I'm not sure about current practice, but awhile back the protocol only called for a comparison at 10 point. each of those 10 points could have only a very few choices. Call it four, but I'm not certain. So after there are 10^4 samples in the databse you can expect to start getting LOTS of false matches.
That's not how DNA "fingerprinting" works. I assume by "10 points with four choices each" you're referring to ten nucleotides that can each be one of the four bases. But DNA fingerprinting doesn't involve actual sequencing, i.e. the identification of individual nucleotides. That would be too expensive, and largely futile because we are all 99.9+% identical.
Instead, you look a small number of repetitive DNA regions that are known to be highly variable between people as to the number of repeats. For example, a region that contains "CAG" repeated over and over again. Some people have 2 repeats, some people have 10, some people have 30. Thus, for each such region, there is a huge number of possible choices. There are hundreds of such regions in our genomes, and it's very quick and cheap to examine a dozen or so at once. 12 repeat regions, each with 10 possible repeat lengths, quickly gets you into the "1 in 100 million" territory that's typical for DNA fingerprinting. If you need even more discriminatory power, just look at more repeat regions.
DNA fingerprinting is ubiquitous precisely because it *doesn't* require expensive sequencing, and has enormous power to easily discriminiate between essentially infinite numbers of people. Identical twins excluded.
Actually, it's not ALL downhill after 30 anymore, but for a reason peculiar to the times we live in: there's a crossover point for anyone under around 70 years-old where exponentially advancing technology will slow, stop, then reverse the biological aging disease
I see this claim often, but it's really not so simple. And it's very unlikely to happen even for our (young) generation - though maybe for our grandchildren. The process of biological research does not follow Moore's Law, and only a few of its tools do. Even worse, the application of that research to therapies is severely hampered by more-or-less fundamental contraints on how we can manipulate things inside a living body. There are essentially only four very imperfect vectors: surgery, small molecule drugs, custom antibodies, and viral transduction (gene therapy). Our understanding of how to really customize the last three is still shockingly poor, and their capabilities are therefore very, very limited. Our current knowledge about cancer, for example, would stun a researcher put into a time capsule in 1970. But even more stunning to him would be the realization that we aren't yet even close to "curing", or even routinely manageing, cancer, despite this awesome accumulation of knowledge about the disease process. The therapies are a huge bottleneck. And in many diseases, including cancer, natural selection is proving a redoubtable foe, to a degree that we didn't appreciate 30 or 40 years ago. Resistant bugs, resistant viruses, resistant cancers: These are not problems that can be solved by a magic drug. By the very definition, they are problems that can never be absolutely solved.
Basically, what I'm saying is that biology is hard. Really hard. We will get to effective immortality one day, or at least to the limits of what our brains can handle, but I think even those born in 1980 aren't very likely to see it.
I don't want to start a flame war over KDE vs. Mepis but KDE is so damn cluttered looking, it ain't funny. It looks like an intimidating nightmare of icons right on the desktop. This doesn't even begin to describe the amount of programs under the start menu - half of which, in my experience. don't even work even when packaged by the distro (especially KDE IDE programming tools).
I wonder what distro produced such behavior. A default install of KDE on Gentoo leaves only two desktop icons (Home and Trash); and KUbuntu, IIRC, none at all. Maybe Linspire or Mandriva clutter up the desktop, but that's their doing, not KDE's. I've also never experienced the problems with KDE apps that you mention. Again, this might be the result of a faulty "second-messenger" install; i.e. the distro not doing its job, or an in-situ-upgrade gone wrong, rather than an inherent problem with KDE.
Yes, there are a ton of programs in the Kmenu, but there isn't an easy way around that, given how relatively difficult discoverability is on Linux. How many people will browse through/usr/KDE/3.x/share/apps to find an FTP program? Unless you want KDE to install only a handful of the most-used apps by default, but people's idea of "most-used" varies considerably - and then you need users to routinely resort to emerge/apt-get/whatever. Having a central "/Applications" directory violates modularity, since different installed versions of KDE, to say nothing of GNOME and other third parties, might stomp on each other. At least the menu is nicely organized. I really can't imagine that anyone willing to sit down at a Linux box is going to be intimidated by it. If it merely offends you, you can customize every aspect of it to your liking.
So, yes, sadly Fox News is the "most trusted" news source in America, if that sentence is true.
Not exactly, because it was a single-answer free-response question. The fact that Fox News led at only 11% shows that in a nation awash in news sources, Fox News viewers are the most monolithic block of news comsumers in America. In other words, people who get their news from Fox are more likely to only get it from Fox - a finding supported by other surveys. I'm actually surprised that Fox News didn't poll higher, given its near-mythical status among 30% or so of the population, including most of the government.
How about, instead, Windows stops using a keystroke that has meant "kill this process RIGHT NOW" for over 20 years? You know, Control-C ?
Exactly. And I wish KDE/GNOME didn't follow the Windows convention by default, although I realize they largely need to work with keyboards intended for use by Windows. Working in OS X Terminal is just heavenly compared to other platforms, because both the OS-level keystrokes (Cmd-C) *and* the terminal keystrokes (Ctrl-C) work just as expected. None of this "Ctrl-V to paste in every app exept Konsole, which uses Shift-Insert" gobbledegook.
The main problem with that advice: Most people overestimate their intelligence. Unintelligent people are FAR more likely to overestimate it than those who are actually in the top 5-10%. This is borne out by everyday experience, but there have been studies to back up that experience. Most stupid people think they're smart. So if you can't trust a person to decide whether they're smart or not, when you say "all smart people should drop out" you run the risk of convincing many stupid people to drop out, as well. And those stupid people are NOT likely to go much of anywhere in life without an education. They certainly won't be the next Jobs or Gates.
Mod parent up!
This might be a generalization of the dual burden of incompetence, that people lacking in ability also lack the capacity to properly evaluate what ability they have.
But there are other factors. Some recent debates in the US highlights these well: the need for identification to fly, and the need for identification for voter registration. In other words, ID is already necessary to fully participate in the society.
ID is not currently required for either of these things. You do not *need* to show a government ID to fly - but if you don't, you will surely be subjected to more rigorous screening. You absolutely do not need to show ID to register or to vote, anywhere. The recent debate was about proposed laws in a few states (Georgia, I remember for certain) that *would* require an ID, laws which have been attacked as discriminatory. At the moment, all you generally need to register is proof of residence (like a utility bill) and all you need to vote is your signature. Many folks would prefer it stay that way.
Cool, man. I've been using Gentoo for almost two years, OS X for four, and I amazingly had never read man's man. Man -k would have come in handy dozens of times. Thanks, for helping a slowly emerging n00b.
If you survey the top IP firms you'll find that their attorneys typically have, at minimum, a BSc in the field that they practice in. More often they have an MSc or a Phd. Moreover in such firms it is usual that the work of all attorneys be reviewed by partners and senior attorneys who have practiced much longer than their three year JD/LLB would indicate. Be careful with how far you take your analogies... unless you expect that you can so easily master the language and knowledge that comes with such experience.
Is this really true? I surveyed the partners at SeedLaw, a prestigious IP law firm in the Seattle area. Only 3 of 15 had PhDs, and four more had MS's. All were undergrad science majors, but an undergrad biology degree, at least, isn't considered a qualification for more than rote technical work (or further training). In any event, you're describing a different world than the one inhabited by the story poster. JD/PhDs with decades of experience don't come cheap - nothing that requires 10+ years of post-graduate schooling and apprenticeship comes cheap - so if you can afford to retain a firm staffed by such wonders, you're a corporate entity with funding to burn. The invention content of the patent is secondary to the legal status it gives to sue and/or cross-license with similarly deep-pocketed corporate competitors. Heck, I imagine that the cost of filing and pursuing a patent with such a firm can easily exceed the average engineer's annual salary.
There are, for example, plenty of scientists who are not programmers, but who learn enough to adequately accomplish their scientific tasks. The results are rarely pretty, but often functional. Is the same not true for patents? Or should small fry not even bother to play this game? Is the only reason to apply for a patent to sue someone? Should patents only be applied for if the immediate stakes are high enough to justify spending obscene amounts of money on gold-plated lawyers, as a means to generating even more obscene amounts of money through legal action?
Thank you. That was one of the most informative summaries I've read on the subject. There are still some things that seem to be constiutively confusing, though. Willful infringement seems only tangentially related to the patent process - because to be willful, you must have seen the prior patent, and you would likely only have seen it if you were doing a search for a new patent. Trying to patent a product that has already been patented won't get you sued for infringement (although your application should be rejected), but selling that product will. So I understand the point about abstraction and anonymization, but the extrapolation is that no one involved in developing or selling anything should ever look at the patent archive for any reason, lest they open themselves up to willful infringement suits. That's patently (sorry) absurd, since it abrogates the whole constiutional point of having a patent archive.
the main function of patent attorneys and agents is to write precisely, using terms and phrases the definition of which have previously been determined
This is surely true, but in most cases it must be easier for the inventor to learn the language than for the lawyer to understand the science and technology behind the invention. As your analogy suggests, scientists and programmers are no stranger to the precise use of specialized language. How much of the current patent morass is due to patents being written very precisely by people who understand imprecisely what it is they're actually writing about?
I realize I'm tiling at long-vanished windmills here, and I recognize the truth in the advice I received a short time ago that the only reason to file a patent is in order to sue someone - and ergo, a patent should be written by lawyers, for lawyers. But these disucssions, to me, just emphasize how little sense the current patent-lawyer system makes, with respect to the constiutional purpose of patents.
1. If you search for stuff related to A but find B, you're now screwed. You can't proceed without purchasing a license for B.
2. If you hire someone to search for you, even if they find B, they will ignore it as irrelevant to A. They file a patent application, noting prior art. You can later amend the application if necessary.
Even if method A infringes on another patent, it is still possible to proceed. The buyer of your patent can purchase licenses for the prior art for method A as well as the license for method B. You get paid for your work.
If you are a lwayer, please help me understand the distinction here. I see two possibilities. First, using a lawyer is advantageous because he doesn't understand the subject matter, and so will either ignore or misinterpret the significance of B. But this ignorance of the subject won't have any significance consequences down the road, because he's a lawyer. Second, using a lawyer is advantageous because he'll talk your licensee into also licensing B, while if *you* had to talk your licensee into also licensing B, you'd be screwed.
Either way, I don't get it. What does involving a lawyer provide, aside from (useful?) technical ignorance and a target for blame?
(Note on "1", to clear up an ambiguity in your logic - you can most certainly proceed with a patent without licensing dependent methods that have already been patented. You can freely patent A. You can license the patent for A. The only thing you cannot do is commercially develop A, without first licensing B. Whether a lawyer is involved or not is irrelevant to the requirement, or lack thereof, of licensing B.)
Other than confirming the time line, what's your point? Apple launched developer-only releases of Rhapsody in 1996, making it clear that the release was about to happen.
Wait, huh? 1996? Apple's acquisition of NeXT wasn't even announced until December 20, 1996, and wasn't consummated until February 4, 1997. Rhapsody wasn't even demonstrated until later in 1997, and I don't think anyone was led to believe that the prospective late 1998 release would be a polished mass-market product.
Vista is hardly a "service pack update" to Server 2003.
"Vista" is the re-application of the few remaining Longhorn features to Server 2003. By MS's low standards, it may be more than a service pack, but by Apple standards it's looking like an underwhelming point release.
Apple launched developer-only releases of Rhapsody in 1996, making it clear that the release was about to happen.... And as for the mythical "2003" release date: Sure, that was a target date, once, well before the operating system even started major development.
So one vendor's "clear" target date is another vendor's "mythical" target date. If you really want to pretend that Rhapsody was supposed to be OS X, at least admit that when OS X finally emerged a few years later it was almost unrecognizably more developed than Rhapsody. Apple spent its time adding new features and completely transforming the OS. MS has spent its time scaling back features and reducing Vista to, like I said, barely a SP of Server 2003.
If you think Microsoft's going to switch to Mac OS X on the basis that its new operating system is slightly late, and out of envy that Mac OS X came out, bug ridden and barely usable, only five long years after NeXT took over Apple, then you're smoking crack.
Whoa there. I couldn't care less what Microsoft does or doesn't do, and I think anyone who expects them to drop the NT kernel for Mach/Darwin is indeed smoking crack. Just as much crack as anyone who expects the opposite. I'm just pointing out that however badly Apple failed with Copland, once the NeXT folks came on board they did a hell of job getting OS X polished and out the door in just a few years. Given Microsoft's recent history of OS development, they ought to be taking process lessons from Apple. Hell, they've practically admitted as much with their recent shakeup of the Windows division execs, bringing in Office guys who know how to advance a code base (in what direction, I leave to the reader's opinion, but no one can deny that Office has been going somewhere).
You're talking about OS X, which was originally shipped four years late and totally unusable (the OS didn't become usable until 10.2, in 2003)?... "Longhorn" is being delayed by two months.
You're taking some sweet crazy pills there. Apple purchase NeXT in 1996, and rolled out 10.0 in early 2001 - barely four years later. How can that translate to "four years late"? I agree that 10.2 was the first OS X that was really ready for prime-time; but it was out the door barely a year after 10.0, in August 2002.
Apple went from zero to mass-market-ready in less than the six years it's taken Microsoft to scale Vista back to a service pack update to Server 2003. Like the other reply said, Vista is way more than 2 months behind its original schedule - in fact, very nearly your mythical four years, because the original announced target date was late 2003. And Microsoft is only finally getting it out the door because they've stripped every major feature that was going to make it Longhorn in the first place. A prettified and hardware-accelerated GUI is about the only survivor.
Those are important and we should all be grateful for their generosity. However, they contribute only a very, very small percentage of the overall amount of money that goes into bioscience research
Good point. As rich as Bill Gates is, the NIH disperses roughly the equivalent of his entire fortune every year, and this represents only a tiny fraction of the federal budget. No private agency could possibly hope to match a long-term public commitment. Nor should they. This is what governments are for, harnessing the power of society to advance the public good.
And the thousands of American scientists, engineers, technicians and support staff that design and work on these systems. Based on comments like this, you'd think that the government is stuffing shells full of cash and launching them at the enemy. Where do you think these "weapon systems" are designed and built?
True, but one can argue that at least a few of those scientists, engineers, technicians and support staff who are on the government dole building weapons could be better put to use creating new energy sources, curing diseases, advancing our understanding of the universe, etc. Every engineer employed by federal money to study a cutting-edge aspect of missle-defense mechanics is one fewer biologist or doctor funded by federal money to cure cancer. And right now there are many, many more of the former than the latter. Yeah, in a perfect world we'd fund both. But the reality is that defense spending is still booming while the NIH budget (barely a rounding error on the DoD budget anyway) is actually shinking in real terms.
Thus it was not Islamic advances in Europe that directly caused the crusades; it was Islamic advances against the Byzantine empire (which is not in what is currently considered Europe).
Not only that, but Manzikert was a battle between the Byzantines and the Seljuk Turks, a tribe of converted nomads who had swept down into the Middle East from Central Asia in the preceeding decades and overrun most of the Muslim states between Afghanistan and Egypt. The Seljuks spilled far more Arab-Muslim blood than Christian. They were a disaster for the traditional Muslim societies in the Middle East. But they were only the first of a series of nomad conquerers who were to dominate the Middle East for the next, well, millenium. Seljuks, Mongols, and later Ottomans. Not until after WWI was the nomad (Ottoman) control finally broken. This long dominance by nomadic outsiders - converted to Islam though most eventually were - is probably what caused the gradual decline in Islamic culture in the Middle East.
It would, if Admin users didn't still need to enter their password and authorize the iChat trojan.
Actually, the reports were clear that it doesn't require a password. The reason is that it only modifies iChat.app, not any system files. An admin user has read/write access to/Applications, no authentication necessary. Try it yourself (modifying/Applications, that is, not running the worm).
You're absolutely right that admin != root; but nor is it quite as blind, deaf and dumb as an unprivileged user.
Unlike Windows, it's perfectly safe to run full-time as the "Administrator" user, and nearly every OS X user does.
It's mostly safe, not perfectly safe. The iChat virus/trojan suggests one reason why. Since an admin has free access to/Applications, a bug running under that user's permissions can modify apps in that folder, helping the bug to spread itself either locally (next time another user on the machine opens an infected app) or remotely (e.g. via a modified iChat). A second reason is that admin users can sudo with their own password. If the admin account's password is compromised by a bug or hacker, root control of the machine goes with it. This is not the same as running as root, like Windows admins do, and viruses running under the admin user's permissions do not have root access. A regular user must enter an admin's username/password to sudo, making the virus/hacker's job more complicated.
Nearly ever OS X user on a single-person machine runs as admin, and that's what Apple sets up by default. But it's not a bad idea to reocnsider.
In reality, the site is probably most interesting because it is an attempt at an ajax webapp for scientists.
I played with it a bit, and I don't think there's any AJAX in there. It does try to interconnect the various modules, which is neat, but everything is done through regular CGI links and page-loads.
Ultimately, the best solution is to teach programming to biologists as a core subject. Our old categorization of the sciences is out-moded.
Hear hear. I don't know how a molecular biologist can live these days without at least knowing enough of a scripting language to manipulate sequences and high-throughput data. The gulf that's evolving between bench biologists and non-biologist "bioinformaticians" is getting too wide for either group to effectively interact with the other.
OSX sort-of does this already. Whenever you launch an executable for the first time - either directly or indirectly, by opening a dependent data file - OSX puts up a prompt warning you that you are about to execute program XXX for the first time, and asking if you're sure you want to do this. So every executable is considered non-trusted the first time you run it. This trojan ought to trigger this dialog, which I'd think would cause even most novice users to wonder if something odd was going on.
Repost this after they've shown some actual data, gotten it published in a respectable peer-reviewed journal, and had independent investigators replicate it. There isn't a single hit for this "family" of compounds on PubMed, and the compound is named after a frikkin biotech company, so color me extremely skeptical - of both data and motives.
This is why normal people get fed up with science. Their exposure to science is through media stories, PR bullshit like this, which says "Huzzah! Cure for X found!" Later on, we find out that the data is too weak to pass peer-review, that the new compound is toxic, that it only weakly suppresses X in animal models, and that X is not yet, in fact, cured. The real scientists around the world keep at their benchwork, with barely a glance up, steadily and (to the public) inconspicuously advancing our fundamantal understanding of X. But five years later, Mr. Normal Person hears another story like this one and says to himself "Didn't they cure X years ago? What are those ivory-tower leeches spending my $30 billion a year on, anyway?"
Slashdot Mirror
Instead, you look a small number of repetitive DNA regions that are known to be highly variable between people as to the number of repeats. For example, a region that contains "CAG" repeated over and over again. Some people have 2 repeats, some people have 10, some people have 30. Thus, for each such region, there is a huge number of possible choices. There are hundreds of such regions in our genomes, and it's very quick and cheap to examine a dozen or so at once. 12 repeat regions, each with 10 possible repeat lengths, quickly gets you into the "1 in 100 million" territory that's typical for DNA fingerprinting. If you need even more discriminatory power, just look at more repeat regions.
DNA fingerprinting is ubiquitous precisely because it *doesn't* require expensive sequencing, and has enormous power to easily discriminiate between essentially infinite numbers of people. Identical twins excluded.
Basically, what I'm saying is that biology is hard. Really hard. We will get to effective immortality one day, or at least to the limits of what our brains can handle, but I think even those born in 1980 aren't very likely to see it.
(IAAB)
Yes, there are a ton of programs in the Kmenu, but there isn't an easy way around that, given how relatively difficult discoverability is on Linux. How many people will browse through
See Newton's Laws.
This might be a generalization of the dual burden of incompetence, that people lacking in ability also lack the capacity to properly evaluate what ability they have.
Cool, man. I've been using Gentoo for almost two years, OS X for four, and I amazingly had never read man's man. Man -k would have come in handy dozens of times. Thanks, for helping a slowly emerging n00b.
There are, for example, plenty of scientists who are not programmers, but who learn enough to adequately accomplish their scientific tasks. The results are rarely pretty, but often functional. Is the same not true for patents? Or should small fry not even bother to play this game? Is the only reason to apply for a patent to sue someone? Should patents only be applied for if the immediate stakes are high enough to justify spending obscene amounts of money on gold-plated lawyers, as a means to generating even more obscene amounts of money through legal action?
I realize I'm tiling at long-vanished windmills here, and I recognize the truth in the advice I received a short time ago that the only reason to file a patent is in order to sue someone - and ergo, a patent should be written by lawyers, for lawyers. But these disucssions, to me, just emphasize how little sense the current patent-lawyer system makes, with respect to the constiutional purpose of patents.
Either way, I don't get it. What does involving a lawyer provide, aside from (useful?) technical ignorance and a target for blame?
(Note on "1", to clear up an ambiguity in your logic - you can most certainly proceed with a patent without licensing dependent methods that have already been patented. You can freely patent A. You can license the patent for A. The only thing you cannot do is commercially develop A, without first licensing B. Whether a lawyer is involved or not is irrelevant to the requirement, or lack thereof, of licensing B.)
Apple went from zero to mass-market-ready in less than the six years it's taken Microsoft to scale Vista back to a service pack update to Server 2003. Like the other reply said, Vista is way more than 2 months behind its original schedule - in fact, very nearly your mythical four years, because the original announced target date was late 2003. And Microsoft is only finally getting it out the door because they've stripped every major feature that was going to make it Longhorn in the first place. A prettified and hardware-accelerated GUI is about the only survivor.
You're absolutely right that admin != root; but nor is it quite as blind, deaf and dumb as an unprivileged user.
Nearly ever OS X user on a single-person machine runs as admin, and that's what Apple sets up by default. But it's not a bad idea to reocnsider.
Hear hear. I don't know how a molecular biologist can live these days without at least knowing enough of a scripting language to manipulate sequences and high-throughput data. The gulf that's evolving between bench biologists and non-biologist "bioinformaticians" is getting too wide for either group to effectively interact with the other.
OSX sort-of does this already. Whenever you launch an executable for the first time - either directly or indirectly, by opening a dependent data file - OSX puts up a prompt warning you that you are about to execute program XXX for the first time, and asking if you're sure you want to do this. So every executable is considered non-trusted the first time you run it. This trojan ought to trigger this dialog, which I'd think would cause even most novice users to wonder if something odd was going on.
Repost this after they've shown some actual data, gotten it published in a respectable peer-reviewed journal, and had independent investigators replicate it. There isn't a single hit for this "family" of compounds on PubMed, and the compound is named after a frikkin biotech company, so color me extremely skeptical - of both data and motives.
This is why normal people get fed up with science. Their exposure to science is through media stories, PR bullshit like this, which says "Huzzah! Cure for X found!" Later on, we find out that the data is too weak to pass peer-review, that the new compound is toxic, that it only weakly suppresses X in animal models, and that X is not yet, in fact, cured. The real scientists around the world keep at their benchwork, with barely a glance up, steadily and (to the public) inconspicuously advancing our fundamantal understanding of X. But five years later, Mr. Normal Person hears another story like this one and says to himself "Didn't they cure X years ago? What are those ivory-tower leeches spending my $30 billion a year on, anyway?"