Let us suppose this lawsuit also eats four lawyers and the equivalent of one administrator year, for a total of 50 * 5 (fifty weeks, five days a week) * 40 (forty hours a week) * 5 (five people) = 50,000 thousand man-hours. Now, that may sound like a lot...... but it's only 62.5% of ONE single day for the aforementioned 10,000 person company (of which there are many).
You double counted the conversion from weeks to days, and double stated the thousands on man-hours. This should have been: 50(fifty weeks) * 40(forty hours/week) * 5(five people) = 10,000 man-hours. While this is still more than the 365*16*1=5840 man-hours that you could spend in a year if you spent every waking moment on the case, the difference in man-hours spent on the case is not as substantial as you make it out to be. In the end they probably can spend 5-10 times as much time as you.
However, it's only 12.5% of a single day for them (not 62.5%), while it's every waking hour of a year for you., so your point still stands.
Funny, I was thinking "what else would AIX mean than IBM's flavor of unix." (Although I think I've seen some of these T-shirts and wondered what the heck it stood for. Thanks for letting me know.)
If this were just a semantic argument, then the people who keep saying it's stealing would just call it infringement.
As you noted, calling copyring infringement stealing invariably results in a bunch of "it's not 'stealing'" posts. The response to these posts is always "that's a semantic argument, and they're essentially the same."
My question is, if they're essentially the same, why do people insist on calling it stealing, even though they recognize that it's technically incorrect? The only answer I can come up with is that they're NOT the same, and people who call it stealing are choosing inaccurate language for a reason, probably to evoke an undue emotional response.
So, if you want to get rid of the "it's not 'stealing'" posts, stop calling it stealing, since we both know it isn't.
Optical mark readers are no better than electronic voting with a paper trail.
Just as the electronic voting machines can be modified to misreport votes, so can the optical readers.
The advantage of a semi-closed electronic system, is that it can get to the point where the electronic system should be 100% verifiable, such that a difference of even 1 vote is suspect. This is impossible with optical scanners.
With random (and targeted) hand recounts, it should be possible to verify the quality of the recorded vote.
Assuming this were actually possible, which it probably isn't (trusted client problem)...
If you really consider the stalker case a problem, you include a "Panic button" on every car that calls the police and allows you to go as fast as your car will go. That way the police KNOW that you need their help, they know where you are, and can send a squad car to help you. Remember, if you're being chased by a stalker, you WANT the police to come, so this isn't a problem. Of course this will typically be used when people get stranded on the side of the road, but if the public is already safe because noone's speeding, then we presumably will have a lot of policemen with nothing to do.:)
All other cases: the trick is to allow cars to go 5-10 MPH over the speed limit, at least for a limited time. That way you have speed in bursts when you need it, just not for extended cruising. In some ways, this can actually make passing easier.
A.) He trusts the webmail server's connection to be more free from eavesdropping than his own. This is in a sense true as his mail will just be one among many, instead of all mail sent from his machine being his. This wouldn't be such an issue if most ISPs offered encrypted SMTP.
B.) With webmail your password is never sent cleartext, which is another advantage over the typical non-ssl encrypted imap/pop3 servers. This again is more a statement about lack of encryption at most ISPs than about a webmail technical advantage.
C.) Using webmail allows him to firewall off outbound port 25. This provides some security advantages, espescially in the area of intrusion detection.
While I don't think these marginal security advantages outweigh the benefits of downloadable mail, they are real.
I freely admit that I have no idea what he said. I was responding to the earlier statement about what types of speech the first amendment supposedly doesn't protect, and arguing that it should and in some cases does protect those socially unacceptable types of speech.
<conspiracy mode on> As for this specific case, it looks to me like they were looking to set a precedent that hyperlinking qualifies as disseminating bomb-making information, and found this convenient defendant who they could say "hey, we won't charge you for hacking these web sites if you'll plead guilty to dissemination of bomb-making information." </conspiracy>
Either that or the evidence might not have been enough to convict on the other charges, so they agreed not to try if he would plead guilty to this one.
We should all go drive the lawyers and people with green eyes off a cliff... I sure hate lawyers and people with green eyes.
Lawyers and people with green eyes are dorks, and they're all short too.
The best way to blow up a major government building/person is: a.) With explosives. b.) Where they are. c.) When they are there.
Yup, I can say all of the above, and most would agree that I'm protected by the First Amendment, provided I DO SO PEACEABLY. (The last one is a little questionable, not because it is a government building/official, but because it can go beyond merely advocating criminal behavior to actually planning a specific criminal act.)
The problem is that the govenment has made it such that there are certain groups that I can't say these things about. One group you mentioned was government officials, but there are many others. It just so happens that I CAN say it about lawyers (because they're generally disliked, even by my uncle who IS a lawyer) and about people with green eyes (because everyone can tell it's a joke.)
NOTE: for the duration of this response I use the definition of "swap" used in the windows "swap space" description, and recognize that the term "page" might be more appropriate. As this didn't appear to be a semantics argument, this seems appropriate.
Uhm, you're (deliberately, I hope) misunderstanding the context. Swapping anytime is never, ever an improvement over never swapping. FreeBSD swaps out pages early and makes them available for other uses so it doesn't have to swap when it's under load. That's better than swapping when load hits the system, but far slower than just not swapping, i.e. having enough RAM in the first place.
That's true, if you're willing to accept that the definition of "having enough RAM in the first place" is the sum of the memory requirements of all typically running applications as well as enough memory to store all filesystem data used by those applications. That's probably more than 4GB.
Otherwise, there often is "a little bit" of almost totally unused program data that could efficiently replaced with disk cache. Things like `login` may always be running, but they're almost never actually doing anything on a server as they spend most of their time waiting on input or network connections. During that wait time, a little bit of swap would allow those programs to be moved off of memory to make more efficient use of that space.
The easiest way to look at this is to think of a database server. If you've got a database with thousands of TB of data in it, then there is no way that you can fit the entire database in memory, even with several GB of memory. Sure, adding 256MB of memory is better than adding 256MB of swap, but the cost of adding 256MB of swap to such a system is basically zero, and allows the OS to make efficient use of existing memory resources. Adding more than that probably won't improve performance because most of memory will be disk cache anyways, but why have no swap space and deny the OS of the chance to make such optimizations?
What it comes down to is: moving an active (CPU or disk bound) program to disk to make room for another active program is bad; get more memory in this case. Moving an inactive program to disk to make room for an active program's disk cache is good; this is just using existing resources more efficiently.
I overstate my case somewhat, as there are reliability, development cost, and minor performance tradeoffs to implementing swap space, but saying it is NEVER a benefit is just not true.
I was just looking at roystgnr's reply to your earlier post and it looks like that solution solves both the speed problem and the cyphertext/cleartext problem. This is probably why PCs do it like that.
By encrypting just a random symmetric key you significantly decrease the amount of encrypted data available while at the same time increasing the cost of testing a decryption.
As I said in my other post there's plenty of randomness available in a digital camera to use for creating the random symmetric keys, so this sounds like a good solution.
This is more of a cryptography issue, and is clearly more difficult to deal with. Would adding some random data to the mix help? There's no reason you couldn't add some random bits to the cleartext if that makes cryptoanalysis more difficult. A digital camera certainly has plenty of sources to get randomness from.
Either way, what does it take to not qualify as "weak-to-medium" encryption?
Another option is to have a different encryption key for every camera that is put on the chip during wafer test, and whose pins aren't available on the package. This somewhat weakens the private key distribution security, but makes it so that one camera crack doesn't break all of them.
Having a digital camera changes the way you take pictures.
Since there's essentially no processing cost, you can take pictures EVERY DAY, and keep an album of the good ones. These are the kinds of shots you don't bother to take with a film camera.
I suspect that after a few rentals most people would decide that they want one of their own, so I doubt there's much of a long-term market for this.
I don't know how many people would fit this description (and note the single-chip CCD/encryption option.)
My guess is that the attacker they're really worried about is the one who just wants to use it as a camera of some sort without modifying the hardware much (if at all.) I doubt the market for used CCD chips that have to be pulled from a camera in order to be used is that large.
So you do the encryption between the storage and the USB device.
OR (for lower performance requirements)
Every time a shot is taken you write it to storage un-encrypted. When the camera isn't busy taking shots, it works on encrypting any photos that have been taken but aren't encrypted yet. This way you have to protect the storage or you're still distributing free CCDs, but that's not really the attacker you're worried about anyways.
You don't allow file transfers until a file is finished being encrypted, with an error something like "Please wait while camera finishes processing your photos."
Unless they do the decryption on a USB dongle device that STAYS IN THE STORE and that said clerk would be promptly fired for losing. The decryption device could even be built into the cable that they connect to the camera with.
That would make the cameras a LOT harder to break into.
The other option would be to do the decryption at some central site, but this seems less likely due to transport/bandwith issues.
Not necessarily. This is an obvious candidate for public key encryption.
Encrypt the pictures before you store them, and if you use a good encrytion algorithm, there's not much an attacker can do to reverse engineer the device. They could put a USB connection on the outside and it still wouldn't let an attacker get at the pictures.
All that reverse engineering a well designed one of these devices will give a hacker is either of two things:
A.) A cheap CCD and some optics. This is what happens if it's a two-chip design with the CCD on one chip and everything else outside the CCD chip.
B.) Just some optics. This is what you get if it's a one-chip design with the CCD on the same chip as the encryption circuits. In this case the pictures go in the CCD and come out of the chip encrypted so there's not much the attacker can do.
Either way there may be some other tricks to pull (like overwriting the encryption key), but there's nothing that prevents this from being hard-wired into the device and changed periodically as upgrades come out.
Now, all of this neglects social engineering at the company, which may be the real weak point of attacking these devices. If an insider gives out the private key, then that could compromise all of them.
I don't understand how this is any better than a typical cheap disposable which you can get for about the same price, including developing. 2MP is a little worse than film quality, and all you get is the ability to delete prints (blindly) and the photo CD, which isn't worth much if you're getting prints.
BTW Can any lawyers argue whether conflicts between the first amendment and the copyright clause can be said to favor the first amendment?
Note: IANAL
In general, amendments are considered to overrule all earlier law, but the Bill of Rights is a special case. The Framers were the same ones who wrote the Bill of Rights, and they were in some ways a planned part of the original constitution even though they weren't enacted as such. As a result, the Bill of Rights is generally viewed as on par legally with the original constitution, and when there are conflicts justices usually try and view the work as a whole. This is part of why the first amendment argument was so weak in the Eldred case.
The patent system is anti-competitive BY DESIGN. Can you think of a single product (medical products exempted) that society as a whole cannot do without for 20 years? If the patent holder makes unreasonable requirements for granting a patent license then you just wait 20 years for the patent to expire and THEN build one.
In cases where selling the product is a natural monopoly, then it makes sense that the first developer should be the one to get that monopoly. In other cases, they get monopoly power for 20 years, after which prices will drop and availability will go up dramatically as new manufacturers enter the market. Either way, a product is invented and made available to the public that might not have otherwise been invented. This is how the system is supposed to work, and it IS in the public interest.
The problem is when you have a product (like SOME medical developments) that certain customers quite literally CANNOT do without. AIDS drugs are a good example of this. In these cases the inventor has a captive market and can charge quite literally whatever they want, while the customer doesn't have the option of saying no and walking away or waiting 20 years. In those cases, and only in those cases, does some sort of compulsory licensing make sense. The licensing has to be done in such a way as to allow the inventor to recoup design costs plus a reasonable profit, while making the invention as widely available as possible. As such, compulsory licensing would probably have to be handled on a case-by-case basis.
And, as I said, litigation is a totally different matter, as it DOES stifle legitimate innovation, due mainly to the granting of a lot of obvious patents that are then infringed in ignorance by others.
Our office is currently working with the United States Attorney for the Eastern District of Virginia on the extradition from Australia of Hew Raymond Griffiths, a.k.a. bandido, the former leader of various warez groups, including DrinkOrDie and RiSC. In March 2003, a Federal Grand Jury sitting in the Eastern District of Virginia indicted Griffiths on charges of conspiracy to violate U.S. copyright laws; his extradition is being sought to face these charges. This is the first extradition of a foreign national for online copyright piracy.
Although I don't know all the details of this case, this description seems like the case is setting a disturbing precedent. I don't like the idea that I can break another country's laws while sitting in my own living room, and that that country can then have me extradited from where I live, without my ever setting foot on their soil. This sounds a lot like the Skylarov case, only in this one they're extraditing him without his ever coming to the US.
It's one thing if the US attorney's office says "hey, we noticed that this guy is committing crimes in your country, and you might want to prosecute. Here's some evidence." It's quite another to say "this citizen of yours is doing what is considered a crime in our country, and we want you to hand him over to us for prosecution."
While I'd agree that patents are a good idea, there are particular problems with patent system as it works now, especially in the USA.
* It is too easy for someone to get a "bad" patent; i.e. one that is overly broad, for an obvious "invention", or for an "invention" for which prior art exists. You could blame USPTO for this, but in reality they are just trying to operate within their legal and budgetary constraints.
* It is too hard to get a "bad" patent overturned. You have to resort to the Court system, which is far too expensive and time consuming for most SMOs or individuals to contemplate.
Agreed
* Once a patent has been granted, the patent holder is in a position to demand license fees that are out of all proportion to the true R&D cost of developing the invention.
There's no particular reason why license fees should be in proportion to R&D costs. They should simply be based on the value of the patent, which is best determined by what an informed buyer is willing to pay. The problem here is when a competitor infringes a patent unknowingly, and when called on it is forced to pay exhorbitant fees to the patent holder. To me, not knowing (in one form or another) should be a valid defense, with reasonable expectations for when to pull products from the market (or set license fees) once you DO know.
Here are some possible fixes:
* A fast, zero-cost system for getting dubious patents reviewed and (if appropriate) overturned. The process should be triggered by complaints from the general public; e.g. based on reports of prior art, etc. Questions of "obviousness" should be assessed by independent panels of domain experts, rather than run-of-the-mill patent officers.
I think the patent office is already doing this somewhat by opening patent applications up for external review.
* Financial or other penalties for patent applicants who fail to mention prior art, or otherwise try to rort the patent system. Also sanctions for patent clerks who approve patents that are later overturned.
Reluctantly agreed on the first part (really repeat offenders you're after here.) I disagree on the second part. That could just push things too far in the other direction.
* Patent lifetimes that match the timescales of the industry in question.
This is very hard to quantify. RSA was in the computer field, which arguably should have made it a very short-lived patent. However, it was also EXTREMELY innovative, so the 17-20 year timespan seems reasonable. Timespan really isn't the problem with patent law like it is with copyrights. 20 years is a reasonable timeframe for any truly non-obvious patent.
* A legal cap on patent license fees (or damages) which is based on the audited R&D cost for the invention... NOT the "loss of earnings" due to having a competitor.
I disagree completely on license fees. When the two parties enter into an agreement in good faith, it tends to be an equitable one. This way the person wishing to use the patent can decide whether to use it or not in their product, based on the fees that the patent holder is willing to license for.
On damages, the best thing to do is probably to allow ignorance in some form to be a valid defense, with a reasonable timeframe to remove products from the market once the violation has been discovered. If implemented correctly (I.E. it doesn't work to remain intentionally ignorant) this could be a good way of handling these types of problems.
Re:Interesting technology
on
RFID Explained
·
· Score: 1
Like I said:
1.) The w*m data SHOULD require a warrant. This doesn't mean the police can't get it, it's just harder.
2.) I've heard mixed statements on whether or not they will disable rfid tags at the store. If rfid tags are disabled at the store, then there aren't any long-term privacy issues. This is what many privacy advocates are pushing for, and they don't feel that they've got assurrance that they will be disabled by all retailers yet.
Re:Interesting technology
on
RFID Explained
·
· Score: 1
No, they can't prove that "Joe" was at the KKK(choose your undesirable group) rally. What they can prove is:
A.) Device 1234567... (a pair of blue jeans) was purchased with Joe's credit card on whatever date.
B.) The blue jeans passed through the toll booth between his house and the KKK rally at 3pm on Tuesday. He was carrying, among other things, the rfid tag from a bolt of white cloth (Device 2345678...) and a $20 bill (Device 3456789...).
C.) Joe's blue jeans and his $20 bill entered the Wal-mart 20 minutes from the KKK rally (Wal-mart tracks goods entering and goods exiting, to prevent theft) at 6pm on Wednesday. The blue jeans and a piece of lumber (Device 4567890...) purchased with the aforementioned $20 bill left at 7:30pm.
D.) Joe's blue jeans and the bolt of white cloth passed back through the toll booth at 1pm the following Monday.
None of this PROVES Joe was at the KKK rally, but circumstatial evidence adds up rapidly. Much more information than that above would be possible if rfid tags are in most consumer products. Some of this data would (hopefully) take a warrant to get. It's even better if it isn't available in the first place.
This isn't to say I'm against rfid tags. I just think they need to be used wisely. Putting them in an inconspicuous place in the packaging (or disabling them at the store) is the best way to do that.
Slashdot Mirror
However, it's only 12.5% of a single day for them (not 62.5%), while it's every waking hour of a year for you., so your point still stands.
Funny, I was thinking "what else would AIX mean than IBM's flavor of unix." (Although I think I've seen some of these T-shirts and wondered what the heck it stood for. Thanks for letting me know.)
If this were just a semantic argument, then the people who keep saying it's stealing would just call it infringement.
As you noted, calling copyring infringement stealing invariably results in a bunch of "it's not 'stealing'" posts. The response to these posts is always "that's a semantic argument, and they're essentially the same."
My question is, if they're essentially the same, why do people insist on calling it stealing, even though they recognize that it's technically incorrect? The only answer I can come up with is that they're NOT the same, and people who call it stealing are choosing inaccurate language for a reason, probably to evoke an undue emotional response.
So, if you want to get rid of the "it's not 'stealing'" posts, stop calling it stealing, since we both know it isn't.
It's the staircase closest to Doherty, on the side of the building that faces Porter Hall.
Optical mark readers are no better than electronic voting with a paper trail.
Just as the electronic voting machines can be modified to misreport votes, so can the optical readers.
The advantage of a semi-closed electronic system, is that it can get to the point where the electronic system should be 100% verifiable, such that a difference of even 1 vote is suspect. This is impossible with optical scanners.
With random (and targeted) hand recounts, it should be possible to verify the quality of the recorded vote.
Assuming this were actually possible, which it probably isn't (trusted client problem)...
:)
If you really consider the stalker case a problem, you include a "Panic button" on every car that calls the police and allows you to go as fast as your car will go. That way the police KNOW that you need their help, they know where you are, and can send a squad car to help you. Remember, if you're being chased by a stalker, you WANT the police to come, so this isn't a problem. Of course this will typically be used when people get stranded on the side of the road, but if the public is already safe because noone's speeding, then we presumably will have a lot of policemen with nothing to do.
All other cases: the trick is to allow cars to go 5-10 MPH over the speed limit, at least for a limited time. That way you have speed in bursts when you need it, just not for extended cruising. In some ways, this can actually make passing easier.
What it sounds like he's trying to say is that:
A.) He trusts the webmail server's connection to be more free from eavesdropping than his own. This is in a sense true as his mail will just be one among many, instead of all mail sent from his machine being his. This wouldn't be such an issue if most ISPs offered encrypted SMTP.
B.) With webmail your password is never sent cleartext, which is another advantage over the typical non-ssl encrypted imap/pop3 servers. This again is more a statement about lack of encryption at most ISPs than about a webmail technical advantage.
C.) Using webmail allows him to firewall off outbound port 25. This provides some security advantages, espescially in the area of intrusion detection.
While I don't think these marginal security advantages outweigh the benefits of downloadable mail, they are real.
I freely admit that I have no idea what he said. I was responding to the earlier statement about what types of speech the first amendment supposedly doesn't protect, and arguing that it should and in some cases does protect those socially unacceptable types of speech.
<conspiracy mode on>
As for this specific case, it looks to me like they were looking to set a precedent that hyperlinking qualifies as disseminating bomb-making information, and found this convenient defendant who they could say "hey, we won't charge you for hacking these web sites if you'll plead guilty to dissemination of bomb-making information."
</conspiracy>
Either that or the evidence might not have been enough to convict on the other charges, so they agreed not to try if he would plead guilty to this one.
Hmmmm,
So I can't say:
We should all go drive the lawyers and people with green eyes off a cliff... I sure hate lawyers and people with green eyes.
Lawyers and people with green eyes are dorks, and they're all short too.
The best way to blow up a major government building/person is:
a.) With explosives.
b.) Where they are.
c.) When they are there.
Yup, I can say all of the above, and most would agree that I'm protected by the First Amendment, provided I DO SO PEACEABLY. (The last one is a little questionable, not because it is a government building/official, but because it can go beyond merely advocating criminal behavior to actually planning a specific criminal act.)
The problem is that the govenment has made it such that there are certain groups that I can't say these things about. One group you mentioned was government officials, but there are many others. It just so happens that I CAN say it about lawyers (because they're generally disliked, even by my uncle who IS a lawyer) and about people with green eyes (because everyone can tell it's a joke.)
That's true, if you're willing to accept that the definition of "having enough RAM in the first place" is the sum of the memory requirements of all typically running applications as well as enough memory to store all filesystem data used by those applications. That's probably more than 4GB.
Otherwise, there often is "a little bit" of almost totally unused program data that could efficiently replaced with disk cache. Things like `login` may always be running, but they're almost never actually doing anything on a server as they spend most of their time waiting on input or network connections. During that wait time, a little bit of swap would allow those programs to be moved off of memory to make more efficient use of that space.
The easiest way to look at this is to think of a database server. If you've got a database with thousands of TB of data in it, then there is no way that you can fit the entire database in memory, even with several GB of memory. Sure, adding 256MB of memory is better than adding 256MB of swap, but the cost of adding 256MB of swap to such a system is basically zero, and allows the OS to make efficient use of existing memory resources. Adding more than that probably won't improve performance because most of memory will be disk cache anyways, but why have no swap space and deny the OS of the chance to make such optimizations?
What it comes down to is: moving an active (CPU or disk bound) program to disk to make room for another active program is bad; get more memory in this case. Moving an inactive program to disk to make room for an active program's disk cache is good; this is just using existing resources more efficiently.
I overstate my case somewhat, as there are reliability, development cost, and minor performance tradeoffs to implementing swap space, but saying it is NEVER a benefit is just not true.
I was just looking at roystgnr's reply to your earlier post and it looks like that solution solves both the speed problem and the cyphertext/cleartext problem. This is probably why PCs do it like that.
By encrypting just a random symmetric key you significantly decrease the amount of encrypted data available while at the same time increasing the cost of testing a decryption.
As I said in my other post there's plenty of randomness available in a digital camera to use for creating the random symmetric keys, so this sounds like a good solution.
This is more of a cryptography issue, and is clearly more difficult to deal with. Would adding some random data to the mix help? There's no reason you couldn't add some random bits to the cleartext if that makes cryptoanalysis more difficult. A digital camera certainly has plenty of sources to get randomness from.
Either way, what does it take to not qualify as "weak-to-medium" encryption?
Another option is to have a different encryption key for every camera that is put on the chip during wafer test, and whose pins aren't available on the package. This somewhat weakens the private key distribution security, but makes it so that one camera crack doesn't break all of them.
Having a digital camera changes the way you take pictures.
Since there's essentially no processing cost, you can take pictures EVERY DAY, and keep an album of the good ones. These are the kinds of shots you don't bother to take with a film camera.
I suspect that after a few rentals most people would decide that they want one of their own, so I doubt there's much of a long-term market for this.
I don't know how many people would fit this description (and note the single-chip CCD/encryption option.)
My guess is that the attacker they're really worried about is the one who just wants to use it as a camera of some sort without modifying the hardware much (if at all.) I doubt the market for used CCD chips that have to be pulled from a camera in order to be used is that large.
So you do the encryption between the storage and the USB device.
OR (for lower performance requirements)
Every time a shot is taken you write it to storage un-encrypted. When the camera isn't busy taking shots, it works on encrypting any photos that have been taken but aren't encrypted yet. This way you have to protect the storage or you're still distributing free CCDs, but that's not really the attacker you're worried about anyways.
You don't allow file transfers until a file is finished being encrypted, with an error something like "Please wait while camera finishes processing your photos."
Unless they do the decryption on a USB dongle device that STAYS IN THE STORE and that said clerk would be promptly fired for losing. The decryption device could even be built into the cable that they connect to the camera with.
That would make the cameras a LOT harder to break into.
The other option would be to do the decryption at some central site, but this seems less likely due to transport/bandwith issues.
Not necessarily. This is an obvious candidate for public key encryption.
Encrypt the pictures before you store them, and if you use a good encrytion algorithm, there's not much an attacker can do to reverse engineer the device. They could put a USB connection on the outside and it still wouldn't let an attacker get at the pictures.
All that reverse engineering a well designed one of these devices will give a hacker is either of two things:
A.) A cheap CCD and some optics. This is what happens if it's a two-chip design with the CCD on one chip and everything else outside the CCD chip.
B.) Just some optics. This is what you get if it's a one-chip design with the CCD on the same chip as the encryption circuits. In this case the pictures go in the CCD and come out of the chip encrypted so there's not much the attacker can do.
Either way there may be some other tricks to pull (like overwriting the encryption key), but there's nothing that prevents this from being hard-wired into the device and changed periodically as upgrades come out.
Now, all of this neglects social engineering at the company, which may be the real weak point of attacking these devices. If an insider gives out the private key, then that could compromise all of them.
I don't understand how this is any better than a typical cheap disposable which you can get for about the same price, including developing. 2MP is a little worse than film quality, and all you get is the ability to delete prints (blindly) and the photo CD, which isn't worth much if you're getting prints.
(Not that I'm actually expressing an opinion here, this is just a hypothetical.)
Note: IANAL
In general, amendments are considered to overrule all earlier law, but the Bill of Rights is a special case. The Framers were the same ones who wrote the Bill of Rights, and they were in some ways a planned part of the original constitution even though they weren't enacted as such. As a result, the Bill of Rights is generally viewed as on par legally with the original constitution, and when there are conflicts justices usually try and view the work as a whole. This is part of why the first amendment argument was so weak in the Eldred case.
The patent system is anti-competitive BY DESIGN. Can you think of a single product (medical products exempted) that society as a whole cannot do without for 20 years? If the patent holder makes unreasonable requirements for granting a patent license then you just wait 20 years for the patent to expire and THEN build one.
In cases where selling the product is a natural monopoly, then it makes sense that the first developer should be the one to get that monopoly. In other cases, they get monopoly power for 20 years, after which prices will drop and availability will go up dramatically as new manufacturers enter the market. Either way, a product is invented and made available to the public that might not have otherwise been invented. This is how the system is supposed to work, and it IS in the public interest.
The problem is when you have a product (like SOME medical developments) that certain customers quite literally CANNOT do without. AIDS drugs are a good example of this. In these cases the inventor has a captive market and can charge quite literally whatever they want, while the customer doesn't have the option of saying no and walking away or waiting 20 years. In those cases, and only in those cases, does some sort of compulsory licensing make sense. The licensing has to be done in such a way as to allow the inventor to recoup design costs plus a reasonable profit, while making the invention as widely available as possible. As such, compulsory licensing would probably have to be handled on a case-by-case basis.
And, as I said, litigation is a totally different matter, as it DOES stifle legitimate innovation, due mainly to the granting of a lot of obvious patents that are then infringed in ignorance by others.
It's one thing if the US attorney's office says "hey, we noticed that this guy is committing crimes in your country, and you might want to prosecute. Here's some evidence." It's quite another to say "this citizen of yours is doing what is considered a crime in our country, and we want you to hand him over to us for prosecution."
On damages, the best thing to do is probably to allow ignorance in some form to be a valid defense, with a reasonable timeframe to remove products from the market once the violation has been discovered. If implemented correctly (I.E. it doesn't work to remain intentionally ignorant) this could be a good way of handling these types of problems.
Like I said:
1.) The w*m data SHOULD require a warrant. This doesn't mean the police can't get it, it's just harder.
2.) I've heard mixed statements on whether or not they will disable rfid tags at the store. If rfid tags are disabled at the store, then there aren't any long-term privacy issues. This is what many privacy advocates are pushing for, and they don't feel that they've got assurrance that they will be disabled by all retailers yet.
No, they can't prove that "Joe" was at the KKK(choose your undesirable group) rally. What they can prove is:
A.) Device 1234567... (a pair of blue jeans) was purchased with Joe's credit card on whatever date.
B.) The blue jeans passed through the toll booth between his house and the KKK rally at 3pm on Tuesday. He was carrying, among other things, the rfid tag from a bolt of white cloth (Device 2345678...) and a $20 bill (Device 3456789...).
C.) Joe's blue jeans and his $20 bill entered the Wal-mart 20 minutes from the KKK rally (Wal-mart tracks goods entering and goods exiting, to prevent theft) at 6pm on Wednesday. The blue jeans and a piece of lumber (Device 4567890...) purchased with the aforementioned $20 bill left at 7:30pm.
D.) Joe's blue jeans and the bolt of white cloth passed back through the toll booth at 1pm the following Monday.
None of this PROVES Joe was at the KKK rally, but circumstatial evidence adds up rapidly. Much more information than that above would be possible if rfid tags are in most consumer products. Some of this data would (hopefully) take a warrant to get. It's even better if it isn't available in the first place.
This isn't to say I'm against rfid tags. I just think they need to be used wisely. Putting them in an inconspicuous place in the packaging (or disabling them at the store) is the best way to do that.