First, numerous functional brain imaging studies have shown patterns of decreased activation with practice, especially in the PFC. It's sometimes informally described as a sign of effortful performance. So more activation is not always good.
Second, there are lots of reasons to be cautious interpreting these kinds of results. Assuming the authors did the stats reasonably well (about a 10% chance in this relatively new field), understanding the roles of interacting brain regions is not straightforward. The press release doesn't mention that the earlier article in Neuroreport reported apparently opposite results (reduced activation with sleep deprivation). I think that article is available free online at: http://www.neuroreport.com
On a more practical note, if we believed that UCITA would become law tomorrow, what should poor shareware authors do? In particular, suppose one wanted to be legally protected at all costs. I'm concerned about this because I have a GPL'ed package that could, if someone were exceptionally stupid, be used in a medical context. Not being a lawyer, I can imagine a few things it might be good to do now, to avoid having my assets seized. Of course, I hate the idea of doing any of these. But I don't think I would make an effective martyr.
Incidentally, I would really enjoy seeing GNU/Linux systems built from shrink-wrapped components. Can you imagine if every user had to type out the full phrase "yes, i agree to these terms," for each executable before running it the first time? System installation alone would take months.
Anyway, supposing I wanted to limit my liability at all costs. Would the following be good ideas?
1. Withdraw my package, try to identify anyone who might have downloaded it, and do what I can to make sure nobody uses it now or in the future, at least until I can shrink-wrap it.
2. Alter the documentation to indicate that the package's express purpose is to consume disk space by merit of being composed of information. Note that I consider treatment of any of this data as "executable" a severely dangeous step that may result in crashes, the destruction of data, or other undesirable consequences, including death, dismemberment, and destruction. Hope that someone else writes documentation describing the "off-label" uses.
3. Change my name and appearance, convert all my assets to cash, build a remote cabin.
I have to admit, if all this is characterized correctly (again, I'm not a lawyer), it's easily the scariest thing I've read in recent memory. I haven't seen a compelling argument in favor of it -- there seems to be no way anyone could support legislation like this with both reasonable knowledge of the software industry and good intentions. So my guess is that the people who support it are both ignorant and evil. But I'm willing to be enlightened.
Not everything that plays DVDs has a click-through license, right? I don't recall that I had to sign anything or break any printed seals to open my DVD player (component, not in a computer). Wouldn't most of this issue be moot if someone could find something without a click-through license to reverse engineer? And (speaking here as a non-lawyer), it seems like it would be best if this could be done quickly, so as to minmize the window of time during which it could be claimed that the availability of DeCSS violated a trade secret.
I thought the injunction order seemed pretty reasonable, anyway. Like everyone else around here, I have some reservations about the law, though.
(Feel free to moderate this down into oblivion if I'm missing some obvious point. We non-lawyers are dumb that way.)
Okay, so these two mild-mannered bacteria have what it takes to survive a trip from Mars. The real question is what else they could have survived. Without extrapolating too much, we can also guess that they have what it takes to survive trips from all kinds of other places! Populated planets on other solar systems! The future! Alternate dimensions! Middle Earth! If I don't see an item on CNN by this evening with a title something like "Slashdot reports bacteria from the future" I'm going to be very disappointed.
Absolutely. Driver's license testing is an even better example, where I think we can all agree my viscera have an excellent point. On general principle I wouldn't trust the government to tell me who should and shouldn't be allowed to drive. But if I give them the benefit of the doubt and take a look at their actual tests, they don't make a compelling case. Based on the tests I had to pass to get my driver's license 17 years ago in Pennsylvania, I would guess that these certification exams are perhaps just a hair better than completely worthless. And this particular certification authority is extremely conservative about revoking certification.
So yes, I would say that certification exams are even less appealing as a screening tool for drivers than they are as a screening tool for sysadmins. When you screw up the former, people are more likely to die as a result. With sysadmin testing, I was wondering aloud in my earlier comment whether or not there's an obvious better solution. With driving certification, I think there almost certainly is, although I know too little about the subject to comment knowledgeably.
I should add that I don't know what the other 49 states do for driver's license testing, but I know they can't all be as lame as Pennsylvania. But I would still guess that anything that did a good job of weeding out potentially dangerous drivers would have to go beyond anything one would describe as an exam.
As an aside, can anyone explain to me why everywhere I go, people say, "drivers here are the worst?" Compared to where?
Viscerally, I don't like the idea of certification exams -- I wouldn't want to have to take one, and I certainly wouldn't want to lose a job to someone vastly less capable, just because they happened to have the approval of some dubious skill certification authority.
But I wonder what the alternative is. I'm no expert on hiring, but I've been asked to do some certification on similarly limited information -- that is, someone handed me a few resumes and asked me to express some opinions. You can't always tell much from a resume. At some point I'll have to interview people. Interviews are notoriously unreliable. If I hadn't been around, the interview and resume reading would have to have been done by someone with no relevant technical knowledge whatsoever. It's hard to imagine them getting any useful information about competence. Hiring someone to do your hiring might be a good idea, or it might be a painful experience in the difficulty of bootstrapping. So is the certification exam an example of the worst form of certification, except for all the others?
I'd have to guess that a certification exam score is, if only slightly, better than nothing (or, rather, better than just a resume). The question then is more one of weighting. How bad does your other information have to be before the certification exam scores should carry significant weight? And how often does the typical interview rise above that threshold of informativeness? I have the unjustified feeling that, with some forethought, I could outperform an exam. But I don't think everyone is in my position.
Yes, it's trivially true that you can make copies of anything you can see with your eyes, without DeCSS. But if you do that, you only get the benefit of whatever some particular player app produces, which could fall well short of the richness of the information on the disk. Using this kind of technique, it would be a ton of work reconstructing a (hypothetical at this point) DVD with a complicated branching structure. And you'd lose all the careful work the studio was supposed to have done in choosing key frames. And it's conceivable that the player might degrade the image somewhat.
As far as I can tell, copy protection schemes only ever affect convenience. So, assuming that ESR wasn't making an incredibly trivial point, for his statement (that encryption offers absolutely zero protection) to be false, it only has to be true that the DeCSS code makes some form of copying or redistribution easier.
More directly, duplication is not the only thing from which content producers might want protection, contra the implication of ESR's article.
Let me jump on this bandwagon with one more question about ESR's article. He writes that, "The amount of protection content producers get from DVD is exactly zero." Is this true? I'm going to go with "no" on that. More specifically:
It seems like there are lots of things you can't do with a DVD movie without cracking the encoding/encryption, and some of these things might be the things from which content producers would like protection. The examples that leap to mind all involve transforming the data -- e.g., extracting stills, or making available reduced-resolution versions for easy distribution or streaming. I can't imagine it will be long before you can walk into dorm rooms around the world and find streaming movie servers with excellent selections of major studio releases. Probably running off a laptop. Although I can think of reasons to be both happy and sad about this, it doesn't seem like the kind of thing that can be prevented.
1. My boss was told he should worry about our computers getting damaged by a Y2K-related power failure. So he unplugged them. If they should happen to have trashed filesystems or something, would this count as a Y2K bug?
2. In the same way that politicians start lining up for the next election right after the last one ends, I think it's never too soon to start planting the seeds for the next Y2K-style bug. I'm sure there are lots of gems out there, but it would be worth thinking up something really artful to impress future generations and ensure a nice retirement package. I nominate the Y2K bug (not to be confused with the Y2k bug), although I'm sure others will have better ideas.
3. My bank claims everything will work fine in the new year. Where were they last year? More seriously, has anyone yet written about an effect I will call the Y2K-backlash, the effect of Y2K-necessitated code cleanup causing broken things to start working?
I think the fact that conservative businesses (those that shy away from innovation) will eventually fail is almost tautological. The alternative (only slightly extreme) seems to be that there will be no more substantial innovation ever. I suppose it's also possible that big companies will suck up all useful innovation with their financial muscle, but things haven't gotten quite that bad yet.
Not to beat a dead horse, but it does seem pretty intuitive why conservative strategies won't be successful against a lot of competition. Shakespeare would be justly nervous against a sufficiently large fleet of monkeys with typewriters. This is not necessarily good news for each individual competitor. It does suggest that Shakespeare would do well to experiment a bit more and rest on his laurels a bit less, more so the more monkeys are out there.
I think if you had to plot the probability of long-term survival for big companies and small innovative companies vs time, the guess would have to be that big companies would start off with a huge advantage, and end up with a small disadvantage. By embracing innovation, one would imagine the initial downwards slope for the big companies would increase, but that they shouldn't cross over the line. I guess this just suggests a graphical representation of shortsightedness, but also I think suggests why shortsightedness is not such a bad idea sometimes.
Slashdot Mirror
A few things worth keeping in mind.
First, numerous functional brain imaging studies have shown patterns of decreased activation with practice, especially in the PFC. It's sometimes informally described as a sign of effortful performance. So more activation is not always good.
Second, there are lots of reasons to be cautious interpreting these kinds of results. Assuming the authors did the stats reasonably well (about a 10% chance in this relatively new field), understanding the roles of interacting brain regions is not straightforward. The press release doesn't mention that the earlier article in Neuroreport reported apparently opposite results (reduced activation with sleep deprivation). I think that article is available free online at:
http://www.neuroreport.com
On a more practical note, if we believed that UCITA would become law tomorrow, what should poor shareware authors do? In particular, suppose one wanted to be legally protected at all costs. I'm concerned about this because I have a GPL'ed package that could, if someone were exceptionally stupid, be used in a medical context. Not being a lawyer, I can imagine a few things it might be good to do now, to avoid having my assets seized. Of course, I hate the idea of doing any of these. But I don't think I would make an effective martyr.
Incidentally, I would really enjoy seeing GNU/Linux systems built from shrink-wrapped components. Can you imagine if every user had to type out the full phrase "yes, i agree to these terms," for each executable before running it the first time? System installation alone would take months.
Anyway, supposing I wanted to limit my liability at all costs. Would the following be good ideas?
1. Withdraw my package, try to identify anyone who might have downloaded it, and do what I can to make sure nobody uses it now or in the future, at least until I can shrink-wrap it.
2. Alter the documentation to indicate that the package's express purpose is to consume disk space by merit of being composed of information. Note that I consider treatment of any of this data as "executable" a severely dangeous step that may result in crashes, the destruction of data, or other undesirable consequences, including death, dismemberment, and destruction. Hope that someone else writes documentation describing the "off-label" uses.
3. Change my name and appearance, convert all my assets to cash, build a remote cabin.
I have to admit, if all this is characterized correctly (again, I'm not a lawyer), it's easily the scariest thing I've read in recent memory. I haven't seen a compelling argument in favor of it -- there seems to be no way anyone could support legislation like this with both reasonable knowledge of the software industry and good intentions. So my guess is that the people who support it are both ignorant and evil. But I'm willing to be enlightened.
dan
Not everything that plays DVDs has a click-through license, right? I don't recall that I had to sign anything or break any printed seals to open my DVD player (component, not in a computer). Wouldn't most of this issue be moot if someone could find something without a click-through license to reverse engineer? And (speaking here as a non-lawyer), it seems like it would be best if this could be done quickly, so as to minmize the window of time during which it could be claimed that the availability of DeCSS violated a trade secret.
I thought the injunction order seemed pretty reasonable, anyway. Like everyone else around here, I have some reservations about the law, though.
(Feel free to moderate this down into oblivion if I'm missing some obvious point. We non-lawyers are dumb that way.)
Okay, so these two mild-mannered bacteria have what it takes to survive a trip from Mars. The real question is what else they could have survived. Without extrapolating too much, we can also guess that they have what it takes to survive trips from all kinds of other places! Populated planets on other solar systems! The future! Alternate dimensions! Middle Earth! If I don't see an item on CNN by this evening with a title something like "Slashdot reports bacteria from the future" I'm going to be very disappointed.
Absolutely. Driver's license testing is an even better example, where I think we can all agree my viscera have an excellent point. On general principle I wouldn't trust the government to tell me who should and shouldn't be allowed to drive. But if I give them the benefit of the doubt and take a look at their actual tests, they don't make a compelling case. Based on the tests I had to pass to get my driver's license 17 years ago in Pennsylvania, I would guess that these certification exams are perhaps just a hair better than completely worthless. And this particular certification authority is extremely conservative about revoking certification.
So yes, I would say that certification exams are even less appealing as a screening tool for drivers than they are as a screening tool for sysadmins. When you screw up the former, people are more likely to die as a result. With sysadmin testing, I was wondering aloud in my earlier comment whether or not there's an obvious better solution. With driving certification, I think there almost certainly is, although I know too little about the subject to comment knowledgeably.
I should add that I don't know what the other 49 states do for driver's license testing, but I know they can't all be as lame as Pennsylvania. But I would still guess that anything that did a good job of weeding out potentially dangerous drivers would have to go beyond anything one would describe as an exam.
As an aside, can anyone explain to me why everywhere I go, people say, "drivers here are the worst?" Compared to where?
Viscerally, I don't like the idea of certification exams -- I wouldn't want to have to take one, and I certainly wouldn't want to lose a job to someone vastly less capable, just because they happened to have the approval of some dubious skill certification authority.
But I wonder what the alternative is. I'm no expert on hiring, but I've been asked to do some certification on similarly limited information -- that is, someone handed me a few resumes and asked me to express some opinions. You can't always tell much from a resume. At some point I'll have to interview people. Interviews are notoriously unreliable. If I hadn't been around, the interview and resume reading would have to have been done by someone with no relevant technical knowledge whatsoever. It's hard to imagine them getting any useful information about competence. Hiring someone to do your hiring might be a good idea, or it might be a painful experience in the difficulty of bootstrapping. So is the certification exam an example of the worst form of certification, except for all the others?
I'd have to guess that a certification exam score is, if only slightly, better than nothing (or, rather, better than just a resume). The question then is more one of weighting. How bad does your other information have to be before the certification exam scores should carry significant weight? And how often does the typical interview rise above that threshold of informativeness? I have the unjustified feeling that, with some forethought, I could outperform an exam. But I don't think everyone is in my position.
Yes, it's trivially true that you can make copies of anything you can see with your eyes, without DeCSS. But if you do that, you only get the benefit of whatever some particular player app produces, which could fall well short of the richness of the information on the disk. Using this kind of technique, it would be a ton of work reconstructing a (hypothetical at this point) DVD with a complicated branching structure. And you'd lose all the careful work the studio was supposed to have done in choosing key frames. And it's conceivable that the player might degrade the image somewhat.
As far as I can tell, copy protection schemes only ever affect convenience. So, assuming that ESR wasn't making an incredibly trivial point, for his statement (that encryption offers absolutely zero protection) to be false, it only has to be true that the DeCSS code makes some form of copying or redistribution easier.
More directly, duplication is not the only thing from which content producers might want protection, contra the implication of ESR's article.
Let me jump on this bandwagon with one more question about ESR's article. He writes that, "The amount of protection content producers get from DVD is exactly zero." Is this true? I'm going to go with "no" on that. More specifically:
It seems like there are lots of things you can't do with a DVD movie without cracking the encoding/encryption, and some of these things might be the things from which content producers would like protection. The examples that leap to mind all involve transforming the data -- e.g., extracting stills, or making available reduced-resolution versions for easy distribution or streaming. I can't imagine it will be long before you can walk into dorm rooms around the world and find streaming movie servers with excellent selections of major studio releases. Probably running off a laptop. Although I can think of reasons to be both happy and sad about this, it doesn't seem like the kind of thing that can be prevented.
dan
1. My boss was told he should worry about our computers getting damaged by a Y2K-related power failure. So he unplugged them. If they should happen to have trashed filesystems or something, would this count as a Y2K bug?
2. In the same way that politicians start lining up for the next election right after the last one ends, I think it's never too soon to start planting the seeds for the next Y2K-style bug. I'm sure there are lots of gems out there, but it would be worth thinking up something really artful to impress future generations and ensure a nice retirement package. I nominate the Y2K bug (not to be confused with the Y2k bug), although I'm sure others will have better ideas.
3. My bank claims everything will work fine in the new year. Where were they last year? More seriously, has anyone yet written about an effect I will call the Y2K-backlash, the effect of Y2K-necessitated code cleanup causing broken things to start working?
dan
I think the fact that conservative businesses (those that shy away from innovation) will eventually fail is almost tautological. The alternative (only slightly extreme) seems to be that there will be no more substantial innovation ever. I suppose it's also possible that big companies will suck up all useful innovation with their financial muscle, but things haven't gotten quite that bad yet.
Not to beat a dead horse, but it does seem pretty intuitive why conservative strategies won't be successful against a lot of competition. Shakespeare would be justly nervous against a sufficiently large fleet of monkeys with typewriters. This is not necessarily good news for each individual competitor. It does suggest that Shakespeare would do well to experiment a bit more and rest on his laurels a bit less, more so the more monkeys are out there.
I think if you had to plot the probability of long-term survival for big companies and small innovative companies vs time, the guess would have to be that big companies would start off with a huge advantage, and end up with a small disadvantage. By embracing innovation, one would imagine the initial downwards slope for the big companies would increase, but that they shouldn't cross over the line. I guess this just suggests a graphical representation of shortsightedness, but also I think suggests why shortsightedness is not such a bad idea sometimes.
dan