Next thing you know DOJ discovers incriminating emails on Gates' machine from the MS internal network. Of course, more work would be required than just that one little act, but the philosophical point is that email is just bits on hard drives, and is therefore no more reliable than heresay, which is inadmissable.
Yes, and written letters are just bits of ink on pieces of paper, but using them is quite common in legal circles. Fairly reasonably, if I ask you for your records and I find something incriminating in them (and bear in mind that you also have to provide copies to the court, so I can't change them and claim that they're original) it should be your burden to prove that the incriminating comments were forged, rather than mine to prove that they're genuine! If anything, people should be suspicious if they show something unusually exculpatory, since you're far more likely to modify them in a way that reflects well on you than to forge records that incriminate you. In any case, IIRC these aren't emails from Gates's desktop machine; they're from the corporate email archive.
Getting back to something closer to the article that triggered the discussion, the FBI isn't talking about either of these things. They're talking about intercepting email in transit, so my original interpretation of the more conventional approach to header forging is more of what the FBI would be interested in. In thise case, though, the FBI's tap is actually less likely to be forged than a random email, since they're going to be tapping his immediate upstream connection, so a forger would need to insert their forgeries exactly there rather than at any random point in the network. As for the FBI being able to forge the email, they could potentially do that no matter what system you used, so you're going to have to trust them to be honest in any case.
One interesting aspect of this is that it suggests that if you're a criminal you shouldn't PGP sign your incriminating emails. If they're PGP signed, it provides the FBI with excellent evidence to use in court that they're not forged; unsurprising since proving authenticity is the intent of signing them. If they're unsigned, though, it'll be a lot easier to claim that the FBI forged them. You can probably enhance the effect by signing all of your non-incriminating emails (which you figure that even the most hardened criminal would have) so that you can intimate that the FBI forged the incriminating ones but were unable to forge the signature since they didn't have your private key.
've wondered about this one for a while. In the MS v. DOJ thing, apparently they used a bunch of emails from Billy G. as evidence.
Admittedly, I didn't follow it all that closely, (by them time I had first heard about it, I was sick of hearing about it) but why didn't he just say "I didn't write that."
It should be virtually impossible to prove that email was written by any particular person. I could set my "Real Name" to Bill Gates and send out an email, or if I really wanted to put effort into it I could even make it look like it really came from bgates@microsoft.com. It's not that hard to create a file with a certain set of text in it, so an email header that says "this is from person X" doesn't at all guarantee that it actually is.
While it's true that it's easy to forge email on the internet, that's not where the billg mail came from in the Microsoft case. In that case, the email was from Microsoft's internal email system. It had been turned over to the government as part of the pre-trial discovery phase, which is basically when the lawyers for the two sides are allowed to demand that the other side turn over information that might be relevant to the case.
Furthermore, the emails weren't just random mails from billg to the rest of the world. They were part of multiparty email correspondance on particular issues. IOW for Gates to disavow the emails, he would have had to claim that someone was not only forging his name but was also intercepting his personal emails and forging a conversation on his behalf. Not only that, but they were doing so not on some leaky internet system but on Microsoft's presumably secure internal system, and that the other people he was corresponding with, who presumably encountered him at least occasionally in person never brought up the topic of the emails in non-email conversation so that the forgery never came to light. That claim would be so obviously bogus that all it would do is damaged Gates's credibility as a witness and not impeach the credibility of the email at all.
Actually, they are under no GPL obligation to make their software available on the net at all. The GPL only says that they have to make the source available anyone who gets the binaries, and that anyone who gets the source and binaries is free to redistribute them. IOW, you're free to buy the full price distribution and make all of the GPLed software available for free download, but you have no right to demand they make it available for free download.
FWIW, this is exactly what the Free Software Foundation advocates. To quote (from Selling Free Software on the FSF web site.):
Actually we encourage people who redistribute free software to charge as much as they wish or can.
Free programs are sometimes distributed gratis, and sometimes for a substantial price. Often the same program is available in both ways from different places. The program is free regardless of the price, because users have freedom in using it.
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!
(Emphasis is theirs.) IOW, the people who wrote the GPL don't just accept the idea of selling GPLed software for what the market will bear, they actually advocate it. Of course the fact that anyone who can get a copy of GPL software is free to redistribute it inherently limits the price anyone can charge, but that's for the market to decide.
It's not the DNA you need to worry about
on
Walk-By DNA Testing
·
· Score: 3
Honestly, scanning your identity this way is about the last thing you should be worried about. The main goal of testers like these is to be able to scan people rapidly, like the metal detectors at airports. They want to be able to tell if someone is trying to smuggle bombs or drugs onto an airplane. That means that you need to know the answer from your test now, not in an hour or two when the guy's already had a chance to pass his stuff to some third party.
At the present, and for the forseeable future, it's just not possible to make a DNA-based individual ID in anything like real time. Even in the lab with nearly ideal samples doing that kind of thing takes time, and a lot of that is not something that can be easily reduced; certain chemical and physical reactions take time and can't be sped up. That puts a pretty strong damper on using this as a DNA vacum to violate people's rights.
OTOH, you can bet that the war against drugs and the war against terrorism will be used as excuses. Pretty soon you won't be able to get on a plane without being subjected to a battery of tests to make sure that you're not trying to put anything illegal onto the plane. Oops, you're a mining engineer who uses explosives at work? Prepare to be hassled every time you try to fly. Your pot smoking brother came over to visit? Prepare to be stopped and have your luggage examined. In the long term those kinds of minor erosions of personal protection are a much more dangerous threat to privacy than some hypothetical DNA screening.
Test them on price/performance instead of performance; for general-purpose memory, I see no compelling reason to use RDRAM except to say that you're using it. (As in, "Wow, RDRAM, that's new, isn't it? I bet that set you back quite a bit...")
While many end users are actually more interested in price/performance than they are in performance per se, the idea of listing price per performance is still a bad one. There are two main reasons for this:
Different users have different willingnesses to spend extra for more performance. By making a composite yourself, you deprive the reader the ability to make that choice himself.
Prices for computer components are well known for being unstable both in time and location, while performance fluctuates less. By factoring in the price at the time you ran the test and where you bought the components, you muddy the comparison for users buying components in a different environment.
Both of these factors suggest that rating by price/performance is a bad idea, and that rating just by performance is much better.
The point is, most people will look at a new technology and not try to understand how it works, they'll just use it the way they're told to. We take something like a PC and use it for doing spreadsheets, because even though the machine is capable of far more, our thinking has become limited to things we use every day. Kids don't usually yet have those mental barriers in place, so they're not afraid to take it apart and mess with it, and they don't have the mentality that things can only be used a certain way, so they come up with more creative uses for what the technology can do.
There's a bit more to it than that. IMO the real distinction is that (for reasons that don't make sense to me personally) many, if not most, people are perfectly happy to use use equipment they don't really understand. In fact, they seem to prefer not to understand it because they think that it will be too much effort to learn. This is, as Larry Wall would say, false laziness. Understanding in at least a general sense how things work at one level deeper than you actually use them is a huge time saver in diagnosing the inevitable problems you'll encounter.
The point is that with any complex technology there always seems to be a small group of people who gain real power through a greater understanding of it. One thing that I dislike about the article is that it focuses on the handful of kids who are going out and doing really outstanding things rather than the more interesting pattern of kids with more computer savy than their parents gaining household power. The latter is actually such a well accepted part of the system that it's a popular topic for newspaper cartoons and the like.
But what do you do about this? Take away all of the DEA's powers? Make everything go in front of a judge? Think of what that will do for red-tape.
I'm not defending them at all, but you have to think of alternatives to give them the tools they need to solve their task. Either that or legalize all drugs and have the country turn into a cesspool.
I certainly do think that the DEA should have to go in front of a judge to get a warrant for a wiretap, search, etc. The purpose of that "Red Tape" is to protect your rights. Going through the trouble to justify a warrant isn't some stupid formality that's there to get in the way. It's a vitally important step in preventing abuse of police powers. If there's actually a problem processing these things promptly that means you need more judges to deal with the workload, not that you should start ripping up the Bill of Rights to save time.
The key is that allowing a minor loophole in an important protection is like being a little bit pregnant. Once there's a single agency that can violate your rights, all of the other groups that want to can just call them up, trump up some bogus grounds for a search, and "happen" to find whatever else they were looking for.
The big thing to remember is that someone involved in a big, organized activity like large scale drug traffic will inherently produce a trail of physical and documentary evidence. Sooner or later some of that is going to show up in a way that will convince a judge that it's worth while letting the police have their warrant. It's people who haven't done anything wrong but just happen to look suspicious (or have pissed off a member of the police) who are the real potential victims here. When you let the police barge in wherever and whenever they want, you'll increase the number of innocent people harmed much more than the number of guilty people caught.
If it is not legal for a company to categorize an employee as exempt unless they are "in management or do work involving significant creativity and indepenence" then I'd sure love it if someone could post a URL or some other reference to the law that spells this out.
The relevant law is the Fair Labor Standards Act. Unfortunately, it appears that they included a big fat exception specifically for computer workers. Otherwise, you might find this document interesting. It mentions four classes of exempt employees: Executive, Administrative, Professional, and Outside Sales. For most of the people on Slashdot, the following definition about which professionals qualify for exemption is the most interesting:
Professional Exemption
Applicable to employees who perform work requiring advanced knowledge and education, work in an artistic field which is original and creative, work as a teacher, or work as a computer system analyst, programmer, software engineer, or similarly skilled worker in the computer software field; who regularly exercise discretion and judgment; who perform work which is intellectual and varied in character, the accomplishment of which cannot be standardized as to time; who receive a salary which meets the requirements of the exemption (except doctors, lawyers, teachers and certain computer occupations); and who do not devote more than 20% of their time to work other than that described above.
The salary mentioned above means that:
Subject to certain exceptions set forth in the regulations, in order to be considered "salaried", employees must receive their full salary for any workweek in which they perform any work without regard to the number of days or hours worked. This rule applies to each exemption that has a salary requirement (outside sales employees, and certain licensed or certified doctors, lawyers and teachers have no salary requirement. For certain computer-related occupations under the professional exemption, they need not be paid a salary if they are paid on an hourly basis at a rate not less than $27.63 per hour). The special requirements which apply to each category of employees are summarized below.
The part about being paid in full for any week in which you work, regardless of the number of hours actually worked, is probably the most important thing here. IOW, if your employer can dock your pay for working part days (or apparently even part weeks!) you are not an exempt employee, unless you're an outside salesperson, doctor, lawyer, or teacher or a computer specialist paid hourly and earning at least $27.63 per hour.
The big thing is that if you really want to know your rights as an employee, you should really take a careful look at the Department of Labor web site.
55 hours a week? Is that considered hard time in Kansas City? I live in Boston and that is what everyone around here does, for about the same money, in a city that is much more expensive.
I don't always love my job either, but long workweeks aren't going to get any sympathy from me. Sounds like good ol' Worldcom just trimmed a little fat.
It sounds to me as if they're engaged in a bit of classic (and illegal) abuse of exempt employee status. Legally, employees must be paid overtime for work beyond specified limits (generally either 40 hours per week or 8 hours per day) unless they're legally exempt. Exempt status is supposed to be reserved for work that falls into two categories: 1) management and 2) work involving significant creativity and independence. Working in QA does not sound as though it falls into either category.
Unfortunately, there are lots of businesses out there that are eager to abuse the system. They try to convince people that because they are on a salary, they don't have any right to overtime pay. Legally this is complete bunk; if you don't fit one of the legal categories for exemption, you should get overtime. In practice, though, it's all too effective. Employees think that being on salary means no overtime, so they don't even know to demand it and their employers get away with drastically underpaying them for their work.
1.VC's have too much cash, and are too quick to throw it away at any old IT startup.
2.People have become so excited by the whole Internet and dotcom "bubble economy" that they will risk a whole lotta cash on the slim chance of making a quick buck.
That's the whole stock market bubble in a nutshell. The problem (if you want to call it that) is that there's more money out there and it has about the same number of outlets for investment. It winds up going into one of two places:
Chasing the limited number of existing stocks that are already overvalued. This is what Fed Chairman Greenspan calls "irrational exhuberance".
Investing in the limited number of new startups. This leads to everyone and their brother in law thinking that they can get VC money if only they have a convincing story. Unfortunately, many of them are right!
What it really boils down to is that a combination of greater real wealth and less even distribution of it make the equities market a less attractive investment.
don't have a problem with people using my code for profit. A lot of people who use GPL do.
This is a huge distortion. I've never heard any of the people who are really serious about the GPL complain about making a profit off GPLed code. Hell, RMS made money for a long time by selling tapes of GPL programs. On the FSF web site, in fact, they say:
Actually we encourage people who redistribute free software to charge as much as they wish or can.
That hardly sounds like people who don't want anyone to be able to make a profit from their code!
What people like the FSF do have a problem with is the idea of taking free/open source software and making it non-free/closed source. The GPL doesn't exist to prevent people from making a profit. The existence of companies that market GPLed software is proof that it doesn't accomplish that, even incidentally. The goal is to prevent people from taking software that is available in source form and releasing it without source.
But can you reapply a license to software that's already been released under a license? I think that is probably the biggest worry here. Not, can we fix it if there's a problem, but what will happen to everything that's been released already.
This isn't as much of a problem as you might think. Most of the software that's released under the GPL includes the clause that the software may be licensed under version whatever of the GPL or, at the licencee's option, any later version. That means that if problems are found in the GPL in a way that makes it invalid, the software should revert to standard copyright (i.e. copying forbidden). Then the FSF only has to release a new version of the GPL that fixes whatever was wrong before and everything is OK again.
What if media to me represents a load of paper ? At 0.05$ a sheet, I'll let you have a "copy" (print-out) of my code (Which will amount to 10000 sheets (500$). This would be considered reasonable under the GPL. Now, just imagine how many pages a printout of the kernel might take. What if they run it through a swahili translator before hand ? Or use a non-ocr'able font type ? You have a copy of the source code, which is useless unless you type it back in manually. The license allows for that. Fun heh ?
It's a good thing that this actually isn't allowed. The GPL does specify that the code must be A) on a medium customarilly used for transmitting code, B) in machine readable format, and C) in the form preferred for development. That means that a paper printout might be OK for a small bit of code, but your hypothetical shelf worth is pretty clearly a violation (particularly if it can't be read by OCR). It also means that code that is deliberately obfuscated is a direct violation.
The more I see people spouting this kind of stuff, the more I think that those people should actually stop to read the licenses under which their free software is distributed.
Well, it's both, to the effect that it would take a few lines of perl (maybe you could do it with tr?) to convert from one to the other. Three pairs of nucleic acids code for one amino acid. And arguably, the "amino acid sequence" is the one that people would be more concerned with. Once you have that, who cares what the nucleic acids were?
Actually, the translation can get a bit tricky, and there are reasons to stick with the DNA form rather than translating. With an unanotated DNA sequence you have to translate in all six relevant frames in order to be sure to get the one frame that's actually used. Even worse, with eukaryotic DNA you have to worry about the whole intron-exon structure. That's a little bit tough to handle with tr, or even a short perl script. The software I'm used to using will auto-translate in all frames, though, so I'm used to thinking of DNA and protein sequences as essentially the same thing.
On a deeper note, there are some interesting things that you can spot by looking at the untranslated DNA sequences. For homology matching, for instance, you can spot mutations that leave the AA sequence unchanged. More interestingly, you can look at things like codon bias. Codon bias is really interesting. There are, of course, multiple DNA triplets that code for the same amino acid. The more heavily translated a protein is, though, the more it tends to use only one of the available triplets to code for each amino acid. That means that given the DNA sequence for a protein you can make a reasonably accurate prediction of its abundance in the cell. Try that from the amino acid sequence alone!
Actually, I think he's right. Big business is not inherently evil. There are evil businesses, but that does not make big business, in itself, evil. I don't think you're successfully picking up on the distinction.
It's true that big businesses are not necessarily evil. Ben and Jerry's, for instance, got to be quite a big business while deliberately using goodness as a strategy. IBM, after decades of being the evil empire, now seems to be making a pretty good effort to be a good corporate citizen.
The problem is that it takes a lot of time and effort to grow a business up to be a big business. There are various shortcuts along the way, and many of them are generally regarded as being evil. Thus it's much easier to get to be a big business if you are evil than if you make an effort not to be.
Here comes in the cDNA microarrays which measure just that. (cDNA microarrays work by figuring out how much mRNA (the template for proteins) for a given gene is in a certain type of cell, and do this for 5000 or more genes at a time)
The problem with the genechips (and IIRC they normally look at mRNA, not cDNA) is that there's not that strong of a relationship between transcriptional level and translational level. If you really want to know about levels of protein expression (and more importantly, differences in level of expression) you're going to have to look at the proteins themselves. Good thing that's what pays my bills. Of course then you have to realize that the level of protein expression doesn't necessarily equate with protein activity and you have to look at post-translational modifications...
Another problem I see is that even if we are able to sequence the genetic code for all the proteins, what are we going to do with them. Identifying genetic diseases before they occur is all well and good but is it really that valuable if all we can tell people right now is that twenty years down the line you're going to get Hunington's disease or someother incurable ailment and die?
What you have to understand is that this is really very basic research. Just knowing the sequences alone is of comparatively little direct value. The real value comes from the fact that this will make all kinds of biological research tons easier. Knowing the sequences, for instance, makes it much easier to identify an "unknown" protein in minute quantities, which is critical to a huge number of experiments. The genome is basically a low level building block for generations of future biologists and medical researchers.
Knowing the amino acid sequences is a big key to being able to figure out how things work. Some examples:
You want to know what part of the genome makes us uniquely human rather than, say, a mouse. You will soon be able to compare the whole human genome to the whole mouse genome (which will be out in a couple of years) and see where they're similar and where they're different.
You want to know what things are really important for making organisms tick at a basic level. You can compare the whole genome of humans, mice, yeast, bacteria, etc. and find what genes in all of them are very similar. If it's close to the same in humans and bacteria, chances are it's really, really important.
You find a protein that's implicated in some disease or other. You correlate data generated from the unknown protein with the sequences for all human proteins to identify it. There's an excellent chance that you'll be able to figure out what it does by comparing it to known genes in other organisms.
You don't know what the protein above does. You can do experiments to see which other proteins it associates with (there are several ways of doing this) and that will often give you excellent information about what it does.
Coming soon You have identified a protein but can't figure out what it does. Using its sequence, you will soon be able to predict its 3-D structure, which can give you clues about what it does.
The real work will be turning the sequence into useful information. First what and where are the actual genes, then what proteins do the genes code for and what role the protein plays in metabolism or regulating other genes. Some idea of how much work needs to be done can be gathered from the fact that we don't even know how many genes there are - the most recent estimates for the total number of genes range from about 40,000 to 120,000. This process is called "annotating" and will take years.
I just got back from a conference where there was some very, very interesting work on data assisted annotation efforts. The basic idea is that you can look at the actual proteins produced by an organism and work your way back to finding the genes that specified them. This kind of approach could make the annotation effort a lot easier and speed the whole process up a lot. Those of us who work with proteins for a living also find it funny as hell that after hearing DNA folks brag about how everything is really in the DNA, they may need our help to finish up their work. Of course it's also great that once they're done with the genome, our work can really get started.
Easy; the one from Celera. Why? Because the scientific effort didn't make any attempt to apply something like the GPL to their data. That means that Celera is ahead and always will be ahead because they can combine their privately generated data with the publically generated data to get a more complete picture. The result is the Celera will be able to make a big profit by selling data half of which was funded by government sources.
A strong license might have been able to force Celera to release data that incorporated the publically funded results under less restrictive terms. Instead they can grab all that public effort, combine it with their own work (which is admittedly pretty impressive) and sell it back to people. It hardly seems fair.
I think that an important part of what you have to understand on this issue is that he's refering to a very formal concept of a trusted system. If you read the government guidelines on building trusted computer systems (e.g. the Orange book), one of the specific factors that is involved in designing systems at level B3 (IIRC) and above is that they be formally specified and proven to meet that specification.
While it's easy to gloss over this kind of requirement, there's some reason to think that it's actually a good idea. By the time you get to a class B system, you have to think about things like mandatory access controls, covert channel analysis, and the like. A formal demonstration that A) your system specification succeeds in meeting those goals and B) the system as built successfully implements the specification seems like a reasonable basis for reaching a high state of trust.
It's not impossible that you could build a free software project that could achieve this kind of goal. It certainly seems to be the case that every time someone says that free software can't do this, that, or the other thing, they've been proven wrong. But it's going to be tough to attract people to a project where you have to do stuff like keeping records of what you've done to meet design specifications, which is actually a requirement of high level trusted systems.
The ruling today was on a ACLU request for a preliminary injunction to prevent the application of the law while waiting for it to undergo review. They won an earlier, lower court ruling for an injuction, which is what was appealed. The government now has the option of trying to appeal the preliminary injuction further or to stop fighting it and go to the full trial that would shut down COPA completely.
It's pretty clear that a preliminary injunction has to be granted in this case; the ACLU has a pretty damn strong case, and they can show a real risk of harm if the COPA is enforced. That means that an appeal to the Supreme Court on the injunction is unlikely to get anywhere, which is why Wired is reporting that the government is leaning toward going to a full trial. After all, the sooner the appeals are over, the sooner the case can be brought to a conclusion.
Actually, they were looking at diesel engines, not gas ones. There are a fair number of diesel applications where the lifetime fuel costs are larger than the entire vehicle cost, much less the cost of the engine alone. A big rig may pile up 1 million miles, and at 5-7 mpg, that adds up to a lot of diesel fuel.
I'd suspect that the population size is probably pretty closely related to the complexity of the sample space. This particular example was looking at a system with only 6 parameters, so it may not have needed as large a generation size to get acceptable results. Of course they were also able to start out with the best known design rather than a random starting location (as many GA's use) so their search space may have been even more constrained than the number of parameters alone suggests.
Slashdot Mirror
Yes, and written letters are just bits of ink on pieces of paper, but using them is quite common in legal circles. Fairly reasonably, if I ask you for your records and I find something incriminating in them (and bear in mind that you also have to provide copies to the court, so I can't change them and claim that they're original) it should be your burden to prove that the incriminating comments were forged, rather than mine to prove that they're genuine! If anything, people should be suspicious if they show something unusually exculpatory, since you're far more likely to modify them in a way that reflects well on you than to forge records that incriminate you. In any case, IIRC these aren't emails from Gates's desktop machine; they're from the corporate email archive.
Getting back to something closer to the article that triggered the discussion, the FBI isn't talking about either of these things. They're talking about intercepting email in transit, so my original interpretation of the more conventional approach to header forging is more of what the FBI would be interested in. In thise case, though, the FBI's tap is actually less likely to be forged than a random email, since they're going to be tapping his immediate upstream connection, so a forger would need to insert their forgeries exactly there rather than at any random point in the network. As for the FBI being able to forge the email, they could potentially do that no matter what system you used, so you're going to have to trust them to be honest in any case.
One interesting aspect of this is that it suggests that if you're a criminal you shouldn't PGP sign your incriminating emails. If they're PGP signed, it provides the FBI with excellent evidence to use in court that they're not forged; unsurprising since proving authenticity is the intent of signing them. If they're unsigned, though, it'll be a lot easier to claim that the FBI forged them. You can probably enhance the effect by signing all of your non-incriminating emails (which you figure that even the most hardened criminal would have) so that you can intimate that the FBI forged the incriminating ones but were unable to forge the signature since they didn't have your private key.
While it's true that it's easy to forge email on the internet, that's not where the billg mail came from in the Microsoft case. In that case, the email was from Microsoft's internal email system. It had been turned over to the government as part of the pre-trial discovery phase, which is basically when the lawyers for the two sides are allowed to demand that the other side turn over information that might be relevant to the case.
Furthermore, the emails weren't just random mails from billg to the rest of the world. They were part of multiparty email correspondance on particular issues. IOW for Gates to disavow the emails, he would have had to claim that someone was not only forging his name but was also intercepting his personal emails and forging a conversation on his behalf. Not only that, but they were doing so not on some leaky internet system but on Microsoft's presumably secure internal system, and that the other people he was corresponding with, who presumably encountered him at least occasionally in person never brought up the topic of the emails in non-email conversation so that the forgery never came to light. That claim would be so obviously bogus that all it would do is damaged Gates's credibility as a witness and not impeach the credibility of the email at all.
Mod this up! (Score:5 Funny)
Actually, they are under no GPL obligation to make their software available on the net at all. The GPL only says that they have to make the source available anyone who gets the binaries, and that anyone who gets the source and binaries is free to redistribute them. IOW, you're free to buy the full price distribution and make all of the GPLed software available for free download, but you have no right to demand they make it available for free download.
FWIW, this is exactly what the Free Software Foundation advocates. To quote (from Selling Free Software on the FSF web site.):
(Emphasis is theirs.) IOW, the people who wrote the GPL don't just accept the idea of selling GPLed software for what the market will bear, they actually advocate it. Of course the fact that anyone who can get a copy of GPL software is free to redistribute it inherently limits the price anyone can charge, but that's for the market to decide.
Honestly, scanning your identity this way is about the last thing you should be worried about. The main goal of testers like these is to be able to scan people rapidly, like the metal detectors at airports. They want to be able to tell if someone is trying to smuggle bombs or drugs onto an airplane. That means that you need to know the answer from your test now, not in an hour or two when the guy's already had a chance to pass his stuff to some third party.
At the present, and for the forseeable future, it's just not possible to make a DNA-based individual ID in anything like real time. Even in the lab with nearly ideal samples doing that kind of thing takes time, and a lot of that is not something that can be easily reduced; certain chemical and physical reactions take time and can't be sped up. That puts a pretty strong damper on using this as a DNA vacum to violate people's rights.
OTOH, you can bet that the war against drugs and the war against terrorism will be used as excuses. Pretty soon you won't be able to get on a plane without being subjected to a battery of tests to make sure that you're not trying to put anything illegal onto the plane. Oops, you're a mining engineer who uses explosives at work? Prepare to be hassled every time you try to fly. Your pot smoking brother came over to visit? Prepare to be stopped and have your luggage examined. In the long term those kinds of minor erosions of personal protection are a much more dangerous threat to privacy than some hypothetical DNA screening.
While many end users are actually more interested in price/performance than they are in performance per se, the idea of listing price per performance is still a bad one. There are two main reasons for this:
Both of these factors suggest that rating by price/performance is a bad idea, and that rating just by performance is much better.
There's a bit more to it than that. IMO the real distinction is that (for reasons that don't make sense to me personally) many, if not most, people are perfectly happy to use use equipment they don't really understand. In fact, they seem to prefer not to understand it because they think that it will be too much effort to learn. This is, as Larry Wall would say, false laziness. Understanding in at least a general sense how things work at one level deeper than you actually use them is a huge time saver in diagnosing the inevitable problems you'll encounter.
The point is that with any complex technology there always seems to be a small group of people who gain real power through a greater understanding of it. One thing that I dislike about the article is that it focuses on the handful of kids who are going out and doing really outstanding things rather than the more interesting pattern of kids with more computer savy than their parents gaining household power. The latter is actually such a well accepted part of the system that it's a popular topic for newspaper cartoons and the like.
I certainly do think that the DEA should have to go in front of a judge to get a warrant for a wiretap, search, etc. The purpose of that "Red Tape" is to protect your rights. Going through the trouble to justify a warrant isn't some stupid formality that's there to get in the way. It's a vitally important step in preventing abuse of police powers. If there's actually a problem processing these things promptly that means you need more judges to deal with the workload, not that you should start ripping up the Bill of Rights to save time.
The key is that allowing a minor loophole in an important protection is like being a little bit pregnant. Once there's a single agency that can violate your rights, all of the other groups that want to can just call them up, trump up some bogus grounds for a search, and "happen" to find whatever else they were looking for.
The big thing to remember is that someone involved in a big, organized activity like large scale drug traffic will inherently produce a trail of physical and documentary evidence. Sooner or later some of that is going to show up in a way that will convince a judge that it's worth while letting the police have their warrant. It's people who haven't done anything wrong but just happen to look suspicious (or have pissed off a member of the police) who are the real potential victims here. When you let the police barge in wherever and whenever they want, you'll increase the number of innocent people harmed much more than the number of guilty people caught.
The relevant law is the Fair Labor Standards Act. Unfortunately, it appears that they included a big fat exception specifically for computer workers. Otherwise, you might find this document interesting. It mentions four classes of exempt employees: Executive, Administrative, Professional, and Outside Sales. For most of the people on Slashdot, the following definition about which professionals qualify for exemption is the most interesting:
The salary mentioned above means that:
The part about being paid in full for any week in which you work, regardless of the number of hours actually worked, is probably the most important thing here. IOW, if your employer can dock your pay for working part days (or apparently even part weeks!) you are not an exempt employee, unless you're an outside salesperson, doctor, lawyer, or teacher or a computer specialist paid hourly and earning at least $27.63 per hour.
The big thing is that if you really want to know your rights as an employee, you should really take a careful look at the Department of Labor web site.
It sounds to me as if they're engaged in a bit of classic (and illegal) abuse of exempt employee status. Legally, employees must be paid overtime for work beyond specified limits (generally either 40 hours per week or 8 hours per day) unless they're legally exempt. Exempt status is supposed to be reserved for work that falls into two categories: 1) management and 2) work involving significant creativity and independence. Working in QA does not sound as though it falls into either category.
Unfortunately, there are lots of businesses out there that are eager to abuse the system. They try to convince people that because they are on a salary, they don't have any right to overtime pay. Legally this is complete bunk; if you don't fit one of the legal categories for exemption, you should get overtime. In practice, though, it's all too effective. Employees think that being on salary means no overtime, so they don't even know to demand it and their employers get away with drastically underpaying them for their work.
That's the whole stock market bubble in a nutshell. The problem (if you want to call it that) is that there's more money out there and it has about the same number of outlets for investment. It winds up going into one of two places:
What it really boils down to is that a combination of greater real wealth and less even distribution of it make the equities market a less attractive investment.
This is a huge distortion. I've never heard any of the people who are really serious about the GPL complain about making a profit off GPLed code. Hell, RMS made money for a long time by selling tapes of GPL programs. On the FSF web site, in fact, they say:
That hardly sounds like people who don't want anyone to be able to make a profit from their code!
What people like the FSF do have a problem with is the idea of taking free/open source software and making it non-free/closed source. The GPL doesn't exist to prevent people from making a profit. The existence of companies that market GPLed software is proof that it doesn't accomplish that, even incidentally. The goal is to prevent people from taking software that is available in source form and releasing it without source.
This isn't as much of a problem as you might think. Most of the software that's released under the GPL includes the clause that the software may be licensed under version whatever of the GPL or, at the licencee's option, any later version. That means that if problems are found in the GPL in a way that makes it invalid, the software should revert to standard copyright (i.e. copying forbidden). Then the FSF only has to release a new version of the GPL that fixes whatever was wrong before and everything is OK again.
It's a good thing that this actually isn't allowed. The GPL does specify that the code must be A) on a medium customarilly used for transmitting code, B) in machine readable format, and C) in the form preferred for development. That means that a paper printout might be OK for a small bit of code, but your hypothetical shelf worth is pretty clearly a violation (particularly if it can't be read by OCR). It also means that code that is deliberately obfuscated is a direct violation.
The more I see people spouting this kind of stuff, the more I think that those people should actually stop to read the licenses under which their free software is distributed.
Actually, the translation can get a bit tricky, and there are reasons to stick with the DNA form rather than translating. With an unanotated DNA sequence you have to translate in all six relevant frames in order to be sure to get the one frame that's actually used. Even worse, with eukaryotic DNA you have to worry about the whole intron-exon structure. That's a little bit tough to handle with tr, or even a short perl script. The software I'm used to using will auto-translate in all frames, though, so I'm used to thinking of DNA and protein sequences as essentially the same thing.
On a deeper note, there are some interesting things that you can spot by looking at the untranslated DNA sequences. For homology matching, for instance, you can spot mutations that leave the AA sequence unchanged. More interestingly, you can look at things like codon bias. Codon bias is really interesting. There are, of course, multiple DNA triplets that code for the same amino acid. The more heavily translated a protein is, though, the more it tends to use only one of the available triplets to code for each amino acid. That means that given the DNA sequence for a protein you can make a reasonably accurate prediction of its abundance in the cell. Try that from the amino acid sequence alone!
It's true that big businesses are not necessarily evil. Ben and Jerry's, for instance, got to be quite a big business while deliberately using goodness as a strategy. IBM, after decades of being the evil empire, now seems to be making a pretty good effort to be a good corporate citizen.
The problem is that it takes a lot of time and effort to grow a business up to be a big business. There are various shortcuts along the way, and many of them are generally regarded as being evil. Thus it's much easier to get to be a big business if you are evil than if you make an effort not to be.
The problem with the genechips (and IIRC they normally look at mRNA, not cDNA) is that there's not that strong of a relationship between transcriptional level and translational level. If you really want to know about levels of protein expression (and more importantly, differences in level of expression) you're going to have to look at the proteins themselves. Good thing that's what pays my bills. Of course then you have to realize that the level of protein expression doesn't necessarily equate with protein activity and you have to look at post-translational modifications ...
What you have to understand is that this is really very basic research. Just knowing the sequences alone is of comparatively little direct value. The real value comes from the fact that this will make all kinds of biological research tons easier. Knowing the sequences, for instance, makes it much easier to identify an "unknown" protein in minute quantities, which is critical to a huge number of experiments. The genome is basically a low level building block for generations of future biologists and medical researchers.
Knowing the amino acid sequences is a big key to being able to figure out how things work. Some examples:
I just got back from a conference where there was some very, very interesting work on data assisted annotation efforts. The basic idea is that you can look at the actual proteins produced by an organism and work your way back to finding the genes that specified them. This kind of approach could make the annotation effort a lot easier and speed the whole process up a lot. Those of us who work with proteins for a living also find it funny as hell that after hearing DNA folks brag about how everything is really in the DNA, they may need our help to finish up their work. Of course it's also great that once they're done with the genome, our work can really get started.
Easy; the one from Celera. Why? Because the scientific effort didn't make any attempt to apply something like the GPL to their data. That means that Celera is ahead and always will be ahead because they can combine their privately generated data with the publically generated data to get a more complete picture. The result is the Celera will be able to make a big profit by selling data half of which was funded by government sources.
A strong license might have been able to force Celera to release data that incorporated the publically funded results under less restrictive terms. Instead they can grab all that public effort, combine it with their own work (which is admittedly pretty impressive) and sell it back to people. It hardly seems fair.
I think that an important part of what you have to understand on this issue is that he's refering to a very formal concept of a trusted system. If you read the government guidelines on building trusted computer systems (e.g. the Orange book), one of the specific factors that is involved in designing systems at level B3 (IIRC) and above is that they be formally specified and proven to meet that specification.
While it's easy to gloss over this kind of requirement, there's some reason to think that it's actually a good idea. By the time you get to a class B system, you have to think about things like mandatory access controls, covert channel analysis, and the like. A formal demonstration that A) your system specification succeeds in meeting those goals and B) the system as built successfully implements the specification seems like a reasonable basis for reaching a high state of trust.
It's not impossible that you could build a free software project that could achieve this kind of goal. It certainly seems to be the case that every time someone says that free software can't do this, that, or the other thing, they've been proven wrong. But it's going to be tough to attract people to a project where you have to do stuff like keeping records of what you've done to meet design specifications, which is actually a requirement of high level trusted systems.
The ruling today was on a ACLU request for a preliminary injunction to prevent the application of the law while waiting for it to undergo review. They won an earlier, lower court ruling for an injuction, which is what was appealed. The government now has the option of trying to appeal the preliminary injuction further or to stop fighting it and go to the full trial that would shut down COPA completely.
It's pretty clear that a preliminary injunction has to be granted in this case; the ACLU has a pretty damn strong case, and they can show a real risk of harm if the COPA is enforced. That means that an appeal to the Supreme Court on the injunction is unlikely to get anywhere, which is why Wired is reporting that the government is leaning toward going to a full trial. After all, the sooner the appeals are over, the sooner the case can be brought to a conclusion.
Actually, they were looking at diesel engines, not gas ones. There are a fair number of diesel applications where the lifetime fuel costs are larger than the entire vehicle cost, much less the cost of the engine alone. A big rig may pile up 1 million miles, and at 5-7 mpg, that adds up to a lot of diesel fuel.
I'd suspect that the population size is probably pretty closely related to the complexity of the sample space. This particular example was looking at a system with only 6 parameters, so it may not have needed as large a generation size to get acceptable results. Of course they were also able to start out with the best known design rather than a random starting location (as many GA's use) so their search space may have been even more constrained than the number of parameters alone suggests.