But like I said, this is an artificial segmentation of the market. There is no reason that you can't have "business" grade 802.11b equipment with say, hardware encryption, connectors for external antenna's, support for thousands of simultaneous clients, a high price point (like Cisco). While at the same time having "home" grade 802.11b with no external antenna provision, software encryption, and a low price point (like Apple).
It's like saying that we should have two differnet processor standards because businesses like to buy higher quality PC's than most home users. It doesn't make sense. You can make cheap x86 PC's, and you can make expensive ones. Just like you can make "business grade" 802.11b and "consumer grade" 802.11b. But don't you want your software to work in both places? Does it make sense for me to have two wireless cards for my laptop? One for home and one for work? Of course not.
Like I said, this is just Intel and some other folks trying to segment the market so that they can dominate that segement. My opinion is that rather than compete in the 802.11 segment, they want to create their own segment. The fact that this strategy offers no advantage to the consumer whatsoever is irrelevant to them.
HomeRF is just going to cause problems. Who needs it? Another incompatable standard running in the same frequency band which offers no advantages whatsoever. At best, it will work just as well as 802.11b. There is nothing about HomeRF vs. 802.11b which makes one cheaper than the other. Apple is pushing 802.11b access points for under $300. Is HomeRF going to be less? I doubt it.
What makes this stuff cheap is economies of scale. If the same folks that are spending all this effort competing with 802.11b just started making and selling 802.11b equipment we would all be better off. They could help grow the 802.11b market which would push prices down. But no, it's another example of companies trying to segment the market and screw the consumer to jack up their bottom line.
This article does not do a good job of talking about what products are available, or comparing them.
For one, Lucent WaveLAN cards run at 11Mbps, not 2 as the article states. In addition, Lucent provides binary x86 Linux drivers on their web site.
For another, it's not really important whether or not the card's vendors provide Linux drivers. What's most important is whether or not they work under Linux. The fact is most of these cards work using pcmcia-cs.
No mention is made of the access points, which are important if you want to easily hook up to your wired LAN. Of the ones I've tested, it comes down to two: Apple and Cisco. The Apple Airport can be configured via a third party Java client which works fine under Linux. It's also dirt cheap (~$280). The downside is that it is not a high performance solution, and it can only use one encryption key. If you want performance and flexibility, buy a Cisco. It can be configured via web, telnet, and serial interfaces. All encryption is done in hardware so it is far and away the fasted 802.11b solution I've seen. Most vendors do the encryption in software, and take about a 20% performance hit with it. Which would be fine, but you pretty much have to use encryption. If you're not, you are opening up your network connection to all your neighbors.
All the other access points I've tried use Windoze-only configuration clients (3Com and Lucent). Some of them try to prevent compatability by not allowing you to set the encryption keys directly (3Com and Apple(only applies if you use Apple's software instead of the Java client))
Overall, this stuff rocks though. I have an Apple Airport at home. Combined with my DSL line, it is awesome. I can use my laptop anywhere in the house. Sit on the sofa and play games, check email in the kitchen, whatever. The 802.11b is much faster than DSL, so you don't notice a performance hit at all.
I would also recommend only buying 802.11b 11Mbps hardware if you want to use it other places. There are two coding techniqes allowed in 802.11. One is Direct Sequence Spread Spectrum (DSSS), which runs from 1Mbps to 11Mbps. The other is Frequency Hopping Spread Spectrum (FHSS) which maxes out at 2Mbps. The two are 100% incompatable. They will not communicate in any way at all (except possibly some interference.) So, if you buy cheap 2Mbps cards, there is a good chance that they are FHSS and will not work with 11Mbps DSSS equipment. Spend the extra bucks and get 802.11b equipment. You will appreciate the extra speed and compatability. If you can find 2Mbps DSSS cards, they should talk to 11Mbps equipment fine (at 2Mbps of course.)
Why do you have to analyze all the results? If the first file you check has had the watermark removed, and sounds fine, you can bet you are screwed. The files are all independent. The extent to which the first file has been cracked does not depend on the next 449 files.
If I recall correctly, the hacksdmi.org website said that they would be performing some sort of quantitative test to determine how much the uploaded song differs from the original. Given that this test and the test for the watermark can both be easily automated, they have a pretty good idea of how "cracked" their algorithms are.
Sure, they will still want to do more analysis of the samples. Any they probably want to try and wrangle the details out of the people who submitted the samples. But they know. And that's why they're pissed. They know which uploads have the watermark, and they know how similar to the original they all are. If this crap worked at all, they would have said something positive.
Curiously, at almost the same time, the RIAA announced a new program to prevent piracy. See the C|Net story here. The idea is to create a "digital bar code" for songs to allow tracking of copyright information, etc. On the face of it, this sounds like a whole new tack for the RIAA. Any maybe it is. However, the article says:
Nevertheless, if it proves to be difficult to tamper with, the system could be a potential way of identifying authorized and unauthorized songs being traded through services such as Napster.
It sounds to me like the exec's at the RIAA know that their SDMI scheme is going down in flames, and are looking for a backup. However, it sounds like they are going to head right back down the watermark path. Only this time, instead of trying to encode a single bit (play / don't play), they will be trying to encode hundreds or thousands of bits. Whoever brought up Don Quixote in an earlier SDMI thread was right on the money. These folks aren't ever going to learn.
Who ever decided to call this technology watermarking anyway? It's really a horribly inaccurate name. The whole point to a real watermark is that it's difficult to create, and easy for everyone to see. But with "digital watermarks" the idea is to make it invisible. An invisible watermark it what they should call it. Maybe then people would understand what a stupid idea it is.
I've heard, but don't know for a fact, that Sun does not allow M$ Office in their operations. All of their employee's were already using StarOffice. So, they have a strong incentive to see it improve. I don't see them getting "bored", as this isn't just an academic exercise.
I also don't know what kind of bulk discount M$ gives on Office licenses, but it is quite possible that Sun has considerably less invested in OpenOffice than they would pay to legally license Office2000 for all of their employees. I'm pretty sure that the ongoing costs of running OpenOffice, even with several dozen programmers assigned to it, would be less than paying M$ for upgrades every couple of years.
I think this just might be a testament to how greedy M$ licensing practices are. It's quite possible that it's cheaper for Sun to write their own office suite than it is to license M$'s "mass market" office suite.
I didn't say they were going to break into your local CO and physically take over your DSL line. But, if I figure out your IP address, I can ping it from anywhere in the world and you get to pay the bills for it. And depending on how well your ISP verifies the traffic and exactly what they bill on, there might not be a damn thing you can do about it. If they just count how many bytes flow to your modem from their DSLAM, I can send you all sorts of traffic and it is impossible for you to stop me. Like I said, it's an irritation if some kiddie floods you with traffic and makes your connection unusable for a few hours. It's a whole different game if they send a trickle of traffic all month long and you don't notice. Suddenly, you have a huge DSL bill, and no idea why.
My laptop has a ATI Rage Mobility in it that can apply what I think is a bi-linear filter to the image before it displays it. The result is that the image does not have any of the horrible aliasing that results with most video cards. The result is a little fuzzy, but rather than annoying it has an almost attractive "magnified" look to it.
The result is that the lower resolution are 100% usable. To be honest, I don't even know why vendors implemented the old stretching method at all. It looked so crappy as to be useless. I always turned it off, and just had the display use the center 640x480 pixels with a huge black border. At least then I could read things.
However, I don't know if this solution is available outside of laptops. It is a function of the video chipset, and I don't know if the desktop ATI's implement it. I also don't know how it would work with the VGA connection that most LCD displays use. I wouldn't be suprised if it didn't. But, if you get a laptop with the ATI Rage Mobility, you won't have this problem. However, you might want to check for a BIOS setting to let you turn the stretching off on your laptop. All the one's I've used have had an option for this. The 1:1 mapping isn't great because it's so small, but at least it looks OK.
One of the biggest problems with metered billing is what do you do when someone decides to rack up your bill? Let's say I'm some script kiddie whose on an unmetered line (say at college.) Let's say I do a ping with 10K packet size to your home DSL address at the default frequency of once a second for a month. By my math, that adds up to almost 25GB of traffic. One ping a second won't even get noticed by 99% of DSL users, until they get that $200 dollar ISP bill.
That's one of the big problems with metered billing. It's one thing when a script kiddie gets upset at you and floods you with traffic for a few hours. It's a whole different story if you get a huge bill from your ISP because of it.
If everyone goes to metered billing you will see all sorts of abuses as crackers try to set up servers on other peoples machines to avoid paying the bills for their traffic. Add that up to the aforementioned harrasment traffic jacking up peoples bills. Plus the dishonest users who blame their traffic spikes on "hackers".
I just don't see it being worth the headache for an ISP to charge by the byte. You can bet that any user that is hit by the above problems is going to run screaming to the nearest flat rate ISP. Besides, the rates are metered to a certain extent. Dialup access is not the same cost as OC-3 by a long shot. So all the dialup users are in the same cost pool. So what? They are in a different cost pool from the DSL users, who are in a different pool from the T3 users, etc.
There are way too many zeros in that price. It should be in scientific notation or Mega. Or they should at least add the appropriate commas. 2e07 dollars, or 20M$, or $20,000,000. But 20000000.00? You'd think for 20 million bucks they could buy a comma for the ad. And a JPEG of the whole thing.;-)>
I also disagree that this is what every geek girl and boy needs. It's actually what every drug-running girl and boy needs. Which raises the interesting question of what Federal laws you might be violating cruising around in a submarine. I can't imagine that you could sail up the Potomic with this without raising some sort of ruckus.
And what exactly does the license agreement have to do with anything? There is no evidence that any of the people getting the letters have installed or are using the software. Hence, the license agreement has no relevance whatsoever. Many of the cuecats were sent out as gifts via the post office. Attempting to attach a license onto these gifts is actually a violation of federal law. In my case, I went to RatShack and asked for one. I was never asked to sign anything, nor asked for my name. The clerk simply handed it to me.
At what point do you think any of us entered into a contract with Digital Convergence? That's what a license agreement is: a contract. Did people enter into a contract simply by removing the package from their mailbox? Did they enter into a contract by opening the package? Did I enter into a contract by asking the RatShack clerk for a free scanner? Did I enter into a contract when he handed me the bag? I was never asked for my name, and certainly never signed anything. None of us have installed or used the software. Why bother? If the license agreement said everybody born in odd years owes Digital Convergence a dollar, would that mean I owe them a dollar? Of course not.
So why exactly are any of us bound by the license agreement on the software? Did I enter into a contract when you read the license agreement? The simple fact of the matter is these lawyers are sending out scary-sounding but completely worthless letters. I don't think they even qualify as a cease and desist letter. There is no IP infringement going on. The lawyers know that. They are just hoping that they can scare everyone with their worthless threats.
I am tired of this hot potato fantasy. Everyone uses BGP4 to determine which route should be taken by a packet. BGP4 only considers how many AS's are in the route. It doesn't know anything about hops, latency, etc. If I'm an ISP, and I have two routes for a packet with the same AS path length, of course I'm going to "get it off my network at the earliest possible time." Would you rather I ran it back and forth across the country a few times? The two paths are the same length, you should get it to whichever exit point is closest in your network.
Or do you think that the backbone ISP's are sending packets out of longer AS path points because they are closer in their network? I am almost positive that they are not doing this because your packet would probably never arrive at it's destination. If everybody was choosing longer AS paths, your packet would probably never arrive. It would just run around it loops as everyone shunts it off to someone else ignoring the AS path.
If you have specific details about what hot potato routing is, and how it differs from the correct routing that everyone does, please inform me. I would love to know. But I think this belongs firmly in the urban legend category.
This whole idea of voting against somebody is bullshit. There is no such thing. If you vote for Al Gore and the Democrats, you are voting for the DMCA. You are voting for the War on Drugs. You are voting for a political system in which bribes (aka soft money) are routinely exchanged in public.
You can tell yourself whatever lies you want, but the fact is that by voting for a Republicrat you are showing your support for the current corrupt system. Our current system is quickly failing. As long as people continue to buy into the FUD and vote for candidates which they know are corrupted by soft money and which they don't really want to see elected, it's going to keep getting worse. Every time somebody endorses the Republicrats with their vote, they are endorsing the current decline of our democracy into a cesspool of big-money corruption.
Actually, it's tougher than you might think to block Napster. You see, if you block Napster entirely, the students will use different port numbers to get around your block. It's a loosing battle for the network admins.
Now, I happen to know from first hand experience of one university which is limiting it's Napster traffic. Using various QoS tricks, they lower the priority of Napster traffic so that it doesn't consume too much of their bandwidth. A Napster "knob" if you will that lets them tweak the bandwidth it gets. They crank it down some, and the students just think that the link is more congested than it used to be. But they know if they crank it down too far, the students will realize that something is up and move to a different port. It's an interesting bit of network engineer / psychology going on.
If you actually go and download the files for the contest, you won't find much. Rather than any sort of description of the watermark technology, or any software that checks for the watermark, you get three.wav files. File 1 has no watermark. File 2 is the same audio as file 1 with a watermark applied. File 3 is a different song with a watermark applied. Your "challenge" is to remove the watermark from file 3. To check the file, you have to upload it to their server, and they will send you email with the results of the check.
So, from a cryptographic point of view, this is pretty worthless. It's along the lines of the newbies who post to sci.crypt saying "I've developed a new algorythm. Here is some ciphertext, crack it!". Of course, to do any valid analysis you need to know how the algorithm works.
My guess is that either the people setting up the "contest" are pretty clueless, or they have no faith in their algorithm, or both. Or this is just a publicity stunt to reassure the record labels. My money is on the latter.
Any hacker who attacks SDMI after it's released will certainly have access to a software implementation, or the algorithm, or both. So, to leave both of those out of the "contest" just makes it a sham.
Actually, this is a bad example. A good friend of mine owns a Thompson machine gun, and it is 100% legal. He had to get a license for it, which is not possible to do for fully automatic weapons manufactured after a certain date. But, when the "ban" on fully automatic weapons went into effect, existing guns were grandfathered. Not that this invalidates your point about ex post facto, I just wanted to point out that machine guns are a bad example.
Actually, I looked around some more on the companies web site. From the content there, they make it quite clear that their market plan is to force these books on the students:
---quote---
Publishers are guaranteed 100% market penetration at partner schools who opt to implement the Vital Source system. Purchase of all included titles is mandated by the universities.
---/quote--- emphasis in original
Later on, and this is the best part, it talks about how they make you license the book even if you don't need it:
---quote---
Publishers receive a mandated, preset fee for every student for every title chosen by professors. Because the service is a global curriculum application, the fee comes in from each student each of the four years of their studies, regardless of whether they are taking that course that year.
---/quote---
Nothing like paying for that advanced quantum physics books when you're a freshman enrolled in basic mechanics, eh?
First, in the case of textbooks, the students usually have little to no say in what textbooks are used. So, I would expect that in many cases, the consumer will be forced to buy the product.
Second, it certainly does not appear that the time-limited version will be cheaper. Given that the company is talking about it's vastly increased revenue, they can't be charging a whole lot less. In fact, the Wired article states that the subscription fee is $1200/annum. Not exactly dirt cheap.
Third, this has nothing to do with contracts. The DMCA allows companies to prohibit any and all rights that a consumer had under copyright law. While I agree that the market should decide which products sink and which swim, the law should not allow the makers of bad products to have people thrown in jail on criminal charges for trying to assert their Constitutionally protected rights under existing copyright law. Make no bones about it, that's exactly what the DMCA does.
So, as someone who considers themselves libertarian, my problem is not with e-texts or DVD per se. It is with the fact that the DMCA allows the makers of these bad products to get other people thrown in jail for exposing the weaknesses in their products. I'm not asking for more government regulation. I'm asking for less, in the form of repealing the DMCA. It is truly the purest form of bad lawmaking I have ever seen.
I'm 100% positive the 6502 is an 8-bit processor, and I'm 99% positive that the Z-80 is too. Since endian refers to the order of bytes in a word, it doesn't make sense to refer to them as being big or little endian.
The old copies of Mosaic were never taken off of NCSA's FTP server You can find them here. This is a little more complete than Evolt.org since it includes the Unix binaries along with source for several versions. Don't forget to pick up your copy of NCSA HTTPD while you're there!
Actually, I don't believe this is the way SDMI will work . There is actually a lot of specification for it on their web site. Rather than not play songs that are not watermarked, they will begin to watermark CD's with a special "do not play" mark. So, in theory, the SDMI players will still play your existing MP3's but not MP3's made from "new" CD's (i.e., ones after they begin watermarking.) The players should also play any MP3 by a band that didn't watermark without any problem. In theory.
I wouldn't worry too much about SDMI though. It will sink faster that DiVX. In fact, it's already pretty much dead in the water. Lot's of companies signed up, but no products. They wanted to have them out last Xmas, but as far as I can tell right now there is only a single Sony SDMI unit available. More importantly, lot's of non-SDMI players are already on the market. So I think this battle has already been lost.
We need a hardware player that uses DeCSS!!!
on
NYT On DeCSS Case
·
· Score: 4
What would really be great to make the real issues apparent is for some consumer electronics manufacturer to produce a standard DVD player. But, instead of getting a CSS license, they could use the source to DeCSS.
Right now, the reality is that a program like DeCSS can be used to pirate DVD's. That makes it easy to paint it with the "evil" brush. However, if there was a hardware DVD player that was functionally just like an "authorized" DVD player, they couldn't use that tact. Provide an Apex-like menu to disable the region controls (but NOT macrovision), and the MPAA would be forced to admit that CSS and DMCA go far beyond copying and piracy.
Let's take the fictional scenario a bit further. Acme Inc makes a DVD player that is just like all the other home players, but doesn't include region controls. It also doesn't have a CSS license. Let's further suppose that the MPAA takes them to court for DMCA violations. What would the arguments look like in this case? If the MPAA said that CSS was an anti-copying mechanism, the question becomes why does Acme Inc. have to use it if you aren't copying the DVD? If they try to claim that the player is a piracy tool, you could point out that it doesn't enable piracy any more than a Sony or RCA DVD player.
So what would this player be doing that's "wrong"? Nothing. Well, almost nothing. By not having region controls, it's a way around the artificial trade barriers that the MPAA are enforcing to enrich themselves. And so the real issue would come to light before the customers. It's not about piracy. It's not about copying. It's about the DMCA being used as a shelf to place arbitrary rules and restrictions on. If this trial was about the MPAA's right to disable the FF button on your DVD player during ads, and the MPAA's right to restrict free trade via region encoding, the public's opinion of the MPAA and the DMCA might be a little less rosy.
It seems that the judge and others in this case have a hard time thinking of code as being expressive. It seems to me that the courts and the law have already overwhelmingly decided that it is expressive. Why? We allow copyright's on computer programs. Do we allow copyrights on objects which are purely functional? No. Copyright is reserved for things which are artistic or expressive, correct?
If the court feels that code is simply a tool, like a hammer or a screwdriver, why is it that code is deserving of copyright protection when hammers are not? And isn't a DVD just a program? Why does it get first amendment protection, even though by the judges own logic it isn't artistic or expressive?
The simple fact is that virtually 100% of the market and legal realities surrounding computer code treat it as expressive content deserving of both first amendment and copyright protection. It's only when a powerful interest decides to do something just plain wrong for it's own benefit does this stupid argument come up. Last time it was in the case of export restrictions on encryption. Now with the MPAA trying desparately to eliminate fair-use.
If code in non-artistic and non-expressive, I'm sure the MPAA won't mind me distributing an mpeg decoder source which #includes "matrix-vob.h", right?
Where did you get the idea that the Internet is run off of static routes? Sure, a lot of ISP's on the edge use statics. Primarily because they only need one default route to their upstream provider. But the backbone?
Have you heard of BGP? Border Gateway Protocol? That's what runs routing on the backbone, and it is the dynamic routing protocol. It's the duct tape that holds this thing together, and it's quite dynamic. It would probably take any backbone ISP (C&W, UUNet, Qwest, etc.) a week to statically configure that routes that work for one day. Never mind that the only way you would figure out how to configure the routes would be to use a dynamic routing protocol. The Internet is far too big and complex to ever manually configure it.
While it's true that BGP will still sometimes black-hole traffic by sending it down a broken link, that doesn't change the fact that it's dynamic. The problem stems from the fact that BGP can't always tell that a given route doesn't work. Usually it can, but not in all circumstances.
So, if you think one of the options is to boycott sites that accept ads from doubleclick, what are you doing here? Slashdot not only serves up doubleclick ads, sometimes the ads include the infamous cookies, too. Other alternative news sites feature doubleclick ads too. TheRegister seems to get most of it's ads from doubleclick (although to their credit, they've never sent me a cookie, unlike Slashdot).
This is what really disturbs me about doubleclick. They have become so ubiquitous that even sites whose creators are almost certainly not doubleclick fans are carrying the ads. Either in spite of or because of their outragous behavior, they are far and away the most successful ad banner company.
You say that if we boycott doubleclick.net sites, we will force doubleclick to rethink their policies. Will they? Or are we just going to cut off the air supply for the sites we DO care about? I have to say, it's hell of a dilema, and I don't see a clear way out.
Slashdot Mirror
But like I said, this is an artificial segmentation of the market. There is no reason that you can't have "business" grade 802.11b equipment with say, hardware encryption, connectors for external antenna's, support for thousands of simultaneous clients, a high price point (like Cisco). While at the same time having "home" grade 802.11b with no external antenna provision, software encryption, and a low price point (like Apple).
It's like saying that we should have two differnet processor standards because businesses like to buy higher quality PC's than most home users. It doesn't make sense. You can make cheap x86 PC's, and you can make expensive ones. Just like you can make "business grade" 802.11b and "consumer grade" 802.11b. But don't you want your software to work in both places? Does it make sense for me to have two wireless cards for my laptop? One for home and one for work? Of course not.
Like I said, this is just Intel and some other folks trying to segment the market so that they can dominate that segement. My opinion is that rather than compete in the 802.11 segment, they want to create their own segment. The fact that this strategy offers no advantage to the consumer whatsoever is irrelevant to them.
HomeRF is just going to cause problems. Who needs it? Another incompatable standard running in the same frequency band which offers no advantages whatsoever. At best, it will work just as well as 802.11b. There is nothing about HomeRF vs. 802.11b which makes one cheaper than the other. Apple is pushing 802.11b access points for under $300. Is HomeRF going to be less? I doubt it.
What makes this stuff cheap is economies of scale. If the same folks that are spending all this effort competing with 802.11b just started making and selling 802.11b equipment we would all be better off. They could help grow the 802.11b market which would push prices down. But no, it's another example of companies trying to segment the market and screw the consumer to jack up their bottom line.
This article does not do a good job of talking about what products are available, or comparing them.
For one, Lucent WaveLAN cards run at 11Mbps, not 2 as the article states. In addition, Lucent provides binary x86 Linux drivers on their web site.
For another, it's not really important whether or not the card's vendors provide Linux drivers. What's most important is whether or not they work under Linux. The fact is most of these cards work using pcmcia-cs.
No mention is made of the access points, which are important if you want to easily hook up to your wired LAN. Of the ones I've tested, it comes down to two: Apple and Cisco. The Apple Airport can be configured via a third party Java client which works fine under Linux. It's also dirt cheap (~$280). The downside is that it is not a high performance solution, and it can only use one encryption key. If you want performance and flexibility, buy a Cisco. It can be configured via web, telnet, and serial interfaces. All encryption is done in hardware so it is far and away the fasted 802.11b solution I've seen. Most vendors do the encryption in software, and take about a 20% performance hit with it. Which would be fine, but you pretty much have to use encryption. If you're not, you are opening up your network connection to all your neighbors.
All the other access points I've tried use Windoze-only configuration clients (3Com and Lucent). Some of them try to prevent compatability by not allowing you to set the encryption keys directly (3Com and Apple(only applies if you use Apple's software instead of the Java client))
Overall, this stuff rocks though. I have an Apple Airport at home. Combined with my DSL line, it is awesome. I can use my laptop anywhere in the house. Sit on the sofa and play games, check email in the kitchen, whatever. The 802.11b is much faster than DSL, so you don't notice a performance hit at all.
I would also recommend only buying 802.11b 11Mbps hardware if you want to use it other places. There are two coding techniqes allowed in 802.11. One is Direct Sequence Spread Spectrum (DSSS), which runs from 1Mbps to 11Mbps. The other is Frequency Hopping Spread Spectrum (FHSS) which maxes out at 2Mbps. The two are 100% incompatable. They will not communicate in any way at all (except possibly some interference.) So, if you buy cheap 2Mbps cards, there is a good chance that they are FHSS and will not work with 11Mbps DSSS equipment. Spend the extra bucks and get 802.11b equipment. You will appreciate the extra speed and compatability. If you can find 2Mbps DSSS cards, they should talk to 11Mbps equipment fine (at 2Mbps of course.)
Why do you have to analyze all the results? If the first file you check has had the watermark removed, and sounds fine, you can bet you are screwed. The files are all independent. The extent to which the first file has been cracked does not depend on the next 449 files.
If I recall correctly, the hacksdmi.org website said that they would be performing some sort of quantitative test to determine how much the uploaded song differs from the original. Given that this test and the test for the watermark can both be easily automated, they have a pretty good idea of how "cracked" their algorithms are.
Sure, they will still want to do more analysis of the samples. Any they probably want to try and wrangle the details out of the people who submitted the samples. But they know. And that's why they're pissed. They know which uploads have the watermark, and they know how similar to the original they all are. If this crap worked at all, they would have said something positive.
Curiously, at almost the same time, the RIAA announced a new program to prevent piracy. See the C|Net story here. The idea is to create a "digital bar code" for songs to allow tracking of copyright information, etc. On the face of it, this sounds like a whole new tack for the RIAA. Any maybe it is. However, the article says:
Nevertheless, if it proves to be difficult to tamper with, the system could be a potential way of identifying authorized and unauthorized songs being traded through services such as Napster.
It sounds to me like the exec's at the RIAA know that their SDMI scheme is going down in flames, and are looking for a backup. However, it sounds like they are going to head right back down the watermark path. Only this time, instead of trying to encode a single bit (play / don't play), they will be trying to encode hundreds or thousands of bits. Whoever brought up Don Quixote in an earlier SDMI thread was right on the money. These folks aren't ever going to learn.
Who ever decided to call this technology watermarking anyway? It's really a horribly inaccurate name. The whole point to a real watermark is that it's difficult to create, and easy for everyone to see. But with "digital watermarks" the idea is to make it invisible. An invisible watermark it what they should call it. Maybe then people would understand what a stupid idea it is.
I've heard, but don't know for a fact, that Sun does not allow M$ Office in their operations. All of their employee's were already using StarOffice. So, they have a strong incentive to see it improve. I don't see them getting "bored", as this isn't just an academic exercise.
I also don't know what kind of bulk discount M$ gives on Office licenses, but it is quite possible that Sun has considerably less invested in OpenOffice than they would pay to legally license Office2000 for all of their employees. I'm pretty sure that the ongoing costs of running OpenOffice, even with several dozen programmers assigned to it, would be less than paying M$ for upgrades every couple of years.
I think this just might be a testament to how greedy M$ licensing practices are. It's quite possible that it's cheaper for Sun to write their own office suite than it is to license M$'s "mass market" office suite.
Anybody notice that green thing on page 6? What is that, some sort of digital bottle opener? Weird.
I didn't say they were going to break into your local CO and physically take over your DSL line. But, if I figure out your IP address, I can ping it from anywhere in the world and you get to pay the bills for it. And depending on how well your ISP verifies the traffic and exactly what they bill on, there might not be a damn thing you can do about it. If they just count how many bytes flow to your modem from their DSLAM, I can send you all sorts of traffic and it is impossible for you to stop me. Like I said, it's an irritation if some kiddie floods you with traffic and makes your connection unusable for a few hours. It's a whole different game if they send a trickle of traffic all month long and you don't notice. Suddenly, you have a huge DSL bill, and no idea why.
My laptop has a ATI Rage Mobility in it that can apply what I think is a bi-linear filter to the image before it displays it. The result is that the image does not have any of the horrible aliasing that results with most video cards. The result is a little fuzzy, but rather than annoying it has an almost attractive "magnified" look to it.
The result is that the lower resolution are 100% usable. To be honest, I don't even know why vendors implemented the old stretching method at all. It looked so crappy as to be useless. I always turned it off, and just had the display use the center 640x480 pixels with a huge black border. At least then I could read things.
However, I don't know if this solution is available outside of laptops. It is a function of the video chipset, and I don't know if the desktop ATI's implement it. I also don't know how it would work with the VGA connection that most LCD displays use. I wouldn't be suprised if it didn't. But, if you get a laptop with the ATI Rage Mobility, you won't have this problem. However, you might want to check for a BIOS setting to let you turn the stretching off on your laptop. All the one's I've used have had an option for this. The 1:1 mapping isn't great because it's so small, but at least it looks OK.
One of the biggest problems with metered billing is what do you do when someone decides to rack up your bill? Let's say I'm some script kiddie whose on an unmetered line (say at college.) Let's say I do a ping with 10K packet size to your home DSL address at the default frequency of once a second for a month. By my math, that adds up to almost 25GB of traffic. One ping a second won't even get noticed by 99% of DSL users, until they get that $200 dollar ISP bill.
That's one of the big problems with metered billing. It's one thing when a script kiddie gets upset at you and floods you with traffic for a few hours. It's a whole different story if you get a huge bill from your ISP because of it.
If everyone goes to metered billing you will see all sorts of abuses as crackers try to set up servers on other peoples machines to avoid paying the bills for their traffic. Add that up to the aforementioned harrasment traffic jacking up peoples bills. Plus the dishonest users who blame their traffic spikes on "hackers".
I just don't see it being worth the headache for an ISP to charge by the byte. You can bet that any user that is hit by the above problems is going to run screaming to the nearest flat rate ISP. Besides, the rates are metered to a certain extent. Dialup access is not the same cost as OC-3 by a long shot. So all the dialup users are in the same cost pool. So what? They are in a different cost pool from the DSL users, who are in a different pool from the T3 users, etc.
There are way too many zeros in that price. It should be in scientific notation or Mega. Or they should at least add the appropriate commas. 2e07 dollars, or 20M$, or $20,000,000. But 20000000.00? You'd think for 20 million bucks they could buy a comma for the ad. And a JPEG of the whole thing. ;-)>
I also disagree that this is what every geek girl and boy needs. It's actually what every drug-running girl and boy needs. Which raises the interesting question of what Federal laws you might be violating cruising around in a submarine. I can't imagine that you could sail up the Potomic with this without raising some sort of ruckus.
And what exactly does the license agreement have to do with anything? There is no evidence that any of the people getting the letters have installed or are using the software. Hence, the license agreement has no relevance whatsoever. Many of the cuecats were sent out as gifts via the post office. Attempting to attach a license onto these gifts is actually a violation of federal law. In my case, I went to RatShack and asked for one. I was never asked to sign anything, nor asked for my name. The clerk simply handed it to me.
At what point do you think any of us entered into a contract with Digital Convergence? That's what a license agreement is: a contract. Did people enter into a contract simply by removing the package from their mailbox? Did they enter into a contract by opening the package? Did I enter into a contract by asking the RatShack clerk for a free scanner? Did I enter into a contract when he handed me the bag? I was never asked for my name, and certainly never signed anything. None of us have installed or used the software. Why bother? If the license agreement said everybody born in odd years owes Digital Convergence a dollar, would that mean I owe them a dollar? Of course not.
So why exactly are any of us bound by the license agreement on the software? Did I enter into a contract when you read the license agreement? The simple fact of the matter is these lawyers are sending out scary-sounding but completely worthless letters. I don't think they even qualify as a cease and desist letter. There is no IP infringement going on. The lawyers know that. They are just hoping that they can scare everyone with their worthless threats.
I am tired of this hot potato fantasy. Everyone uses BGP4 to determine which route should be taken by a packet. BGP4 only considers how many AS's are in the route. It doesn't know anything about hops, latency, etc. If I'm an ISP, and I have two routes for a packet with the same AS path length, of course I'm going to "get it off my network at the earliest possible time." Would you rather I ran it back and forth across the country a few times? The two paths are the same length, you should get it to whichever exit point is closest in your network.
Or do you think that the backbone ISP's are sending packets out of longer AS path points because they are closer in their network? I am almost positive that they are not doing this because your packet would probably never arrive at it's destination. If everybody was choosing longer AS paths, your packet would probably never arrive. It would just run around it loops as everyone shunts it off to someone else ignoring the AS path.
If you have specific details about what hot potato routing is, and how it differs from the correct routing that everyone does, please inform me. I would love to know. But I think this belongs firmly in the urban legend category.
This whole idea of voting against somebody is bullshit. There is no such thing. If you vote for Al Gore and the Democrats, you are voting for the DMCA. You are voting for the War on Drugs. You are voting for a political system in which bribes (aka soft money) are routinely exchanged in public.
You can tell yourself whatever lies you want, but the fact is that by voting for a Republicrat you are showing your support for the current corrupt system. Our current system is quickly failing. As long as people continue to buy into the FUD and vote for candidates which they know are corrupted by soft money and which they don't really want to see elected, it's going to keep getting worse. Every time somebody endorses the Republicrats with their vote, they are endorsing the current decline of our democracy into a cesspool of big-money corruption.
Actually, it's tougher than you might think to block Napster. You see, if you block Napster entirely, the students will use different port numbers to get around your block. It's a loosing battle for the network admins.
Now, I happen to know from first hand experience of one university which is limiting it's Napster traffic. Using various QoS tricks, they lower the priority of Napster traffic so that it doesn't consume too much of their bandwidth. A Napster "knob" if you will that lets them tweak the bandwidth it gets. They crank it down some, and the students just think that the link is more congested than it used to be. But they know if they crank it down too far, the students will realize that something is up and move to a different port. It's an interesting bit of network engineer / psychology going on.
If you actually go and download the files for the contest, you won't find much. Rather than any sort of description of the watermark technology, or any software that checks for the watermark, you get three .wav files. File 1 has no watermark. File 2 is the same audio as file 1 with a watermark applied. File 3 is a different song with a watermark applied. Your "challenge" is to remove the watermark from file 3. To check the file, you have to upload it to their server, and they will send you email with the results of the check.
So, from a cryptographic point of view, this is pretty worthless. It's along the lines of the newbies who post to sci.crypt saying "I've developed a new algorythm. Here is some ciphertext, crack it!". Of course, to do any valid analysis you need to know how the algorithm works.
My guess is that either the people setting up the "contest" are pretty clueless, or they have no faith in their algorithm, or both. Or this is just a publicity stunt to reassure the record labels. My money is on the latter.
Any hacker who attacks SDMI after it's released will certainly have access to a software implementation, or the algorithm, or both. So, to leave both of those out of the "contest" just makes it a sham.
Actually, this is a bad example. A good friend of mine owns a Thompson machine gun, and it is 100% legal. He had to get a license for it, which is not possible to do for fully automatic weapons manufactured after a certain date. But, when the "ban" on fully automatic weapons went into effect, existing guns were grandfathered. Not that this invalidates your point about ex post facto, I just wanted to point out that machine guns are a bad example.
Actually, I looked around some more on the companies web site. From the content there, they make it quite clear that their market plan is to force these books on the students:
;-)>
---quote---
Publishers are guaranteed 100% market penetration at partner schools who opt to implement the Vital Source system. Purchase of all included titles is mandated by the universities.
---/quote--- emphasis in original
Later on, and this is the best part, it talks about how they make you license the book even if you don't need it:
---quote---
Publishers receive a mandated, preset fee for every student for every title chosen by professors. Because the service is a global curriculum application, the fee comes in from each student each of the four years of their studies, regardless of whether they are taking that course that year.
---/quote---
Nothing like paying for that advanced quantum physics books when you're a freshman enrolled in basic mechanics, eh?
So I guess I'll be seeing you at the lynching
First, in the case of textbooks, the students usually have little to no say in what textbooks are used. So, I would expect that in many cases, the consumer will be forced to buy the product.
Second, it certainly does not appear that the time-limited version will be cheaper. Given that the company is talking about it's vastly increased revenue, they can't be charging a whole lot less. In fact, the Wired article states that the subscription fee is $1200/annum. Not exactly dirt cheap.
Third, this has nothing to do with contracts. The DMCA allows companies to prohibit any and all rights that a consumer had under copyright law. While I agree that the market should decide which products sink and which swim, the law should not allow the makers of bad products to have people thrown in jail on criminal charges for trying to assert their Constitutionally protected rights under existing copyright law. Make no bones about it, that's exactly what the DMCA does.
So, as someone who considers themselves libertarian, my problem is not with e-texts or DVD per se. It is with the fact that the DMCA allows the makers of these bad products to get other people thrown in jail for exposing the weaknesses in their products. I'm not asking for more government regulation. I'm asking for less, in the form of repealing the DMCA. It is truly the purest form of bad lawmaking I have ever seen.
I'm 100% positive the 6502 is an 8-bit processor, and I'm 99% positive that the Z-80 is too. Since endian refers to the order of bytes in a word, it doesn't make sense to refer to them as being big or little endian.
The old copies of Mosaic were never taken off of NCSA's FTP server You can find them here. This is a little more complete than Evolt.org since it includes the Unix binaries along with source for several versions. Don't forget to pick up your copy of NCSA HTTPD while you're there!
Actually, I don't believe this is the way SDMI will work . There is actually a lot of specification for it on their web site. Rather than not play songs that are not watermarked, they will begin to watermark CD's with a special "do not play" mark. So, in theory, the SDMI players will still play your existing MP3's but not MP3's made from "new" CD's (i.e., ones after they begin watermarking.) The players should also play any MP3 by a band that didn't watermark without any problem. In theory.
I wouldn't worry too much about SDMI though. It will sink faster that DiVX. In fact, it's already pretty much dead in the water. Lot's of companies signed up, but no products. They wanted to have them out last Xmas, but as far as I can tell right now there is only a single Sony SDMI unit available. More importantly, lot's of non-SDMI players are already on the market. So I think this battle has already been lost.
What would really be great to make the real issues apparent is for some consumer electronics manufacturer to produce a standard DVD player. But, instead of getting a CSS license, they could use the source to DeCSS.
Right now, the reality is that a program like DeCSS can be used to pirate DVD's. That makes it easy to paint it with the "evil" brush. However, if there was a hardware DVD player that was functionally just like an "authorized" DVD player, they couldn't use that tact. Provide an Apex-like menu to disable the region controls (but NOT macrovision), and the MPAA would be forced to admit that CSS and DMCA go far beyond copying and piracy.
Let's take the fictional scenario a bit further. Acme Inc makes a DVD player that is just like all the other home players, but doesn't include region controls. It also doesn't have a CSS license. Let's further suppose that the MPAA takes them to court for DMCA violations. What would the arguments look like in this case? If the MPAA said that CSS was an anti-copying mechanism, the question becomes why does Acme Inc. have to use it if you aren't copying the DVD? If they try to claim that the player is a piracy tool, you could point out that it doesn't enable piracy any more than a Sony or RCA DVD player.
So what would this player be doing that's "wrong"? Nothing. Well, almost nothing. By not having region controls, it's a way around the artificial trade barriers that the MPAA are enforcing to enrich themselves. And so the real issue would come to light before the customers. It's not about piracy. It's not about copying. It's about the DMCA being used as a shelf to place arbitrary rules and restrictions on. If this trial was about the MPAA's right to disable the FF button on your DVD player during ads, and the MPAA's right to restrict free trade via region encoding, the public's opinion of the MPAA and the DMCA might be a little less rosy.
It seems that the judge and others in this case have a hard time thinking of code as being expressive. It seems to me that the courts and the law have already overwhelmingly decided that it is expressive. Why? We allow copyright's on computer programs. Do we allow copyrights on objects which are purely functional? No. Copyright is reserved for things which are artistic or expressive, correct?
If the court feels that code is simply a tool, like a hammer or a screwdriver, why is it that code is deserving of copyright protection when hammers are not? And isn't a DVD just a program? Why does it get first amendment protection, even though by the judges own logic it isn't artistic or expressive?
The simple fact is that virtually 100% of the market and legal realities surrounding computer code treat it as expressive content deserving of both first amendment and copyright protection. It's only when a powerful interest decides to do something just plain wrong for it's own benefit does this stupid argument come up. Last time it was in the case of export restrictions on encryption. Now with the MPAA trying desparately to eliminate fair-use.
If code in non-artistic and non-expressive, I'm sure the MPAA won't mind me distributing an mpeg decoder source which #includes "matrix-vob.h", right?
Where did you get the idea that the Internet is run off of static routes? Sure, a lot of ISP's on the edge use statics. Primarily because they only need one default route to their upstream provider. But the backbone?
Have you heard of BGP? Border Gateway Protocol? That's what runs routing on the backbone, and it is the dynamic routing protocol. It's the duct tape that holds this thing together, and it's quite dynamic. It would probably take any backbone ISP (C&W, UUNet, Qwest, etc.) a week to statically configure that routes that work for one day. Never mind that the only way you would figure out how to configure the routes would be to use a dynamic routing protocol. The Internet is far too big and complex to ever manually configure it.
While it's true that BGP will still sometimes black-hole traffic by sending it down a broken link, that doesn't change the fact that it's dynamic. The problem stems from the fact that BGP can't always tell that a given route doesn't work. Usually it can, but not in all circumstances.
So, if you think one of the options is to boycott sites that accept ads from doubleclick, what are you doing here? Slashdot not only serves up doubleclick ads, sometimes the ads include the infamous cookies, too. Other alternative news sites feature doubleclick ads too. TheRegister seems to get most of it's ads from doubleclick (although to their credit, they've never sent me a cookie, unlike Slashdot).
This is what really disturbs me about doubleclick. They have become so ubiquitous that even sites whose creators are almost certainly not doubleclick fans are carrying the ads. Either in spite of or because of their outragous behavior, they are far and away the most successful ad banner company.
You say that if we boycott doubleclick.net sites, we will force doubleclick to rethink their policies. Will they? Or are we just going to cut off the air supply for the sites we DO care about? I have to say, it's hell of a dilema, and I don't see a clear way out.