You have some excellent points, but you might be addressing a totally different problem. These drugs that you're mentioning actually fit the bill of the bottom-feeders of the industry - the ones making generics. They LOVE to sit around, look for "new and innovative" ways to basically circumvent patents made by larger corporations. Most research pharmaceuticals will not mess with drug packaging quite frankly because selling the various packaging is worth only pennies.
As for the stuff about the various pain killer formulations, you're again confusing generic over the counter drugs with larger research drugs. A much better example would be Vioxx and Celebrex. Admittedly, these two drugs work on the same target - the cox2 receptor. However, they involve vastly different structures, since they were effectively found independently by two large corporations, namely Pfizer and Merck. That is why Celebrex was pulled off the market by Merck for having nasty cardiovascular side effects, while Vioxx was able to stick around, even though they were both related painkillers. Once a drug goes over the counter, it's effectively worthless - the company is no longer selling the "licenses" to use the drug anymore, and the pennies it costs to make the drug no longer contribute to research funding.
Also, "mirroring" a drug, aka manufacturing a pure enantiomer, is a decidedly non-trivial process. See the infamous example of thalidomide. One enantiomer is great for treating morning sickness - the other is a highly potent mutagen. The initial manufacturing was of a racemic mixture of both, which created the infamous thalidomide babies. If a corporation discovered a manufacturing method that could effectively make a pure enantiomer, that is definitely worth money since it is so hard to do so. One thing you have to remember about the chemistry of medicines - minute, seemingly pointless changes to the structure of the medicine have VASTLY different effects on the body. It is very different from microprocessors - you add an extra instruction, it's not going to make much of a difference in performance. You add a flouride to a benzene ring on an opoid agonist, and you turn heroine into a successful non-addictive painkiller.
One last point - do not confuse the crazy copyrights and patents found in the computer realm with those in drug research. Drug patents is almost strictly 17 years. If Prilosec's patent runs out, the generics can make it to their heart's content - nothing AstraZeneca can do will stop them. Patenting Nexium has no effect on their sales of prilosec- it's still going down the tubes. If you're pissed that they didn't release nexium sooner, that's really basic economics - they don't want to cannibalize sales of their own successful drug. This is found in the computer industry too - Nvidia didn't release the 8800 ultra until recently because they wanted to cash out on the 8800 GTX first. HOWEVER, if a competitor happened to come around (namely the rather lame ati 2900), you can bet they'd release something new. This happened with Claritin and Allegra in the pharmaceutical market - Allegra was nearly squeezed out of the market when claritin went OTC.
IMHO, pharmaceutical research should really be limited to government-funded research. A corporation has too much overhead to be successful in a field where you literally need an up-front investment in the BILLIONS of dollars to even have a HOPE of finding a single drug that might break even. The news of various countries actively proclaiming that they were violating patents for anti-retrovirals (AIDS medications) was rather painful for me. This sent a very clear message to the industry - don't research AIDS medications! These countries shot themselves in the foot by not allowing the corporations to make a return on investment. This tiny bit of news effectively stopped the flow of money to AIDS medication research in corporations. This is why government should be doing this type of research, not corporations...
Just to clear something up - drug patents last 17 years. All stages of clinical tests put together lasts about 10 years. Company has about 7 years to break even before the drug becomes a useless generic, assuming some other drug doesn't come up and steal their market (see Claritin vs. Allegra). Break even point for most drugs is about 1 billion in sales. Aside from the odd "blockbuster" drugs like Viagra and Allegra, sales for most drugs barely top the hundred millions. In other words, it's fairly hard to make money off drug research. I've worked in the pharmaceutical industry before... The combination of absurd (but necessary!) FDA tests and public stigma make pharma an awful industry to work in. That's why I'm not in it anymore...
Why not just use the simplest one - a black box? Information can't be extrapolated from something that doesn't exist. Also, I doubt you'd have problems with recognition - if anyone sees a body, neck, and a black box, I doubt they'll think you have decapitated people running around...
It's passed the parent test for me, and I've been slowly switching my friends over to ubuntu now. For one thing, it's blindly easy now to fix things for my parents. SSH is a godsend for remote administration, since my mother has a bad habit of breaking things (she randomly clicks around, changes things, you know - naturally curious). I'll tell you one thing - ubuntu has VASTLY better hardware support out of box than any install of winxp or vista I've ever done. I can't comment on OSX install because I've never done one. Why haven't I done one? Well because the closest I've come to installing OSX was ripping it off of a broken ibook and installing the latest ubuntu, hehe.
Hmm... it seems that you might not be very familiar with the intricacies of raytracing... Having just recently finished a raytracing homework assignment, I can tell you that it's no magic bullet. I can definitely see it replacing shadow rendering, for example, but poly count? That's even more of a problem for raytracing than raster graphics. There still exists the need to send rays out to EVERY single polygon in the scene, (BSP trees improve performance of that, of course). Texture sampling is still a problem - in fact the algorithm we used for rendering textures on raytraced objects is extremely similar to the one we're using in the next assignment for rendering raster graphics. In fact, it might be even worse because aliasing becomes a pretty big problem in raytraced textures. The blurriness will still be there if don't want nasty discontinuities everywhere.
I haven't quite RTFA yet, but I can guess at what they're using - probably a pretty complex raytracer based on proximity and continuity (as in, the color of a poly at the intersection would project to roughly the same color as in adjacent pixels). This still doesn't improve the complexity however - in the simplest terms, a single raytraced pixel requires about the same computation as an entire frame in raster. Obviously, some day in the future we might all be typing on 80-core machines, and we might be able to pump out fully-raytraced images at real time. However, as of right now, unless the resolution is fairly low, i doubt it would very usable...
I'd have to contest that. Sure horses can't run to the moon, but neither can our automobiles. I'm sure in 1870, they would have known that the fun fireworks that they were firing off would have the best chance of getting all the way up there. What's the modern day equivalent of that?
Ehhh, might actually prove to be a pretty fun project for some bored google engineer. I'm guessing that the program picks from a dictionary of possible words. I'd guess that it might actually be possible to run some sort of a statistical analysis over your data set and drop the queries that match the statistical profile of randomly selected words from this dictionary. Either that or try some sort of a semantic analysis - those queries don't make much sense semantically, and can be dropped. Nothing is foolproof:)
Umm.... I'd doubt that an OLPC will survive much bashing either. If one had to pour the compaq out of the paper bag, well the best the OLPC can probably do is be picked piece by piece from a box. Either way, the OLPC's software is open source, and sticking that on the low end compaq would be fine too.
I'm lost, what exactly are you trying to imply? Yes, some of these techniques did exist in the 1970's, but Monsanto sure wasn't making GM corn around then, that was late 80's/early 90's. The manipulations that I used as examples are pretty common genetic manipulation techniques used in labs today.
What makes you think that? Monsanto does it frequently for their GM corn to avoid dangerously spreading the crop. Granted it's probably a lot harder for an animal instead of a plant, but genes do exist for autotermination. One exists in humans already, instead in the opposite direction. Phenylketonuria, or PKU patients cannot eat foods with phenylalanine. It's standard practice to couple transfection studies of bacteria with an antibiotic resistance gene, then grow them on media containing antibiotics. In fact, I did that one in high school biology (albeit unsuccessfully). Sterility can be easily induced through a variety of genetic modifications. Perhaps it's your chance to review on modern genetic manipulation techniques?
The consequences are indeed quite SEVERE if this GM mosquito doesn't work like we expect it to. It's stronger than natural mosquitoes, making it harder to control. This alone makes it a very dangerous feature.
If anything, they should do a test using a specially designed mosquito. This breed must be sexually incompatible with natural mosquitos and they should lack some critical non-natural nutrient in order to survive. By releasing them into an area and only providing this nutrient source in that one area, it's possible to safely test to see how an ecosystem would react to a GM mosquito. Of course this will still not eliminate the possibility of a mutation that could disable the "death gene" and cause a dangerous spreading of the GM mosquito...
Yeah, taskbar grouping is definitely one of the first things I turn off. However, barring that (utterly retarded) feature, the taskbar is by far the most intuitive task indicator possible. Again, I'm a linux person anyways, so having multiple desktops are also necessary for completely fluid work. Grouping a set of tasks into a desktop greatly reduces the number of buttons on the taskbar. However, desktop icons and the taskbar still indicates what exactly is running at a glance (without the need to press a button). For example, desktop 1 is my browser desktop, so I can immediately see what windows I have open. IMs are grouped into tabs already - untabbed clients are for uncivilized people;) .
Except Merck and Pfizer would not EXIST if it weren't for the patents. Have you heard of Andrx or Ivax? They are generic drug companies. They do exactly what you say should happen - they sell drugs that are out of patent, and they have equal access over any (old) compounds. They do little or no research of their own, because they don't have the money to do so.
The actual COST of a pill is almost nothing - no more than the cost of generic acetaminophen in your local grocery store. Any organic chem graduate could take almost any of the major drugs on the market and design a workable and mostly efficient synthesis method. Any chemical engineering graduate could probably take one of these methods and upscale it to million-pill quantities. Ensuring that these drugs are safe is where most of the cost is.
You'd also have a pretty nice big influx of borderline drugs/cocktails like Fen-phen and the Cox-2 inhibitors. The general population is a very hypocritical crowd. They want cheap drugs without paying for its development. They want drugs to come out quickly, but they must be 100% safe. Many argued that the FDA wasn't strict enough, and that they let the Cox-2 inhibitors come out. And you're sitting here suggesting that they should be completely abolished, allowing any and all drugs to be tested. So, which way would you like it? High priced but safe drugs, low priced and unsafe drugs, or no drugs at all (when there's no development)?
Two issues with what you've said so far - alt-tab doesn't switch individual windows by default, and there's no good place to find out. Thankfully, you've actually shown me a very useful bit of info, and that will greatly lessen the hassle of using mac. Also, that program that you showed me still requires a key be pressed to see exactly what's running. What exactly is wrong with the taskbar? Was it too ugly for your saintly OS of beauty?
Also, don't confused me with a windows user. The lack of multiple desktops becomes especially annoying for me after a while. But at least the presence of a taskbar keeps me sane.
For one thing, changing between the 6-10 windows with a single keyboard shortcut, or knowing at a glance exactly what each console is running at the time. And no, expose is not the answer, because it requires a lot more than a single smooth motion (alt-tab) to get to any specific window.
Developing stuff on OSX is a pain. I can't SEE what's running, I can't switch windows conveniently or without the mouse. And who the hell thought of the idea of switching the alt and ctrl in the console?! I get used to alt-everything for shortcuts, go into terminal, and WHAM nothing works anymore. Get used to ctrl-everything, switch out of console, WHAM nothing works again. It's a constant battle with the operating system to do what I want.
Slashdot Mirror
You have some excellent points, but you might be addressing a totally different problem. These drugs that you're mentioning actually fit the bill of the bottom-feeders of the industry - the ones making generics. They LOVE to sit around, look for "new and innovative" ways to basically circumvent patents made by larger corporations. Most research pharmaceuticals will not mess with drug packaging quite frankly because selling the various packaging is worth only pennies.
As for the stuff about the various pain killer formulations, you're again confusing generic over the counter drugs with larger research drugs. A much better example would be Vioxx and Celebrex. Admittedly, these two drugs work on the same target - the cox2 receptor. However, they involve vastly different structures, since they were effectively found independently by two large corporations, namely Pfizer and Merck. That is why Celebrex was pulled off the market by Merck for having nasty cardiovascular side effects, while Vioxx was able to stick around, even though they were both related painkillers. Once a drug goes over the counter, it's effectively worthless - the company is no longer selling the "licenses" to use the drug anymore, and the pennies it costs to make the drug no longer contribute to research funding.
Also, "mirroring" a drug, aka manufacturing a pure enantiomer, is a decidedly non-trivial process. See the infamous example of thalidomide. One enantiomer is great for treating morning sickness - the other is a highly potent mutagen. The initial manufacturing was of a racemic mixture of both, which created the infamous thalidomide babies. If a corporation discovered a manufacturing method that could effectively make a pure enantiomer, that is definitely worth money since it is so hard to do so. One thing you have to remember about the chemistry of medicines - minute, seemingly pointless changes to the structure of the medicine have VASTLY different effects on the body. It is very different from microprocessors - you add an extra instruction, it's not going to make much of a difference in performance. You add a flouride to a benzene ring on an opoid agonist, and you turn heroine into a successful non-addictive painkiller.
One last point - do not confuse the crazy copyrights and patents found in the computer realm with those in drug research. Drug patents is almost strictly 17 years. If Prilosec's patent runs out, the generics can make it to their heart's content - nothing AstraZeneca can do will stop them. Patenting Nexium has no effect on their sales of prilosec- it's still going down the tubes. If you're pissed that they didn't release nexium sooner, that's really basic economics - they don't want to cannibalize sales of their own successful drug. This is found in the computer industry too - Nvidia didn't release the 8800 ultra until recently because they wanted to cash out on the 8800 GTX first. HOWEVER, if a competitor happened to come around (namely the rather lame ati 2900), you can bet they'd release something new. This happened with Claritin and Allegra in the pharmaceutical market - Allegra was nearly squeezed out of the market when claritin went OTC.
IMHO, pharmaceutical research should really be limited to government-funded research. A corporation has too much overhead to be successful in a field where you literally need an up-front investment in the BILLIONS of dollars to even have a HOPE of finding a single drug that might break even. The news of various countries actively proclaiming that they were violating patents for anti-retrovirals (AIDS medications) was rather painful for me. This sent a very clear message to the industry - don't research AIDS medications! These countries shot themselves in the foot by not allowing the corporations to make a return on investment. This tiny bit of news effectively stopped the flow of money to AIDS medication research in corporations. This is why government should be doing this type of research, not corporations...
Just to clear something up - drug patents last 17 years. All stages of clinical tests put together lasts about 10 years. Company has about 7 years to break even before the drug becomes a useless generic, assuming some other drug doesn't come up and steal their market (see Claritin vs. Allegra). Break even point for most drugs is about 1 billion in sales. Aside from the odd "blockbuster" drugs like Viagra and Allegra, sales for most drugs barely top the hundred millions. In other words, it's fairly hard to make money off drug research. I've worked in the pharmaceutical industry before... The combination of absurd (but necessary!) FDA tests and public stigma make pharma an awful industry to work in. That's why I'm not in it anymore...
Leopard on a 486?! Oh the HORROR!
Why not just use the simplest one - a black box? Information can't be extrapolated from something that doesn't exist. Also, I doubt you'd have problems with recognition - if anyone sees a body, neck, and a black box, I doubt they'll think you have decapitated people running around...
It might be wrong headed, but it sure would be funny :-D
It's passed the parent test for me, and I've been slowly switching my friends over to ubuntu now. For one thing, it's blindly easy now to fix things for my parents. SSH is a godsend for remote administration, since my mother has a bad habit of breaking things (she randomly clicks around, changes things, you know - naturally curious). I'll tell you one thing - ubuntu has VASTLY better hardware support out of box than any install of winxp or vista I've ever done. I can't comment on OSX install because I've never done one. Why haven't I done one? Well because the closest I've come to installing OSX was ripping it off of a broken ibook and installing the latest ubuntu, hehe.
Never say never... http://www.artlebedev.com/everything/optimus/
Hmm... it seems that you might not be very familiar with the intricacies of raytracing... Having just recently finished a raytracing homework assignment, I can tell you that it's no magic bullet. I can definitely see it replacing shadow rendering, for example, but poly count? That's even more of a problem for raytracing than raster graphics. There still exists the need to send rays out to EVERY single polygon in the scene, (BSP trees improve performance of that, of course). Texture sampling is still a problem - in fact the algorithm we used for rendering textures on raytraced objects is extremely similar to the one we're using in the next assignment for rendering raster graphics. In fact, it might be even worse because aliasing becomes a pretty big problem in raytraced textures. The blurriness will still be there if don't want nasty discontinuities everywhere. I haven't quite RTFA yet, but I can guess at what they're using - probably a pretty complex raytracer based on proximity and continuity (as in, the color of a poly at the intersection would project to roughly the same color as in adjacent pixels). This still doesn't improve the complexity however - in the simplest terms, a single raytraced pixel requires about the same computation as an entire frame in raster. Obviously, some day in the future we might all be typing on 80-core machines, and we might be able to pump out fully-raytraced images at real time. However, as of right now, unless the resolution is fairly low, i doubt it would very usable...
Oh, so it would take Opera only a year and a half, or one year to load, depending on your reference? That is quite an improvement!
I'd have to contest that. Sure horses can't run to the moon, but neither can our automobiles. I'm sure in 1870, they would have known that the fun fireworks that they were firing off would have the best chance of getting all the way up there. What's the modern day equivalent of that?
Electron microscopes image physical properties of an object. Last I checked, magnetism isn't a physical property...
http://www.jamesgao.com/09/
Ehhh, might actually prove to be a pretty fun project for some bored google engineer. I'm guessing that the program picks from a dictionary of possible words. I'd guess that it might actually be possible to run some sort of a statistical analysis over your data set and drop the queries that match the statistical profile of randomly selected words from this dictionary. Either that or try some sort of a semantic analysis - those queries don't make much sense semantically, and can be dropped. Nothing is foolproof :)
I'm sure Amtrak will make that one affordable trip! :-D
You want to put WHAT in your URINARY TRACT?!
Umm.... I'd doubt that an OLPC will survive much bashing either. If one had to pour the compaq out of the paper bag, well the best the OLPC can probably do is be picked piece by piece from a box. Either way, the OLPC's software is open source, and sticking that on the low end compaq would be fine too.
I'm lost, what exactly are you trying to imply? Yes, some of these techniques did exist in the 1970's, but Monsanto sure wasn't making GM corn around then, that was late 80's/early 90's. The manipulations that I used as examples are pretty common genetic manipulation techniques used in labs today.
What makes you think that? Monsanto does it frequently for their GM corn to avoid dangerously spreading the crop. Granted it's probably a lot harder for an animal instead of a plant, but genes do exist for autotermination. One exists in humans already, instead in the opposite direction. Phenylketonuria, or PKU patients cannot eat foods with phenylalanine. It's standard practice to couple transfection studies of bacteria with an antibiotic resistance gene, then grow them on media containing antibiotics. In fact, I did that one in high school biology (albeit unsuccessfully). Sterility can be easily induced through a variety of genetic modifications. Perhaps it's your chance to review on modern genetic manipulation techniques?
The consequences are indeed quite SEVERE if this GM mosquito doesn't work like we expect it to. It's stronger than natural mosquitoes, making it harder to control. This alone makes it a very dangerous feature.
If anything, they should do a test using a specially designed mosquito. This breed must be sexually incompatible with natural mosquitos and they should lack some critical non-natural nutrient in order to survive. By releasing them into an area and only providing this nutrient source in that one area, it's possible to safely test to see how an ecosystem would react to a GM mosquito. Of course this will still not eliminate the possibility of a mutation that could disable the "death gene" and cause a dangerous spreading of the GM mosquito...
Yeah, taskbar grouping is definitely one of the first things I turn off. However, barring that (utterly retarded) feature, the taskbar is by far the most intuitive task indicator possible. Again, I'm a linux person anyways, so having multiple desktops are also necessary for completely fluid work. Grouping a set of tasks into a desktop greatly reduces the number of buttons on the taskbar. However, desktop icons and the taskbar still indicates what exactly is running at a glance (without the need to press a button). For example, desktop 1 is my browser desktop, so I can immediately see what windows I have open. IMs are grouped into tabs already - untabbed clients are for uncivilized people ;) .
Except Merck and Pfizer would not EXIST if it weren't for the patents. Have you heard of Andrx or Ivax? They are generic drug companies. They do exactly what you say should happen - they sell drugs that are out of patent, and they have equal access over any (old) compounds. They do little or no research of their own, because they don't have the money to do so.
The actual COST of a pill is almost nothing - no more than the cost of generic acetaminophen in your local grocery store. Any organic chem graduate could take almost any of the major drugs on the market and design a workable and mostly efficient synthesis method. Any chemical engineering graduate could probably take one of these methods and upscale it to million-pill quantities. Ensuring that these drugs are safe is where most of the cost is.
You'd also have a pretty nice big influx of borderline drugs/cocktails like Fen-phen and the Cox-2 inhibitors. The general population is a very hypocritical crowd. They want cheap drugs without paying for its development. They want drugs to come out quickly, but they must be 100% safe. Many argued that the FDA wasn't strict enough, and that they let the Cox-2 inhibitors come out. And you're sitting here suggesting that they should be completely abolished, allowing any and all drugs to be tested. So, which way would you like it? High priced but safe drugs, low priced and unsafe drugs, or no drugs at all (when there's no development)?
Isn't a triply-linked list just a binary tree? So he just patented a binary tree instead of a linked list?
Two issues with what you've said so far - alt-tab doesn't switch individual windows by default, and there's no good place to find out. Thankfully, you've actually shown me a very useful bit of info, and that will greatly lessen the hassle of using mac. Also, that program that you showed me still requires a key be pressed to see exactly what's running. What exactly is wrong with the taskbar? Was it too ugly for your saintly OS of beauty?
Also, don't confused me with a windows user. The lack of multiple desktops becomes especially annoying for me after a while. But at least the presence of a taskbar keeps me sane.
For one thing, changing between the 6-10 windows with a single keyboard shortcut, or knowing at a glance exactly what each console is running at the time. And no, expose is not the answer, because it requires a lot more than a single smooth motion (alt-tab) to get to any specific window.
Developing stuff on OSX is a pain. I can't SEE what's running, I can't switch windows conveniently or without the mouse. And who the hell thought of the idea of switching the alt and ctrl in the console?! I get used to alt-everything for shortcuts, go into terminal, and WHAM nothing works anymore. Get used to ctrl-everything, switch out of console, WHAM nothing works again. It's a constant battle with the operating system to do what I want.