You might as well have cameras as the waveform shape changes when your drive over the sensor at different positions (i.e. waveform for the sensor at the right side of the car, another for the middle and another for the left side) . On the highway you would get waveform A from the car at position 1 then waveform B at position 2 and so on. Matching up waveforms sort of works but is very fuzzy and works about 10% of the time. We could do it for our small fleet of cars but only because we had captured just about all possible variations of their waveforms.
Been there done that. Worked with a company that did traffic sensors and could see the waveforms from vehicles. We could and did identify makes and models of cars. BUT take that same car and drive north or south over rough road for a while and you get a different waveform ! (Hint the car is now magnetized differently). Anyway yes you could identify specific cars of the set the company owned. But this would not extrapolate to those same models in the wild driven differently or with a different magnetic history. As just driving the car will change the waveform.
Ahem, the Telco's have hordes of lawyers to advise them if something is legal or not. What they did was NOT legal under the law when they did it. Remember this is a country of laws not of men, you cannot be told to do illegal things and then NOT be held responsible. The US government cannot order you to do illegal things. (Remember we hung German officials after WW II for "Only following orders") In any case these suites against the telcos are not for $ they are to discover the truth about what was done to us the citizens of this country.
I remember this from an article in the 60's in Popular Science. I think it was some university in the Midwest (US) doing these types of studies. They had pictures of the shock waves and how far then extended from the actual cause. Also the Long Island expressway in NY (US again) is famous for this. Traffic would slow to a crawl for no apparent reason.
Ah, do you know where the Internet came from ? It was NOT a commercial venture, it was DARPA that created it (and tried to give it to AT&T who refused). Without that start we would still be using Prodigy or something like AOL. Yes the free market and such is great but if they don't see growth in 2 quarters they are outta there. So to start these things you need the.gov to pave the way (with $ usually).
This brings back memories, in the late 80's our company was experimenting with GPS. Since at that time there were so few satellites in orbit you had to calculate when you could have 4 in view. This always turned out to be about 3 AM. So there we were cruising down the highways in Western PA at 3AM in a tricked out van full of computers and other things that go beep in the night... Anyway we dreamed of being pulled over by the state cops. "Ok officer let me play that back for you...."
Right now my group and I are working on an embedded system and the development environment REQUIRES you have about 4-6 IDE project windows open at once and you usually have a number of edit windows, debug windows, connection organizer windows, miscellaneous compile status windows. On top of this you have a couple of serial terminal windows, telnet/ftp windows, MSys windows and then a few special purpose editor windows open. Now this is just the dev environment and doesn't count the e-mail,IM system and browser you probably have open. I have two BIG screens but others manage on just one. I also have a PowerBook with two screens used for a simulator development, subversion system, editors for the embedded system's files and stimulator for the system under test.
With the two running side by side OS X wins hands down on handling multiple windows and just the sort of human multitasking done in this sort of system. Unfortunately the embedded IDE only runs on Windows.
I agree with this sub-thread. I've been on both sides of this as a user and as an author. I can tell there is not much more frustrating that wanting to use some piece of code (that does close to what you want) but being prevented from using because you cannot figure out how to change it and the author has no internal or external documentation. In effect you have to do the whole thing again and get inside his head to get to the point of changing it. At that point you usually bail and do your own kludge.
If you want your stuff received well try to document it so others can understand what you have done. There are many ways of doing this but two I have found most useful are (1) Good external documents (it goes in here and comes out here) documenting how to use the code and some examples are really nice. (some of the best I've seen was in libsndfile http://www.mega-nerd.com/libsndfile/ where I've never needed to change the code and external descriptions make it a breeze to use). (2) Internal documentation where you document how it works and have meaningful headers on your routines so the user can see what the input/output variables are and what the routine does. This has worked on code I have put out. I have been pleasantly surprised by users adding things (which I've incorporated) or fixing things.
Also you have to see it as a gift whenever you get a note about the code from someone, even if they want a bug fix. I try not to get burned out on this and try to see it that someone was interested enough to try to use my code even if it doesn't work and they are giving you another chance.
We are on the second iteration of an embedded system (first pass on Linux, second pass on a propriety OS) in each we start a number of applications. These apps are all the same executable but command line options tell the app where to find info on which.so's (DLL's) it loads to give it a personality. the options also tell a particular app its name and information on where it's file system lives. In this case command line parsing is quite important. We use a simple system which locks argc,argv into globals in each app and then we have a function that can find a particular flag and associated value at any point in the code. This way we don't have to parse the command line in main() although some happens there but wherever we need to get a particular flag we can access it. We come down on the minimize code size/RAM size vs efficiency in our choices as none of these flag searches is done more than once and don't have any visible effect on program speed.
I didn't see it in the article but are there any checks on the data supplied by the transponders ? Could you feed in false GPS data and make your aircraft disappear from where it is and put it somewhere else ? Is there crypto in this system or is everything just trusted ?
Unless there is software needs the old OS, there are USB floppies and USB serial ports that work fine in new hardware. We have a bunch of Tek scopes that have floppies for screen captures. The hardware guys have to have a USB floppy so they can put the images into their reports.
--jim
And another use I've found is to keep all my PDF's in one place. You can drag them to the library and make a smart folder for type PDF and there they are. Real handy and you can sort by title,author and stuff (you do have put that in yourself, no CDDB for these babies.) Double click and you can view them. Keeps my references right at hand.
You might not have any other Mac's in the building but have you used the Distrubuted builds under XCode ? I have a 12" 1 GHz laptop but when I'm at work the graphics guy lets me steal cycles on Dual G5. Plus they are bringing up a couple more G4's soon so all I have to do is install XCode enable the distrubuted build stuff on them. I've cleared with each of the owners and since they don't see any degredation of performance they just fine with this.
Instead of just one compile at a time I get 3 to 6 at once, with a large build this really helps. You still have the link phase which is only on the build box but the compiles happen a LOT faster.
I think there was a note once about using Linux boxes in the distcc net but it's not quite so dead simple as this.
In OS X dual cores always help because you are running a lot of processes and they get spread over two cores not just one. (In the Activity monitor you can show up to 200% processor utilization with dual cores.)
Give every kid a copy of the KNOPPIX CD. (http://www.knopper.net/knoppix/index-en.html) They can take it home and run Open Source stuff on the home PC with little risk of messing up Mom's recipes or Dad's tax forms. There is no installation hassle and it has a very good selection of OS stuff.
(The people with Mac's at home just need to be told about Fink).
--jim
Slashdot Mirror
(1) Convert the machine(s) to virtual images.
(2) Host on Linux boxes/generic hardware
(3) Arrange to boot into the app running in the VM
(4) Backup the images regularly
(5) Buy a spare box
The VM image is effectively immortal if hardware fails just buy whatever is cheap & current install Linux and the VM program.
(2) GnuPlot Plots all that beautiful data....
7000 -> f = lambda/c -> 4.28275E+14
5000 -> f = lambda/c -> 5.99585E+14
Difference -> 1.713E+14 Hz -> 1.713E5 GHZ
About 171,000 GHZ not 17
Ruby doc
You might as well have cameras as the waveform shape changes when your drive over the sensor at different positions (i.e. waveform for the sensor at the right side of the car, another for the middle and another for the left side) . On the highway you would get waveform A from the car at position 1 then waveform B at position 2 and so on. Matching up waveforms sort of works but is very fuzzy and works about 10% of the time. We could do it for our small fleet of cars but only because we had captured just about all possible variations of their waveforms.
Been there done that. Worked with a company that did traffic sensors and could see the waveforms from vehicles. We could and did identify makes and models of cars. BUT take that same car and drive north or south over rough road for a while and you get a different waveform ! (Hint the car is now magnetized differently). Anyway yes you could identify specific cars of the set the company owned. But this would not extrapolate to those same models in the wild driven differently or with a different magnetic history. As just driving the car will change the waveform.
Where I worked then (1992) we got letters from Nigeria with the 419 scams. The advantage then was you got some cool stamps.
Ahem, the Telco's have hordes of lawyers to advise them if something is legal or not. What they did was NOT legal under the law when they did it. Remember this is a country of laws not of men, you cannot be told to do illegal things and then NOT be held responsible. The US government cannot order you to do illegal things. (Remember we hung German officials after WW II for "Only following orders") In any case these suites against the telcos are not for $ they are to discover the truth about what was done to us the citizens of this country.
So give every customer a Live CD of a really locked down Linux and a special purpose browser pointed to the bank.
I remember this from an article in the 60's in Popular Science. I think it was some university in the Midwest (US) doing these types of studies. They had pictures of the shock waves and how far then extended from the actual cause. Also the Long Island expressway in NY (US again) is famous for this. Traffic would slow to a crawl for no apparent reason.
Ah, do you know where the Internet came from ? It was NOT a commercial venture, it was DARPA that created it (and tried to give it to AT&T who refused). Without that start we would still be using Prodigy or something like AOL. Yes the free market and such is great but if they don't see growth in 2 quarters they are outta there. So to start these things you need the .gov to pave the way (with $ usually).
"Speed Trap" by Frederik Pohl (1967)
This brings back memories, in the late 80's our company was experimenting with GPS. Since at that time there were so few satellites in orbit you had to calculate when you could have 4 in view. This always turned out to be about 3 AM. So there we were cruising down the highways in Western PA at 3AM in a tricked out van full of computers and other things that go beep in the night... Anyway we dreamed of being pulled over by the state cops. "Ok officer let me play that back for you...."
Right now my group and I are working on an embedded system and the development environment REQUIRES you have about 4-6 IDE project windows open at once and you usually have a number of edit windows, debug windows, connection organizer windows, miscellaneous compile status windows. On top of this you have a couple of serial terminal windows, telnet/ftp windows, MSys windows and then a few special purpose editor windows open. Now this is just the dev environment and doesn't count the e-mail,IM system and browser you probably have open. I have two BIG screens but others manage on just one. I also have a PowerBook with two screens used for a simulator development, subversion system, editors for the embedded system's files and stimulator for the system under test.
With the two running side by side OS X wins hands down on handling multiple windows and just the sort of human multitasking done in this sort of system. Unfortunately the embedded IDE only runs on Windows.
I agree with this sub-thread. I've been on both sides of this as a user and as an author. I can tell there is not much more frustrating that wanting to use some piece of code (that does close to what you want) but being prevented from using because you cannot figure out how to change it and the author has no internal or external documentation. In effect you have to do the whole thing again and get inside his head to get to the point of changing it. At that point you usually bail and do your own kludge.
If you want your stuff received well try to document it so others can understand what you have done. There are many ways of doing this but two I have found most useful are (1) Good external documents (it goes in here and comes out here) documenting how to use the code and some examples are really nice. (some of the best I've seen was in libsndfile http://www.mega-nerd.com/libsndfile/ where I've never needed to change the code and external descriptions make it a breeze to use). (2) Internal documentation where you document how it works and have meaningful headers on your routines so the user can see what the input/output variables are and what the routine does. This has worked on code I have put out. I have been pleasantly surprised by users adding things (which I've incorporated) or fixing things.
Also you have to see it as a gift whenever you get a note about the code from someone, even if they want a bug fix. I try not to get burned out on this and try to see it that someone was interested enough to try to use my code even if it doesn't work and they are giving you another chance.
We are on the second iteration of an embedded system (first pass on Linux, second pass on a propriety OS) in each we start a number of applications. These apps are all the same executable but command line options tell the app where to find info on which .so's (DLL's) it loads to give it a personality. the options also tell a particular app its name and information on where it's file system lives. In this case command line parsing is quite important. We use a simple system which locks argc,argv into globals in each app and then we have a function that can find a particular flag and associated value at any point in the code. This way we don't have to parse the command line in main() although some happens there but wherever we need to get a particular flag we can access it. We come down on the minimize code size/RAM size vs efficiency in our choices as none of these flag searches is done more than once and don't have any visible effect on program speed.
I didn't see it in the article but are there any checks on the data supplied by the transponders ? Could you feed in false GPS data and make your aircraft disappear from where it is and put it somewhere else ? Is there crypto in this system or is everything just trusted ?
Unless there is software needs the old OS, there are USB floppies and USB serial ports that work fine in new hardware. We have a bunch of Tek scopes that have floppies for screen captures. The hardware guys have to have a USB floppy so they can put the images into their reports. --jim
And another use I've found is to keep all my PDF's in one place. You can drag them to the library and make a smart folder for type PDF and there they are. Real handy and you can sort by title,author and stuff (you do have put that in yourself, no CDDB for these babies.) Double click and you can view them. Keeps my references right at hand.
You might not have any other Mac's in the building but have you used the Distrubuted builds under XCode ? I have a 12" 1 GHz laptop but when I'm at work the graphics guy lets me steal cycles on Dual G5. Plus they are bringing up a couple more G4's soon so all I have to do is install XCode enable the distrubuted build stuff on them. I've cleared with each of the owners and since they don't see any degredation of performance they just fine with this.
Instead of just one compile at a time I get 3 to 6 at once, with a large build this really helps. You still have the link phase which is only on the build box but the compiles happen a LOT faster.
I think there was a note once about using Linux boxes in the distcc net but it's not quite so dead simple as this.
In OS X dual cores always help because you are running a lot of processes and they get spread over two cores not just one. (In the Activity monitor you can show up to 200% processor utilization with dual cores.)
Project Builder ->File -> New Project -> C++ tool Complle (hammer symbol) run (terminal icon) Hello, World !
Give every kid a copy of the KNOPPIX CD. (http://www.knopper.net/knoppix/index-en.html) They can take it home and run Open Source stuff on the home PC with little risk of messing up Mom's recipes or Dad's tax forms. There is no installation hassle and it has a very good selection of OS stuff. (The people with Mac's at home just need to be told about Fink). --jim