I shudder to think what would happen to the MSC if the Chief of Police and Colonel March had "all of the lessons from Columbine" at their disposal. Back then, "boys will be boys" was an actual legal dictum that lawyers could offer to judges and actually get their clients off. These guys were using pyrotechnics in several so-called capers.(They blew up the monster in Strawberry Lake at the end.) They seemed dangerously interested in military surplus. Hacking the radio frequencies was second nature to them. All of these actions are dangerous predictors of future Bad People.
I wish the book publishers would start reprinting more books about good children who sit still and devote themselves to watching Disney cartoons. If kids must get off of the couch, they might devote themselves to collecting Disney beanbag dolls or maybe those plastic action figures for Disney characters.
I agree with everything the review said, but I have some reservations about the book. The most depressing part for me is the lack of appreciation shown for the other contributors. There's only a short mention of Stallman and little about the other contributors to the movement and the kernel. You might get the impression that Torvalds and Torvalds alone was responsible for everything that is Linux. Is the name Alan Cox found anywhere in the book? I don't remember seeing it.
At one point he even seems to confuse the Netcraft surveys of the webservers with a measure of penetration of Linux. The Netcraft surveys measure Apache, a server that also runs on FreeBSD, AIX, MacOS, and dozens of other OSs.
For this reason I think both Rebel Code and Free for All do a better job with the history. The only really novel part here are the stories about growing up in Finland and spending his life coding in a closet. Those parts are great fun. He's a charming guy. The book's great. But I wish he could spread the credit around a bit more.
What's wrong with personal black books.
on
IETF vs. ICANN
·
· Score: 1
I have a black box filled with my own personal links between names and phone numbers. What's wrong with having my own list of matchups between names and IP numbers? ANd what's wrong with circulating this list in DNS format? Ooops. That's stepping on ICANN's so called monopoly.
I say it's a first amendment issue. Network solutions does a good job, but I would love a backup directory service.
I've got a NAT router hooked up to an Earthlink account. The connection gets disconnected every so often and then, voila, a new dialup generates a new IP number.
The animated series rocked, but the buzz about the live-action version isn't so hot. They nuked one of the writers who helped make the animated version everything it ever was. The word is that they want it to be more "Seinfeld with Capes" than "The Tick". Well, that's the price you pay when you switch to film, a medium where creating an explosion is significantly more expensive than creating a wink. Spoon!
Here's the basic solution. Put the music through a narrow bandpass filter. That is, remove all frequencies except those in a narrow band, say 800Hz to 1000Hz. It probably makes more sense to locate this band in the lower octaves.
Then, go through the song looking for the presence or absence of a signal.
Turn this into a string of 1's and 0's where a 1 means that there's some noise with a frequency in the range and 0 if there's only a neglible amount.
Then do string matching. You might need to slide this back and forth a bit to find the best match, but that's not too complicated.
I've really wanted to get a Ricochet modem, but the price is too steep for me. $70/month. I don't travel much for work so it doesn't help me. If there was some travel involved, then I would be signing up immediately. If they bring the price down, I'm there.
They run a network that hides your browsing and mail reading. It's pretty slick. Of course you have to pay to support it, but the price isn't bad. It's something like $50/year.
Who's going to edit the web? Is he volunteering? I think the Founding Fathers passed the First Amendment because they didn't want to sit around and adjudicate disputes over the written word. It's not like the words from the officially editted and approved news outlets are that much better than the Drudge Report or Slashdot.
"I sayeth that he calleth my tax plans nasty, brutish and short and that is surely a seditious act against the United States."
"No, I sayeth that his tax plan is seditious in the breadth and scope of its grasp."
It's quite complicated. If you donate services, well, then you have to declare some kind of income and it all washes out. But MS Windows is a product, an object, and a thing. It cost something to make and they're able to deduct what they would have made selling the product.
What's the fair value? No one knows but the businesses have to guess. So they do. It's certainly fair for them to take of something. Then it's just a big argument and neither the government nor the business has time to do a complete and accurate job. If I was a tax lawyer, which I"m not, I would fudge as high as my conscience allowed. That's probably what many are doing in the business world.
Yes, I can see the trial now. The Nielsen rating system asks you to fill out their diary to accurately reflect your viewing. So everytime you record a show and watch it later, that show gets higher ratings. So you're essentially putting money in the pocketbooks of the broadcasters.
I wonder why they don't sue for negative damages! Oooh.
On the other hand, I think the networks should sue the copyprotection folks for making it harder to record their shows and watch them later. The Copyprotection folks are the real ones who are cutting their revenues by cutting television watching.
I can understand the point of top level domains that mean something, but the.tv domain is bogus. This is not a new version of Fandom that happens to be located in Tuva, or where ever country the.TV domain happens to be attached to. The.TV people are the ones who are corrupting the meaning of TLDs by giving out domains to everyone.
There's nothing wrong with say, "Fandom.il" if it really was a site associated with Fandom in Israel.
To me, I see this as an example of someone starting a company called "Ford Cars" and claiming it was distinct because the famous company is named "Ford Automobiles".
The key point in my mind is confusion. If I grabbed Slashdot.tv and went to slashhosting.com, I could clone Slashdot. Is that wrong? At some point, the confusion is so bad that it hurts people.
So, either enforce some semantic meaning for the TLDs or quit pretending that they are in fact differnt.
While I hate the way that people are going after websites and stealing domains using trademark law, I don't have any sympathy for Fandom.tv. The whole purpose of the law is to prevent confusion. Choosing the exact same domain name and registering in the TV domain is sure to cause some confusion.
It would be very easy for the new site to choose FansGarage, FansPlace, Fantopia, or anything else. There are plenty of names and it's easy to avoid confusion.
I realize this is a problem when companies like Ford go after car owner sites with the name Ford in it, but I think that the Fandom.tv owner should make some attempt to differentiate themselves. There's no reason for anyone to be confused.
Burning a cd-rom amounts to using a laser to create pits in a surface. A microscope reads the device by shining a laser on the spot to determine whether it has been burned or not. The presence or absence of these pits determine the pattern of zeros and ones.
So using a microscope really isn't news. They're just using a different form of microscope.
Watermarks aren't great tools, but they are better and more sophisticated than Schneier understands. He's just behind the loop on them. He constantly brags that he's never gotten a digital image in email. He must be reading mail with PINE on an old VAX.
Seriously, we can do better than he claims. There is at least one good system from NEC (Ingemar Cox, I believe, and others) that hides separate bits in separate places. If someone tries to attack a watermark by averaging together multiple copies, all of the signals come through. The rightful owner can track down the signals and figure out who was the legitimate owner of the copies thrown into the so-called average. Naturally, this is not perfect, but it can withstand some simple attacks.
Schneier also doesn't understand why record companies want to include watermarks. Sure, folks can still copy the music, but the watermark can help the record company track down the guy who purchased the first copy. The point is that every copy comes with a different watermark that points to the legitimate owner. Of course, that guy could always claim that someone stole his computer or something like that.
There are a host of reasons why watermarking won't work. Some of them are political. Some of them are technical. The algorithms are far from perfect. But then, Schneier constantly worries about technology being anything less than perfect.
There are two ways to look at this. In theory, you're a scholar and scholars make their reputation by publishing. Plus, they share their data with everyone. So publishing the source code and sharing it seems to be entirely in this tradition. The GPL seems like a great license to protect scholarship.
You are also a student. If you're truly an employee, as the university would like to claim, then they could become required to provide benefit plans, pension plans, and health plans that are consistent. You might want to ask them whether they truly want to classify you as an employee because it would set so many precidents.
It has never been clear to me that graduate students are employees. The fiction in almost all cases is that the tuition is paid by either a government grant or work teaching classes. You're not hired to do the research, as the professors are.
Of course none of this really matters because your true status is a slave. The professor probably wants to make some money off your labor so that's why the professor wants to own your code. Most professors don't give a rats ass if you actually publish something and get a job later. They just want to make the money for themselves now.
The only real negotiating position you have is switching advisors. This professor owns your ass. If your professor wants your code, you have to give it to them or leave. I don't really recommend trying to "educate" them about free software or teach them anything. They believe they own you. They do not have a collegial relationship with you. You're their source of income.
Unfortunately, you've got a tough decision to make. If you want to GPL the code, you'll probably permanently damage your relationship with the advisor. If you don't, well, you'll get along. It's up to you.
As I write this, Transmeta's stock is up 129%.
What a roller coaster ride. While this is supposed
to look good for the company, it really has the
potential to be quite a disaster. The good news
is that the company raised more than $200 million
in capital, enough to keep rolling for a few more
days.
But I don't think anyone should believe that the folks at Transmeta are seeing this money. The folks at the company won't be able to cash out for some time. Wall Street has always been good at hyping stocks. The shares are allocated to friends who then turn around and sell them in the aftermarket for a quick buck. It's a very sleezy process and it's not surprising that IPO shares are used as grease in many political situations.
This kind of wild ride can often hurt a company. I'm very bullish on Red Hat and VA Linux, but all I hear about is their descent from the stratosphere. Somehow the companies have a cloud hanging over them. The money from the peak didn't end up in the company treasury. It didn't end up in the pockets of the programmers or the workers. It only went into the hands of the speculators. But the company and the programmers must pick up the pieces underneith this cloud of failure.
The Compaq decision is surprising and probably the result of some brassknuckle dealing between the companies. Transmeta probably pushed too hard in negoatiations and didn't have the benchmarks to justify their demands. I can't say I blame Compaq or IBM. They can always come back after the IPO when real chips are shipping and negoatiate a more stable price.
The saddest part about this event is how it shows how the nerds are being used by the Wall Street guys.
The way to hack this can also be found in Peter Wayner's Digital Copyright Protection . The mechanism is simple. Rotate by 45 degrees. Blur twice and then sharpen twice. Boom. The watermark is gone, or at least as gone as the Digimark detector in Photoshop.
Mitnick should spend about ten seconds and then he would realize how to hack DNA databases. All you need to do is plant samples from the person being framed at the scene of the crime. This is an old technique in crime movies.
In the movie "The In Crowd", for instance, one girl borrows a blue dress of a friend, sleeps with a guy, gets his DNA smeered all over it, then frames the girl later.
In the movie "Gattaca", the lead character uses DNA samples from another to constantly fool the establishment.
In Presumed Innocent , the main character is almost set up by DNA>.
The list goes on and on and on. Identity fraud is a serious problem. The only solution is not to rely upon identity. Biometric solutions like DNA or other details can be easily fooled and hacked. Mitnick is an expert at assuming other identities. He should know it by now.
Anonymous techniques like anonymous cash or nyms are a better solution. They have their problems, but they aren't so easily fooled.
The names of the IIT reviewers were initially redacted and only revealed when it turned out that the electronic version was poorly constructed. Why were the names hidden? Do you feel that it's hypocritical to demand privacy for your reviewers while stripping away the privacy of everyone else? If it's so important that our actions be open, why can't yours be open?
Steganography and watermarking are closely related. Steganography is all about hiding information so it can't be found. Ideal watermarks contain information and hide it so it can't be found. It's a proper subset.
Bruce Schneier's discussion of steganography should be taken with a grain of salt. He's not really an expert on the subject and he deliberately excludes the topic from his technical books. In the steganography section of Secrets and Lies he even makes the dubious claim that he's never received nor sent a GIF image via email. He's sure living in the old command line days, I guess.
Steganography can be very hard to defeat. The signal can be hidden with the same strength as a cryptographic function. It can often be harder to suss out steganography than break a cipher, including Blowfish or Rijndael.
Watermarks, on the other hand, can be defeated even if they can't be unlocked. You just act like the mongol hordes and piss all over them. Eventually you'll wipe them out. Getting rid of some watermarks is as easy as writing over the least significant bit.
In fact, it's easy to remove the watermark inserted by Adobe Photoshop. All you do is rotate 45 degrees, blur twice, sharpen twice, and then rotate back. Voila! It's gone or at least unrecognizable.
Transmeta hasn't said much about their processors publically, but they do say that they can do a good job with MPEG without spending too much power. They like to avoid giving raw speed numbers and just give numbers like computations per watt.
That seems to suggest that they don't have great raw speed, but they do execute code rather cheaply.
But others already do the same thing. Mac laptops do a great job with MPEG decoding in software. They can play an entire DVD without a recharge. I mentioned this to a Transmeta engineeer and he said, "Yeah, that's a great demo." But he wouldn't say more.
I think they can beat some of the fat power hogs like the Pentium chips, but they have trouble with the RISC chips like the G3 or the StrongARM. (The G4 is another story.) I just find it hard to believe that a compiler optimized binary for a RISC chip is ever going to take more power to execute than an old x86 binary that needs to be massaged extensively by a pre-processor. The pre-processing has to take power and I doubt it will yield enough savings later to pay for it.
My bet is that Transmeta can beat the Pentiums, but not the RISC chips.
The website for Free for All discusses this and suggests that open source is not really a good model for books. While there are some collaborative books out there, most are the work of one person. It's not exactly clear that letting everyone change a book and reissue it as they want is good. Someone could simply take All the President's Men and rewrite a new ending with Richard Nixon chortling in the White House as IRS agents audit and audit Woodward and Bernstein. That's the ending I'm sure one person wanted to read.
The important thing to remember is that facts are not copyrightable. They're sort of open sourced already. If you want to use a detail or a factoid from a non-fiction book, you can reuse it and reuse it again. You can lift it like you lift a subroutine. Non-fiction is pretty close to open sourced. Yes, you need to rephrase things, but it's probably good for you.
Fiction is a more complicated matter. I think fan fiction is great. There's no question in my mind that the authors should have less control over their characters. If someone wants to write new Star Trek stories, they shouldn't have to pay Paramount. This is a more radical point and I might be wrong. Maybe there's an better way with open licensing? Paramount would get a fixed amount no matter who grabbed the characters. That's how they regulate covers of songs.
The important thing to remember is that open source is good and cool because it simplifies cooperation. The lack of copyright on facts is already a quasi open source solution for books. It's a great way for people to exchange knowledge and work together.
I remember watching the discount of books changed frequently too. Free for All , for instance, bounced all over the place. Now they don't offer any discount at all. I wonder why? FatBrain was offering 40% off the last time I looked.
I think they've been experimenting in many more places. I wonder if they're going to come clean there too.
Slashdot Mirror
I shudder to think what would happen to the MSC if the Chief of Police and Colonel March had "all of the lessons from Columbine" at their disposal. Back then, "boys will be boys" was an actual legal dictum that lawyers could offer to judges and actually get their clients off. These guys were using pyrotechnics in several so-called capers.(They blew up the monster in Strawberry Lake at the end.) They seemed dangerously interested in military surplus. Hacking the radio frequencies was second nature to them. All of these actions are dangerous predictors of future Bad People. I wish the book publishers would start reprinting more books about good children who sit still and devote themselves to watching Disney cartoons. If kids must get off of the couch, they might devote themselves to collecting Disney beanbag dolls or maybe those plastic action figures for Disney characters.
At one point he even seems to confuse the Netcraft surveys of the webservers with a measure of penetration of Linux. The Netcraft surveys measure Apache, a server that also runs on FreeBSD, AIX, MacOS, and dozens of other OSs.
For this reason I think both Rebel Code and Free for All do a better job with the history. The only really novel part here are the stories about growing up in Finland and spending his life coding in a closet. Those parts are great fun. He's a charming guy. The book's great. But I wish he could spread the credit around a bit more.
I have a black box filled with my own personal links between names and phone numbers. What's wrong with having my own list of matchups between names and IP numbers? ANd what's wrong with circulating this list in DNS format? Ooops. That's stepping on ICANN's so called monopoly. I say it's a first amendment issue. Network solutions does a good job, but I would love a backup directory service.
I've got a NAT router hooked up to an Earthlink account. The connection gets disconnected every so often and then, voila, a new dialup generates a new IP number.
The animated series rocked, but the buzz about the live-action version isn't so hot. They nuked one of the writers who helped make the animated version everything it ever was. The word is that they want it to be more "Seinfeld with Capes" than "The Tick". Well, that's the price you pay when you switch to film, a medium where creating an explosion is significantly more expensive than creating a wink. Spoon!
Here's the basic solution. Put the music through a narrow bandpass filter. That is, remove all frequencies except those in a narrow band, say 800Hz to 1000Hz. It probably makes more sense to locate this band in the lower octaves. Then, go through the song looking for the presence or absence of a signal. Turn this into a string of 1's and 0's where a 1 means that there's some noise with a frequency in the range and 0 if there's only a neglible amount. Then do string matching. You might need to slide this back and forth a bit to find the best match, but that's not too complicated.
I've really wanted to get a Ricochet modem, but the price is too steep for me. $70/month. I don't travel much for work so it doesn't help me. If there was some travel involved, then I would be signing up immediately. If they bring the price down, I'm there.
They run a network that hides your browsing and mail reading. It's pretty slick. Of course you have to pay to support it, but the price isn't bad. It's something like $50/year.
"I sayeth that he calleth my tax plans nasty, brutish and short and that is surely a seditious act against the United States."
"No, I sayeth that his tax plan is seditious in the breadth and scope of its grasp."
Founding Fathers: "Whatever."
What's the fair value? No one knows but the businesses have to guess. So they do. It's certainly fair for them to take of something. Then it's just a big argument and neither the government nor the business has time to do a complete and accurate job. If I was a tax lawyer, which I"m not, I would fudge as high as my conscience allowed. That's probably what many are doing in the business world.
I wonder why they don't sue for negative damages! Oooh.
On the other hand, I think the networks should sue the copyprotection folks for making it harder to record their shows and watch them later. The Copyprotection folks are the real ones who are cutting their revenues by cutting television watching.
There's nothing wrong with say, "Fandom.il" if it really was a site associated with Fandom in Israel.
To me, I see this as an example of someone starting a company called "Ford Cars" and claiming it was distinct because the famous company is named "Ford Automobiles".
The key point in my mind is confusion. If I grabbed Slashdot.tv and went to slashhosting.com, I could clone Slashdot. Is that wrong? At some point, the confusion is so bad that it hurts people.
So, either enforce some semantic meaning for the TLDs or quit pretending that they are in fact differnt.
It would be very easy for the new site to choose FansGarage, FansPlace, Fantopia, or anything else. There are plenty of names and it's easy to avoid confusion.
I realize this is a problem when companies like Ford go after car owner sites with the name Ford in it, but I think that the Fandom.tv owner should make some attempt to differentiate themselves. There's no reason for anyone to be confused.
So using a microscope really isn't news. They're just using a different form of microscope.
Seriously, we can do better than he claims. There is at least one good system from NEC (Ingemar Cox, I believe, and others) that hides separate bits in separate places. If someone tries to attack a watermark by averaging together multiple copies, all of the signals come through. The rightful owner can track down the signals and figure out who was the legitimate owner of the copies thrown into the so-called average. Naturally, this is not perfect, but it can withstand some simple attacks.
Schneier also doesn't understand why record companies want to include watermarks. Sure, folks can still copy the music, but the watermark can help the record company track down the guy who purchased the first copy. The point is that every copy comes with a different watermark that points to the legitimate owner. Of course, that guy could always claim that someone stole his computer or something like that.
There are a host of reasons why watermarking won't work. Some of them are political. Some of them are technical. The algorithms are far from perfect. But then, Schneier constantly worries about technology being anything less than perfect.
You are also a student. If you're truly an employee, as the university would like to claim, then they could become required to provide benefit plans, pension plans, and health plans that are consistent. You might want to ask them whether they truly want to classify you as an employee because it would set so many precidents.
It has never been clear to me that graduate students are employees. The fiction in almost all cases is that the tuition is paid by either a government grant or work teaching classes. You're not hired to do the research, as the professors are.
Of course none of this really matters because your true status is a slave. The professor probably wants to make some money off your labor so that's why the professor wants to own your code. Most professors don't give a rats ass if you actually publish something and get a job later. They just want to make the money for themselves now.
The only real negotiating position you have is switching advisors. This professor owns your ass. If your professor wants your code, you have to give it to them or leave. I don't really recommend trying to "educate" them about free software or teach them anything. They believe they own you. They do not have a collegial relationship with you. You're their source of income.
Unfortunately, you've got a tough decision to make. If you want to GPL the code, you'll probably permanently damage your relationship with the advisor. If you don't, well, you'll get along. It's up to you.
But I don't think anyone should believe that the folks at Transmeta are seeing this money. The folks at the company won't be able to cash out for some time. Wall Street has always been good at hyping stocks. The shares are allocated to friends who then turn around and sell them in the aftermarket for a quick buck. It's a very sleezy process and it's not surprising that IPO shares are used as grease in many political situations.
This kind of wild ride can often hurt a company. I'm very bullish on Red Hat and VA Linux, but all I hear about is their descent from the stratosphere. Somehow the companies have a cloud hanging over them. The money from the peak didn't end up in the company treasury. It didn't end up in the pockets of the programmers or the workers. It only went into the hands of the speculators. But the company and the programmers must pick up the pieces underneith this cloud of failure.
The Compaq decision is surprising and probably the result of some brassknuckle dealing between the companies. Transmeta probably pushed too hard in negoatiations and didn't have the benchmarks to justify their demands. I can't say I blame Compaq or IBM. They can always come back after the IPO when real chips are shipping and negoatiate a more stable price.
The saddest part about this event is how it shows how the nerds are being used by the Wall Street guys.
The way to hack this can also be found in Peter Wayner's Digital Copyright Protection . The mechanism is simple. Rotate by 45 degrees. Blur twice and then sharpen twice. Boom. The watermark is gone, or at least as gone as the Digimark detector in Photoshop.
In the movie "The In Crowd", for instance, one girl borrows a blue dress of a friend, sleeps with a guy, gets his DNA smeered all over it, then frames the girl later.
In the movie "Gattaca", the lead character uses DNA samples from another to constantly fool the establishment.
In Presumed Innocent , the main character is almost set up by DNA>.
The list goes on and on and on. Identity fraud is a serious problem. The only solution is not to rely upon identity. Biometric solutions like DNA or other details can be easily fooled and hacked. Mitnick is an expert at assuming other identities. He should know it by now.
Anonymous techniques like anonymous cash or nyms are a better solution. They have their problems, but they aren't so easily fooled.
The names of the IIT reviewers were initially redacted and only revealed when it turned out that the electronic version was poorly constructed. Why were the names hidden? Do you feel that it's hypocritical to demand privacy for your reviewers while stripping away the privacy of everyone else? If it's so important that our actions be open, why can't yours be open?
Steganography and watermarking are closely related. Steganography is all about hiding information so it can't be found. Ideal watermarks contain information and hide it so it can't be found. It's a proper subset.
Steganography can be very hard to defeat. The signal can be hidden with the same strength as a cryptographic function. It can often be harder to suss out steganography than break a cipher, including Blowfish or Rijndael.
Watermarks, on the other hand, can be defeated even if they can't be unlocked. You just act like the mongol hordes and piss all over them. Eventually you'll wipe them out. Getting rid of some watermarks is as easy as writing over the least significant bit.
In fact, it's easy to remove the watermark inserted by Adobe Photoshop. All you do is rotate 45 degrees, blur twice, sharpen twice, and then rotate back. Voila! It's gone or at least unrecognizable.
That seems to suggest that they don't have great raw speed, but they do execute code rather cheaply.
But others already do the same thing. Mac laptops do a great job with MPEG decoding in software. They can play an entire DVD without a recharge. I mentioned this to a Transmeta engineeer and he said, "Yeah, that's a great demo." But he wouldn't say more.
I think they can beat some of the fat power hogs like the Pentium chips, but they have trouble with the RISC chips like the G3 or the StrongARM. (The G4 is another story.) I just find it hard to believe that a compiler optimized binary for a RISC chip is ever going to take more power to execute than an old x86 binary that needs to be massaged extensively by a pre-processor. The pre-processing has to take power and I doubt it will yield enough savings later to pay for it.
My bet is that Transmeta can beat the Pentiums, but not the RISC chips.
The important thing to remember is that facts are not copyrightable. They're sort of open sourced already. If you want to use a detail or a factoid from a non-fiction book, you can reuse it and reuse it again. You can lift it like you lift a subroutine. Non-fiction is pretty close to open sourced. Yes, you need to rephrase things, but it's probably good for you.
Fiction is a more complicated matter. I think fan fiction is great. There's no question in my mind that the authors should have less control over their characters. If someone wants to write new Star Trek stories, they shouldn't have to pay Paramount. This is a more radical point and I might be wrong. Maybe there's an better way with open licensing? Paramount would get a fixed amount no matter who grabbed the characters. That's how they regulate covers of songs.
The important thing to remember is that open source is good and cool because it simplifies cooperation. The lack of copyright on facts is already a quasi open source solution for books. It's a great way for people to exchange knowledge and work together.
I think they've been experimenting in many more places. I wonder if they're going to come clean there too.