Well, actually, they'd just have to vet the source and compile it themselves, then distribute the binaries they compiled. Then they'd just have to worry about the compilers they used.
No. Given that both schedulers probably use the same interface as the previous ones, it'd be a matter of linking -- an if statement in a makefile somewhere, and a line in the config.
From what I've heard, proposing a new idea to Bill Gates consists of:
- him ignoring you during the first few minutes of your presentation
- him interrupting you randomly and telling you your idea is stupid and not to waste his time
Which has the effect of people who dislike confrontation falling behind. But, at the same time, people who can stand up to him but who have flawed ideas, their ideas don't make it.
Linus is portrayed as being more approachable. This means he has less time to deal with new ideas. However, Con Kolivas has been around for years and been reliable. I don't know why Linus would suddenly lash out at him.
We need a standards-based, open FS for all to use FREELY - get enough momentum outside of Windows-land and Windows will have to eventually support it, too. I, and Microsoft's 90% market share, beg to differ.
The only reliable, interoperable way of exporting files from a drive to arbitrary operating systems is to carry around an embedded FTP/DHCP server. Yank the network cable out, replace it, and you're good to go. And then who cares what filesystem it's running?
Windows 2000 uses the same NTFS filesystem as Windows XP, but XP experiences far less fragmentation. It's the fault of poorly written drivers, not the filesystem itself.
That said, I've only dealt with NTFS on Windows 2000 and XP, so I'm not sure if ntfs-3g is any better.
These people are selling products and services to prosecutors. Defense attorneys only need to be aware of flaws in forensics software and practices that can result in false positives.
Pleading the fifth in front of a jury when you're the defendant is tantamount to an admission of guilt. But there was an encryption/steganography system called Rubberhose ( http://iq.org/~proff/rubberhose.org/ ) that allowed you to create an arbitrary number of encrypted volumes in one disk segment, where each volume took up a random sequence of blocks. You could have four or five encrypted volumes, one of which contained the incriminating material and the rest of which contained plausibly embarrassing and private material. Then you can comply; nobody can prove that you haven't decrypted everything, since the entire disk segment is filled with random-seeming data.
TrueCrypt does almost as well as Rubberhose, and it's maintained. It allows you to create nested encrypted volumes, but defaults to two volumes deep, and I'm not sure whether it supports any more than that.
When it appeared, the BSDs were in court with the Santa Cruz Organization. SCO used a lot of BSD-copyrighted work; the BSD kernel used about six files full of assembly from the old, old Bell Labs UNIX kernel; when SCO sued BSD386, it was revealed that they were suing in regards to intellectual property that mostly belonged to other people.
But in that period, about 1990-1994, the future of BSD UNIX was rather shaky, so adoption slowed. In stepped Linux, built completely from scratch, guaranteed not to have copyright issues. And being released under the GNU license rather than the BSD license, academics and FSFers alike (though there wasn't a Free Software Foundation yet) were able to submit code to it without feeling guilty.
Your time per day is limited, with no hope of increasing. Your desires are not limited and consume both time and resources. Gaining resources consumes time.
So, you prioritize your desires and use some sort of scheduling algorithm to trade between time gaining resources and time satisfying desires. But you can't simply assert that saving resources is more important than saving time.
The CPU schedulers are all designed for enterprise systems with 16 cores or more. Say that again.
That's a hypothetical situation, of course; Linux has several schedulers, and I'm sure some of them are reasonable for desktop use. But there are probably a few non-optional subsystems that, needlessly and by design, work well on high-end systems but do poorly on low-end systems.
That said, people don't care too much about performance on their computers. If it were the case that, say, Ubuntu's standard install was snappy on a 600MHz processor with 128MB RAM, then more people would likely switch to save their old computers. But there's currently such an upgrade mentality in computing that that wouldn't affect many people.
Don't blame the books; blame the education system and the kids. I think, currently, a pretty good percentage of kids aged 11 would be able to understand the concepts and plot in 1984, though I think most of them wouldn't like it and plenty of them would choke on vocabulary and writing style.
In the movies, people sometimes use magic without wands, but it's rare.
If the movies are not canon, we could assume that, like Griphook implies, both species can use magic with and without wands, but wands are an easy way to gain power and control. I believe something similar was mentioned in a previous book about casting with and without words.
Probably, with easy access to wands, nobody has bothered studying magic without wands, aside from Potions.
If 2% of your DNA is actually useful, that would be a good defense against cancer and mutations. If there's no way to control the rate of mutation, that would be a way to control the rate of mutations that actually affect the organism.
Well, yes, there are a lot more diseases, but it's probably cheaper to maintain a focus on AIDS than to switch that research group to a new disease. Not to mention $15k per year per patient is better than what the vaccine will likely go for. (Though on the other hand, in ten years it'll probably be standard for everyone to get the HIV vaccine, so that's a lot more sales.)
Having multiple drug companies helps. One might be afraid that one of the others will cure a disease, slashing their treatment profits. That gives incentive to find a cure themselves. Then, when one company finds that another is researching a cure, that provides an incentive to send their cure to the FDA and clinical trials.
It seems to me that the real "injustice" is a life expectancy of less than thirty years, a belief that raping virgins cures AIDS, religious civil wars, genocide...
Africa has oil, diamonds, and copper as some of its major resources.
Diamonds? De Beers owns most African diamond mines.
Copper? Zimbabwe sold off its copper mines recently because they were losing money. The price of copper has quadrupled in the last ten years. (And people are emigrating to Zimbabwe...)
Oil? Foreign companies own most African oil wells and pipelines. In some countries, these oil companies hire the military to defend the pipelines and wells. In some of these, the military has stolen the pipelines. In others, people steal high-pressure bolts from the pipelines, which is worse. The pipeline explodes; the thief dies; the oil leaks out; everyone loses. Things are better in the Ukraine, where an entire oil pipeline was stolen but functioning, and made certain individuals large amounts of money. At least there, someone profited.
What would happen if all foreign aid was removed from Africa for two years? The death toll would be in the hundreds of millions.
Kenya's economy is going up, slightly, though I have no idea how long that will last. Morocco and Egypt are stable, and South Africa's major concern at the moment is immigration.
So, why is it that only Mediterranean Africa and South Africa have any internet connectivity to speak of?
There are limits to the degree to which you can abstract an issue while retaining functionality. In point of fact, almost every abstraction is a loss of power. C doesn't allow you to write programs that are as optimal as assembly language (unless you actually use inline assembly language); most other programming languages don't allow you to do pointer arithmetic. What would you have to give up in order to get a programming language that is as easy to use as a calculator?
And why are calculators easy to use? Because they only have a limited number of functions. A typical calculator can add, subtract, multiply, or divide the last number calculated with the number inputted, turn on and off, and clear its input. A scientific calculator adds another fifteen or twenty functions.
A programming language does input and output through the console, a GUI, a network, or a disk; stores an arbitrary number of arbitrary values; and does arbitrary operations on them. (Yes, there's a very limited number of operations that the language gives; but you'll have a standard library to learn, and that is much larger. Besides which, the user defines their own operations in any non-trivial work.)
There's a lot to be done to make programming easier, but the only thing that could bring programming to the masses is people getting used to increased complexity, especially at an early age.
As for your example with Arabic versus Roman numerals, it's having ranked digits where each rank could be computed separately, independent of which symbols surrounded it. If you see 1, it always means 1, never -1. (Though I've seen texts that use IIII rather than IV.) And because the same symbols are used for the same value regardless of rank. If the Romans had had a symbol for zero, well, they could have simply written 'nullus' and be done with it. But the system of digits and ranks was the advantage.
So...you don't want to use Linux because it's harder to program for it than to use a calculator, and thus misses one of your requirements for being free software.
WTF?
I see the barrier to entry argument, but while that can be lowered slightly, programming is inherently difficult. Not everyone could do it, and some people can program much better than others.
Be happy when there are no *artificial* barriers to entry. There's jack you can do about natural barriers to entry.
The other common change is init systems. I sometimes use Debian, but I can't get my head around its weird/etc/rc.d system with symlinks and strange names. I realize that that's a more traditional setup than anything else, but still, Debian's the only Linux I've used that kept that system.
It could be valid, still, even though both the number of Linux users and the number of Linux distros is rising. It's quite possible that, if there were only two or three Linux distros, or even just one, Linux would have double its current market share.
My opinion is that, in that case, those few distros would not be as good as the major distros today, so Linux would have less market share. But I don't have any experimental data to back that.
Slashdot Mirror
'will their be'? Are you talking about BeOS? And whose BeOS? Last I heard, Palm had sold it, but I don't know to whom.
Well, actually, they'd just have to vet the source and compile it themselves, then distribute the binaries they compiled. Then they'd just have to worry about the compilers they used.
No. Given that both schedulers probably use the same interface as the previous ones, it'd be a matter of linking -- an if statement in a makefile somewhere, and a line in the config.
From what I've heard, proposing a new idea to Bill Gates consists of:
- him ignoring you during the first few minutes of your presentation
- him interrupting you randomly and telling you your idea is stupid and not to waste his time
Which has the effect of people who dislike confrontation falling behind. But, at the same time, people who can stand up to him but who have flawed ideas, their ideas don't make it.
Linus is portrayed as being more approachable. This means he has less time to deal with new ideas. However, Con Kolivas has been around for years and been reliable. I don't know why Linus would suddenly lash out at him.
The only reliable, interoperable way of exporting files from a drive to arbitrary operating systems is to carry around an embedded FTP/DHCP server. Yank the network cable out, replace it, and you're good to go. And then who cares what filesystem it's running?
Windows 2000 uses the same NTFS filesystem as Windows XP, but XP experiences far less fragmentation. It's the fault of poorly written drivers, not the filesystem itself.
That said, I've only dealt with NTFS on Windows 2000 and XP, so I'm not sure if ntfs-3g is any better.
These people are selling products and services to prosecutors. Defense attorneys only need to be aware of flaws in forensics software and practices that can result in false positives.
Pleading the fifth in front of a jury when you're the defendant is tantamount to an admission of guilt. But there was an encryption/steganography system called Rubberhose ( http://iq.org/~proff/rubberhose.org/ ) that allowed you to create an arbitrary number of encrypted volumes in one disk segment, where each volume took up a random sequence of blocks. You could have four or five encrypted volumes, one of which contained the incriminating material and the rest of which contained plausibly embarrassing and private material. Then you can comply; nobody can prove that you haven't decrypted everything, since the entire disk segment is filled with random-seeming data.
TrueCrypt does almost as well as Rubberhose, and it's maintained. It allows you to create nested encrypted volumes, but defaults to two volumes deep, and I'm not sure whether it supports any more than that.
When it appeared, the BSDs were in court with the Santa Cruz Organization. SCO used a lot of BSD-copyrighted work; the BSD kernel used about six files full of assembly from the old, old Bell Labs UNIX kernel; when SCO sued BSD386, it was revealed that they were suing in regards to intellectual property that mostly belonged to other people.
But in that period, about 1990-1994, the future of BSD UNIX was rather shaky, so adoption slowed. In stepped Linux, built completely from scratch, guaranteed not to have copyright issues. And being released under the GNU license rather than the BSD license, academics and FSFers alike (though there wasn't a Free Software Foundation yet) were able to submit code to it without feeling guilty.
Your time per day is limited, with no hope of increasing. Your desires are not limited and consume both time and resources. Gaining resources consumes time.
So, you prioritize your desires and use some sort of scheduling algorithm to trade between time gaining resources and time satisfying desires. But you can't simply assert that saving resources is more important than saving time.
Sorry; I meant as adults. That was the original question.
You can favor any one of [smaller, faster, cheaper] in a given chip. The current trend is to go faster for most PCs and laptops.
I'd prefer going cheaper. The resulting processors would probably stay cooler, too.
The CPU schedulers are all designed for enterprise systems with 16 cores or more. Say that again.
That's a hypothetical situation, of course; Linux has several schedulers, and I'm sure some of them are reasonable for desktop use. But there are probably a few non-optional subsystems that, needlessly and by design, work well on high-end systems but do poorly on low-end systems.
That said, people don't care too much about performance on their computers. If it were the case that, say, Ubuntu's standard install was snappy on a 600MHz processor with 128MB RAM, then more people would likely switch to save their old computers. But there's currently such an upgrade mentality in computing that that wouldn't affect many people.
Don't blame the books; blame the education system and the kids. I think, currently, a pretty good percentage of kids aged 11 would be able to understand the concepts and plot in 1984, though I think most of them wouldn't like it and plenty of them would choke on vocabulary and writing style.
ed? You wuss. I use TECO.
In the movies, people sometimes use magic without wands, but it's rare.
If the movies are not canon, we could assume that, like Griphook implies, both species can use magic with and without wands, but wands are an easy way to gain power and control. I believe something similar was mentioned in a previous book about casting with and without words.
Probably, with easy access to wands, nobody has bothered studying magic without wands, aside from Potions.
If 2% of your DNA is actually useful, that would be a good defense against cancer and mutations. If there's no way to control the rate of mutation, that would be a way to control the rate of mutations that actually affect the organism.
Well, yes, there are a lot more diseases, but it's probably cheaper to maintain a focus on AIDS than to switch that research group to a new disease. Not to mention $15k per year per patient is better than what the vaccine will likely go for. (Though on the other hand, in ten years it'll probably be standard for everyone to get the HIV vaccine, so that's a lot more sales.)
Having multiple drug companies helps. One might be afraid that one of the others will cure a disease, slashing their treatment profits. That gives incentive to find a cure themselves. Then, when one company finds that another is researching a cure, that provides an incentive to send their cure to the FDA and clinical trials.
"Fear will keep the local systems in line..."
It seems to me that the real "injustice" is a life expectancy of less than thirty years, a belief that raping virgins cures AIDS, religious civil wars, genocide...
Africa has oil, diamonds, and copper as some of its major resources.
Diamonds? De Beers owns most African diamond mines.
Copper? Zimbabwe sold off its copper mines recently because they were losing money. The price of copper has quadrupled in the last ten years. (And people are emigrating to Zimbabwe...)
Oil? Foreign companies own most African oil wells and pipelines. In some countries, these oil companies hire the military to defend the pipelines and wells. In some of these, the military has stolen the pipelines. In others, people steal high-pressure bolts from the pipelines, which is worse. The pipeline explodes; the thief dies; the oil leaks out; everyone loses. Things are better in the Ukraine, where an entire oil pipeline was stolen but functioning, and made certain individuals large amounts of money. At least there, someone profited.
What would happen if all foreign aid was removed from Africa for two years? The death toll would be in the hundreds of millions.
Kenya's economy is going up, slightly, though I have no idea how long that will last. Morocco and Egypt are stable, and South Africa's major concern at the moment is immigration.
So, why is it that only Mediterranean Africa and South Africa have any internet connectivity to speak of?
There are limits to the degree to which you can abstract an issue while retaining functionality. In point of fact, almost every abstraction is a loss of power. C doesn't allow you to write programs that are as optimal as assembly language (unless you actually use inline assembly language); most other programming languages don't allow you to do pointer arithmetic. What would you have to give up in order to get a programming language that is as easy to use as a calculator?
And why are calculators easy to use? Because they only have a limited number of functions. A typical calculator can add, subtract, multiply, or divide the last number calculated with the number inputted, turn on and off, and clear its input. A scientific calculator adds another fifteen or twenty functions.
A programming language does input and output through the console, a GUI, a network, or a disk; stores an arbitrary number of arbitrary values; and does arbitrary operations on them. (Yes, there's a very limited number of operations that the language gives; but you'll have a standard library to learn, and that is much larger. Besides which, the user defines their own operations in any non-trivial work.)
There's a lot to be done to make programming easier, but the only thing that could bring programming to the masses is people getting used to increased complexity, especially at an early age.
As for your example with Arabic versus Roman numerals, it's having ranked digits where each rank could be computed separately, independent of which symbols surrounded it. If you see 1, it always means 1, never -1. (Though I've seen texts that use IIII rather than IV.) And because the same symbols are used for the same value regardless of rank. If the Romans had had a symbol for zero, well, they could have simply written 'nullus' and be done with it. But the system of digits and ranks was the advantage.
So...you don't want to use Linux because it's harder to program for it than to use a calculator, and thus misses one of your requirements for being free software.
WTF?
I see the barrier to entry argument, but while that can be lowered slightly, programming is inherently difficult. Not everyone could do it, and some people can program much better than others.
Be happy when there are no *artificial* barriers to entry. There's jack you can do about natural barriers to entry.
No, the demo machine will have 2GB RAM and a dedicated video card.
The other common change is init systems. I sometimes use Debian, but I can't get my head around its weird /etc/rc.d system with symlinks and strange names. I realize that that's a more traditional setup than anything else, but still, Debian's the only Linux I've used that kept that system.
It could be valid, still, even though both the number of Linux users and the number of Linux distros is rising. It's quite possible that, if there were only two or three Linux distros, or even just one, Linux would have double its current market share.
My opinion is that, in that case, those few distros would not be as good as the major distros today, so Linux would have less market share. But I don't have any experimental data to back that.