The wavelength is 532nm (about 563.52 THz). There is little variance between individual devices. A narrowband filter could do the job. But such filters can be expensive over a large surface, tend to not be 100% blocking, and won't address a laser based on a different technology that would be at a different wavelength). But the cool looking bright green ones are 532nm.
Thousands of businesses outsource their IT security every day. Lots of it goes overseas, too. And the best part of it is that it's free. The bad part is they don't know they are outsourcing it at all.
If you wipe your disk after being notified you are being sued in a matter related to what is on your disk, then you can certainly get into trouble. I don't know if a judge would accept the excuse "My computer was running that insecure operating system from the company in Redmond, and needs to be reinstalled every few week or months just to keep working, and it is necessary to wipe off every file to do the reinstall cleanly. Having to retain these files means a cost to me in terms of additional storage space that I simply cannot afford. But neither can I simply quit using my computer since it is my source of income."
Every time your audio level goes down by 6 db, you lose a bit. Once you are down by 48 db, you only have 8 bits to resolve the waveform. With 24 bits, you'd still have 16 bits left working for you at -48db.
A higher sampling rate also gives you are more accurate sound, especially when the waveform is complex. Sure, a sampling rate of X can give you a frequency response to X/2. But when you have many frequencies in the range X/F where F is less than 10, they can easily get jumbled together and result in distortions to each other. If you had an infinite number of bits that would not happen. But with both a finite sample rate and a finite number of bits, samples get pushed up and down due to this intermodulation aliasing. All the frequencies are there. But some extra ones will be as well (products of the intermodulation).
I've never actually heard 24/96. But I have compared 16/44.1 to the original analog it was encoded from, and I can most definitely hear the difference, despite professional recording and encoding involved. I can't say if 24/96 would completely fix it. But there is a lot there to be fixed. I've proposed 32/192 myself.
The fundamental problem as to why there is so little drive to make a big switch to IPv6 is because what IPv6 offers... and this is important... over IPv4 is relatively small compared to what IPv4 offered over its predecessor, which was essentially going from no internet at all to having what we have today. There needs to be some kind of real motivating force to make it happen. IPv4 happened because having an internet was a motivating force. What does IPv6 offer? Very little as long as we still get IPv4 addresses. Other kinds of motivations are also possible. Take a look at how much the over-the-air TV broadcasters dragged their feet in deploying digital transmission at full capacity. Now we have a pretty solid analog shutoff date, so they better get those digital transmitters going (most have, to at least some degree, now). The biggest encouragement to getting IPv6 rolling is to schedule a definite, but very doable, IPv4 cutoff date for at least some critical piece of the net most people want. But we have to choose what that is. Access to the government? Access to routers going across national borders? Access to porn?
Does a thermocouple have a tube filled with slightly positive pressure gas, and perform a measurement of something that can be measured in terms of a mechanical movement, by means of sensing the changed position by acoustic impuses, and analyze the reflected waveform much like a time domain reflectomoter, but with a direct calculation of the sensor measurement from the impulse timing? If yes, that's not the kind of thermocouple I've ever heard of.
Maybe you were thinking of a bi-metal strip, such as found in a thermostat, that would bend according to the temperature. That would be a component of what I suggested in cases of measuring temperature.
My younger nephew wants a Wii. But he said he'd take something else instead. He's too young (8) to be in much of a financial position to buy these things on his own. So I must decide if I get him some other game console, or cash.
Yes, this can be a serious problem. If there is any level of shortage of a product (as many of the latest products coming into the holidays are), EBay will have the effect of making the shortage worse. And it's not the manufacturers that profit from it in this case (apparently); it's whoever has the connections to get those bulk deliveries redirected to them (possibly even stolen).
I have the solution to solving the holiday product crunch: spread the holidays out over the calendar. In the USA, divide the country up into 12 regions of about the same population and economics. Then assign each region a different month to have the gift giving holidays. Most people don't celebrate Christmas religiously, anymore, so this shouldn't be much of a problem.
The above does still leave a big crunch at stores and malls within a region. So maybe it's better to divide things up on a micro-scale instead of a macro-scale. So, how about celebrating the gift giving holiday based on (zipcode % 12), where you celebrate gift giving based on your zip code modulo 12 to choose the month.
This still means a big crunch for families and neighbors in the same zip code. So I have a better idea. Let's use the date of birth to determine when to celebrate the gift giving holiday, based on who the gift is for. And instead of having it all on one day of the month, let's spread it out further and use the actual date in the date of birth for everyone's own personalized gift giving holiday.
Ooops. I didn't take into account February 29. Never mind.
A small tube is lightly pressurized with a known gas. A mechanical sensor at the far end moves a tiny piston in or out of the tube to measure fuel level or temperature. At the near end, a device emits an acoustic pulse into the gas and measures the return reflection timing. This timing gives the length the piston has moved in the tube. The tube can be made of metal (well grounded to the tank frame at many points) or other reasonably rigid materials.
One tube can even be used for multiple sensors. This would be sorted out by the timings of multiple pulses returned from the various sensors at different distances along the tube. Some of those sensors measure temperature of the gas in the tube itself to maintain calibration. A low angle tube splitter can be used to greatly minimize the impulse reflections between sensors (which would otherwise appears as ghost sensors to the firmware scanning the pulse train, which can easily eliminate them at low levels).
If this has never been done before, this posting hereby constitutes public disclosure of the idea on Monday, 17 December, 2007.
So we need to be running more encryption. IPsec might be the simplest to start with. And web sites can go with HTTPS (HTTP redirects to the HTTPS). Let them dream of filtering that. It will be hard enough (and horrendously expensive) to have ISPs do the filtering on all TCP connections. While limited encryption (e.g. no pre-shared key or PKC to authenticate the peer) would be vulnerable to MitM attacks, it would be many times more expensive for ISPs to deploy that kind of filtering.
Do not construe this to mean I support copyright infringement as I do not. But I sure as hell do not want my non-infringing traffic to be messed with in any way, nor be expected to pay for it in the form of higher ISP costs, just because a bunch of incompetent buffoons, and their stupid congressman, in Hollywood, want to keep a lame and doomed business model going long after its practical end of life.
My dad was always having trouble understanding what I would tell him was wrong, or what to do to fix it. Eventually I gave him one simple solution, which was easy for him to understand and do, which he does all the time, now: stay away from all computers!
You still have to do the key exchange and authentication, though, so all you save is the low level encryption of the content stream beyond the checksum. For large content, that might be a big gain. For small content, though, it's not that much of a gain.
Doesn't Microsoft want to convert Linux users over to Windows? Sure, this would need a different kind of program. But still, monitoring what you do in Linux could give them some insight into what what Linux users do and maybe they can make their next products something people would want to stay with Windows for.
Web sites need to enable HTTPS properly over their entire site. Then your ISP can do nothing more than just prevent the secure connection from being established. And if they do that, they break all kinds of stuff like shopping checkout and access to bank accounts.
Right now, Slashdot's own HTTPS URL just redirects to the HTTP URL. This needs to be changed to just leave things in the HTTPS mode. Eventually this should be changed so that HTTP redirects to HTTPS. Google does the same boneheaded redirection.
Tell me how it is that you can check the bar codes on the sponges that are left behind in the patient, especially after the patient is closed up. At least RFID adds this check which can be done afterwards. You just make sure the chips are working both during the operation as new items are being selected for use, as well as during the preparation for the surgery in advance. Any device not working must be discarded and statistics recorded. Any manufacturer with more than 0.1% failure rate is to be fined.
When it comes to medicine and surgical operations, we expect a level of perfection beyond the level we consider needed for computer software and hardware (except for that used for direct medical care). And I believe your counting method for surgical sponges just doesn't fly well enough. That is because these mistakes still happen with so many hospitals using such a procedure. Bar code scanning I believe would not be any better. Either way, someone could enter the operation process with a failure to be counted or scanned. Then you can do all the post-operation checks, and something could still be left behind. At least with the RFID, the patient can be scanned to see if anything responds, and it would identify what it is. You would then just need to do the RFID scan on things going into the procedure if for nothing else but to be sure the RFID chip is functional (if it fails in pre-scan, discard it and get another).
You might be more qualified in medicine. But this is not a true medical issue. This is an issue of logistics that happens to be connected to medicine in a very crucial way. If you medical guys can't get it solved on your own (and you have not done so, yet, because this is an area where the goal of zero errors is more than just the direction to aim, it is the destination that must be reached), then maybe we do need to step in and solve it for you. Things like RFID do look promising. IMHO, any hospital with a history of having left anything in any patient in the past 5 years should be required to do an X-ray and/or other appropriate scan of every patient to check for leftovers at no cost to the patient.
I have an older Seagate 400G drive from the previous "Pushbutton" series. It works fine. When it spins down due to idleness, it remains logically connected via USB. At the very next operation, the drive spins back up. That takes about 6.9 seconds. That I/O request that spins it back up then completes and a response is sent back. Everything works fine if you are not annoyed by that occaisional delay after idleness.
My point: someone (possibly formerly) at Seagate does know how to implement a USB connected drive and spin down logic correctly. But the latest drives do not do this properly. I bought one of those Free Agent portable drives. I'm guessing it was intended for laptop users as it does the spin down rather quickly, in about 20 seconds or so (it seems to be inconsistent... perhaps a peridic check to see if the last operation was long enough ago). That makes most of the workarounds proposed for the Free Agent series rather "abusive" since they would have to run at least every 10 seconds.
I have 4 Western Digital 500GB MyBook drives (USB only) and 1 Western Digital 1TB MyBook drive (USB, Firewire, eSATA), and they all work fine.
So the question comes down to simply why Seagate is doing this: Is it because development people are incompetent and implement USB incorrectly, or did Microsoft pay them off to change the design to one that would only work with the supplied Windows-only software.
Seagate now owns Maxtor. Maxtor's older USB drives work fine. Being quite satisfied with the WD drives I have, I have not considered buying the new Maxtor drives. But I would be quite hesitant to do so given that they, too, may have been compromised by Seagate's incompetence.
Use IPsec. Not only can they not tell what your packets mean (only where they are going and came from), but they cannot forge an RST since that also needs to be encrypted with the association key.
So they could do a man-in-the-middle attack on a simplistic key exchange done over IPsec. But that would require far more resources (they have to get in the middle of each connection) than they appear to be willing to use (RST forgery is about the cheapest form of net interference there is). So I think even minimal IPsec would bring this blocking to and end until such time as they want to invest in whatever it takes to mount an attack on IPsec. Then we just use a strong key infrastructure and end that.
If the protocol involved understood the work to be done (e.g. how many bytes to be transferred), it could also re-establish a new connection if the existing one got dropped, and resume the transfer... until done or one end decides to not do this anymore.
This decision is a correct one. Trademark law traditionally has, with few exceptions, protected a name only within a specific market context. Now "Spam" is a name in a new context Hormel apparently had not registered the name in (or conducted any business in). Their lawyer(s) need(s) to go back to school.
Slashdot Mirror
A filter can be made to be quite narrow to the point of people not really noticing it, except for a very slight pink tone to it. But it's not cheap.
The wavelength is 532nm (about 563.52 THz). There is little variance between individual devices. A narrowband filter could do the job. But such filters can be expensive over a large surface, tend to not be 100% blocking, and won't address a laser based on a different technology that would be at a different wavelength). But the cool looking bright green ones are 532nm.
Thousands of businesses outsource their IT security every day. Lots of it goes overseas, too. And the best part of it is that it's free. The bad part is they don't know they are outsourcing it at all.
If you wipe your disk after being notified you are being sued in a matter related to what is on your disk, then you can certainly get into trouble. I don't know if a judge would accept the excuse "My computer was running that insecure operating system from the company in Redmond, and needs to be reinstalled every few week or months just to keep working, and it is necessary to wipe off every file to do the reinstall cleanly. Having to retain these files means a cost to me in terms of additional storage space that I simply cannot afford. But neither can I simply quit using my computer since it is my source of income."
Every time your audio level goes down by 6 db, you lose a bit. Once you are down by 48 db, you only have 8 bits to resolve the waveform. With 24 bits, you'd still have 16 bits left working for you at -48db.
A higher sampling rate also gives you are more accurate sound, especially when the waveform is complex. Sure, a sampling rate of X can give you a frequency response to X/2. But when you have many frequencies in the range X/F where F is less than 10, they can easily get jumbled together and result in distortions to each other. If you had an infinite number of bits that would not happen. But with both a finite sample rate and a finite number of bits, samples get pushed up and down due to this intermodulation aliasing. All the frequencies are there. But some extra ones will be as well (products of the intermodulation).
I've never actually heard 24/96. But I have compared 16/44.1 to the original analog it was encoded from, and I can most definitely hear the difference, despite professional recording and encoding involved. I can't say if 24/96 would completely fix it. But there is a lot there to be fixed. I've proposed 32/192 myself.
Why not use a normal standard format? Then we won't need Flash.
The fundamental problem as to why there is so little drive to make a big switch to IPv6 is because what IPv6 offers ... and this is important ... over IPv4 is relatively small compared to what IPv4 offered over its predecessor, which was essentially going from no internet at all to having what we have today. There needs to be some kind of real motivating force to make it happen. IPv4 happened because having an internet was a motivating force. What does IPv6 offer? Very little as long as we still get IPv4 addresses. Other kinds of motivations are also possible. Take a look at how much the over-the-air TV broadcasters dragged their feet in deploying digital transmission at full capacity. Now we have a pretty solid analog shutoff date, so they better get those digital transmitters going (most have, to at least some degree, now). The biggest encouragement to getting IPv6 rolling is to schedule a definite, but very doable, IPv4 cutoff date for at least some critical piece of the net most people want. But we have to choose what that is. Access to the government? Access to routers going across national borders? Access to porn?
Does a thermocouple have a tube filled with slightly positive pressure gas, and perform a measurement of something that can be measured in terms of a mechanical movement, by means of sensing the changed position by acoustic impuses, and analyze the reflected waveform much like a time domain reflectomoter, but with a direct calculation of the sensor measurement from the impulse timing? If yes, that's not the kind of thermocouple I've ever heard of.
Maybe you were thinking of a bi-metal strip, such as found in a thermostat, that would bend according to the temperature. That would be a component of what I suggested in cases of measuring temperature.
My younger nephew wants a Wii. But he said he'd take something else instead. He's too young (8) to be in much of a financial position to buy these things on his own. So I must decide if I get him some other game console, or cash.
Yes, this can be a serious problem. If there is any level of shortage of a product (as many of the latest products coming into the holidays are), EBay will have the effect of making the shortage worse. And it's not the manufacturers that profit from it in this case (apparently); it's whoever has the connections to get those bulk deliveries redirected to them (possibly even stolen).
I have the solution to solving the holiday product crunch: spread the holidays out over the calendar. In the USA, divide the country up into 12 regions of about the same population and economics. Then assign each region a different month to have the gift giving holidays. Most people don't celebrate Christmas religiously, anymore, so this shouldn't be much of a problem.
The above does still leave a big crunch at stores and malls within a region. So maybe it's better to divide things up on a micro-scale instead of a macro-scale. So, how about celebrating the gift giving holiday based on (zipcode % 12), where you celebrate gift giving based on your zip code modulo 12 to choose the month.
This still means a big crunch for families and neighbors in the same zip code. So I have a better idea. Let's use the date of birth to determine when to celebrate the gift giving holiday, based on who the gift is for. And instead of having it all on one day of the month, let's spread it out further and use the actual date in the date of birth for everyone's own personalized gift giving holiday.
Ooops. I didn't take into account February 29. Never mind.
A small tube is lightly pressurized with a known gas. A mechanical sensor at the far end moves a tiny piston in or out of the tube to measure fuel level or temperature. At the near end, a device emits an acoustic pulse into the gas and measures the return reflection timing. This timing gives the length the piston has moved in the tube. The tube can be made of metal (well grounded to the tank frame at many points) or other reasonably rigid materials.
One tube can even be used for multiple sensors. This would be sorted out by the timings of multiple pulses returned from the various sensors at different distances along the tube. Some of those sensors measure temperature of the gas in the tube itself to maintain calibration. A low angle tube splitter can be used to greatly minimize the impulse reflections between sensors (which would otherwise appears as ghost sensors to the firmware scanning the pulse train, which can easily eliminate them at low levels).
If this has never been done before, this posting hereby constitutes public disclosure of the idea on Monday, 17 December, 2007.
And there's nothing stopping Microsoft from using an open standard ... except the guy wielding the chair.
... getting a DCMA takedown notice from aliens because of the copying of their space signals over the internet.
Yes, the satire runneth over.
So we need to be running more encryption. IPsec might be the simplest to start with. And web sites can go with HTTPS (HTTP redirects to the HTTPS). Let them dream of filtering that. It will be hard enough (and horrendously expensive) to have ISPs do the filtering on all TCP connections. While limited encryption (e.g. no pre-shared key or PKC to authenticate the peer) would be vulnerable to MitM attacks, it would be many times more expensive for ISPs to deploy that kind of filtering.
Do not construe this to mean I support copyright infringement as I do not. But I sure as hell do not want my non-infringing traffic to be messed with in any way, nor be expected to pay for it in the form of higher ISP costs, just because a bunch of incompetent buffoons, and their stupid congressman, in Hollywood, want to keep a lame and doomed business model going long after its practical end of life.
My dad was always having trouble understanding what I would tell him was wrong, or what to do to fix it. Eventually I gave him one simple solution, which was easy for him to understand and do, which he does all the time, now: stay away from all computers!
You still have to do the key exchange and authentication, though, so all you save is the low level encryption of the content stream beyond the checksum. For large content, that might be a big gain. For small content, though, it's not that much of a gain.
Doesn't Microsoft want to convert Linux users over to Windows? Sure, this would need a different kind of program. But still, monitoring what you do in Linux could give them some insight into what what Linux users do and maybe they can make their next products something people would want to stay with Windows for.
Web sites need to enable HTTPS properly over their entire site. Then your ISP can do nothing more than just prevent the secure connection from being established. And if they do that, they break all kinds of stuff like shopping checkout and access to bank accounts.
Right now, Slashdot's own HTTPS URL just redirects to the HTTP URL. This needs to be changed to just leave things in the HTTPS mode. Eventually this should be changed so that HTTP redirects to HTTPS. Google does the same boneheaded redirection.
Now can the dogs determine the gender of the other dog ... without having to resort to sniffing the other dog's butt?
Tell me how it is that you can check the bar codes on the sponges that are left behind in the patient, especially after the patient is closed up. At least RFID adds this check which can be done afterwards. You just make sure the chips are working both during the operation as new items are being selected for use, as well as during the preparation for the surgery in advance. Any device not working must be discarded and statistics recorded. Any manufacturer with more than 0.1% failure rate is to be fined.
When it comes to medicine and surgical operations, we expect a level of perfection beyond the level we consider needed for computer software and hardware (except for that used for direct medical care). And I believe your counting method for surgical sponges just doesn't fly well enough. That is because these mistakes still happen with so many hospitals using such a procedure. Bar code scanning I believe would not be any better. Either way, someone could enter the operation process with a failure to be counted or scanned. Then you can do all the post-operation checks, and something could still be left behind. At least with the RFID, the patient can be scanned to see if anything responds, and it would identify what it is. You would then just need to do the RFID scan on things going into the procedure if for nothing else but to be sure the RFID chip is functional (if it fails in pre-scan, discard it and get another).
You might be more qualified in medicine. But this is not a true medical issue. This is an issue of logistics that happens to be connected to medicine in a very crucial way. If you medical guys can't get it solved on your own (and you have not done so, yet, because this is an area where the goal of zero errors is more than just the direction to aim, it is the destination that must be reached), then maybe we do need to step in and solve it for you. Things like RFID do look promising. IMHO, any hospital with a history of having left anything in any patient in the past 5 years should be required to do an X-ray and/or other appropriate scan of every patient to check for leftovers at no cost to the patient.
I have an older Seagate 400G drive from the previous "Pushbutton" series. It works fine. When it spins down due to idleness, it remains logically connected via USB. At the very next operation, the drive spins back up. That takes about 6.9 seconds. That I/O request that spins it back up then completes and a response is sent back. Everything works fine if you are not annoyed by that occaisional delay after idleness.
My point: someone (possibly formerly) at Seagate does know how to implement a USB connected drive and spin down logic correctly. But the latest drives do not do this properly. I bought one of those Free Agent portable drives. I'm guessing it was intended for laptop users as it does the spin down rather quickly, in about 20 seconds or so (it seems to be inconsistent ... perhaps a peridic check to see if the last operation was long enough ago). That makes most of the workarounds proposed for the Free Agent series rather "abusive" since they would have to run at least every 10 seconds.
I have 4 Western Digital 500GB MyBook drives (USB only) and 1 Western Digital 1TB MyBook drive (USB, Firewire, eSATA), and they all work fine.
So the question comes down to simply why Seagate is doing this: Is it because development people are incompetent and implement USB incorrectly, or did Microsoft pay them off to change the design to one that would only work with the supplied Windows-only software.
Seagate now owns Maxtor. Maxtor's older USB drives work fine. Being quite satisfied with the WD drives I have, I have not considered buying the new Maxtor drives. But I would be quite hesitant to do so given that they, too, may have been compromised by Seagate's incompetence.
Use IPsec. Not only can they not tell what your packets mean (only where they are going and came from), but they cannot forge an RST since that also needs to be encrypted with the association key.
So they could do a man-in-the-middle attack on a simplistic key exchange done over IPsec. But that would require far more resources (they have to get in the middle of each connection) than they appear to be willing to use (RST forgery is about the cheapest form of net interference there is). So I think even minimal IPsec would bring this blocking to and end until such time as they want to invest in whatever it takes to mount an attack on IPsec. Then we just use a strong key infrastructure and end that.
If the protocol involved understood the work to be done (e.g. how many bytes to be transferred), it could also re-establish a new connection if the existing one got dropped, and resume the transfer ... until done or one end decides to not do this anymore.
This decision is a correct one. Trademark law traditionally has, with few exceptions, protected a name only within a specific market context. Now "Spam" is a name in a new context Hormel apparently had not registered the name in (or conducted any business in). Their lawyer(s) need(s) to go back to school.