click buying.
Another possibility is for the credit card verification company to issue a unique token which can be used only for billing to the same merchant. This data is stored in the merchant's databases and should be quite useless to anyone else.
Unlimited exponential growth requires unlimited resources. Sooner or later you start running low on certain resources and when that happens competition kicks in - competing for the limited resource. Your competitors may try to grab that resource more better than you (more efficient herbivores) or simply try to kill you and get the resources you have successfully gathered (carnivores).
In the case of media I believe the resource that runs out the most quickly is mindshare. There are only so many names that the public can remember. When you have a bigger audience it increases mindshare because not everyone needs to remember all the names, there can be some specialization and interest groups.
The important point, though, is that the public's mindshare capacity grows at a sublinear rate. This means that an audience twice as large can remember less than twice as many different names. The reason for this is that in order to grab mindshare a name does not need to reach a certain number of minds, it needs to reach a certain density, or percentage of the minds because names tend to fade away quite quickly unless you keep hearing them mentioned by your peers. The name needs to reach a certain percentage of your peer group in order to remain in your mind.
These are relatively recent (1995?) changes the USPTO made to help independent inventors:
1. Provisional patents. These submissions include only the disclosure part, not the claims. They give you a patent number and one year to secure your investment and submit the full patent. They are much cheaper since they are not examined by a patent examiner and do not require a patent attorney to write the claims.
2. Patent maintenance fees. Unless a periodic fee is payed to the USPTO the patent expires. The idea is that if the patent is used in actual business these fees will be insignificant, but they help reduce the cost of initial applications, again to help the independent inventor.
A note about obviousness: I am still not sure what my opinion is on the subject of software patents, but it is quite obvious (...) that most of the software patent problems arise from the fact that there are not enough patent examiners with software background.
Courts seem to be very reluctant to rule a patent as obvious. How would you prove a patent is obvious in court? If you approach 5 engineers skilled in the art and present them with the problem and 4 of them come up with an identical solution would it be enough proof?
It compresses a high resolution full color scan of a magazine page by 1:200. And I am talking about real-life performance here, not ideal cases.
The trick is an algorithm which automatically separates text and line art from continuous tone images and compresses each one with a different algorithm. The continous tone algorithm is wavelet based, of course. This is mentioned in the JPEG2000 article as a possible future extension but DjVu has been doing it for almost two years now.
They have a Netscape plugin for viewing this stuff and the compressor is free for noncommercial use. It supports linux and many other operating systems.
There are many compression schemes better than JPEG being promoted by their inventors. I believe JPEG2000 will probably be the winner for a very simple reason - the name JPEG.
Many people wouldn't forgive the racist discrimination against Wookies which resulted in Chewie not getting a medal at the end of A New Hope.
Well, Chewie got his medal. At the MTV awards ceremony Leia gave him his long due medal and he gave a speech in Wookie. No doubt he was thanking his parents which we now got the chance to know...
Patent Number 5,982,276 Magnetic field based power transmission line communication method and system
They are not exactly transmitting the data by injecting it into the line - that method is limited by bandwidth, SNR and regulations to a few megabits per second. They are using a MASER to transmit a microwave RF signal that uses the magnetic field AROUND the powerline as a waveguide. This will not work all the way down to your socket but powerlines hanging from pylons could be used by utilities to compete with fiber.
I remember the first time I saw the Lucent logo: it was on a USB developers conference and a speaker from the newly-named Lucent stepped up to the podium. Behind him the projector showed his name and the huge Lucent condom logo. He looked so ambarrased... he spent the first few minutes of his speech apologizing about the logo.
How can you compare a mobile 128kbps system to a fixed 44Mbps link?
And why is everyone against Cisco, saying "oh, this is not new" or "it's just marketing BS".
Well, for your information it IS new. There is no other comparable solution in the market for fixed long-range high speed non-line-of-sight communication.
Configuring sendmail is notoriously difficult, but with linuxconf it was very easy and it works like a charm. I'm sure postfix, qmail and exim are all worthy mailers but I just don't have the time to go through their manuals.
This may be a good reason for me to use sendmail but it definitely isn't a reason why redhat should use it as the default mailer in the first place.
BTW. compared to Mozilla, sendmail is an extremely lean monster - it's memory footprint is actually quite small.
Who said it was a linux distribution? The term "linux distribution" has come to mean "a unix like system with all the unix-like tools, based on the Linux kernel". It is entirely possible to create an OS which looks nothing like unix but using the Linux kernel.
Remember that Microsoft's Paul Allen is one of the major investors in Transmeta. And I definitely remember seeing somewhere that the Transmeta CPU is supposed to run Windows NT as its primary target.
My guess would be that they are writing a new HAL and recompiling the performance-critical parts to native code. You can afford to run MSWord in emulation. Even your soundcard driver won't mind too much.
Now all that remains is to get a few CPU-hogging killer apps like Lightwave or Adobe Premiere to recompile to Transmeta native code. A really fast JVM would make a Transmeta box an attractive middleware application server, too.
But I am definitely looking forward to a linux kernel that can execute both i386 and transmeta executables... ----
I don't know about the film industry, but I certainly expected this. I was making bets on it with some friends. When I heard there will be software implementations I gave the DVD format 6 months before it's reverse-engineered. I lost... it took a little longer.
Why is everyone so convinced that oil and gas are from a biological source?
Hydrocarbons are found in comets, outer solar system planet, some theories say that in Titan it might be raining petrol - but here on Earth? oh, no, the old textbooks can't be wrong. It must be biological in source.
Too often in academic circles, certain views can be taken as correct, without being proven.
Here are a few examples of things that "everybody knows" with the responses from heretics in the field. (btw, I don't necessarily agree with these views, but can you imagine what would happen if just one of them were true?)
1. A virus called HIV causes AIDS. AIDS is certain death. Yup. that's right. There are quite a few respected scientists and even nobel prize winners who think this is either an oversimplification or even a complete myth. Unlike most hypotheses, the AIDS-HIV hypothesis wasn't published first in a scientific magazine. It was announced on a press conference, in a rush, before someone on the other side of the atlantic had the chance to do it first. It became "instant truth" and trying to claim otherwise was a good recipe to get ridiculed by your peers an even ruin a career. Some of the AIDS heretic's claims: * HIV isn't a sufficient condition for AIDS. Some people have been living with HIV for over 10 years with no treatment and apparent symptoms. The response of AIDS mainstream research has been to update the AIDS incubation period several times, from a few months to 3 years, 5 years and even more. * HIV isn't even a NECESSARY condition for AIDS. People have been diagnozed with AIDS without testing as HIV antidote positive. * AZT and other drugs used in AIDS treatment attack the immune system, creating the exact same symptoms that HIV is alleged to cause. The justification for giving such highly toxic drugs to AIDS patients is that AIDS is invariably fatal.
The radicals like Peter Duesberg claim that AIDS is caused solely by behavioural and environmental factors and that HIV is harmless. The more moderate claim that it might be a contributing factor, but has not been proven to the the only cause.
2. CFCs deplete ozone. Less ozone=more UV. More UV=more skin cancer.
Some of the heretic's claims: *Volcanos and other natural sources release several orders of magnitude more ozone-killing chlorine and bromides into the atmosphere that all of man-made CFCs. *The correlation between ozone and UV levels is in fact POSITIVE, not negative as might be expected. People often forget that ozone not only blocks UV, it is GENERATED by UV. *The antrarctic ozone hole is a periodic phenomenon which has been observed at similar levels back in the 50s. *The patents on the original CFCs alleged to destroy the ozone have expired. The same companies now own the patents on the ozone friendly replacements. *Changes in UV levels due to altitude and latitude are much larger than those predicted by worst-case ozone scenarios and yet skin cancer levels closer to the equator aren't noticably higher.
3. Cold fusion was either a hoax or gross measurement error. *Many scientists around the world, including researches in US government national labs have been generating significant amounts of unexplained excess heat, are they all charlatans or totally incompetent in operating a calorimeter? They claim to get much more consistent results than the original Fleischman and Pons experiment. *Helium and other fusion byproducts have been measured and are accumulating over time.
4. No amount of radiation is safe. It is a purely accumulated risk. * Some evidence shows that low levels of radiation (still significantly higher than ambient) are actually good for your health * It's a well known response of the immune system where small amounts of something are actually beneficial in preventing the symptoms caused by large amounts of the same thing. (This is called hormesis) * Statistics of cancer in nuclear industry workers are actually much lower than average, especially after subtracting those which can be traced to high exposures in accidents. * Cancer in US states with higher levels of average background radiation due to altitude or radioactive materials in the grounds is noticably lower.
In all of these cases it's easy to see the role of the media, popular hysteria, politics or political correctness in determining what is considered an inappropriate viewpoint. It's much easier to believe in something which won't get you in trouble.
One method which Schneier did not mention for getting better security out the amount of entropy you can remember is using a workload factor - apply a function which wastes CPU and possibly memory to the passphrase before using it as a key. You can probably spare 512k and 200ms any time you decrypt your private key for use, but it will make brute-forcing your passphrase much harder.
My question is - what function would you use? The trivial answer would be to run the hash function over and over again, but this is not necessarily a good idea: these functions would run very fast on dedicated hardware cracker. I believe it may be better to use an algorithm which is known to be the most efficient solution for a certain problem (something from Knuth?), an algorithm which makes good use of the fetures of a general purpose CPU so an FPGA will not really be any faster. Add a cryptographic hash function here and there just to make sure there are no shortcuts or invertible stages.
Slashdot Mirror
A small number of samples is enough to estimate the traffic state using their doppler shift. For example, if there is even one cellphone at 60MPH out of hundreds of vehicles on a given section of road it is very unlikely that traffic is jammed. While it is possible that some vehicles will drive much slower than others if you get just two or three 8MPH doppler signatures and no higher ones you probably have slow traffic.
----
One of the serious problems with massively parallel supercumputers is heat dissipation. I'll leave the rest of the calculation as an exercise for the reader.
----
This idea has been discussed to death over the last years. There are a few devices like this already shipping but they are very expensive. They used for special application like the healthcare industry (yes, they replace the doctor's clipboard).
The problem is not coming up with ideas, any slashdotter can do that. The problem is implementing them in a way that works and at the right price.
----
Well, isn't that exactly what they are already doing? How can a "ubiquitous consumer machine" get any more ubiquitous than a cordless webpad?
If you really want a kiosk form factor so badly you can take the webpad and stick it on the front of a big, empty box.
----
Of course they did nothing wrong. You don't get to be a monopoly by doing anything wrong. You only get to be a monopoly if you do everything right. If Microsoft behaved any differently than it did it should have been sued by its stockholders, and rightly so.
But that is not the point.
Antitrust law is not about "you have been a bad boy, now you will be punished". It's not about right and wrong. It's about what works, and monopolys don't work for consumers. The free market works remarkably well most of the time but when it doesn't antittrust laws kick in.
----
One way to avoind storing card numbers is to have the client reenter them at each purchase. Unfortunately, people don't like that. They like the convenience of one-click buying.
Another possibility is for the credit card verification company to issue a unique token which can be used only for billing to the same merchant. This data is stored in the merchant's databases and should be quite useless to anyone else.
----
Unlimited exponential growth requires unlimited resources. Sooner or later you start running low on certain resources and when that happens competition kicks in - competing for the limited resource. Your competitors may try to grab that resource more better than you (more efficient herbivores) or simply try to kill you and get the resources you have successfully gathered (carnivores).
In the case of media I believe the resource that runs out the most quickly is mindshare. There are only so many names that the public can remember. When you have a bigger audience it increases mindshare because not everyone needs to remember all the names, there can be some specialization and interest groups.
The important point, though, is that the public's mindshare capacity grows at a sublinear rate. This means that an audience twice as large can remember less than twice as many different names. The reason for this is that in order to grab mindshare a name does not need to reach a certain number of minds, it needs to reach a certain density, or percentage of the minds because names tend to fade away quite quickly unless you keep hearing them mentioned by your peers. The name needs to reach a certain percentage of your peer group in order to remain in your mind.
----
...but there is no guarantee for how long it will take. Bad science can survive for a generation or two and cause much grief on the way.
----
Next we'll be hearing about "Why isn't there a way to just 'get a domain name' by changing your machine's name under windows?"
Actually there is - just not a second level domain name.
At work I have a zone automatically updated from the DHCP database. Just change your machine's name under windows and it shows up on DNS.
----
These are relatively recent (1995?) changes the USPTO made to help independent inventors:
1. Provisional patents. These submissions include only the disclosure part, not the claims. They give you a patent number and one year to secure your investment and submit the full patent. They are much cheaper since they are not examined by a patent examiner and do not require a patent attorney to write the claims.
2. Patent maintenance fees. Unless a periodic fee is payed to the USPTO the patent expires. The idea is that if the patent is used in actual business these fees will be insignificant, but they help reduce the cost of initial applications, again to help the independent inventor.
A note about obviousness:
I am still not sure what my opinion is on the subject of software patents, but it is quite obvious (...) that most of the software patent problems arise from the fact that there are not
enough patent examiners with software background.
Courts seem to be very reluctant to rule a patent as obvious. How would you prove a patent is obvious in court? If you approach 5 engineers skilled in the art and present them with the problem and 4 of them come up with an identical solution would it be enough proof?
----
Check out DjVu (pronounced deja-vu) by AT&T.
It compresses a high resolution full color scan of a magazine page by 1:200. And I am talking about real-life performance here, not ideal cases.
The trick is an algorithm which automatically separates text and line art from continuous tone images and compresses each one with a different algorithm. The continous tone algorithm is wavelet based, of course. This is mentioned in the JPEG2000 article as a possible future extension but DjVu has been doing it for almost two years now.
They have a Netscape plugin for viewing this stuff and the compressor is free for noncommercial use. It supports linux and many other operating systems.
There are many compression schemes better than JPEG being promoted by their inventors. I believe JPEG2000 will probably be the winner for a very simple reason - the name JPEG.
----
Goddard was a rare combination of a theoretical thinker AND a practical engineer and tinkerer.
He was also a victim of journalistic FUD and mistrustful of governments.
This quote from goddard has been my sig for quite a long time now.
----
Many people wouldn't forgive the racist discrimination against Wookies which resulted in Chewie not getting a medal at the end of A New Hope.
Well, Chewie got his medal. At the MTV awards ceremony Leia gave him his long due medal and he gave a speech in Wookie. No doubt he was thanking his parents which we now got the chance to know...
----
Patent Number 5,982,276 Magnetic field based power transmission line communication method and system
They are not exactly transmitting the data by injecting it into the line - that method is limited by bandwidth, SNR and regulations to a few megabits per second. They are using a MASER to transmit a microwave RF signal that uses the magnetic field AROUND the powerline as a waveguide. This will not work all the way down to your socket but powerlines hanging from pylons could be used by utilities to compete with fiber.
----
I remember the first time I saw the Lucent logo: it was on a USB developers conference and a speaker from the newly-named Lucent stepped up to the podium. Behind him the projector showed his name and the huge Lucent condom logo. He looked so ambarrased... he spent the first few minutes of his speech apologizing about the logo.
----
How can you compare a mobile 128kbps system to a fixed 44Mbps link?
And why is everyone against Cisco, saying "oh, this is not new" or "it's just marketing BS".
Well, for your information it IS new. There is no other comparable solution in the market for fixed long-range high speed non-line-of-sight communication.
----
Configuring sendmail is notoriously difficult, but with linuxconf it was very easy and it works like a charm. I'm sure postfix, qmail and exim are all worthy mailers but I just don't have the time to go through their manuals.
This may be a good reason for me to use sendmail but it definitely isn't a reason why redhat should use it as the default mailer in the first place.
BTW. compared to Mozilla, sendmail is an extremely lean monster - it's memory footprint is actually quite small.
----
Who said it was a linux distribution? The term "linux distribution" has come to mean "a unix like system with all the unix-like tools, based on the Linux kernel". It is entirely possible to create an OS which looks nothing like unix but using the Linux kernel.
----
Remember that Microsoft's Paul Allen is one of the major investors in Transmeta. And I definitely remember seeing somewhere that the Transmeta CPU is supposed to run Windows NT as its primary target.
My guess would be that they are writing a new HAL and recompiling the performance-critical parts to native code. You can afford to run MSWord in emulation. Even your soundcard driver won't mind too much.
Now all that remains is to get a few CPU-hogging killer apps like Lightwave or Adobe Premiere to recompile to Transmeta native code. A really fast JVM would make a Transmeta box an attractive middleware application server, too.
But I am definitely looking forward to a linux kernel that can execute both i386 and transmeta executables...
----
I don't know about the film industry, but I certainly expected this. I was making bets on it with some friends. When I heard there will be software implementations I gave the DVD format 6 months before it's reverse-engineered. I lost... it took a little longer.
----
Lulling our anxiety over the extinction of fossil fuels is a dangerous effect of somthing that may be nothing more than just such gibberish.
So if a hypothesis is dangerous in your opinion it doesn't deserve the same treatment as any other?
Since when do the social effects of a theory have any relationship to it being either true or untrue?
This is first step on the same road that leads to political correctness hell.
----
Why is everyone so convinced that oil and gas are from a biological source?
Hydrocarbons are found in comets, outer solar system planet, some theories say that in Titan it might be raining petrol - but here on Earth? oh, no, the old textbooks can't be wrong. It must be biological in source.
----
Too often in academic circles, certain views can be taken as correct, without being proven.
Here are a few examples of things that "everybody knows" with the responses from heretics in the field.
(btw, I don't necessarily agree with these views, but can you imagine what would happen if just one of them were true?)
1. A virus called HIV causes AIDS. AIDS is certain death.
Yup. that's right. There are quite a few respected scientists and even nobel prize winners who think this is either an oversimplification or even a complete myth.
Unlike most hypotheses, the AIDS-HIV hypothesis wasn't published first in a scientific magazine. It was announced on a press conference, in a rush, before someone on the other side of the atlantic had the chance to do it first. It became "instant truth" and trying to claim otherwise was a good recipe to get ridiculed by your peers an even ruin a career.
Some of the AIDS heretic's claims:
* HIV isn't a sufficient condition for AIDS.
Some people have been living with HIV for over 10 years with no treatment and apparent symptoms. The response of AIDS mainstream research has been to update the AIDS incubation period several times, from a few months to 3 years, 5 years and even more.
* HIV isn't even a NECESSARY condition for AIDS.
People have been diagnozed with AIDS without testing as HIV antidote positive.
* AZT and other drugs used in AIDS treatment attack the immune system, creating the exact same symptoms that HIV is alleged to cause. The justification for giving such highly toxic drugs to AIDS patients is that AIDS is invariably fatal.
The radicals like Peter Duesberg claim that AIDS is caused solely by behavioural and environmental factors and that HIV is harmless. The more moderate claim that it might be a contributing factor, but has not been proven to the the only cause.
2. CFCs deplete ozone. Less ozone=more UV. More UV=more skin cancer.
Some of the heretic's claims:
*Volcanos and other natural sources release several orders of magnitude more ozone-killing chlorine and bromides into the atmosphere that all of man-made CFCs.
*The correlation between ozone and UV levels is in fact POSITIVE, not negative as might be expected. People often forget that ozone not only blocks UV, it is GENERATED by UV.
*The antrarctic ozone hole is a periodic phenomenon which has been observed at similar levels back in the 50s.
*The patents on the original CFCs alleged to destroy the ozone have expired. The same companies now own the patents on the ozone friendly replacements.
*Changes in UV levels due to altitude and latitude are much larger than those predicted by worst-case ozone scenarios and yet skin cancer levels closer to the equator aren't noticably higher.
3. Cold fusion was either a hoax or gross measurement error.
*Many scientists around the world, including researches in US government national labs have been generating significant amounts of unexplained excess heat, are they all charlatans or totally incompetent in operating a calorimeter?
They claim to get much more consistent results than the original Fleischman and Pons experiment.
*Helium and other fusion byproducts have been measured and are accumulating over time.
4. No amount of radiation is safe. It is a purely accumulated risk.
* Some evidence shows that low levels of radiation (still significantly higher than ambient) are actually good for your health
* It's a well known response of the immune system where small amounts of something are actually beneficial in preventing the symptoms caused by large amounts of the same thing. (This is called hormesis)
* Statistics of cancer in nuclear industry workers are actually much lower than average, especially after subtracting those which can be traced to high exposures in accidents.
* Cancer in US states with higher levels of average background radiation due to altitude or radioactive materials in the grounds is noticably lower.
In all of these cases it's easy to see the role of the media, popular hysteria, politics or political correctness in determining what is considered an inappropriate viewpoint. It's much easier to believe in something which won't get you in trouble.
----
To prevent a precomputation attack all you need is a small salt value per user.
----
One method which Schneier did not mention for getting better security out the amount of entropy you can remember is using a workload factor - apply a function which wastes CPU and possibly memory to the passphrase before using it as a key. You can probably spare 512k and 200ms any time you decrypt your private key for use, but it will make brute-forcing your passphrase much harder.
My question is - what function would you use? The trivial answer would be to run the hash function over and over again, but this is not necessarily a good idea: these functions would run very fast on dedicated hardware cracker. I believe it may be better to use an algorithm which is known to be the most efficient solution for a certain problem (something from Knuth?), an algorithm which makes good use of the fetures of a general purpose CPU so an FPGA will not really be any faster.
Add a cryptographic hash function here and there just to make sure there are no shortcuts or invertible stages.
Any suggestions?
----