Seems to me the OS made it too easy to change the defaults to something less secure. At the very least, you should have to read a man page to figure out how to turn off your security so thoroughly. Or is it just that the OS was insecure by default?
By defining your criteria to include "doesn't sue [potential] customers" you can legitimately exclude Diebold. The only recourse Diebold would then have would be to contest the criteria themselves as illegal or discriminatory, which would be absurd. And fun to watch.
So the throbber and status bar are not enough? Submit a request for enhancement to bugzilla. And call it an accessibility option. It shouldn't be hard to add an optional click noise.
That is by no means universal. On a lot of mainstream hardware, a modern Linux will support power management as well as or better than Win2k or even XP. Add to that the fact that Linux software is often more efficient (mostly server apps). And then consider that you don't need to upgrade the hardware in order to upgrade to a newer Linux. Then (and this is the biggie) realize that your Linux workstations will never be compromised by spammers and turned into a network-crippling botnet. All of those add up to significantly less energy used than what most large organizations are currently using.
In other words: switching to Linux has huge up front costs, but it pretty much pays for itself from then on, in a lot of ways.
Not all that earth-shattering. Or even non-obvious. Switching to a fundamentally better system is usually a good idea. Convincing people (users, government, etc.) to invest a little bit of time to learn about the new system it the impossible part.
I really doubt that a chord keyboard could rival the original palm graffiti in usability. With graffiti, there is very little chance of dropping the pda or having it slip out of your hand because you were squeezing it in a strange way. Also, with graffiti, you had a decent chance of guessing correctly.
That said, the ideal text input method for a handheld is probably a projection keyboard plus a touch screen keyboard or graffiti for the times when you can't sit it down.
Anybody running Linux on a non-x86 box will know how to install from a distro's repositories. The same can no longer be said for Linux on x86, because of people installing ubuntu on their grnadmother's computer, etc. However, in this case it really doesn't matter. This is alpha software. If you are messing with it, you should know how to compile it yourself. It is not really worth Mozilla's time to acquire obsolete non-x86 desktop hardware for building nightlies and alpha releases.
I find your comment about hardware manufacturers odd. Which companies are currently ignoring the non-x86 market that would actually profit from making hardware for that very small market? The way I see it, market forces have decided that x86 is the only architecture that matters for desktops.
That ARM board stores the entire operating system in flash. It uses busybox and pretty much nothing else, to get a shell up that fast. It doesn't have to wait for any hard drives or peripherals to initalize. LinuxBIOS can do similar things, but only on some machines. TFA is all about getting services to start quickly. Turning off everything is not an acceptable alternative.
It's been a while since I analyzed the boot process of one of my Linux systems, but this is what I recall noticing:
Immediately after kernel startup, there is a lot of hardware probing and configuration going on. On older systems, setting up ISA pnp cards can take quite a while. Along with that, udev uses a lot of cpu time getting things going. Dbus was also one of the slower services to start. Also, if your init scripts are really borked, you can waste several seconds waiting for a dhcp response.
Also, the bars in a chart are all labeled. The sample in the article was cropped. Check out the bootchart website to see a complete chart.
I live in an area where most commuters drive in from an adjacent county. As to the analogy, standard transmissions are no longer "standard" and automatic transmissions are pretty much universally availible. So in that sense, they are pretty much obsolete. Similarly, when programming, there is hardly ever any justification for not using the standard library's math routines. The problems come when you start to put together complex procedures with relatively low level operators. When you are doing that, it is easy to end up with a few extra calls that actually slow down your code by an order of magnitude. One of the other posters provided a good example of when not to use strcat.
Gears matter, yes. But most people had a bike as a kid. However, a car, unlike a bike, can have certain parts melt or bend out of shape or shatter when you redline it on the interstate for miles. The actual gear ratios don't matter much. It is the engine speed you need to be careful of.
You need to understand quite a bit about internal combustion just to drive a stick shift on a highway without blowing the engine or the clutch. You just can't go 65mph in first gear in a Civic and expect to get to work.
That's why most people get automatic transmissions. They can't be bothered to understand the concepts of "rpm" and "torque". And very few cars can handle a commute all in first gear.
To get back to computers, yes, most apps written in an interpreted language or just written badly will still be fast enough. But the libraries and operating systems have to be quite fast, otherwise the lags will add up to an unresponsive system. People notce the difference between 150ms and 500ms. In many cases, 150ms is acceptably close to instant, but a 500ms pause is so slow that it is distracting. A responsive gui makes any product seem much more professional.
Go look at the recent changes to Cairo, from 1.2 to 1.4. It is astounding how many operations can be sped up by a factor of ten or a hundred just by spending a few extra days or weeks improving a few routines. You can't accomplish that much in so short a time span without knowing a lot about what is going on under the hood. Optimizing code through iteratively profiling and changing code with trial and error will not produce the kind of quality that Cairo has achieved.
Look at how Spitzer got to be govenor of NY - as state attorney general, he did what the feds (specifically, the SEC) refused to do. There is plenty of room for more people like him.
You seem to be offering evidence in support of his position. You act as though the performance difference between threads and processes is the primary factor contributing to the difference you have percieved between Windows and Linux desktops. It is not. Most of the time, when windows programs are faster, it is because they do less. On Linux, simple apps are run by slow, complex language interpreters (eg. cpython) and at any given moment there can be half a dozen gui libraries in use, with all the necessary abstraction layers to arbitrate between them. Every step of the way, there are multiple ways of doing things. That comes at a cost, especially compared with an operating system that essentially can't do many things, or can only do them with compromised security.
I think it is more the younger members of the slashdot crowd. The term "gifted" used to be used far more sparingly. These days, anybody who can ace a standardized test in spite of having bad teachers is labeled "gifted". It is reasonable that a larger proportion of slashdotters would qualify as "gifted" than in the population at large.
However, the proportion of slashdotters who qualify as "highly gifted" or better in the precise medical sense (ie having an IQ of at least 145) is still very small.
I think it is more that gifted children deal with a very different set of emotional challenges. For example, not many pop artists sing about the excruciating boredom of being stuck in a middle school when you are academically ready for college. Nor do you hear much about the angst of being eight years younger than the dumbest people you would consider peers. Also, gifted kids are usually far less interested in the dating scene during the teenage years. All of the really gifted kids I know have developed really weird or dangerous hobbies to cope with the relative monotony of suburban American life.
When you are dealing with profoundly gifted kids, it is a safe bet that any antisocial behavior is mostly intentional and the kid is fully aware of how other people interpret those actions. I've known at least one kid who could fool most any psychologist into thinking he had Asperger's, at first glance.
Your comment did exactly nothing to clarify the situation. In American English, double negatives are always ambiguous (though that is more an issue with the users of the language than the language itself).
This is IIT we are talking about. These are engineering and computer science type students. People like that never conform to the socially accepted sleep patterns. Using censorship to try to force them to sleep from 11PM to 6AM is going to be counter-productive.
Slashdot Mirror
Seems to me the OS made it too easy to change the defaults to something less secure. At the very least, you should have to read a man page to figure out how to turn off your security so thoroughly. Or is it just that the OS was insecure by default?
Come on! That one has been making the rounds since the announcement! I modded it up last time I saw it on slashdot. Make up a new joke.
By defining your criteria to include "doesn't sue [potential] customers" you can legitimately exclude Diebold. The only recourse Diebold would then have would be to contest the criteria themselves as illegal or discriminatory, which would be absurd. And fun to watch.
So the throbber and status bar are not enough? Submit a request for enhancement to bugzilla. And call it an accessibility option. It shouldn't be hard to add an optional click noise.
Two wiimotes (or one with nunchuk) and a DDR pad should suffice. In other words, now.
That is by no means universal. On a lot of mainstream hardware, a modern Linux will support power management as well as or better than Win2k or even XP. Add to that the fact that Linux software is often more efficient (mostly server apps). And then consider that you don't need to upgrade the hardware in order to upgrade to a newer Linux. Then (and this is the biggie) realize that your Linux workstations will never be compromised by spammers and turned into a network-crippling botnet. All of those add up to significantly less energy used than what most large organizations are currently using.
In other words: switching to Linux has huge up front costs, but it pretty much pays for itself from then on, in a lot of ways.
Not all that earth-shattering. Or even non-obvious. Switching to a fundamentally better system is usually a good idea. Convincing people (users, government, etc.) to invest a little bit of time to learn about the new system it the impossible part.
I really doubt that a chord keyboard could rival the original palm graffiti in usability. With graffiti, there is very little chance of dropping the pda or having it slip out of your hand because you were squeezing it in a strange way. Also, with graffiti, you had a decent chance of guessing correctly.
That said, the ideal text input method for a handheld is probably a projection keyboard plus a touch screen keyboard or graffiti for the times when you can't sit it down.
Anybody running Linux on a non-x86 box will know how to install from a distro's repositories. The same can no longer be said for Linux on x86, because of people installing ubuntu on their grnadmother's computer, etc. However, in this case it really doesn't matter. This is alpha software. If you are messing with it, you should know how to compile it yourself. It is not really worth Mozilla's time to acquire obsolete non-x86 desktop hardware for building nightlies and alpha releases.
I find your comment about hardware manufacturers odd. Which companies are currently ignoring the non-x86 market that would actually profit from making hardware for that very small market? The way I see it, market forces have decided that x86 is the only architecture that matters for desktops.
No. http://www.penny-arcade.com/comic/2006/11/13
I expect Celebrex to regain all of its previous popularity, and then some. Senior citizens are going to want their range of motion back.
That ARM board stores the entire operating system in flash. It uses busybox and pretty much nothing else, to get a shell up that fast. It doesn't have to wait for any hard drives or peripherals to initalize. LinuxBIOS can do similar things, but only on some machines. TFA is all about getting services to start quickly. Turning off everything is not an acceptable alternative.
It's been a while since I analyzed the boot process of one of my Linux systems, but this is what I recall noticing:
Immediately after kernel startup, there is a lot of hardware probing and configuration going on. On older systems, setting up ISA pnp cards can take quite a while. Along with that, udev uses a lot of cpu time getting things going. Dbus was also one of the slower services to start. Also, if your init scripts are really borked, you can waste several seconds waiting for a dhcp response.
Also, the bars in a chart are all labeled. The sample in the article was cropped. Check out the bootchart website to see a complete chart.
I live in an area where most commuters drive in from an adjacent county. As to the analogy, standard transmissions are no longer "standard" and automatic transmissions are pretty much universally availible. So in that sense, they are pretty much obsolete. Similarly, when programming, there is hardly ever any justification for not using the standard library's math routines. The problems come when you start to put together complex procedures with relatively low level operators. When you are doing that, it is easy to end up with a few extra calls that actually slow down your code by an order of magnitude. One of the other posters provided a good example of when not to use strcat.
These days, most of the computation done in undergrad math classes is done with Matlab. FORTRAN no longer dominates those kind of classes.
Gears matter, yes. But most people had a bike as a kid. However, a car, unlike a bike, can have certain parts melt or bend out of shape or shatter when you redline it on the interstate for miles. The actual gear ratios don't matter much. It is the engine speed you need to be careful of.
You need to understand quite a bit about internal combustion just to drive a stick shift on a highway without blowing the engine or the clutch. You just can't go 65mph in first gear in a Civic and expect to get to work.
That's why most people get automatic transmissions. They can't be bothered to understand the concepts of "rpm" and "torque". And very few cars can handle a commute all in first gear.
To get back to computers, yes, most apps written in an interpreted language or just written badly will still be fast enough. But the libraries and operating systems have to be quite fast, otherwise the lags will add up to an unresponsive system. People notce the difference between 150ms and 500ms. In many cases, 150ms is acceptably close to instant, but a 500ms pause is so slow that it is distracting. A responsive gui makes any product seem much more professional.
Go look at the recent changes to Cairo, from 1.2 to 1.4. It is astounding how many operations can be sped up by a factor of ten or a hundred just by spending a few extra days or weeks improving a few routines. You can't accomplish that much in so short a time span without knowing a lot about what is going on under the hood. Optimizing code through iteratively profiling and changing code with trial and error will not produce the kind of quality that Cairo has achieved.
Look at how Spitzer got to be govenor of NY - as state attorney general, he did what the feds (specifically, the SEC) refused to do. There is plenty of room for more people like him.
It's funny. Laugh. But look up "verizon math" first.
You seem to be offering evidence in support of his position. You act as though the performance difference between threads and processes is the primary factor contributing to the difference you have percieved between Windows and Linux desktops. It is not. Most of the time, when windows programs are faster, it is because they do less. On Linux, simple apps are run by slow, complex language interpreters (eg. cpython) and at any given moment there can be half a dozen gui libraries in use, with all the necessary abstraction layers to arbitrate between them. Every step of the way, there are multiple ways of doing things. That comes at a cost, especially compared with an operating system that essentially can't do many things, or can only do them with compromised security.
I think it is more the younger members of the slashdot crowd. The term "gifted" used to be used far more sparingly. These days, anybody who can ace a standardized test in spite of having bad teachers is labeled "gifted". It is reasonable that a larger proportion of slashdotters would qualify as "gifted" than in the population at large.
However, the proportion of slashdotters who qualify as "highly gifted" or better in the precise medical sense (ie having an IQ of at least 145) is still very small.
Is that $0.02 cents per page or $0.02 dollars per page?
I think it is more that gifted children deal with a very different set of emotional challenges. For example, not many pop artists sing about the excruciating boredom of being stuck in a middle school when you are academically ready for college. Nor do you hear much about the angst of being eight years younger than the dumbest people you would consider peers. Also, gifted kids are usually far less interested in the dating scene during the teenage years. All of the really gifted kids I know have developed really weird or dangerous hobbies to cope with the relative monotony of suburban American life.
When you are dealing with profoundly gifted kids, it is a safe bet that any antisocial behavior is mostly intentional and the kid is fully aware of how other people interpret those actions. I've known at least one kid who could fool most any psychologist into thinking he had Asperger's, at first glance.
Your comment did exactly nothing to clarify the situation. In American English, double negatives are always ambiguous (though that is more an issue with the users of the language than the language itself).
This is IIT we are talking about. These are engineering and computer science type students. People like that never conform to the socially accepted sleep patterns. Using censorship to try to force them to sleep from 11PM to 6AM is going to be counter-productive.