That was a well-written roundup of the current wave of threats from the IP cartel. Some minor points, though:
Rivest's question was the basic libertarian one: Why shouldn't Alice be allowed to sell Bob a device that's restrained by a copy protection scheme? In answering this, Gilmore dips into the bad motivations and bad consequences of these schemes. But to stay within Rivest's framework, the answer is simply that Alice should be allowed to sell the box with or without said restraints. The root problem here is not companies hatching restrictive schemes. It's the government intervening and giving these bad schemes the force of law. Without government interference, these schemes would be rejected by consumers.
The stuff about misleading the consumer, while valid, doesn't address Rivest's question. Technology vendors always omit the downside of their products when advertising. I too find it infuriating but it is not specific to these issues.
"Companies that can't adjust should disappear and be replaced by those who can." I heartily agree. But let's translate that one honestly: "Please fuck off and die, dinosaurs." Remember evolution does not work via the adaptation of organisms, but rather of species. The existing IP companies have nothing to offer in the new world, and no hope of adapting. Therefore they must fight to the death to maintain the old one. There has been too much mealy-mouthed talk on both sides of the fence about IP giants adapting to the internet age.
The 'straightdope' URL was interesting, but I wonder about its accuracy. Towards the end, the author writes:
...of the 161,000 people who wrote to the DMA last year, 116,000 wanted more junk mail. They were sent a booklet entitled "How To Get More Interesting Mail" (as God is my witness, I am not making this up)...
How can this person know this statistic with the degree of certainty he's implying? It sounds to me like he took the word of the DMA as gospel.
I can't believe I'm entering a gun debate. Especially because your larger point about data protection is valid. But here goes:
Going a step further, let's say I can shoot a mugger first (a big and highly unlikely assumption) 90% of the time. By the seventh mugging, my overall survival rate has dropped to less than 50%.
This reasonings assumes that potential muggers are not affected by information about the outcomes of muggings. In fact, when Bernard Goetz shot his would-be assailants, violent street crime in New York decreased substantially. If you actually managed to shoot six muggers, I don't think you'd be able to find a seventh. Concealed carry laws decrease violent crime, not because every victim is guarranteed to be armed (they aren't) but because the risk of an armed victim rises to an unacceptable level compared to the small rewards. Anyone willing to take such serious risks could rob a liquor store instead.
Heck, if you assume the average mugger shoots right at the 50th percentile (which is far too low)...
Leaving aside the fact that muggers rarely use guns, I disagree with your implied assertion that criminals have better gun skills. I think that legitimate gun owners are more likely to practice at the range and become thoroughly familiar with their weapons. We'd be much better off if criminals learned to use their guns correctly, because we wouldn't have so many instances of gangsters shooting bystanders by accident.
Couldn't Public key cryptography or one way hash functions be considered security through obscurity?
Yes; there's a fine line. Good security means that the variable, secret part of the algorithm is small and well-defined, and that there are no secrets elsewhere in the system. The first step towards an analysable system is to separate key from algorithm. All the secrecy is concentrated in the key. And the key should be 100% random - pure entropy. Among many other benefits, this enables us to answer the question, "How many bits of entropy does the key have?", which also says, "How long would it take to brute-force this system?" Many failures of real-world security occur because users didn't know which parts of the system needed to be secret - they were told 'everything is secret'.
I use the mouse quite a bit. I'm always pasting subnets or hostnames or usernames between netscape, shell and mutt. I had a two-button mouse for a while with Emulate3Buttons and I nearly threw my computer out the window before I got a three-button.
This isn't about Mac users wanting to run Linux, it's about Linux users looking for a good notebook computer to run Linux.
UNIX/Linux people, on the other hand, can not justify buying Apple hardware just to run GNU/Linux, because you can build a dual-processor SMP Pentium III box with a shiteload of RAM for half of what an SMP G4 will cost you.
That would be true if we were talking about desktop machines. The topic here is notebooks. I think a lot of Linux users perceive, rightly or wrongly, that Intel notebooks have poor battery life as a result of all the transistors doing backwards-compatibility things so everyone's favorite obsolete OS can keep working. Thus the interest in RISC notebooks.
MacOS X will make Linux PPC on Macs obselete, by the way.
I doubt that. If I bought a Powerbook to run Linux on, I would have little interest in MacOS X. No matter what great features it has, it's proprietary and tied in to the Apple mindset, which I don't like. I wish Apple all the best in marketing Os X to their core market, but they know and I know that it's not aimed at Unix geeks.
I just skimmed through Kaplan's decision again, and you're right - I overlooked the fact that Kaplan was judging the constitutionality of the DMCA. Kaplan decided that source code is no more free speech than a car key is. While I disagree with this idea, I think we need congress to validate the protection of source code as free speech - asking a judge, working from legal precedent, to make that leap may be asking too much. I don't rule out the possibility that Kaplan was corrupt or biased. Certainly, he's a former attorney for the IP interests. But I think it's too easy and too satisfying to write him off as corrupt because he ruled against us. We have to recognize that the ideas which prevail on the internet are not yet widely accepted by the ruling class. And I'm not sure how we're going to change that.
There is some logic to what you say, but you may be underestimating human stupidity. The creators of the vote-o-matic probably thought that if someone failed to punch all the way through the card, he'd notice the error after withdrawing the card from the slot, and then reinsert the card and correct the problem. Also, one of the errors in Palm Beach was, given this ballot: Alice @ @ Zorro Bob @ Some people accidentally voted for Zorro because they were (unconsciously?) trying to split the difference between Alice and Bob. Just my interpretation. Slashdotters sometimes get caught up in the technological aspects of a problem. The real problem here is "you can lead a horse to water but you can't make him drink." There will always be some fraction of voters who just aren't competent to vote. We gave up on competency tests because they are always politically biased.
First of all, I'm not persuaded that the Canadian system is trouble-free. I think it hasn't been scrutinised as closely as the Floridian system. Just to cite one part of the Canadian system that arouses curiousity, a previous poster stated that the ballots have black backgrounds with white circles where the voter places his mark. This would automatically reject ambiguous marks made between the circles. I'm sure the Palm Beachers are quite capable of making such ambiguous marks.
From another angle, the cost isn't high if you count by batches, eliminate the ones that everyone agrees upon...
Here's how I see that playing out in the election of Alice vs Bob:
Of 100 batches of votes, 97 have been counted and agreed upon by Alice's volunteers and Bob's volunteers.
Seeing that it's a close race, Alice demands a recount of certain 'agreed-upon' batches on the pretext that volunteer Annie, who represented Alice in the counting of these batches, had once worked for Bob.
Seeing the success of Alice's tactic, Bob (after condemning that tactic) demands a recount of several other batches on the pretext that Bob's volunteer Berverly had poor eyesight and just went along with the Alicites in the count.
Repeat with minor variations until some court intervenes.
By the way, your ad hominem remarks do little for your argument.
I hope that we won't start using Kaplan as a synonym for bad judges or bad judging. I read most of the transcripts and Kaplan was quite intelligent and had a much better grasp of the case than the lawyers for either side. Kaplan was impressed by Turetsky's demonstration that code could be speech. In the end, Kaplan enforced the DMCA as written because the defenses Congress had allowed in the DMCA simply aren't usable in the real world. The blame for Kaplan's judgement rests on Congress, which passed a terrible law. Kaplan did his job as a judge in interpreting the law.
Unix: You think it won't work, but if you find the right guru he'll make it work.
Re:...but will it keep up with the upgrades?
on
Laser-equipped 747
·
· Score: 2
I'll take a guess. During the cold war the US developed a series of early warning satellites called Vela. The Vela satellites use infrared sensors to spot the thermal plume of a rocket taking off. It is extremely hot, fairly vertical, and very distinctive. Maybe this plane uses similar technology to identify missiles during launch. By the way, Vela and related technologies played a major role in anti-test treaties - they enabled the US and USSR to verify that the others weren't testing missiles.
First, you're right that most of the responses are idiotic, particularly the 'genetic-engineered corn' type. Also the ones that confuse this weapon with strategic missile defense. However, the post you're quoting was not saying either of those things. The post was claiming that it's illegal to paint a military aircraft as a civilian aircraft.
And of course Clinton took the selling of the presidency to new heights, with an implicit fee schedule: for example $500K let you sleep over at the white house for one night. Still wondering what he charged China for the ICBM rocketry technical assistance.
No, they should not run anything as complex as Linux. A voting machine does not need multi-tasking or networking. They should run a minimal program, written in assembly language, which would be available on the internet for public scrutiny months before the election.
count the votes until you get the same answer twice, something any two-toothed cretin who runs a cash register knows.
That may work for two-toothed cretins and cash registers, but it doesn't scale. When it comes to accounting, the larger the sums involved, the larger the acceptable margin of error. There is a cost to accuracy, which in this case is borne by unpaid volunteers, and it's not reasonable to ask them to redo their task forever in search of perfection.
And yet I find it hard to trust any of these systems. I wonder why the ballot has a black background? I'd say it's to hide ambiguous votes - marks made between two of the circles. So the vote counters can say with a clear conscience, 'there was no mark.' Assuming that their are Canadians as stupid as the ambiguous voters in Florida, there are probably many Canadian ballots with an X half way between the white circles.
Another reason to support it if you're a slashdot fan: Slashdot got where it is by experimenting. It isn't obvious in advance what will work. But the pace of experimentation has slowed since the buyout and vastly increased popularity. Slashdot is no longer likely to break new ground. This is fine - they now have an important role as a well-understood service. Smaller projects like half-empty can afford to innovate because they have less to lose. And if they find some new principle to apply to online communities, maybe slashdot will implement the idea after it's been proven.
Any Unix and certainly W2K can do anything Linux can do at least as well as Linux can do it.
And you say:
In short, judging NT by how it performs on some metric which is only critical to Unix architectures is (a) unfair and (b) meaningless.
While I realize what you're getting at (I wouldn't want Unix judged on how well it 'edited the registry' or something), it's inherently tricky to counter that claim without running up against the charge of unfairness. For example, before Winsock a Unix geek could have said "Unix does TCP/IP better - in fact Windows doesn't do TCP/IP at all!" And you could have responded, "TCP/IP is a Unix thing. Why judge Windows by something irrelevant like that?" So what do you consider a fair basis of comparison? Shall I dredge up a list of Windows horror stories, only to be told that each one is caused by bad application programming or system administration? Because if that's the case, you're arguing that there's a grain of wheat under a mountain of dung - something which might interest academics, but not hungry people.
Damn, there goes my theory. And I was so hoping to free our language of one awkward construct. Damn you, wolves, for injecting real-world complexity into a nice theory!
If MS starts an all-out attack on Linux, I think it will backfire in their faces.
M$ are smart, and I'm sure they've thought about that. In fact, I'm sure they've been debating for at least a year when to open a full-fledged assault on Linux. Obviously, if they do it too early they simply broadcast the existence of Linux to the masses. If they wait too late, it seems like they can't think of anything to say on the subject - in other words, 'they know they're licked'. But I agree, if M$ attacks Linux full force they will hurt themselves. If they use some sneaky and careful tactics, though, they can sabotage the acceptance of Linux by the clueless majority.
Any Unix and certainly W2K can do anything Linux can do at least as well as Linux can do it.
Are you sure? Which forks a new process faster on equivalent hardware? While I haven't used W2K, I know that NT is abysmally slow at forking. Why is disk I/O so slow under Windows? Just my subjective recollection, but saving or loading a file under Windows seems to take much longer than its size would warrant. I'm not going to pursue this, because the phrase 'can do' covers a multitude of expert tweaking, as witnessed by various Microsoftian benchmarks. I will remark that out of the box Linux performs a wide range of tasks nicely; commercial Unix a somewhat smaller range of tasks, as the vendors are less generous in packaging useful software, and Win2K very little that I can see.
Otherwise, get real; a program loader, scheduler, etc circa 1970 isnt a threat to anyone, technologically.
But Ballmer is not an academic concerned with threats to his theories; he's a businessman concerned with threats to his profits. Linux may strike you as old technology, but its the kind of technology that gets things done. Your statement is equivalent to: 'a CPU, RAM, disk, etc. circa 1970 isn't a threat to anyone, technologically.' And yet these elements have evolved, and vendors compete in selling them. As for the reality of the Linux threat, remember when Microsoft was poised to 'inevitably' take over the web server market? Consult netcraft to see what became of that dream.
To survive the breakup. Thus,.NET..NET will become the true Microsoft platform, hence the seat of their power. The OS will become unimportant. Before breakup: M$ = apps + OS + (Gates, Ballmer, etc.) After breakup: M$ = AppsGroup = apps +.NET + (Gates, Ballmer, etc.). OS_Group = ( OS ). The Apps group will have all the power because it controls the real platform (.NET). The OS Group will be worthless because it can't act independently. The Apps group can always jettison the OS by porting.NET to linux or whatever. Just a theory.
The fact that most of us are essentially at the whim of the big players of the system in which we choose to participate is an uncomfortable one.
Your mailbox is at the whim of any idiot with an internet connection. As long as a person's emotional well-being is linked to the contents of his mailbox, he is doomed to frequent bouts of anger. That's why I think the 'war on spam' is ultimately a dead end - there's always one more idiot ready to spam. The guy who wrote '7 Habits of Highly Effective People' (what's his name?) talks about reducing your circle of concern to match your circle of influence. You can control your mail server and filtering software; you can't control the internet. The MPAA executive has a heart attack because he can't control the spread of information on the internet. The geek has a heart attack because he can't control the flow of mail into his system. Maybe there's a parallel with household phones. In the 1970's it was normal that if you dialled a house a loud alarm-like bell went off, and the occupants dropped whatever they were doing and answered the caller. That worked until it was systematically abused by pranksters and telemarketers. Now the norm is for an answering machine to screen the call, and increasingly caller ID is required. Any protocol which allows you to make me jump via remote control is broken and will be exploited eventually.
Slashdot Mirror
- Rivest's question was the basic libertarian one: Why shouldn't Alice be allowed to sell Bob a device that's restrained by a copy protection scheme? In answering this, Gilmore dips into the bad motivations and bad consequences of these schemes. But to stay within Rivest's framework, the answer is simply that Alice should be allowed to sell the box with or without said restraints. The root problem here is not companies hatching restrictive schemes. It's the government intervening and giving these bad schemes the force of law. Without government interference, these schemes would be rejected by consumers.
- The stuff about misleading the consumer, while valid, doesn't address Rivest's question. Technology vendors always omit the downside of their products when advertising. I too find it infuriating but it is not specific to these issues.
- "Companies that can't adjust should disappear and be replaced by those who can." I heartily agree. But let's translate that one honestly: "Please fuck off and die, dinosaurs." Remember evolution does not work via the adaptation of organisms, but rather of species. The existing IP companies have nothing to offer in the new world, and no hope of adapting. Therefore they must fight to the death to maintain the old one. There has been too much mealy-mouthed talk on both sides of the fence about IP giants adapting to the internet age.
All in all, a really good piece.He's certainly not supposed to append "As God is my witness" to dubious factoids from unnamed sources.
How can this person know this statistic with the degree of certainty he's implying? It sounds to me like he took the word of the DMA as gospel.
This reasonings assumes that potential muggers are not affected by information about the outcomes of muggings. In fact, when Bernard Goetz shot his would-be assailants, violent street crime in New York decreased substantially. If you actually managed to shoot six muggers, I don't think you'd be able to find a seventh. Concealed carry laws decrease violent crime, not because every victim is guarranteed to be armed (they aren't) but because the risk of an armed victim rises to an unacceptable level compared to the small rewards. Anyone willing to take such serious risks could rob a liquor store instead.
Leaving aside the fact that muggers rarely use guns, I disagree with your implied assertion that criminals have better gun skills. I think that legitimate gun owners are more likely to practice at the range and become thoroughly familiar with their weapons. We'd be much better off if criminals learned to use their guns correctly, because we wouldn't have so many instances of gangsters shooting bystanders by accident.
Yes; there's a fine line. Good security means that the variable, secret part of the algorithm is small and well-defined, and that there are no secrets elsewhere in the system. The first step towards an analysable system is to separate key from algorithm. All the secrecy is concentrated in the key. And the key should be 100% random - pure entropy.
Among many other benefits, this enables us to answer the question, "How many bits of entropy does the key have?", which also says, "How long would it take to brute-force this system?"
Many failures of real-world security occur because users didn't know which parts of the system needed to be secret - they were told 'everything is secret'.
I use the mouse quite a bit. I'm always pasting subnets or hostnames or usernames between netscape, shell and mutt. I had a two-button mouse for a while with Emulate3Buttons and I nearly threw my computer out the window before I got a three-button.
That would be true if we were talking about desktop machines. The topic here is notebooks. I think a lot of Linux users perceive, rightly or wrongly, that Intel notebooks have poor battery life as a result of all the transistors doing backwards-compatibility things so everyone's favorite obsolete OS can keep working. Thus the interest in RISC notebooks.
I doubt that. If I bought a Powerbook to run Linux on, I would have little interest in MacOS X. No matter what great features it has, it's proprietary and tied in to the Apple mindset, which I don't like. I wish Apple all the best in marketing Os X to their core market, but they know and I know that it's not aimed at Unix geeks.
I just skimmed through Kaplan's decision again, and you're right - I overlooked the fact that Kaplan was judging the constitutionality of the DMCA. Kaplan decided that source code is no more free speech than a car key is. While I disagree with this idea, I think we need congress to validate the protection of source code as free speech - asking a judge, working from legal precedent, to make that leap may be asking too much.
I don't rule out the possibility that Kaplan was corrupt or biased. Certainly, he's a former attorney for the IP interests.
But I think it's too easy and too satisfying to write him off as corrupt because he ruled against us. We have to recognize that the ideas which prevail on the internet are not yet widely accepted by the ruling class. And I'm not sure how we're going to change that.
There is some logic to what you say, but you may be underestimating human stupidity. The creators of the vote-o-matic probably thought that if someone failed to punch all the way through the card, he'd notice the error after withdrawing the card from the slot, and then reinsert the card and correct the problem.
Also, one of the errors in Palm Beach was, given this ballot: Alice @ @ Zorro Bob @ Some people accidentally voted for Zorro because they were (unconsciously?) trying to split the difference between Alice and Bob. Just my interpretation.
Slashdotters sometimes get caught up in the technological aspects of a problem. The real problem here is "you can lead a horse to water but you can't make him drink."
There will always be some fraction of voters who just aren't competent to vote. We gave up on competency tests because they are always politically biased.
Here's how I see that playing out in the election of Alice vs Bob:
By the way, your ad hominem remarks do little for your argument.
I hope that we won't start using Kaplan as a synonym for bad judges or bad judging. I read most of the transcripts and Kaplan was quite intelligent and had a much better grasp of the case than the lawyers for either side.
Kaplan was impressed by Turetsky's demonstration that code could be speech. In the end, Kaplan enforced the DMCA as written because the defenses Congress had allowed in the DMCA simply aren't usable in the real world.
The blame for Kaplan's judgement rests on Congress, which passed a terrible law. Kaplan did his job as a judge in interpreting the law.
I'll take a guess. During the cold war the US developed a series of early warning satellites called Vela. The Vela satellites use infrared sensors to spot the thermal plume of a rocket taking off. It is extremely hot, fairly vertical, and very distinctive.
Maybe this plane uses similar technology to identify missiles during launch.
By the way, Vela and related technologies played a major role in anti-test treaties - they enabled the US and USSR to verify that the others weren't testing missiles.
First, you're right that most of the responses are idiotic, particularly the 'genetic-engineered corn' type. Also the ones that confuse this weapon with strategic missile defense.
However, the post you're quoting was not saying either of those things. The post was claiming that it's illegal to paint a military aircraft as a civilian aircraft.
And of course Clinton took the selling of the presidency to new heights, with an implicit fee schedule: for example $500K let you sleep over at the white house for one night.
Still wondering what he charged China for the ICBM rocketry technical assistance.
No, they should not run anything as complex as Linux. A voting machine does not need multi-tasking or networking. They should run a minimal program, written in assembly language, which would be available on the internet for public scrutiny months before the election.
That may work for two-toothed cretins and cash registers, but it doesn't scale. When it comes to accounting, the larger the sums involved, the larger the acceptable margin of error. There is a cost to accuracy, which in this case is borne by unpaid volunteers, and it's not reasonable to ask them to redo their task forever in search of perfection.
And yet I find it hard to trust any of these systems. I wonder why the ballot has a black background? I'd say it's to hide ambiguous votes - marks made between two of the circles. So the vote counters can say with a clear conscience, 'there was no mark.'
Assuming that their are Canadians as stupid as the ambiguous voters in Florida, there are probably many Canadian ballots with an X half way between the white circles.
Another reason to support it if you're a slashdot fan: Slashdot got where it is by experimenting. It isn't obvious in advance what will work. But the pace of experimentation has slowed since the buyout and vastly increased popularity. Slashdot is no longer likely to break new ground. This is fine - they now have an important role as a well-understood service.
Smaller projects like half-empty can afford to innovate because they have less to lose. And if they find some new principle to apply to online communities, maybe slashdot will implement the idea after it's been proven.
And you say:
While I realize what you're getting at (I wouldn't want Unix judged on how well it 'edited the registry' or something), it's inherently tricky to counter that claim without running up against the charge of unfairness.
For example, before Winsock a Unix geek could have said "Unix does TCP/IP better - in fact Windows doesn't do TCP/IP at all!" And you could have responded, "TCP/IP is a Unix thing. Why judge Windows by something irrelevant like that?"
So what do you consider a fair basis of comparison? Shall I dredge up a list of Windows horror stories, only to be told that each one is caused by bad application programming or system administration? Because if that's the case, you're arguing that there's a grain of wheat under a mountain of dung - something which might interest academics, but not hungry people.
Damn, there goes my theory. And I was so hoping to free our language of one awkward construct. Damn you, wolves, for injecting real-world complexity into a nice theory!
But I agree, if M$ attacks Linux full force they will hurt themselves. If they use some sneaky and careful tactics, though, they can sabotage the acceptance of Linux by the clueless majority.
I'm not going to pursue this, because the phrase 'can do' covers a multitude of expert tweaking, as witnessed by various Microsoftian benchmarks. I will remark that out of the box Linux performs a wide range of tasks nicely; commercial Unix a somewhat smaller range of tasks, as the vendors are less generous in packaging useful software, and Win2K very little that I can see.
But Ballmer is not an academic concerned with threats to his theories; he's a businessman concerned with threats to his profits. Linux may strike you as old technology, but its the kind of technology that gets things done. Your statement is equivalent to: 'a CPU, RAM, disk, etc. circa 1970 isn't a threat to anyone, technologically.' And yet these elements have evolved, and vendors compete in selling them.
As for the reality of the Linux threat, remember when Microsoft was poised to 'inevitably' take over the web server market? Consult netcraft to see what became of that dream.
To survive the breakup. Thus, .NET. .NET will become the true Microsoft platform, hence the seat of their power. The OS will become unimportant. .NET + (Gates, Ballmer, etc.). .NET to linux or whatever.
Before breakup: M$ = apps + OS + (Gates, Ballmer, etc.)
After breakup: M$ = AppsGroup = apps +
OS_Group = ( OS ).
The Apps group will have all the power because it controls the real platform (.NET). The OS Group will be worthless because it can't act independently. The Apps group can always jettison the OS by porting
Just a theory.
Your mailbox is at the whim of any idiot with an internet connection. As long as a person's emotional well-being is linked to the contents of his mailbox, he is doomed to frequent bouts of anger. That's why I think the 'war on spam' is ultimately a dead end - there's always one more idiot ready to spam.
The guy who wrote '7 Habits of Highly Effective People' (what's his name?) talks about reducing your circle of concern to match your circle of influence. You can control your mail server and filtering software; you can't control the internet.
The MPAA executive has a heart attack because he can't control the spread of information on the internet. The geek has a heart attack because he can't control the flow of mail into his system.
Maybe there's a parallel with household phones. In the 1970's it was normal that if you dialled a house a loud alarm-like bell went off, and the occupants dropped whatever they were doing and answered the caller. That worked until it was systematically abused by pranksters and telemarketers. Now the norm is for an answering machine to screen the call, and increasingly caller ID is required.
Any protocol which allows you to make me jump via remote control is broken and will be exploited eventually.