The biggest issue for home users is support. They want someone who will answer dumb questions about their computer when they don't have any idea what the problem is. They don't know what documentation to read, and they might not have the context to understand the right documentation anyway. Currently, they use Windows, and they get support from the place they bought the computer, and these places tend to be really bad at Linux support.
This is rather different from the business desktop, where the clueless users can be limited to the software that IT wants them using.
As for peripherals, they get generally be detected and just work out of the box, but the documentation that comes with them doesn't apply at all, making them somewhat confusing for new users. I have three USB devices: a camera, a Visor cradle, and a Zaurus cradle. These are, respectively, a storage device, a serial device, and a network device. None of them have manuals that tell you this, because the Windows software for each makes them simply devices you can sync.
People don't want to pay to browse a site; HTML and HTTP just aren't very good if there are costs associated with them. People complain about needing subscriptions even when those subscriptions are entirely free, just because the subscription mechanism interrupts their access to information. They can't necessarily post links and cite the work or get at it from a different computer seemlessly.
On the other hand, people are perfectly willing to buy things over the web (consider Amazon or eBay), and they even seem willing to buy things over the web which will be delivered to them over the web (consider Magnatune).
I think that payments will prove to work fine for items which can be tried in advance or returned, and which are not part of the normal interaction with the site. And purchases depend exclusively on the good will of the purchaser; the content will be available elsewhere, and making it less convenient to get from the original source just reduces the chance that people will spend money on the original site.
Ironically, I think that the best model is to give away your content in a neatly indexed fashion on your site, and allow people to pay their choice of price to download exactly the same content in a zip file. Document who gets the money. People operate both as donors and as consumers, but not really both at the same time. On the web, it's hard to engage people as consumers (unless you promise to send them physical objects), so your best bet is to hook them as donors.
The same company does a number of other "reduced" shows; I've seen American History, which was brilliant (and quite informative).
One of the interesting things about the technique is that it involves picking out the important parts of the work and omitting everything that merely serves to make the work comprehensible to people who are not familar with it, and therefore gets out the important structure of a long work. I wouldn't be surprised if bards learned the order of sections of the Illiad by picking an exemplary line from each section and memorizing the story thus summarized. Considering how often the same bits are repeated in the Illiad (the whole rosy-fingered dawn bit, for example), it would be necessary to have an idea of the whole story so you could remember your place in it. So I wouldn't be surprised if the bards would sit around coming up with the minimal number of lines necessary to convey the important parts of the epic.
The outline of the idea, however, is as given in the claims, which need to be sufficiently specific to allow a working implementation to be produced based on the description. This is because the point of the patent office is to make the techniques necessary to make an invention work commonly known (patens), as the price of giving the inventor the right to limit implementations of the techniques. This is to contrast with trade secrets, where the inventor gets no protection from the law, but does not have to reveal the technique.
Chances are that IBM has some innovation here, and isn't claiming the master password idea (which has been used for decades), but some refinement on it. Of course, they're probably not going to reveal what the innovation is just yet, since they aren't supposed to. (One of the issues with the patent which applied to GIF was that the technique had been published by the inventor for more than a year before the patent application was submitted.) There are, of course, plenty of issues with current master password systems, any of which IBM might have found a way to overcome, and this method would then be patentable.
We don't actually even know if IBM will release the system to the public at all; it could be exclusively internal. Or, for that matter, they could be adopting someone else's system, which they'll deploy for their own use. Since they're looking at 2008, that's plenty of time for someone else to develop a suitable free software VoIP application, which they can just run.
It could be IBM who would decide to buy them, in which case their coutersuit would be moot. In fact, it might turn out to be the case that IBM could get more value out of buying SCO than having SCO squander the last of their money fighting the lawsuit, leaving IBM to try to collect from the dead shell.
In any case, the deal for Boies seems to be while any litigation is pending, and it's unlikely that SCO will ever not have litigation pending.
It uses an assignment in a condition, but it was designed such that it makes sense with an equality test there and doesn't make sense with an assignment there. A number of people (including Larry McVoy, for one, and myself as a random interested person) didn't notice that it was actually a conditional, despite knowing that these lines were suspiciously inserted. (I don't know if Linus noticed or would have; someone pointed out what they would actually do before Linus posted anything on the subject).
Linus is good at finding code which doesn't work right due to the algorithm not working, or which only covers some of the cases, or something like that. I'm not sure he's catch something which looks right but is actually something slightly different which can never have its appearent function but actually does something different. It's just not what he normally catches.
The next day, the CVS repository is regenerated from the clean sources. It's not a "real" repository, it's the output from a gateway. This means that the next day, your version of the file would include changes which CVS has no record of (because the revisions have disappeared, which is different from the changes being reverted). So CVS would report those as changes, because the current state of the file is unlike the last state it knows about. Therefore, your "cvs diff -u" will include them as if you had made those changes.
If the case gets dismissed, SCO's stock becomes worthless, because they're now primarily a lawsuit company, and they would be on the wrong end of all of the pending lawsuits. Nobody would be interested in licensing anything from them if it seemed they were unable to demonstrate that they actually had any assets, and spinning themselves as a IP company doesn't work very well if the litigation in progress unambiguously implies that they tried to steal all of it.
They don't have to win the IBM lawsuit any time soon, but it's the end of the road for them if they lose it. Of course, IBM won't actually make a motion to dismiss until they'd get it (they don't want the press that SCO would put out over IBM's motion being denied). But if they never mention dismissal, SCO will just drag their feet forever.
Of course, I suspect that SCO actually does have some IP of value, and that SCO will be bought while the case is still pending. It's not anything like the IP they're claiming to have, and it's probably not relevant to the case, but they probably own something, and someone will probably buy them when their value is low enough. If nothing else, they've probably got something that can compel Microsoft to pay a couple million on occasion for the sake of the appearance of propriety.
The actual lines of code and the method by which they got there were far too clever for either Microsoft or SCO. In particular, it looked like a check for an invalid combination of flags by root, but would actually set the process to root in the case of the invalid combination of flags (and the error return value would be overwritten).
The intent was probably that a CVS user get the bad version, work on other stuff, and send the diff (including the bad lines) to a maintainer in an otherwise good patch. However, the BKCVS gateway got confused by someone other than it changing the CVS, and complained, and Larry McVoy pointed out the issue, someone asked what the lines were, and other people figured out what they'd do. Now, of course, if someone had gotten that bit accidentally and submitted it to a maintainer, they'd notice, so the attempt seems to have failed.
Linus pointed out a benefit to using BK: even if the official BK repository were changed, he doesn't pull from it (because his local copy has all of his changes), and he would get an error the next time he pushed to it. The repository that would have to be attacked is actually his local disk, behind a firewall and not set up for anyone else to access at all.
If RMS wants to rant about revision control systems, he'll need to say that CVS needs to be replaced with a more functional alternative (Subversion, perhaps), not BK.
The third section of this article actually talks about Cambridge MA, which does exactly that. Somerville, next door (where I voted yesterday), does that as well. And Cambridge actually did have a hardware problem and rescanned the ballots.
As for the capabilities of this system, you can actually do ranking quite easily (and Cambridge actually does use ranking in their voting, although Somerville does not); you don't write numbers, but you fill in an oval in the grid for each rank, like on the SAT.
Cambridge City Council elections have an "instant runoff" system, which uses some algorithm that eliminates candidates who have either won or lost, and moves votes to candidates the voter ranked next until the right number of candidates have been chosen. It's what you'd expect for the city that has both MIT and Harvard in it.
It certainly seemed like the first movie in a trilogy to me; it hints at long-range trends, and ends with only a partial and temporary resolution. Unfortunately, this doesn't mean that the other two thirds of the story is good material for movies.
In general, if you have a trilogy, the first one is the best, the last second best, and the second is the worst, because of where the interesting parts of a story normally go. Of course, any of these can fail to live up to their dramatic potentials. But it is generally the case that, even if something is intended to be a trilogy from the beginning, it's likely to go downhill after the first part, and parts two and three are best viewed in terms of the complete work.
Note that this is not true of works designed as a long series formed of episodes, where the first one is unusual in introducing the characters and can have additional novelty, but afterwards the episodes are individual.
It seems to me that the problem he's talking about isn't with Linux at all, but that local retailers don't have a clue about Linux, and tend to mess up the installation. So you get the machine home and find that you're nearly out of space on some partition, the IP is fixed to one that doesn't work, and everything is in Chinese. Of course, if you bring it in to get it fixed, the retailer will stare at it for a while and then change the locale to Korea or change root's name in/etc/shadow.
It's not that retailers are stupid, but they tend to be trained on Windows and not be interested in learning other systems. And they're who the home user knows to turn to for tech support.
Linux is perfectly good for home use, but the support infrastructure for home users is stuck on Windows, and unlike the corporate IT department, doesn't have a cost motive to switch their expertise to Linux, because they're not the ones who have to pay for Windows.
I suppose, in theory, we might be able to destroy the coordinator's communications equipment, which would cause problems for any attack which depends on coordination, and it might be easier to use this kind of weapon over a large area or where you might accidentally hit civilians (since you can replace electronic devices belonging to the people you hit accidentally, but you can't replace the people).
A quick calculation gives 2000 dots on the 19" diagonal, which is pretty close to 100 DPI. Luckily, 100 DPI is a standard X font scale, so just switching from your -75-75- fonts to -100-100- fonts should make all your otherwise familiar bitmap fonts look right (or, at least, look only a little wierd).
Most vendor-released drivers are pretty bad, even when they bother to provide them. It's almost always better to have kernel hackers write drivers than get the hardware vendor to do it, anyway. Using a Win32 driver is a reasonable stopgap while native drivers are under development, if you've got some hardware that's not supported. In fact, if you're going to use vendor-supplied code, you might as well use a Win32 driver, since it's a fixed API that the vendor is likely familiar with, so they're less likely to make stupid mistakes with it.
If you actually read that link, it doesn't actually say anywhere that the portion of people who are dehydrated is not large. It says that only 6 glasses of water are necessary (or maybe less), and that coffee counts for 2/3 of the volume, and mentions a number of other myths. But it doesn't actually say that people actually have enough water in general, and it stands to reason that people would lose more water in an office with central heat and air conditioning (both of which dry the air) than they would normally.
Of course, the amount of water that people get is clearly enough to maintain their hydration level; otherwise they'd quickly die of thirst. However, this doesn't mean that people aren't generally maintaining a level below what is best for them, or that they aren't getting dehydrated during the day and recovering at night.
On the other hand, I think the study which I was remembering which found a large portion of people to be dehydrated was this summer in France, where thousands of people died due to a heat wave.
Tomatos are clearly vegetables in a culinary sense; there it matters more how sweet something is than what part of the plant it came from. It is the duty of the Supreme Court to interpret laws, and what they were deciding in this case is that it is the culinary sense, not the botanical sense, that the legislature intended in the tariff law.
Furthermore, the use in cooking matters more for imported goods which are obviously no longer on a plant at all.
So, in effect, they were saying that legislators, like most people, care more about food than plants, and therefore, unless the legislature actually specifies "botanical fruit", they probably intend that you ask a grocer rather than a botanist.
Constant ketosis is a treatment for epilipsy, and from the records of patients under that treatment, there is evidence of kidney damage. This can be effectively treated with increased hydration (maybe that initial loss of water weight isn't such a good thing). Of course, the epilepsy treatment is much more extreme than Atkins, but it points out the potential risks and mitigating factors.
Considering that a large portion of people are constantly dehydrated, it's probably worth drinking more on Atkins.
I know several people who have had similar success on Atkins, a number of whom have reached the weight they wanted to be, and have gone back to eating carbohydrates to avoid getting too thin. One actually started because she was constantly hungry when she was eating carbohydrates.
Personally, I make sure to drink a big glass of cherry juice every day, because I'm not that fond of grains or sweets, and I'm frighteningly thin.
Carbohydrates aren't actually good for anything except for calories. Complex carbohydrates are good for short-term storage (like overnight), sugars are good for immediate use. If you manage to get enough calories to do whatever you're planning to do, it doesn't matter if you use carbohydrates in your diet to do it. Of course, there's no way you're going to run a marathon without carboloading the night before, since you need the short-term storage, but most people don't run marathons all that often.
The main thing that's different between carbohydrates and fats is that getting energy out of fats uses a process which suppresses hunger, and getting energy out of carbohydrates doesn't. This is why it's easy to overeat on carbohydrates: your body doesn't notice the caloric intake until you've physically stuffed your digestive tract.
Of course, protein is important, because the human body can't synthesize most of them. And there are a number of other things you need to get from your diet.
So, as far as diets go, cutting out protein is really bad, and cutting out fat makes you hungry. But if you cut out carbohydrates, you'll be perfectly fine.
Caveats: if your body is accustomed to a low-fat diet, it may not have the enzymes in place to get energy out of fat; if you just drop carbohydrates, it can be bad until your body figures out the whole fat thing. If you just cut out everything from your present diet that has carbohydrates, you may be missing some important vitamins which would normally come from the foods you're no longer eating. Some people have a hard time digesting meat (more people than you'd expect, in fact), and it's difficult to have a plausible vegetarian, low-carb, nutritious diet.
The overclocking analogy doesn't make much sense, of course. It's actually more like running your computer off of battery instead of wall current. It's a different power source with somewhat different characteristics, but it's no worse and possibly better in the long run. And it will get the "battery depleted" light to go off.
The thing is, this wasn't free music. MIT was actually trying to pay for music, and the RIAA stopped them. People won't be pissed off about it, but it won't encourage them to buy CDs, either.
I suspect it won't be too long before Congress realized that the RIAA is a bunch of foriegn companies using dodgy interpretations of copyright law to screw over American consumers and artists, and outlaws their business model. In the meantime, I'll be buying my music from people who are willing to sell it to me.
Suing SCO is probably still not worthwhile; it would take too long, and SCO probably won't have any assets let to forfeit by them. Getting a summary judgement will only force them to stop doing this, and it doesn't matter much if they distribute improperly licensed code with is available properly licensed as well.
On the other hand, SCO has made themselves the world's biggest warez site, and has refused to take down the infringing content. It's time their network connection got shut off, and kept off until they can demonstrate that they have the right to distribute the content. In addition to the insignificant effect of preventing them from distributing Linux with their improper license, it would probably kill their stock price if they couldn't get an internet connection due to their blatent disregard for copyright law.
Re:Much of this could be done in linux...
on
Microsoft's new CLI
·
· Score: 2, Informative
I often have to solve complicated problems. But not once have I wanted to solve a complex problem and been sufficiently sure of my shell scripting skills to want to type the command interactively. I recently wrote a 60 line script (with lots of comments) to choose a template file based on pattern matching the desired filename against a directory listing of template files, and then expand the patterns in the file based on the names.
I wrote it in bash, because I'm masochistic that way, but I would never have considered actually typing it directly at a shell prompt. I think I've used a loop in an interactive shell on only one occasion. The fundamental problem is that as soon as you run a command in an interactive shell, it is thrown away. If you're solving a complex problem, it's just a lot easier to do it through an editor with syntax highlighting, useful messages, and the ability to change the first line of a ten-line script.
As for the reason that the shell should be apart from a strong scripting language like Python, have you actually ever tried to use Python as your shell? It is perfectly capable, since it has all of the needed functionality. But it doesn't handle the simple case efficiently, and 99% of the tasks anyone does are simple.
I have to solve complex problems sometimes, and for those I use a suitable tool. But I am constantly doing things to solve simple problems, and any extra keystrokes for the simple case are too many.
Slashdot Mirror
The biggest issue for home users is support. They want someone who will answer dumb questions about their computer when they don't have any idea what the problem is. They don't know what documentation to read, and they might not have the context to understand the right documentation anyway. Currently, they use Windows, and they get support from the place they bought the computer, and these places tend to be really bad at Linux support.
This is rather different from the business desktop, where the clueless users can be limited to the software that IT wants them using.
As for peripherals, they get generally be detected and just work out of the box, but the documentation that comes with them doesn't apply at all, making them somewhat confusing for new users. I have three USB devices: a camera, a Visor cradle, and a Zaurus cradle. These are, respectively, a storage device, a serial device, and a network device. None of them have manuals that tell you this, because the Windows software for each makes them simply devices you can sync.
People don't want to pay to browse a site; HTML and HTTP just aren't very good if there are costs associated with them. People complain about needing subscriptions even when those subscriptions are entirely free, just because the subscription mechanism interrupts their access to information. They can't necessarily post links and cite the work or get at it from a different computer seemlessly.
On the other hand, people are perfectly willing to buy things over the web (consider Amazon or eBay), and they even seem willing to buy things over the web which will be delivered to them over the web (consider Magnatune).
I think that payments will prove to work fine for items which can be tried in advance or returned, and which are not part of the normal interaction with the site. And purchases depend exclusively on the good will of the purchaser; the content will be available elsewhere, and making it less convenient to get from the original source just reduces the chance that people will spend money on the original site.
Ironically, I think that the best model is to give away your content in a neatly indexed fashion on your site, and allow people to pay their choice of price to download exactly the same content in a zip file. Document who gets the money. People operate both as donors and as consumers, but not really both at the same time. On the web, it's hard to engage people as consumers (unless you promise to send them physical objects), so your best bet is to hook them as donors.
The same company does a number of other "reduced" shows; I've seen American History, which was brilliant (and quite informative).
One of the interesting things about the technique is that it involves picking out the important parts of the work and omitting everything that merely serves to make the work comprehensible to people who are not familar with it, and therefore gets out the important structure of a long work. I wouldn't be surprised if bards learned the order of sections of the Illiad by picking an exemplary line from each section and memorizing the story thus summarized. Considering how often the same bits are repeated in the Illiad (the whole rosy-fingered dawn bit, for example), it would be necessary to have an idea of the whole story so you could remember your place in it. So I wouldn't be surprised if the bards would sit around coming up with the minimal number of lines necessary to convey the important parts of the epic.
The outline of the idea, however, is as given in the claims, which need to be sufficiently specific to allow a working implementation to be produced based on the description. This is because the point of the patent office is to make the techniques necessary to make an invention work commonly known (patens), as the price of giving the inventor the right to limit implementations of the techniques. This is to contrast with trade secrets, where the inventor gets no protection from the law, but does not have to reveal the technique.
Chances are that IBM has some innovation here, and isn't claiming the master password idea (which has been used for decades), but some refinement on it. Of course, they're probably not going to reveal what the innovation is just yet, since they aren't supposed to. (One of the issues with the patent which applied to GIF was that the technique had been published by the inventor for more than a year before the patent application was submitted.) There are, of course, plenty of issues with current master password systems, any of which IBM might have found a way to overcome, and this method would then be patentable.
We don't actually even know if IBM will release the system to the public at all; it could be exclusively internal. Or, for that matter, they could be adopting someone else's system, which they'll deploy for their own use. Since they're looking at 2008, that's plenty of time for someone else to develop a suitable free software VoIP application, which they can just run.
It could be IBM who would decide to buy them, in which case their coutersuit would be moot. In fact, it might turn out to be the case that IBM could get more value out of buying SCO than having SCO squander the last of their money fighting the lawsuit, leaving IBM to try to collect from the dead shell.
In any case, the deal for Boies seems to be while any litigation is pending, and it's unlikely that SCO will ever not have litigation pending.
It uses an assignment in a condition, but it was designed such that it makes sense with an equality test there and doesn't make sense with an assignment there. A number of people (including Larry McVoy, for one, and myself as a random interested person) didn't notice that it was actually a conditional, despite knowing that these lines were suspiciously inserted. (I don't know if Linus noticed or would have; someone pointed out what they would actually do before Linus posted anything on the subject).
Linus is good at finding code which doesn't work right due to the algorithm not working, or which only covers some of the cases, or something like that. I'm not sure he's catch something which looks right but is actually something slightly different which can never have its appearent function but actually does something different. It's just not what he normally catches.
The next day, the CVS repository is regenerated from the clean sources. It's not a "real" repository, it's the output from a gateway. This means that the next day, your version of the file would include changes which CVS has no record of (because the revisions have disappeared, which is different from the changes being reverted). So CVS would report those as changes, because the current state of the file is unlike the last state it knows about. Therefore, your "cvs diff -u" will include them as if you had made those changes.
If the case gets dismissed, SCO's stock becomes worthless, because they're now primarily a lawsuit company, and they would be on the wrong end of all of the pending lawsuits. Nobody would be interested in licensing anything from them if it seemed they were unable to demonstrate that they actually had any assets, and spinning themselves as a IP company doesn't work very well if the litigation in progress unambiguously implies that they tried to steal all of it.
They don't have to win the IBM lawsuit any time soon, but it's the end of the road for them if they lose it. Of course, IBM won't actually make a motion to dismiss until they'd get it (they don't want the press that SCO would put out over IBM's motion being denied). But if they never mention dismissal, SCO will just drag their feet forever.
Of course, I suspect that SCO actually does have some IP of value, and that SCO will be bought while the case is still pending. It's not anything like the IP they're claiming to have,
and it's probably not relevant to the case, but they probably own something, and someone will probably buy them when their value is low enough. If nothing else, they've probably got something that can compel Microsoft to pay a couple million on occasion for the sake of the appearance of propriety.
The actual lines of code and the method by which they got there were far too clever for either Microsoft or SCO. In particular, it looked like a check for an invalid combination of flags by root, but would actually set the process to root in the case of the invalid combination of flags (and the error return value would be overwritten).
The intent was probably that a CVS user get the bad version, work on other stuff, and send the diff (including the bad lines) to a maintainer in an otherwise good patch. However, the BKCVS gateway got confused by someone other than it changing the CVS, and complained, and Larry McVoy pointed out the issue, someone asked what the lines were, and other people figured out what they'd do. Now, of course, if someone had gotten that bit accidentally and submitted it to a maintainer, they'd notice, so the attempt seems to have failed.
Linus pointed out a benefit to using BK: even if the official BK repository were changed, he doesn't pull from it (because his local copy has all of his changes), and he would get an error the next time he pushed to it. The repository that would have to be attacked is actually his local disk, behind a firewall and not set up for anyone else to access at all.
If RMS wants to rant about revision control systems, he'll need to say that CVS needs to be replaced with a more functional alternative (Subversion, perhaps), not BK.
The third section of this article actually talks about Cambridge MA, which does exactly that. Somerville, next door (where I voted yesterday), does that as well. And Cambridge actually did have a hardware problem and rescanned the ballots.
As for the capabilities of this system, you can actually do ranking quite easily (and Cambridge actually does use ranking in their voting, although Somerville does not); you don't write numbers, but you fill in an oval in the grid for each rank, like on the SAT.
Cambridge City Council elections have an "instant runoff" system, which uses some algorithm that eliminates candidates who have either won or lost, and moves votes to candidates the voter ranked next until the right number of candidates have been chosen. It's what you'd expect for the city that has both MIT and Harvard in it.
It certainly seemed like the first movie in a trilogy to me; it hints at long-range trends, and ends with only a partial and temporary resolution. Unfortunately, this doesn't mean that the other two thirds of the story is good material for movies.
In general, if you have a trilogy, the first one is the best, the last second best, and the second is the worst, because of where the interesting parts of a story normally go. Of course, any of these can fail to live up to their dramatic potentials. But it is generally the case that, even if something is intended to be a trilogy from the beginning, it's likely to go downhill after the first part, and parts two and three are best viewed in terms of the complete work.
Note that this is not true of works designed as a long series formed of episodes, where the first one is unusual in introducing the characters and can have additional novelty, but afterwards the episodes are individual.
It seems to me that the problem he's talking about isn't with Linux at all, but that local retailers don't have a clue about Linux, and tend to mess up the installation. So you get the machine home and find that you're nearly out of space on some partition, the IP is fixed to one that doesn't work, and everything is in Chinese. Of course, if you bring it in to get it fixed, the retailer will stare at it for a while and then change the locale to Korea or change root's name in /etc/shadow.
It's not that retailers are stupid, but they tend to be trained on Windows and not be interested in learning other systems. And they're who the home user knows to turn to for tech support.
Linux is perfectly good for home use, but the support infrastructure for home users is stuck on Windows, and unlike the corporate IT department, doesn't have a cost motive to switch their expertise to Linux, because they're not the ones who have to pay for Windows.
I suppose, in theory, we might be able to destroy the coordinator's communications equipment, which would cause problems for any attack which depends on coordination, and it might be easier to use this kind of weapon over a large area or where you might accidentally hit civilians (since you can replace electronic devices belonging to the people you hit accidentally, but you can't replace the people).
I believe we're looking for a third part of the sun to be snuffed out. Big as these flares are, they're trivial in terms of the size of the sun.
A quick calculation gives 2000 dots on the 19" diagonal, which is pretty close to 100 DPI. Luckily, 100 DPI is a standard X font scale, so just switching from your -75-75- fonts to -100-100- fonts should make all your otherwise familiar bitmap fonts look right (or, at least, look only a little wierd).
Most vendor-released drivers are pretty bad, even when they bother to provide them. It's almost always better to have kernel hackers write drivers than get the hardware vendor to do it, anyway. Using a Win32 driver is a reasonable stopgap while native drivers are under development, if you've got some hardware that's not supported. In fact, if you're going to use vendor-supplied code, you might as well use a Win32 driver, since it's a fixed API that the vendor is likely familiar with, so they're less likely to make stupid mistakes with it.
If you actually read that link, it doesn't actually say anywhere that the portion of people who are dehydrated is not large. It says that only 6 glasses of water are necessary (or maybe less), and that coffee counts for 2/3 of the volume, and mentions a number of other myths. But it doesn't actually say that people actually have enough water in general, and it stands to reason that people would lose more water in an office with central heat and air conditioning (both of which dry the air) than they would normally.
Of course, the amount of water that people get is clearly enough to maintain their hydration level; otherwise they'd quickly die of thirst. However, this doesn't mean that people aren't generally maintaining a level below what is best for them, or that they aren't getting dehydrated during the day and recovering at night.
On the other hand, I think the study which I was remembering which found a large portion of people to be dehydrated was this summer in France, where thousands of people died due to a heat wave.
Tomatos are clearly vegetables in a culinary sense; there it matters more how sweet something is than what part of the plant it came from. It is the duty of the Supreme Court to interpret laws, and what they were deciding in this case is that it is the culinary sense, not the botanical sense, that the legislature intended in the tariff law.
Furthermore, the use in cooking matters more for imported goods which are obviously no longer on a plant at all.
So, in effect, they were saying that legislators, like most people, care more about food than plants, and therefore, unless the legislature actually specifies "botanical fruit", they probably intend that you ask a grocer rather than a botanist.
Constant ketosis is a treatment for epilipsy, and from the records of patients under that treatment, there is evidence of kidney damage. This can be effectively treated with increased hydration (maybe that initial loss of water weight isn't such a good thing). Of course, the epilepsy treatment is much more extreme than Atkins, but it points out the potential risks and mitigating factors.
Considering that a large portion of people are constantly dehydrated, it's probably worth drinking more on Atkins.
I know several people who have had similar success on Atkins, a number of whom have reached the weight they wanted to be, and have gone back to eating carbohydrates to avoid getting too thin. One actually started because she was constantly hungry when she was eating carbohydrates.
Personally, I make sure to drink a big glass of cherry juice every day, because I'm not that fond of grains or sweets, and I'm frighteningly thin.
Carbohydrates aren't actually good for anything except for calories. Complex carbohydrates are good for short-term storage (like overnight), sugars are good for immediate use. If you manage to get enough calories to do whatever you're planning to do, it doesn't matter if you use carbohydrates in your diet to do it. Of course, there's no way you're going to run a marathon without carboloading the night before, since you need the short-term storage, but most people don't run marathons all that often.
The main thing that's different between carbohydrates and fats is that getting energy out of fats uses a process which suppresses hunger, and getting energy out of carbohydrates doesn't. This is why it's easy to overeat on carbohydrates: your body doesn't notice the caloric intake until you've physically stuffed your digestive tract.
Of course, protein is important, because the human body can't synthesize most of them. And there are a number of other things you need to get from your diet.
So, as far as diets go, cutting out protein is really bad, and cutting out fat makes you hungry. But if you cut out carbohydrates, you'll be perfectly fine.
Caveats: if your body is accustomed to a low-fat diet, it may not have the enzymes in place to get energy out of fat; if you just drop carbohydrates, it can be bad until your body figures out the whole fat thing. If you just cut out everything from your present diet that has carbohydrates, you may be missing some important vitamins which would normally come from the foods you're no longer eating. Some people have a hard time digesting meat (more people than you'd expect, in fact), and it's difficult to have a plausible vegetarian, low-carb, nutritious diet.
The overclocking analogy doesn't make much sense, of course. It's actually more like running your computer off of battery instead of wall current. It's a different power source with somewhat different characteristics, but it's no worse and possibly better in the long run. And it will get the "battery depleted" light to go off.
The thing is, this wasn't free music. MIT was actually trying to pay for music, and the RIAA stopped them. People won't be pissed off about it, but it won't encourage them to buy CDs, either.
I suspect it won't be too long before Congress realized that the RIAA is a bunch of foriegn companies using dodgy interpretations of copyright law to screw over American consumers and artists, and outlaws their business model. In the meantime, I'll be buying my music from people who are willing to sell it to me.
Suing SCO is probably still not worthwhile; it would take too long, and SCO probably won't have any assets let to forfeit by them. Getting a summary judgement will only force them to stop doing this, and it doesn't matter much if they distribute improperly licensed code with is available properly licensed as well.
On the other hand, SCO has made themselves the world's biggest warez site, and has refused to take down the infringing content. It's time their network connection got shut off, and kept off until they can demonstrate that they have the right to distribute the content. In addition to the insignificant effect of preventing them from distributing Linux with their improper license, it would probably kill their stock price if they couldn't get an internet connection due to their blatent disregard for copyright law.
I often have to solve complicated problems. But not once have I wanted to solve a complex problem and been sufficiently sure of my shell scripting skills to want to type the command interactively. I recently wrote a 60 line script (with lots of comments) to choose a template file based on pattern matching the desired filename against a directory listing of template files, and then expand the patterns in the file based on the names.
I wrote it in bash, because I'm masochistic that way, but I would never have considered actually typing it directly at a shell prompt. I think I've used a loop in an interactive shell on only one occasion. The fundamental problem is that as soon as you run a command in an interactive shell, it is thrown away. If you're solving a complex problem, it's just a lot easier to do it through an editor with syntax highlighting, useful messages, and the ability to change the first line of a ten-line script.
As for the reason that the shell should be apart from a strong scripting language like Python, have you actually ever tried to use Python as your shell? It is perfectly capable, since it has all of the needed functionality. But it doesn't handle the simple case efficiently, and 99% of the tasks anyone does are simple.
I have to solve complex problems sometimes, and for those I use a suitable tool. But I am constantly doing things to solve simple problems, and any extra keystrokes for the simple case are too many.