Every piece of my computer is new as of a year ago (almost to the day) except for the keyboard that I've had since it came with the 8088 my dad brought home from his office. Maybe it was over-engineered or something, but this keyboard has been used continuously on an 8088-10, 286-16, 386-33, 486-133, Cyrix-183 (PR 233), K6-350, K6-500, and now Athlon-1667 (2000+). It gives me exactly the tactile feedback I want, and it doesn't have any of those @$#^ windows keys on it. I've got an AT->PS/2 convertor on it now (it has a built-in XT/AT switch), and I'll keep putting convertors on it as long as I need to. Most of the letters are still there too!
Compared to that, the 635 MB hard drive in the DSL router I built for my parents is state-of-the-art.
No one in their right mind would pay money on a per-download basis for peer-to-peer access. The cost of developing a client-server system, in which you know you're getting some standard of quality, is fairly low when amortized over the millions upon millions of downloads that you'd have, so the fee would be only slightly higher for the same royalties. If the labels were smart, they'd read the writing that's been on the wall for the past few years and actually do this. The only possible justification for allowing a fee-based pay-per-download would be that people who get crappy downloads would end up paying again, which is not something the labels want, since customer frustration over pricing is what got them into this mess in the first place.
If they're trying to do business with people located in Missouri, and they send commercial transmissions to people located in Missouri, they sure as hell fall under Missouri jurisdiction. Jurisdiction is a non-issue here.
The Supreme Court has always held political speech to be more protected than commercial speech. Political organizations obviously qualify. Making the case for non-profits is slightly more difficult, but because of how they operate they are obviously not commercial and are at the least implicitly political. As long as this distinction is made to the satisfaction of the court, the FTC is fully entitled to discriminate.
If the court believes that the FTC also has the right to block political and non-profit calls as well, this does not compel the FTC to apply the list to them as well. That would be a policy decision.
If this was really a service that everyone needs, all they need to do is inform the ISPs that they would be glad to accept DNS redirects for all invalid domains. ISPs could, even on a per-user basis, deliver the wildcard address to their users whenever an invalid lookup is made. Someone would develop software to take advantage of this, an xml schema for encoding the return data to non-browser apps would be published, software developers would implement it, and the public would demand that their ISPs support it. As long as it didn't affect me, I wouldn't care.
I'd be curious what his response to such a suggestion would be. Probably that nobody would bother and this "innovation" would go unused. And that's the point.
As an employee of any company, you are obligated to not represent your employer without either express or implied permission. Implied permission would be whatever your company policy specifically allows. If you're making a statement outside the workplace, you'd better not mention your company without explicit permission, because doing so can imply that your statement is either a creation of, or authorized by, that company. It doesn't matter if you're disparaging MS or you're disparaging Linux, you're doing something you're not supposed to do. Maybe the company will agree with what you say and go easy on you, but they have no obligation to. Failure to receive such mercy after disparaging an important client hardly implies unethical behavior on the part of the employer.
That said, I agree with most of the things he said. Doesn't change the fact that his conduct appears to have been unprofessional.
OOD is a language or meta-language level way of optimizing the way you work on the actual code itself. There is a tremendous amount of engineering that has little to do with the code itself, and this is what XP deals with. It is implicit in XP that you are using OOD, or something tantamount to it, in the case that you're using a language that is not OO.
XP is built on certain assumptions specific to software engineering, so it is an optimization of the engineering practices that every engineer follows, whether they are building a bridge or a database. XP is not very appropriate to very large projects, where the cost of prototyping and debugging after failure is very high, or completely forbidden, in the case of safety-critical applications. Since a great deal of software is neither very large nor safety-critical, these are perfectly reasonable trade-offs to make in many cases.
My software engineering professor, whose research is in software reliability, and told us horror story after horror story about aircraft avionics failures and the like, hates XP. In the fields he's looking at, the cost of prototyping and failure are astronomical. My Robocup team, which is writing code to make AI agents play soccer, has a much less strenuous safety obligation. Our cost of prototyping is equivalent to "make" which takes about as long as a trip down the hall to the snack machine, and watching a simulation, which takes about long enough to eat the bag of chips.
This is not to say that XP encourages you to write bad software. Instead, it assumes (reasonably) that all code will have defects, and attempts to maximize the probability that someone will detect them. Furthermore, it attempts to minimize the disruption to the development cycle that repairing a bug can cause. OOD is XP's preferred method of dealing with this at the source code level, so you're certainly very correct to note the similarity.
Actually, it's a lot easier at a place like MIT to get by without a machine of your own, because they're all over the place. I go to a respectably wired university, and I haven't bothered getting a laptop because there are labs everywhere, and many of the classrooms even have computers. My box probably gets rather lonely, as I spend a lot of my time in labs (like now) ostensibly doing classwork.
On another note, his story is perfectly believable. We have MAC address registration at my university as well, and when the RPC stuff hit, we got several calls at the helpdesk saying "I patched my machine, but I got this e-mail saying I was compromised." In many cases, it was a computer in some shared office that they happened to be the first person to sit down at when we turned on network registration, and their department techs forgot to pre-register them, so they had registered those machines as well.
Again, good to know. Now I know to bitch at the guy who made the image at work (which does several other weird permissions things). Well, I'm glad to see that Joe User has some hope now. I guess we'll have to beat it into their heads, since the windows installer still wants to make them all administrators. Thanks.
My boss often asks me where I hear about the things I know, and thinks I'm joking when I say "slashdot".
Good to know. Now why don't they actually TELL ANYONE about this? If I've never heard of it, Joe User sure as hell hasn't either. That's why Joe User doesn't use it. If I try to run kpackage as a normal user, I get prompted for a root password. If I try to install AIM on windows (which I know can be run off a network share from a computer it has not been installed on) I just get "You need to be logged in as administrator, punk."
As long as they put it on Legion, U.Va. will get to maintain state bragging rights. I don't keep up with the football rivalry, but this is much cooler anyway.
For immediate emergencies this generally doesn't work, but if you've got, say, 20 hours a week available for long-term projects, allocate that evenly among the various departments. Invariably they will want much more than you can do, so you give them first pick. If a department is dissatisfied with the wait time for some project of theirs, figure out how much of their budget they'd need to give you for you to outsource it or hire more staff. If that goes well, set up some kind of an auction model for your time. It will then become very clear which projects need to happen now, and which ones do not.
I work for in-house tech support at my university and that person:workload ratio you've got there is just obscene. If management won't give you the budget directly, you gotta find a way of getting it out of the departments you're supporting.
Unix is designed under the assumption that there are supposed to be users who can do whatever they please as long as it doesn't interfere with the operation of the system as a whole.
Windows is designed under the assumption that if you're not giving someone full control of the machine, it's because you don't want them to be able to do certain things that have no bearing on the rest of the machine whatsoever.
The result is that a typical Linux installation will create a user account without root privileges that you are expected to use except when you absolutely need to be root. The windows installation will prompt you to create accounts other than Administrator, but they will still be Administrator-level accounts, because the registry and the windows installer are designed to make it difficult for anyone who is not an administrator to install software.
This is why I'm an administrator on my work machine, where I do tech support and thus need to be able to mess around with things to replicate problems, and I'm a non-root user (with sudo privileges) on my home machine. I can screw up the work machine a hell of a lot faster than I can the home machine if I open up the wicked screensaver.
If windows didn't require a completely separate login to do administrator-level stuff, this problem might go away. XP's user-switching is a far cry from this. If Joe User can't copy and paste from his non-admin web browser to some admin system tool, he'll just be admin all the time, and then when he breaks beyond all repair he'll call me along with the other hundred users I talked to today at work. AAAAAAAAAH!
No, it's becoming exactly like 1984. The people who are spying on us know what they're doing. It's just the people who are supposed to be helping us who are shutting down.
The race rules require an 80 kg driver (176 lbs.) Anything below this and you get ballasted up. When I was on the U.Va. team, all of our drivers were considerably below this weight, since there weren't very many people who could fit into the cockpit comfortably. The nice thing was that we could put the extra weight wherever we wanted it, so we got a better weight balance out if it.
The amount of useable surface area that is horizontal is actually much less on a standard sedan than on a solar car. Even less on a Prius. The practical application here is not solar cars for the masses. The practical applications are:
1) Solar Power elsewhere
2) energy-efficient cars
When you consider how fast these cars go and how little power they use to do it, it becomes obvious just how much room for improvement in current cars. For example, if we stopped using engines that produce 7 times as much waste heat as useable energy, we wouldn't need to cool them nearly as much. You could replace the grill on the front of the car with a faring. Highway gas mileage would go through the roof. As a side bonus, you'd have almost zero radar profile, so if you get pulled over, you could argue that they were actually reading the guy in the Hummer behind you.
The great thing about unix is you can write portable code. If you're doing something academic, all the interesting stuff is abstract anyway, unless you're doing OS or security stuff, in which case your advisor should be shot if they try to force you to switch from Windows to Linux.
Slashdot Mirror
Every piece of my computer is new as of a year ago (almost to the day) except for the keyboard that I've had since it came with the 8088 my dad brought home from his office. Maybe it was over-engineered or something, but this keyboard has been used continuously on an 8088-10, 286-16, 386-33, 486-133, Cyrix-183 (PR 233), K6-350, K6-500, and now Athlon-1667 (2000+). It gives me exactly the tactile feedback I want, and it doesn't have any of those @$#^ windows keys on it. I've got an AT->PS/2 convertor on it now (it has a built-in XT/AT switch), and I'll keep putting convertors on it as long as I need to. Most of the letters are still there too!
Compared to that, the 635 MB hard drive in the DSL router I built for my parents is state-of-the-art.
No one in their right mind would pay money on a per-download basis for peer-to-peer access. The cost of developing a client-server system, in which you know you're getting some standard of quality, is fairly low when amortized over the millions upon millions of downloads that you'd have, so the fee would be only slightly higher for the same royalties. If the labels were smart, they'd read the writing that's been on the wall for the past few years and actually do this. The only possible justification for allowing a fee-based pay-per-download would be that people who get crappy downloads would end up paying again, which is not something the labels want, since customer frustration over pricing is what got them into this mess in the first place.
Of course, it is possible they're just stupid.
If they're trying to do business with people located in Missouri, and they send commercial transmissions to people located in Missouri, they sure as hell fall under Missouri jurisdiction. Jurisdiction is a non-issue here.
The Supreme Court has always held political speech to be more protected than commercial speech. Political organizations obviously qualify. Making the case for non-profits is slightly more difficult, but because of how they operate they are obviously not commercial and are at the least implicitly political. As long as this distinction is made to the satisfaction of the court, the FTC is fully entitled to discriminate.
If the court believes that the FTC also has the right to block political and non-profit calls as well, this does not compel the FTC to apply the list to them as well. That would be a policy decision.
If this was really a service that everyone needs, all they need to do is inform the ISPs that they would be glad to accept DNS redirects for all invalid domains. ISPs could, even on a per-user basis, deliver the wildcard address to their users whenever an invalid lookup is made. Someone would develop software to take advantage of this, an xml schema for encoding the return data to non-browser apps would be published, software developers would implement it, and the public would demand that their ISPs support it. As long as it didn't affect me, I wouldn't care.
I'd be curious what his response to such a suggestion would be. Probably that nobody would bother and this "innovation" would go unused. And that's the point.
The wonderful thing about open-source software is that someone can rewrite init in python, and the rest of us can laugh at them. Everyone wins.
Billions, actually, from what I've heard. Try saying it with your pinky to the corner of your mouth.
As an employee of any company, you are obligated to not represent your employer without either express or implied permission. Implied permission would be whatever your company policy specifically allows. If you're making a statement outside the workplace, you'd better not mention your company without explicit permission, because doing so can imply that your statement is either a creation of, or authorized by, that company. It doesn't matter if you're disparaging MS or you're disparaging Linux, you're doing something you're not supposed to do. Maybe the company will agree with what you say and go easy on you, but they have no obligation to. Failure to receive such mercy after disparaging an important client hardly implies unethical behavior on the part of the employer.
That said, I agree with most of the things he said. Doesn't change the fact that his conduct appears to have been unprofessional.
OOD is a language or meta-language level way of optimizing the way you work on the actual code itself. There is a tremendous amount of engineering that has little to do with the code itself, and this is what XP deals with. It is implicit in XP that you are using OOD, or something tantamount to it, in the case that you're using a language that is not OO.
XP is built on certain assumptions specific to software engineering, so it is an optimization of the engineering practices that every engineer follows, whether they are building a bridge or a database. XP is not very appropriate to very large projects, where the cost of prototyping and debugging after failure is very high, or completely forbidden, in the case of safety-critical applications. Since a great deal of software is neither very large nor safety-critical, these are perfectly reasonable trade-offs to make in many cases.
My software engineering professor, whose research is in software reliability, and told us horror story after horror story about aircraft avionics failures and the like, hates XP. In the fields he's looking at, the cost of prototyping and failure are astronomical. My Robocup team, which is writing code to make AI agents play soccer, has a much less strenuous safety obligation. Our cost of prototyping is equivalent to "make" which takes about as long as a trip down the hall to the snack machine, and watching a simulation, which takes about long enough to eat the bag of chips.
This is not to say that XP encourages you to write bad software. Instead, it assumes (reasonably) that all code will have defects, and attempts to maximize the probability that someone will detect them. Furthermore, it attempts to minimize the disruption to the development cycle that repairing a bug can cause. OOD is XP's preferred method of dealing with this at the source code level, so you're certainly very correct to note the similarity.
"Verisign did not respond [to] Requests For Comment" (emphasis added)
This technology will revolutionize the porn industry!
Actually, it's a lot easier at a place like MIT to get by without a machine of your own, because they're all over the place. I go to a respectably wired university, and I haven't bothered getting a laptop because there are labs everywhere, and many of the classrooms even have computers. My box probably gets rather lonely, as I spend a lot of my time in labs (like now) ostensibly doing classwork.
On another note, his story is perfectly believable. We have MAC address registration at my university as well, and when the RPC stuff hit, we got several calls at the helpdesk saying "I patched my machine, but I got this e-mail saying I was compromised." In many cases, it was a computer in some shared office that they happened to be the first person to sit down at when we turned on network registration, and their department techs forgot to pre-register them, so they had registered those machines as well.
I think you've just invented a new technique for perforation.
Did BBspot just hack slashdot?
Again, good to know. Now I know to bitch at the guy who made the image at work (which does several other weird permissions things). Well, I'm glad to see that Joe User has some hope now. I guess we'll have to beat it into their heads, since the windows installer still wants to make them all administrators. Thanks.
My boss often asks me where I hear about the things I know, and thinks I'm joking when I say "slashdot".
Good to know. Now why don't they actually TELL ANYONE about this? If I've never heard of it, Joe User sure as hell hasn't either. That's why Joe User doesn't use it. If I try to run kpackage as a normal user, I get prompted for a root password. If I try to install AIM on windows (which I know can be run off a network share from a computer it has not been installed on) I just get "You need to be logged in as administrator, punk."
As long as they put it on Legion, U.Va. will get to maintain state bragging rights. I don't keep up with the football rivalry, but this is much cooler anyway.
For immediate emergencies this generally doesn't work, but if you've got, say, 20 hours a week available for long-term projects, allocate that evenly among the various departments. Invariably they will want much more than you can do, so you give them first pick. If a department is dissatisfied with the wait time for some project of theirs, figure out how much of their budget they'd need to give you for you to outsource it or hire more staff. If that goes well, set up some kind of an auction model for your time. It will then become very clear which projects need to happen now, and which ones do not.
I work for in-house tech support at my university and that person:workload ratio you've got there is just obscene. If management won't give you the budget directly, you gotta find a way of getting it out of the departments you're supporting.
Unix is designed under the assumption that there are supposed to be users who can do whatever they please as long as it doesn't interfere with the operation of the system as a whole.
Windows is designed under the assumption that if you're not giving someone full control of the machine, it's because you don't want them to be able to do certain things that have no bearing on the rest of the machine whatsoever.
The result is that a typical Linux installation will create a user account without root privileges that you are expected to use except when you absolutely need to be root. The windows installation will prompt you to create accounts other than Administrator, but they will still be Administrator-level accounts, because the registry and the windows installer are designed to make it difficult for anyone who is not an administrator to install software.
This is why I'm an administrator on my work machine, where I do tech support and thus need to be able to mess around with things to replicate problems, and I'm a non-root user (with sudo privileges) on my home machine. I can screw up the work machine a hell of a lot faster than I can the home machine if I open up the wicked screensaver.
If windows didn't require a completely separate login to do administrator-level stuff, this problem might go away. XP's user-switching is a far cry from this. If Joe User can't copy and paste from his non-admin web browser to some admin system tool, he'll just be admin all the time, and then when he breaks beyond all repair he'll call me along with the other hundred users I talked to today at work. AAAAAAAAAH!
No, it's becoming exactly like 1984. The people who are spying on us know what they're doing. It's just the people who are supposed to be helping us who are shutting down.
The race rules require an 80 kg driver (176 lbs.) Anything below this and you get ballasted up. When I was on the U.Va. team, all of our drivers were considerably below this weight, since there weren't very many people who could fit into the cockpit comfortably. The nice thing was that we could put the extra weight wherever we wanted it, so we got a better weight balance out if it.
The amount of useable surface area that is horizontal is actually much less on a standard sedan than on a solar car. Even less on a Prius. The practical application here is not solar cars for the masses. The practical applications are:
1) Solar Power elsewhere
2) energy-efficient cars
When you consider how fast these cars go and how little power they use to do it, it becomes obvious just how much room for improvement in current cars. For example, if we stopped using engines that produce 7 times as much waste heat as useable energy, we wouldn't need to cool them nearly as much. You could replace the grill on the front of the car with a faring. Highway gas mileage would go through the roof. As a side bonus, you'd have almost zero radar profile, so if you get pulled over, you could argue that they were actually reading the guy in the Hummer behind you.
The great thing about unix is you can write portable code. If you're doing something academic, all the interesting stuff is abstract anyway, unless you're doing OS or security stuff, in which case your advisor should be shot if they try to force you to switch from Windows to Linux.
Perhaps you mean hdparm? Fdisk doesn't seem to do DMA/PIO settings.
didn't they try suing burner manufacturers once?