RIAA has, in some sense, already lost the MP3 suit, because the lost the Rio suit. (Or did they just drop the case? I don't remember exactly.) Since the Rio is legal, so are the MP3s it plays.
It's a lost cause anyway. It's much harder to attack a file format, because it has so many uses. I've seen MP3 being used for so many legal uses that it's highly unlikely it would ever be judged illegal. It's just a compression format.
I haven't read the injunction text (it loads too slow, and I'd probably not understand the legalese anyway), but if Napster is ordered to make sure no copyright violations take place, it is in their best interests to respond by shutting down their servers. This is because their entire line of defense lies in the claim that they're unable to monitor copyright violations because they're nothing more than a common carrier.
Suppose a judge ordered AT&T to make sure no discussion of terrorist activities take place on their networks. The only sane way they could comply is by shutting down their entire phone system. It's in Napster's best interest to appear to be the same.
As far as I remember, it wasn't NSA which invented DES, rather it was IBM. However, the NSA changed some of the coeffecients of the original design and didn't explain why.
30 years down the road, Adi Shamir invented differential cryptoanalysis (a method of attacking crypto systems by 'feeding' them certain inputs and seeing what comes out), and showed how the original design of DES was vulnerable to that method, and that the NSA's changes made DES much less vulnerable.
It was later revealed that NSA had already discovered differential cryptoanalysis in the 60's, and the coeffecient changes were specifically done to protect DES.
The things Microsoft are often accused of - trying to run down competitors, bullying other companies into affiliations, etc. - don't require a conspiracy of any sort. It's not a conspiracy, it's a (stated or unstated) company policy, which gets decided by a small group of people. These people can decide what they want, and no conspiracy is required to get that to happen.
If the senior management of Microsoft decides to create a free Internet browser to gain market dominance, no coercion or conspiracy of employees is required. They just do their job - designing, writing and marketing a Web browser which just happens to be distributed for free.
The upshot of all this is that Microsoft is easily guilt of all the 'conspiracies' it's been accused of - trying to dominate the computer market, in essence (this is not really a conspiracy, and it's not secret - Judge Jackson ruled it quite clearly) - with or without the support of its employees. They just do their day-time jobs.
Don't confuse conspiracy with company policy. I resent Microsoft because of its unstated policy, not because of some vague conspiracy.
I haven't read the novel in question, but if the three planets have a low (<0.1c) relative velocity to one another, realitivistic effects are negligible and simulatenity can be defined, especially when these events take a considerable amount of time (mass extinctinos) so there's a big tolerance factor in the definition of simulatenity.
Variables you don't have to initialize aren't the same as variables you don't have to declare. Think Java or Visual Basic with "Option Explicit" - you have to declare your variables, but you don't have to initialize them (give them an initial value).
Yes, it can hide bugs. But on the other hand, it hides bugs reproducibly - i.e. if your program unwittingly depends on a variable being initialized, it's bad code, but at least it's code that's guaranteed to work. It can't fail in some bizarre situation as with uninitialized variables.
I still prefer the Java approach of having the compiler make sure there are no uninitialized locals (though sometimes it can bite back when you outsmart the compiler and it can't recognize a valid initializing situation). But note that global variables are guaranteed initialized - because Java can't make sure they're always initialized (that would entail a costly program-wide analysis), so it too sacrificies bug unmasking for guaranteed identical and correct exceution by initializing them all itself.
To me, it seems that Bidder's Edge have been unethical, taking away from eBay's ad revenues. The question is what they did wrong that search engines didn't.
I think the answer lies in compilation copyright. This is a special type of copyright granted for an aggregate of information, even if the individual bits of information have others' (or no-one's) copyright. This is what protects, for instance, an anthology or index.
eBay could claim that even though their pages are freely viewable, they own compilation copyright on the whole of the work, i.e. the entire bidding collection and its indexing. There is merit to that claim, as they undoubtedly have to invest a lot of energy in collecting, maintaining and indexing all biddings. In this case, by essentially copying the entire collection and offering their own search facilities, Bidder's Edge have copied and violated eBay's compilation copyright on the work. Note that the violation lies not in downloading the entire archive (they are entitled to it as much as anyone is entitled to remember what she reads), but in offering that archive to other people.
Other search engines do not violate compilation copyright, because they are essentially having their own compilation copyright: they index a much broaded range of works, and do not offer direct access to the individual compilation of eBay. They have the same rights as the 'keyword lookup' index books you occasionally find in libraries, which are an immense help when doing research.
...please please please do not form a FUD army like "Team Linux". The OS/2 Teamers were the worst possible zealots, installing an attitude of defeatism and victimization...
It really didn't seem to me like that at the time, and still doesn't now. Maybe it was just the particular TeamOS/2 Israel I was in and its 'leadership', but back then it was just a bunch of harmless guys who enjoyed diddling with OS/2. We had some support from IBM Israel, especially in organizing meetings, but on the whole it didn't seem more than a hobby of a few dedicated people. There wasn't much zealotry, not even in the mailing list, and when bit-by-bit members of the team resigned to using NT or Linux because they had enough of a dead OS, it was met with understanding, not scorn.
All in all, it was a nice experience, even if 'we didn't win in the end'...
Yes, it might have caused him trouble. But occasionally trouble is what you have to ask for if you want to protest. Just like Greenpeace protesters who want to be arrested so as to attract media attention.
The important thing here is that MPAA has ABSOLUTELY NO right to ask that page to be removed, because it has NO relevance to the court case.
As it is, it portrayed the MPAA for exactly what it is: a bunch of spooks who send their lawyers to shut the mouth of anyone that mentions terms that frighten them. (And it's even more frightening that they're chasing people in the UK. Since when did the DMCA apply outside the USA? It's not a case of international copyright control - the Berne convention, AFAIK, does not prohibit reverse-engineering, which is just why the DMCA was required in the first place.)
I think it'd actually make an excellent publicity stunt if used properly.
I can't seem to understand why this is anything new. Communication satellites have been with us for a very long time, and many of them are in active use as Internet relays. So what's the big difference? That now the satellite itself responds to PINGs, instead of just relaying the packets along? Doesn't seem much of a news...
And by the way, CNET considers viewing the contents of a text file stuck in a program directory "child's play", but turning off or viewing a disk cache is "advanced" and "sophisticated".
Oh, but they know exactly what they're talking about. After all, turning off the disk cache is something that needs to be done by the parent, while viewing of the text file has to be done by the child. 9 out of 10 homes, the parent is the more clueless of the two.
CNET (The Home Of Accurate, Unbiased Reporting TM) has taken some selective examples of blocked sites and attempted to have those indicate some kind of political agenda.
I don't think they were trying to show a political agenda, but only to demonstrate the inherent dangers of these lists. I mean, even without an actual political agenda, the very bias is an inherent free speech issue.
It's a very sad moment when someone who provides a free service to the community has to shut down because part of the community spits back at it.
Anyway, I doubt this would work. A few months ago the Undernet IRC network had announced that because of continuing DoSes and abuses from Israeli users, it would refuse any connections from Israel.
This steered quite a bit of media attention (for, like a day or so). Some KewL DudEZ kiddies were asked for a reply. At least two of them swore revenge on the network for their defacement of Israel, saying they're retaliate by attacking and DoS'ing them. In other words, they said something like "You're calling me violent, you f***ing ****!? I'll hit you until you take it back!".
My point? Script kiddiez never learn. Anyone who thinks taking down voluntary services is a viable option, will just find some way or another to roll the blame for shutting down the networks back onto the IRC. I can already imagine some of them DoS'ing the servers just as soon as they come back online in some sort of perverted revenge.
The salt is irrelevant because we're not actually trying to hide anything - just get the source code (which has to be available because by definition it's under GPL) and you can instantly 'decipher' the hash (that's the intention of the hashed patch, after all). All we're trying is to avoid spelling out the source exactly.
As for getting over multiple occasions of the same line - all these problems have isomorphic problems in the original patch, so we just have to do whatever patch does, only using hashes instead of explicit text. I'm not sure how it's solved there, but you can probably ask Larry Wall:)
A more interesting problem that has occured to me is that even the actual patch text of a line gives it explicitly - so we might actually have to use a sort of 'character-based patching' (i.e. something like "add the word 'i' after character 5") to avoid giving away anything that was in the original text.
Anyway, I think it has all become irrelevant because according to someone else's post, even this sort of patch is considered "derivative work" under the GPL. Ah well.
You don't necessarily need the actual text of the original code in order to provide context. In principle you could use some sort of sophisticated hashing scheme to give context without giving away the actual text.
It's boils down to something like replacing "put '++i;' after 'if (p)'" to "put '++i;' after the line whose hash is 06A204EF". A proper hashing algorithm, like MD5, can give you both a very high certainty (as in, one in 2^128 or so) that you're patching the right line, but give away nothing about the original text (which could still be one of 2^infinity possibilities, any of which would be computantially impratical to find).
The interesting question is whether an MD5 of a portion of GPL code is considered itself GPL... It's the kind of question that makes me want to just forget the whole issue and go to sleep.
Slashdot Mirror
It's a lost cause anyway. It's much harder to attack a file format, because it has so many uses. I've seen MP3 being used for so many legal uses that it's highly unlikely it would ever be judged illegal. It's just a compression format.
Suppose a judge ordered AT&T to make sure no discussion of terrorist activities take place on their networks. The only sane way they could comply is by shutting down their entire phone system. It's in Napster's best interest to appear to be the same.
30 years down the road, Adi Shamir invented differential cryptoanalysis (a method of attacking crypto systems by 'feeding' them certain inputs and seeing what comes out), and showed how the original design of DES was vulnerable to that method, and that the NSA's changes made DES much less vulnerable.
It was later revealed that NSA had already discovered differential cryptoanalysis in the 60's, and the coeffecient changes were specifically done to protect DES.
If the senior management of Microsoft decides to create a free Internet browser to gain market dominance, no coercion or conspiracy of employees is required. They just do their job - designing, writing and marketing a Web browser which just happens to be distributed for free.
The upshot of all this is that Microsoft is easily guilt of all the 'conspiracies' it's been accused of - trying to dominate the computer market, in essence (this is not really a conspiracy, and it's not secret - Judge Jackson ruled it quite clearly) - with or without the support of its employees. They just do their day-time jobs.
Don't confuse conspiracy with company policy. I resent Microsoft because of its unstated policy, not because of some vague conspiracy.
I haven't read the novel in question, but if the three planets have a low (<0.1c) relative velocity to one another, realitivistic effects are negligible and simulatenity can be defined, especially when these events take a considerable amount of time (mass extinctinos) so there's a big tolerance factor in the definition of simulatenity.
Yes, it can hide bugs. But on the other hand, it hides bugs reproducibly - i.e. if your program unwittingly depends on a variable being initialized, it's bad code, but at least it's code that's guaranteed to work. It can't fail in some bizarre situation as with uninitialized variables.
I still prefer the Java approach of having the compiler make sure there are no uninitialized locals (though sometimes it can bite back when you outsmart the compiler and it can't recognize a valid initializing situation). But note that global variables are guaranteed initialized - because Java can't make sure they're always initialized (that would entail a costly program-wide analysis), so it too sacrificies bug unmasking for guaranteed identical and correct exceution by initializing them all itself.
I think the answer lies in compilation copyright. This is a special type of copyright granted for an aggregate of information, even if the individual bits of information have others' (or no-one's) copyright. This is what protects, for instance, an anthology or index.
eBay could claim that even though their pages are freely viewable, they own compilation copyright on the whole of the work, i.e. the entire bidding collection and its indexing. There is merit to that claim, as they undoubtedly have to invest a lot of energy in collecting, maintaining and indexing all biddings.
In this case, by essentially copying the entire collection and offering their own search facilities, Bidder's Edge have copied and violated eBay's compilation copyright on the work. Note that the violation lies not in downloading the entire archive (they are entitled to it as much as anyone is entitled to remember what she reads), but in offering that archive to other people.
Other search engines do not violate compilation copyright, because they are essentially having their own compilation copyright: they index a much broaded range of works, and do not offer direct access to the individual compilation of eBay. They have the same rights as the 'keyword lookup' index books you occasionally find in libraries, which are an immense help when doing research.
It really didn't seem to me like that at the time, and still doesn't now. Maybe it was just the particular TeamOS/2 Israel I was in and its 'leadership', but back then it was just a bunch of harmless guys who enjoyed diddling with OS/2. We had some support from IBM Israel, especially in organizing meetings, but on the whole it didn't seem more than a hobby of a few dedicated people. There wasn't much zealotry, not even in the mailing list, and when bit-by-bit members of the team resigned to using NT or Linux because they had enough of a dead OS, it was met with understanding, not scorn.
All in all, it was a nice experience, even if 'we didn't win in the end'...
The important thing here is that MPAA has ABSOLUTELY NO right to ask that page to be removed, because it has NO relevance to the court case.
As it is, it portrayed the MPAA for exactly what it is: a bunch of spooks who send their lawyers to shut the mouth of anyone that mentions terms that frighten them.
(And it's even more frightening that they're chasing people in the UK. Since when did the DMCA apply outside the USA? It's not a case of international copyright control - the Berne convention, AFAIK, does not prohibit reverse-engineering, which is just why the DMCA was required in the first place.)
I think it'd actually make an excellent publicity stunt if used properly.
I can't seem to understand why this is anything new. Communication satellites have been with us for a very long time, and many of them are in active use as Internet relays. So what's the big difference? That now the satellite itself responds to PINGs, instead of just relaying the packets along? Doesn't seem much of a news...
Oh, but they know exactly what they're talking about. After all, turning off the disk cache is something that needs to be done by the parent, while viewing of the text file has to be done by the child. 9 out of 10 homes, the parent is the more clueless of the two.
CNET (The Home Of Accurate, Unbiased Reporting TM) has taken some selective examples of blocked sites and attempted to have those indicate some kind of political agenda.
I don't think they were trying to show a political agenda, but only to demonstrate the inherent dangers of these lists. I mean, even without an actual political agenda, the very bias is an inherent free speech issue.
Anyway, I doubt this would work. A few months ago the Undernet IRC network had announced that because of continuing DoSes and abuses from Israeli users, it would refuse any connections from Israel.
This steered quite a bit of media attention (for, like a day or so). Some KewL DudEZ kiddies were asked for a reply. At least two of them swore revenge on the network for their defacement of Israel, saying they're retaliate by attacking and DoS'ing them. In other words, they said something like "You're calling me violent, you f***ing ****!? I'll hit you until you take it back!".
My point? Script kiddiez never learn. Anyone who thinks taking down voluntary services is a viable option, will just find some way or another to roll the blame for shutting down the networks back onto the IRC. I can already imagine some of them DoS'ing the servers just as soon as they come back online in some sort of perverted revenge.
As for getting over multiple occasions of the same line - all these problems have isomorphic problems in the original patch, so we just have to do whatever patch does, only using hashes instead of explicit text. I'm not sure how it's solved there, but you can probably ask Larry Wall :)
A more interesting problem that has occured to me is that even the actual patch text of a line gives it explicitly - so we might actually have to use a sort of 'character-based patching' (i.e. something like "add the word 'i' after character 5") to avoid giving away anything that was in the original text.
Anyway, I think it has all become irrelevant because according to someone else's post, even this sort of patch is considered "derivative work" under the GPL. Ah well.
It's boils down to something like replacing "put '++i;' after 'if (p)'" to "put '++i;' after the line whose hash is 06A204EF". A proper hashing algorithm, like MD5, can give you both a very high certainty (as in, one in 2^128 or so) that you're patching the right line, but give away nothing about the original text (which could still be one of 2^infinity possibilities, any of which would be computantially impratical to find).
The interesting question is whether an MD5 of a portion of GPL code is considered itself GPL...
It's the kind of question that makes me want to just forget the whole issue and go to sleep.