I'm not sure how fast a 2 GHz CPU will seem unless we start seeing some really fast RAM and BUSes. Right now the fastest bus I've seen is 133 MHz, and the fastest large scale RAM on the horizon is still RAMBUS. Neither of these will allow a 2 GHz processor to run anywhere near it's full potential. Even with lots of cache you would only approach some upper bound that wouldn't be anywhere as high as 2GHz.
There needs to be some other advances, other then jacking up the CPU speed, before processors like this become useful. Maybe it'll be a new RAM or BUS design (RF anyone?), or a new way of dealing with code internal to the CPU to make memory more effecient then using it for cache. Which ever it's time to start looking at them.
This is a very good point. Such a move could be very effective in surpressing open encryption development in the US. Most companies use proprietary schemes (despite beign told not to) due to the only type of security that management is familiar with, obscurtiy/secrecy(I don't know how many meetings I've been in where management suggests strength by changing one bit in an algorithm). It may also effectivly allow security from companies like Microsoft (if you want to call any of it secure) to overtake any open source solution, because of the technical review process.
While certain parts of the government may be giving it away others may be kicking and screaming. If you go back to the DES release, the NSA was DEAD SET against releasing it in a fasion that could be put in software. It appears that it was released in algorithm form by accident or before they could complain.
It's more likly that the elected officials are listening less to the IC/LE communities and more towards the big contributors (high tech firms and entrpenuers (sp?)) that are playing more of a role in the expensive compaigns for congress and president. You throw enough money around in this town and you'd be surprised how many "converts" appear at your door step.
My guess is yes. Everyone assumes that the NSA won't give away the export restrictions on codes they can't break, but I'd say that they see a bigger threat. With the congress getting more good information from sources outside the goverment on encryption they are probably starting to question the export restrictions the same way that all of us do. Now when that starts happening certain agencies will look bad, very bad. When you look bad to congress you tend to come under all sorts of nasty reviews and hearings (closed door in this case, if it were to happen) and in general budgets get cut and/or restrictions get placed that are forms of buercratic torture.
Maybe I'm just a little to jaded from my proximity to the beltway, but so are all the people making these decisions.
Normally I'm against anything that can block information, but I kinda like this one. It tags shows with a content label (granted the current method of how it does this is horrible) which I can then use as a filter for my shows. The problem is that as it's implemented I can only use it in a way that fits someone else's idea of filtering.
Right now a program that questioned the basis of say a fundamental faith system would probably get some sort of adult tag. Maybe I want that type of program to be available to my child (yes I'm a parent), but not the program that blindly promotes that fundamental faith system. I'm sure we can all think of examples of this.
Now if it had a set of measures, like sex, nudity, violence, drug use, religous content, etc. I could then use it to filter programs as I wanted my child to be exposed to, not just deny everything that might be offensive to a group of people I don't know.
Granted we could add measures until the cows come home, but that might not be a bad thing either. I don't want my child to flip from barney to a porno flick, but I probably don't want her flicking to Videodrome or the angel channel either (well not for a few years). My solution will continue to be no TV, unless it's family time, for years to come.
The other side of this coin is that if parents are allowed to filter content in a way that makes them feel comfortable then we can start having any type of content that you can imagine on the open airwaves. Yes, that local UHF station that has been struggling for years can then show R, or maybe eventually NC-17, movies uncut (except commercials) to bring in the lucrative adult market. More choice without cable or satellite.
So I guess I'd have a set of measures that are from 1 to 5, and I can have a filter that I set like a graphic equalizer, but with a twist. When the little ones go off to bed make it so it filters out the other way. I really don't need to see barney, but debbie...
The government is using Java all over the place right now. In fact with almost all the databases in the government introducing some web frontend probably 10% to 15% are being done with java on the server side (from personal observation of many government projects). This number should increase in the near term as PERL has with the government in the recent past.
One thing to note those is that databases that are to be used by multiple agencies rarely use java or javascript on the client side. This is due to the large number of agencies that block them at their firewalls.
Now if only they wouldn't complain so much when we introduce Python.
C2 certification requires an audit of the code that pertains to those requirements. A vendor has to pay for this audit (when Novell went for it the cost was implied to be quite high in the press), and then control the releases to some degree (audit the final setup with a small application tends to be the way it's done).
I've had Linux used in projects for the verious government agencies for five years now, but I can't get it onto the classified systems because it's not C2 certified. In general NT's lack of current cert is ignored or exempted (as is some other OSs), but Linux is not.
If Redhat could get their distribution of linux C2 certified then the government would have to consider it against NT everytime someone brought it up.
A branch of the NSA does have NT source code. For C2 certification you have to provide source code, and NT 3.51 was C2 certified under some conditions. Plus the official word has been NT 4 will be C2 in 6 months (well that's what I've been hearing for about 2 years). This would signify to me that Microsoft has provided that branch of the NSA (NSCS I think) with some source code.
Reading the headline I thought it was going to include plans:-(. Actually these things are fairly easy to make if you have some experience in the frequencies used. Of course you'll probably have people descending on you fairly quickly due to all the RF services you knocked off of the air (half:-).
I'd like to see a $2 design for a generic firewire interface to a microcontroller. The connector alone would take up a good bit of that. It's a pretty fancy design if you look at it. Now we have to add a physical interface chip to drive the signal (about $5 from lucent in a 64 pin package) and enough smarts to interface from a microcontroller to the device on the other end. Don't forget the power supply/converter.
Now add all of those devices and try and put them in a small package like a drive. Combined with the normal drive controller hardware for a high performance unit and you get a large package without some fancy board layout.
So we have moderate to high recurring costs and high non-recurring costs. Plus a cramped layout to fit the consumer demand for size. In a few years I'd expect to see a single chip solution like USB, but bigger and around $2-$5 without the connector.
This really boils down to cost and application. For most of the applications that a serial bus would be good for, USB is enough, and its damn cheap to implement. You can get a device USB interface for under a dollar. You can realistically make price competitive keyboards, mice, modems, etc..
Firewire does do somethings though that USB can't touch, storage, video and other HIGH bandwidth functions. Although the price on these items are high and for all but the video editing equipment, not justified. Why would I get a firewire drive when for the same price I can get a UW/2 SCSI drive (Note: I haven't checked in a few months).
This is because firewire is just to darn expensive to implement in standard devices. Look at the specs, it's not a walk in the park for the design engineer. Granted if USB 2.0 were here today it would not be much better, but taking the power requirements out will make it the low cost solution that may just work tomorrow when compared to the all in one cable of firewire.
Then again all of our devices may be wireless in five years. Those interfaces are getting dirt cheap (sort of:-).
If I remember back to my Random Processes class not much really is Gaussian. There are two reasons that that assumption is often made. The first is that we have so many tools that assume it, and work OK if we are near it. The second is the abuse of the central limit theorem which says (correct me here if I'm not percise) the sum of a large set of random variables tends toward a gaussian dist as the number of variables approaches infinite. The problem is that people tend to short the infinite part and exagerate the tends to part.
What we really need to do is stop teaching statistics classes that depend on a gaussian distribution. Down with standard deviations:-).
This could be because of a lack of a good ground, but something to watch for is that there isn't actually power coming in on the cable. Cable systems power their amps by sending a 60Hz power signal down the coax. This is done to reduce the number of power hookups in the cable plant. Anyway, the taps are supposed to block power from going to the individual cables. Every now and then the capacitor that does this fails and sends 60Hz power down the cable. It starts out at 60 volts on most systems, with a 15A supply. So be careful.
You could use it to make a basic logic gate. So long as you can then make a not and and or or, or either a nand or nor then you can make any logic circuit.
The telling tale will be what the energy consumption and timing is like.
PC/104 is a perfect solution. You can find Pentium boards with SCSI, IDE, 100BaseT, and flat panel drivers that run on 5V. Just regulate the 5V from the 12 to get rid of the noise. Then you can throw in a solid state drive and away you go.
We have a linux box in this configuration running a gun sight off of a FLIR. Works like a champ. I'd like to see a car that can generate as much abuse:-).
Slashdot Mirror
I'm not sure how fast a 2 GHz CPU will seem unless we start seeing some really fast RAM and BUSes. Right now the fastest bus I've seen is 133 MHz, and the fastest large scale RAM on the horizon is still RAMBUS. Neither of these will allow a 2 GHz processor to run anywhere near it's full potential. Even with lots of cache you would only approach some upper bound that wouldn't be anywhere as high as 2GHz.
There needs to be some other advances, other then jacking up the CPU speed, before processors like this become useful. Maybe it'll be a new RAM or BUS design (RF anyone?), or a new way of dealing with code internal to the CPU to make memory more effecient then using it for cache. Which ever it's time to start looking at them.
This is a very good point. Such a move could be very effective in surpressing open encryption development in the US. Most companies use proprietary schemes (despite beign told not to) due to the only type of security that management is familiar with, obscurtiy/secrecy(I don't know how many meetings I've been in where management suggests strength by changing one bit in an algorithm). It may also effectivly allow security from companies like Microsoft (if you want to call any of it secure) to overtake any open source solution, because of the technical review process.
While certain parts of the government may be giving it away others may be kicking and screaming. If you go back to the DES release, the NSA was DEAD SET against releasing it in a fasion that could be put in software. It appears that it was released in algorithm form by accident or before they could complain.
It's more likly that the elected officials are listening less to the IC/LE communities and more towards the big contributors (high tech firms and entrpenuers (sp?)) that are playing more of a role in the expensive compaigns for congress and president. You throw enough money around in this town and you'd be surprised how many "converts" appear at your door step.
My guess is yes. Everyone assumes that the NSA won't give away the export restrictions on codes they can't break, but I'd say that they see a bigger threat. With the congress getting more good information from sources outside the goverment on encryption they are probably starting to question the export restrictions the same way that all of us do. Now when that starts happening certain agencies will look bad, very bad. When you look bad to congress you tend to come under all sorts of nasty reviews and hearings (closed door in this case, if it were to happen) and in general budgets get cut and/or restrictions get placed that are forms of buercratic torture.
Maybe I'm just a little to jaded from my proximity to the beltway, but so are all the people making these decisions.
Normally I'm against anything that can block information, but I kinda like this one. It tags shows with a content label (granted the current method of how it does this is horrible) which I can then use as a filter for my shows. The problem is that as it's implemented I can only use it in a way that fits someone else's idea of filtering.
Right now a program that questioned the basis of say a fundamental faith system would probably get some sort of adult tag. Maybe I want that type of program to be available to my child (yes I'm a parent), but not the program that blindly promotes that fundamental faith system. I'm sure we can all think of examples of this.
Now if it had a set of measures, like sex, nudity, violence, drug use, religous content, etc. I could then use it to filter programs as I wanted my child to be exposed to, not just deny everything that might be offensive to a group of people I don't know.
Granted we could add measures until the cows come home, but that might not be a bad thing either. I don't want my child to flip from barney to a porno flick, but I probably don't want her flicking to Videodrome or the angel channel either (well not for a few years). My solution will continue to be no TV, unless it's family time, for years to come.
The other side of this coin is that if parents are allowed to filter content in a way that makes them feel comfortable then we can start having any type of content that you can imagine on the open airwaves. Yes, that local UHF station that has been struggling for years can then show R, or maybe eventually NC-17, movies uncut (except commercials) to bring in the lucrative adult market. More choice without cable or satellite.
So I guess I'd have a set of measures that are from 1 to 5, and I can have a filter that I set like a graphic equalizer, but with a twist. When the little ones go off to bed make it so it filters out the other way. I really don't need to see barney, but debbie...
The government is using Java all over the place right now. In fact with almost all the databases in the government introducing some web frontend probably 10% to 15% are being done with java on the server side (from personal observation of many government projects). This number should increase in the near term as PERL has with the government in the recent past.
One thing to note those is that databases that are to be used by multiple agencies rarely use java or javascript on the client side. This is due to the large number of agencies that block them at their firewalls.
Now if only they wouldn't complain so much when we introduce Python.
C2 certification requires an audit of the code that pertains to those requirements. A vendor has to pay for this audit (when Novell went for it the cost was implied to be quite high in the press), and then control the releases to some degree (audit the final setup with a small application tends to be the way it's done).
I've had Linux used in projects for the verious government agencies for five years now, but I can't get it onto the classified systems because it's not C2 certified. In general NT's lack of current cert is ignored or exempted (as is some other OSs), but Linux is not.
If Redhat could get their distribution of linux C2 certified then the government would have to consider it against NT everytime someone brought it up.
A branch of the NSA does have NT source code. For C2 certification you have to provide source code, and NT 3.51 was C2 certified under some conditions. Plus the official word has been NT 4 will be C2 in 6 months (well that's what I've been hearing for about 2 years). This would signify to me that Microsoft has provided that branch of the NSA (NSCS I think) with some source code.
Reading the headline I thought it was going to include plans :-(. Actually these things are fairly easy to make if you have some experience in the frequencies used. Of course you'll probably have people descending on you fairly quickly due to all the RF services you knocked off of the air (half :-).
Let's not forget it's harder to read in the B/W in some lighting.
That 25 cents wasn't even considered in calling it expensive. I was talking only about the hardware implementation and design costs.
I'd like to see a $2 design for a generic firewire interface to a microcontroller. The connector alone would take up a good bit of that. It's a pretty fancy design if you look at it. Now we have to add a physical interface chip to drive the signal (about $5 from lucent in a 64 pin package) and enough smarts to interface from a microcontroller to the device on the other end. Don't forget the power supply/converter.
Now add all of those devices and try and put them in a small package like a drive. Combined with the normal drive controller hardware for a high performance unit and you get a large package without some fancy board layout.
So we have moderate to high recurring costs and high non-recurring costs. Plus a cramped layout to fit the consumer demand for size. In a few years I'd expect to see a single chip solution like USB, but bigger and around $2-$5 without the connector.
This really boils down to cost and application. For most of the applications that a serial bus would be good for, USB is enough, and its damn cheap to implement. You can get a device USB interface for under a dollar. You can realistically make price competitive keyboards, mice, modems, etc..
:-).
Firewire does do somethings though that USB can't touch, storage, video and other HIGH bandwidth functions. Although the price on these items are high and for all but the video editing equipment, not justified. Why would I get a firewire drive when for the same price I can get a UW/2 SCSI drive (Note: I haven't checked in a few months).
This is because firewire is just to darn expensive to implement in standard devices. Look at the specs, it's not a walk in the park for the design engineer. Granted if USB 2.0 were here today it would not be much better, but taking the power requirements out will make it the low cost solution that may just work tomorrow when compared to the all in one cable of firewire.
Then again all of our devices may be wireless in five years. Those interfaces are getting dirt cheap (sort of
If I remember back to my Random Processes class not much really is Gaussian. There are two reasons that that assumption is often made. The first is that we have so many tools that assume it, and work OK if we are near it. The second is the abuse of the central limit theorem which says (correct me here if I'm not percise) the sum of a large set of random variables tends toward a gaussian dist as the number of variables approaches infinite. The problem is that people tend to short the infinite part and exagerate the tends to part.
What we really need to do is stop teaching statistics classes that depend on a gaussian distribution. Down with standard deviations:-).
This could be because of a lack of a good ground, but something to watch for is that there isn't actually power coming in on the cable.
Cable systems power their amps by sending a 60Hz power signal down the coax. This is done to reduce the number of power hookups in the cable plant. Anyway, the taps are supposed to block power from going to the individual cables. Every now and then the capacitor that does this fails and sends 60Hz power down the cable. It starts out at 60 volts on most systems, with a 15A supply. So be careful.
You could use it to make a basic logic gate. So long as you can then make a not and and or or, or either a nand or nor then you can make any logic circuit.
The telling tale will be what the energy consumption and timing is like.
PC/104 is a perfect solution. You can find Pentium boards with SCSI, IDE, 100BaseT, and flat panel drivers that run on 5V. Just regulate the 5V from the 12 to get rid of the noise. Then you can throw in a solid state drive and away you go.
:-).
We have a linux box in this configuration running a gun sight off of a FLIR. Works like a champ. I'd like to see a car that can generate as much abuse