Even then, though, why shouldn't they be free to manually change the ranking if they wanted to?
Because they claim that Page Rank is an automated, objective measure of a site's relevance. If it turns out that they're manually tweaking Page Ranks in a way that they're not telling people about, they might be guilty of false advertizing. I personally think that's a stretch, but then again, IANAL, so somebody who is might not agree.
Note the fact that they are still very much in business, and will be for several more years.
They're only still in business because they've had several rounds of additional funding since the beginning of the trial, most of which looks as though it can be traced to a certain company whose name begins with "M" and which complains about Linux a lot. As of their last financial statement, they're hemorraging money at a rate that will bankrupt them shortly before opening statements in the IBM trial. Even if Bill Gates manages to send SCO some more money, they are going to lose at trial and IBM is going to be awarded damages that substantially exceed the net worth of the company. SCO's defeat won't be a quick one, but it's going to be most painful when it finally arrives.
The problem isn't with SCO bleeding them. The best guess is that SCO wanted to be bought out, which IBM could have done for pocket change. The problem is that caving in to an extortionist sends a bad message- that extortion pays and IBM is signing the checks. That's a message that IBM is not understandably not eager to broadcast. Instead they've decided to squish SCO like a bug, which sends the message that trying to extort IBM is likely to lead to a painful bankruptcy.
The Stallman-esque extremists who want to avoid anything that they think is in some way capitalist are just as bad, though.
Not nearly as bad as the people who try to categorize others incorrectly. Stallman doesn't think that it's wrong to make money selling Free Software. To the contrary, he actively encourages people to do so. Just read the FSF's essay on selling Free Software. For people who can't bother to follow the link, a salient quote is (emphasis is from the original):
Since free software is not a matter of price, a low price isn't more free, or closer to free. So if you are redistributing copies of free software, you might as well charge a substantial fee and make some money. Redistributing free software is a good and legitimate activity; if you do it, you might as well make a profit from it.
Distributing free software is an opportunity to raise funds for development. Don't waste it!
That doesn't seem like somebody who's opposed to capitalism.
Technically the use warnings; and use strict; aren't necessary either; at the very least there are far too many people who don't seem to use them. I've just gotten in the habit of using all of them. The warnings and strict are obvious, and I like to put an exit at the end of my main program both to mark where it ends and the subroutines begin.
If they provide the binary by downloading it from an upstream provider and passing it on unaltered, they can do exactly the same thing with the corresponding source package.
This seems to be the misunderstanding that the article is really all about. By actually distributing a binary, you as a distributor seem to become responsible for making the source available.... Automating the process as you describe would automate the GPL violation, which isn't exactly a good approach to rectifying a mistake.
No, I was just expressing myself poorly. By "do the same thing" I meant that they can treat the source package the same way that they treat the binary package (i.e. download it and make it available unmodified), not that they can deal with the source package the same way that they do when only referencing another provider. IOW, there are three cases:
Packages that they've modified. They should already be distributing the modified source.
Packages that they only reference. Not a problem because they're not actually distributing themselves.
Packages that they download and redistribute unmodified. Currently a problem, but solvable by automatically downloading and providing the corresponding source package whenever downloading the binary package.
That looks as though you don't know Perl very well. There's no need for a separate main routine, as whatever is in the file that's not part of another subroutine is assumed to be part of main. There's also no need to have the program end in a true statement (that's necessary only for modules) or to use an __END__. In the true spirit of Perl (i.e. eliminating needless elements) here's a refined version:
#!/usr/bin/perl
=head1 GOAL Program -- Solve it -- Solves the worlds problems. All of them. At once. This will be a great program when I finish it. =cut
I think the primary concern is, what happens to a distro like MEPIS? Do they need to retain a full and publically available source repository for every package in Ubuntu? That could be an administrative and financial drain.
As another poster said, they only have to provide source for packages for which they provide the binary. If they're providing binaries compiled from modified source code, they'd damn well better be providing the source already. If they provide a binary by refering users to another repository, they can depend on the other repository to provide the source. If they provide the binary by downloading it from an upstream provider and passing it on unaltered, they can do exactly the same thing with the corresponding source package. They only reason that it's a big deal is that they haven't bothered to automate the process; once it's automated it shouldn't be maintainable with minimal effort. And it's unlikely to be a big financial burden because only a small fraction of users will ever download the source anyway.
The organization I was working for (a large international organization that provides relief in disasters, hint hint) keeps data on the people who seek help private.
It's too bad that they don't take an equal interest in protecting the privacy of the people who want to give help in the form of money. If my experience is any guide, they seem to have no compunctions at all about selling their donor list to every other charity in sight.
Currently the mantra If you are not a terrorist/paedophile/Mexican, you have nothing to hide
I'm not doing anything wrong when I take a shower or use the toilet, but that doesn't mean that it's OK for the government to monitor me while I'm doing so.
In response to the problem of symbols losing their meaning: haven't any of these people read "Contact"? Use prime numbers -it doesn't matter what language you speak, prime numbers are the same to everyone!
That's great if the only message you want to send is "this is a deliberate message, not a random chance". The big pillars with pictures and writing on them will be just as effective in that role. If you can come up with an effective, universal way of saying "Danger! Do not enter!" using nothing but prime numbers, I'm sure that the government would be happy to pay you good money for it.
Describing that as risky is a gross understatement. Microsoft would be shooting itself in the foot very badly by trying that approach. Europeans would soon discover how to survive in a Microsoft-free environment, which would lose Microsoft a big market for good. Even worse, it would ensure that there would be a huge group of ITS people skilled in moving from Microsoft to the alternatives and prove to anyone who doubted that life without Microsoft is possible. That's the absolute last thing they want.
Sure, but even if they redact all of the phone numbers, you'll still be able to count the number of redactions and get an idea of the scope of the program. If they're really only spying on people who contacted al Qaida, then the list will probably be short. If it's 10,000 pages of redacted phone numbers, it's pretty easy to figure out that they're going on a fishing expedition.
Apparently including journalists. In a move that should surprise nobody, the reporter who wrote the article doesn't seem to have bothered to look at the data, and relied on the pre-digested summary instead. If you bother to look at the Appendix to the report, which is just a few mouse clicks away for anyone who is interested, it turns out that the college students did substantially better than the population at large.
What's particularly interesting is that they also had a comparison of current college students with college graduates. The current students did better than graduates in all areas of the test. Their average scores were higher, a lower pecentage of them were in the lowest score categories, and a higher percentage were in the highest score categories, with the exception of one test where the 2 year college graduates managed a tie with the current 2 year students.
It would be at least as honest to report the results as saying that current college students are better equipped for daily life than the population at large. But that wouldn't be alarming enough. More importantly, it wouldn't play to the prejudices of the audience, who want to believe that things are going to hell.
ICF schemes currently put the breakeven and ignition regime at around 10^16 - 10^17 neutrons/pulse so I'd say we're more like a factor of nearly a QUADRILLION off.
But you're only looking at one side of the equation, the energy output. Since breakeven is defined by (energy output/energy input) > 1, you also have to look at energy input. I don't have any figures handy, but I'm reasonably confident that ICF requires many orders of magnitude more energy to start than sonofusion does. When you take that into account, you may discover that the 6-7 orders of magnitude figure is correct.
It'd be more intersting to see a larger survey, and with more obscure topics. In my opinion, an encyclopedia should be judged by its weakest entries, not its strongest.
In which case Wikipedia should do much better than Britannica. After all, there are many obscure topics for which Wikipedia has an articles and Britannica doesn't. Any hard copy encyclopedia is going to get trounced by Wikipedia for articles on popular culture and recent events, for instance.
The problem is that the definition of macro being 1:1 was created back when many photographers were making contact prints from 8x10 inch negatives. It doesn't necessarily make sense to preserve that description considering that most digital cameras use sensors smaller than 8x10mm and will always print heavily enlarged pictures. For fixed lens digitals it probably makes more sense to think about the reproduction ratio in a print made at a standard resolution like 300 dpi. I'm sure that this will offend pedants, but it'll be a much more practical measure for the millions of people thinking about buying the cameras.
It would be selected solely because it contains a grammar mistake.
No it doesn't. "Can" is being used as a noun, not as a verb. It would probably be slightly better if written "A pringles can provides limitless food supply", but it's fine the way it is.
Makes me sick. Not to bring up the Bush Corolary of Godwin's Law, but take 9/11. 1 day to happen, 5 years of finger pointing to follow. Why? We can't just say, "Okay, we all screwed up, let's learn something and move on." No no no, we've got to find out exactly whose fault it was that we didn't see it coming, so we can, I don't know, set them on fire or something.
Maybe we need to figure out what happened so that we can fix any systematic problems that it revealed about the way that our intelligence system works. Not every investigation has to concentrate on fixing blame and punishment, you know. Some times there's not a single point of failure, and instead the failure is the result of the interaction of many separate problems. In a case like that the best approach is to find and fix everything that went wrong.eryone who has some control does what he can to fix the problem, then you'll get much better results than if you try to place all the responsibility on any one point.
Consider the case you mention of a kid who died from taking his parent's pills. It may seem logical to place all the blame on the kid and parents, and it may even make sense to do so from a legal standpoint, but if your goal is to reduce poisoning deaths, you have to look at the entire chain:
The pharmaceutical company made the medicine. They might have been able to do something to make it safer, like adding something to the pills that makes them taste awful so that kids don't want to eat a bunch of them.
The doctor prescribed the medicine. He might want to consider whether he really needed to prescribe something in the first place, or if he's just giving a pill to his patient because that's what the patient expects. Depending on the medicine, he might also be able to make a prescription for fewer pills so that there wouldn't be enough in the bottle to cause a deadly OD even if somebody took them all.
The pharmacist dispensed the medicine. He probably ought to make sure that it's in a bottle that's tough for kids to open.
The parents kept the medicine. They can help by storing it somewhere that kids will have a hard time getting into, teaching their kids not to take the parents medicine, and supervising their kids whenever possible. They can also help matters by discarding unused medicines rather than keeping them around the house.
The kid took the medicine. Older kids really ought to know better.
If each person who has some ability to reduce child OD deaths does his part, then you'll get a much more effective solution than if you try to fix the blame on just one person.
I'd expand the exception a little bit to any kind of work where the exact purpose and mechanism isn't obvious. That may be most common in embedded programming, but it can happen anywhere there's something time critical, like inner loops of number crunching programs. One thing that can help in a case like that is referencing a standard reference on the approach you're taking- like mentioning a chapter in Numerical Recipies in C- rather than giving a complete explanation of the approach yourself.
Cases like that also show why it's so important to discuss why you're taking the approach you are. A subsequent maintainer may not understand why you're using a tricky approach instead of a simple one and waste time trying to replace your code with a simpler version. A really good comment might even include benchmark data showing just how much time or space is saved by using the tricky version so that a subsequent maintainer can make an informed decision about whether or not to replace it.
If it's illegal to sell porn in the street, why shouldn't it be illegal to see on the internet?
Because the net isn't the same as the street. It's OK to regulate selling porn in the street because the street is a public place, so putting porn there effectively forces it on people who are offended by it. Web sites, OTOH, are strictly voluntary. Nobody has to visit a porn website if they don't want to, so there's not the same degree of coercion involved. In that respect web sites are like windowless shops; nobody has to see what's going on inside unless they step in through the door.
FWIW, the voluntary/involuntary issue is why I think that it's OK to crack down on porn spam and typosquatting sites. Those are the equivalent of stuffing explicit advertizing into people's doors, or mislabeling a store to lure in people who wouldn't otherwise visit. Neither of those would be acceptable in meatspace, so they shouldn't be acceptable in cyberspace, either.
It's important to remember that visible source code isn't the only requirement for Open Source. For software to be Open Source, it's not only necessary that the source code be available, but also that users are free to modify it and redistribute modified or unmodified copies. There's no real chance that the software in this case will be released under those terms. After all, one of the arguments that the lawyers are using is that the software has been modified without being recertified. It would be much more difficult to ensure that software in use hadn't been modified from a certified version if any user were free to modify it.
But don't you think that if windows code was open sourced that eventually all the leaks would be patched??
Maybe, but I'd bet that the way that it would be done in practice would be to make a Microsoft compatibility layer over an existing, more secure OS. Then you could run each legacy application in a sandbox so that your whole system wouldn't be hosed by the inherent insecurity of Windows's architecture.
Science never has been definite. The defining characteristic of science is that it accepts that all solutions to problems are tenative, and that some piece of information might turn up in the future that will cause us to doubt what we now believe. Intellectual process can't happen without replacing wrong old ideas with better new ones.
Slashdot Mirror
Because they claim that Page Rank is an automated, objective measure of a site's relevance. If it turns out that they're manually tweaking Page Ranks in a way that they're not telling people about, they might be guilty of false advertizing. I personally think that's a stretch, but then again, IANAL, so somebody who is might not agree.
They're only still in business because they've had several rounds of additional funding since the beginning of the trial, most of which looks as though it can be traced to a certain company whose name begins with "M" and which complains about Linux a lot. As of their last financial statement, they're hemorraging money at a rate that will bankrupt them shortly before opening statements in the IBM trial. Even if Bill Gates manages to send SCO some more money, they are going to lose at trial and IBM is going to be awarded damages that substantially exceed the net worth of the company. SCO's defeat won't be a quick one, but it's going to be most painful when it finally arrives.
The problem isn't with SCO bleeding them. The best guess is that SCO wanted to be bought out, which IBM could have done for pocket change. The problem is that caving in to an extortionist sends a bad message- that extortion pays and IBM is signing the checks. That's a message that IBM is not understandably not eager to broadcast. Instead they've decided to squish SCO like a bug, which sends the message that trying to extort IBM is likely to lead to a painful bankruptcy.
Not nearly as bad as the people who try to categorize others incorrectly. Stallman doesn't think that it's wrong to make money selling Free Software. To the contrary, he actively encourages people to do so. Just read the FSF's essay on selling Free Software. For people who can't bother to follow the link, a salient quote is (emphasis is from the original):
That doesn't seem like somebody who's opposed to capitalism.
Technically the use warnings; and use strict; aren't necessary either; at the very least there are far too many people who don't seem to use them. I've just gotten in the habit of using all of them. The warnings and strict are obvious, and I like to put an exit at the end of my main program both to mark where it ends and the subroutines begin.
No, I was just expressing myself poorly. By "do the same thing" I meant that they can treat the source package the same way that they treat the binary package (i.e. download it and make it available unmodified), not that they can deal with the source package the same way that they do when only referencing another provider. IOW, there are three cases:
That looks as though you don't know Perl very well. There's no need for a separate main routine, as whatever is in the file that's not part of another subroutine is assumed to be part of main. There's also no need to have the program end in a true statement (that's necessary only for modules) or to use an __END__. In the true spirit of Perl (i.e. eliminating needless elements) here's a refined version:
As another poster said, they only have to provide source for packages for which they provide the binary. If they're providing binaries compiled from modified source code, they'd damn well better be providing the source already. If they provide a binary by refering users to another repository, they can depend on the other repository to provide the source. If they provide the binary by downloading it from an upstream provider and passing it on unaltered, they can do exactly the same thing with the corresponding source package. They only reason that it's a big deal is that they haven't bothered to automate the process; once it's automated it shouldn't be maintainable with minimal effort. And it's unlikely to be a big financial burden because only a small fraction of users will ever download the source anyway.
It's too bad that they don't take an equal interest in protecting the privacy of the people who want to give help in the form of money. If my experience is any guide, they seem to have no compunctions at all about selling their donor list to every other charity in sight.
I'm not doing anything wrong when I take a shower or use the toilet, but that doesn't mean that it's OK for the government to monitor me while I'm doing so.
That's great if the only message you want to send is "this is a deliberate message, not a random chance". The big pillars with pictures and writing on them will be just as effective in that role. If you can come up with an effective, universal way of saying "Danger! Do not enter!" using nothing but prime numbers, I'm sure that the government would be happy to pay you good money for it.
Describing that as risky is a gross understatement. Microsoft would be shooting itself in the foot very badly by trying that approach. Europeans would soon discover how to survive in a Microsoft-free environment, which would lose Microsoft a big market for good. Even worse, it would ensure that there would be a huge group of ITS people skilled in moving from Microsoft to the alternatives and prove to anyone who doubted that life without Microsoft is possible. That's the absolute last thing they want.
Sure, but even if they redact all of the phone numbers, you'll still be able to count the number of redactions and get an idea of the scope of the program. If they're really only spying on people who contacted al Qaida, then the list will probably be short. If it's 10,000 pages of redacted phone numbers, it's pretty easy to figure out that they're going on a fishing expedition.
Apparently including journalists. In a move that should surprise nobody, the reporter who wrote the article doesn't seem to have bothered to look at the data, and relied on the pre-digested summary instead. If you bother to look at the Appendix to the report, which is just a few mouse clicks away for anyone who is interested, it turns out that the college students did substantially better than the population at large.
What's particularly interesting is that they also had a comparison of current college students with college graduates. The current students did better than graduates in all areas of the test. Their average scores were higher, a lower pecentage of them were in the lowest score categories, and a higher percentage were in the highest score categories, with the exception of one test where the 2 year college graduates managed a tie with the current 2 year students.
It would be at least as honest to report the results as saying that current college students are better equipped for daily life than the population at large. But that wouldn't be alarming enough. More importantly, it wouldn't play to the prejudices of the audience, who want to believe that things are going to hell.
But you're only looking at one side of the equation, the energy output. Since breakeven is defined by (energy output/energy input) > 1, you also have to look at energy input. I don't have any figures handy, but I'm reasonably confident that ICF requires many orders of magnitude more energy to start than sonofusion does. When you take that into account, you may discover that the 6-7 orders of magnitude figure is correct.
In which case Wikipedia should do much better than Britannica. After all, there are many obscure topics for which Wikipedia has an articles and Britannica doesn't. Any hard copy encyclopedia is going to get trounced by Wikipedia for articles on popular culture and recent events, for instance.
The problem is that the definition of macro being 1:1 was created back when many photographers were making contact prints from 8x10 inch negatives. It doesn't necessarily make sense to preserve that description considering that most digital cameras use sensors smaller than 8x10mm and will always print heavily enlarged pictures. For fixed lens digitals it probably makes more sense to think about the reproduction ratio in a print made at a standard resolution like 300 dpi. I'm sure that this will offend pedants, but it'll be a much more practical measure for the millions of people thinking about buying the cameras.
No it doesn't. "Can" is being used as a noun, not as a verb. It would probably be slightly better if written "A pringles can provides limitless food supply", but it's fine the way it is.
Maybe we need to figure out what happened so that we can fix any systematic problems that it revealed about the way that our intelligence system works. Not every investigation has to concentrate on fixing blame and punishment, you know. Some times there's not a single point of failure, and instead the failure is the result of the interaction of many separate problems. In a case like that the best approach is to find and fix everything that went wrong.eryone who has some control does what he can to fix the problem, then you'll get much better results than if you try to place all the responsibility on any one point.
Consider the case you mention of a kid who died from taking his parent's pills. It may seem logical to place all the blame on the kid and parents, and it may even make sense to do so from a legal standpoint, but if your goal is to reduce poisoning deaths, you have to look at the entire chain:
If each person who has some ability to reduce child OD deaths does his part, then you'll get a much more effective solution than if you try to fix the blame on just one person.
I'd expand the exception a little bit to any kind of work where the exact purpose and mechanism isn't obvious. That may be most common in embedded programming, but it can happen anywhere there's something time critical, like inner loops of number crunching programs. One thing that can help in a case like that is referencing a standard reference on the approach you're taking- like mentioning a chapter in Numerical Recipies in C- rather than giving a complete explanation of the approach yourself.
Cases like that also show why it's so important to discuss why you're taking the approach you are. A subsequent maintainer may not understand why you're using a tricky approach instead of a simple one and waste time trying to replace your code with a simpler version. A really good comment might even include benchmark data showing just how much time or space is saved by using the tricky version so that a subsequent maintainer can make an informed decision about whether or not to replace it.
Because the net isn't the same as the street. It's OK to regulate selling porn in the street because the street is a public place, so putting porn there effectively forces it on people who are offended by it. Web sites, OTOH, are strictly voluntary. Nobody has to visit a porn website if they don't want to, so there's not the same degree of coercion involved. In that respect web sites are like windowless shops; nobody has to see what's going on inside unless they step in through the door.
FWIW, the voluntary/involuntary issue is why I think that it's OK to crack down on porn spam and typosquatting sites. Those are the equivalent of stuffing explicit advertizing into people's doors, or mislabeling a store to lure in people who wouldn't otherwise visit. Neither of those would be acceptable in meatspace, so they shouldn't be acceptable in cyberspace, either.
No, no, no. We definitely want to keep spaghetti code out of the kernel, whether it's flying or not.
It's important to remember that visible source code isn't the only requirement for Open Source. For software to be Open Source, it's not only necessary that the source code be available, but also that users are free to modify it and redistribute modified or unmodified copies. There's no real chance that the software in this case will be released under those terms. After all, one of the arguments that the lawyers are using is that the software has been modified without being recertified. It would be much more difficult to ensure that software in use hadn't been modified from a certified version if any user were free to modify it.
Maybe, but I'd bet that the way that it would be done in practice would be to make a Microsoft compatibility layer over an existing, more secure OS. Then you could run each legacy application in a sandbox so that your whole system wouldn't be hosed by the inherent insecurity of Windows's architecture.
Science never has been definite. The defining characteristic of science is that it accepts that all solutions to problems are tenative, and that some piece of information might turn up in the future that will cause us to doubt what we now believe. Intellectual process can't happen without replacing wrong old ideas with better new ones.