'' The car also had built in GPS with maps, and seemed to be able to drive itself anywhere, though not from the exact beginning to the exact end (it relied on humans for such things as parking.) ''
Which is weird, because I would have thought that parking would be one thing an automated system should be able to do better than any human. And it is much easier to make it failsafe than most parts of driving: When in doubt, hit the brakes.
'' The moment you get a human interacting with computer driven cars, all chaos will result. No computer system, radar system, and automated response system can anticipate a drunk human driver swerving across 6 lanes of traffic at 100 mph in order to make an exit. ''
It doesn't have to anticipate anything. A good radar system should be able to keep track of the location, speed and direction of that car with higher precision than you, calculate its course, and take evasive action faster and more precisely than you can, while taking into consideration where other cars are as well. Assuming that some other cars are automated as well and some aren't, and the automated ones know who is automated and who is not, all those cars could act coordinated as well (like everyone not directly involved making space for those closer to the action).
'' This would be a great advancement in driving for the elderly. Just imagine, a car that plows through a farmers market automatically! They should beta test it in Florida. ''
I've seen this attitude being posted here several times. Just yesterday, four children (four, eight, and two fifteen your old) died in a car accident about four miles away from my home. The oldest was driving. They hit the car of an elderly couple frontally; the kids were on the wrong side of the road. The people in the other car had light injuries only, probably because they were wearing seatbelts.
During the year-2000 craze, one insurance company in Britain reported that their computer software had always been able to handle customers who were born in the century before the current century (they had to, because the system was started around 1960 and there were plenty of pensioners born in the previous century at that time), but they figured out that to change it to handle customers two centuries before the current one would be a significant, expensive, and dangerous change.
On the other hand, they only had half a dozen customers born in the 19th century and still alive in the year 2000, so they removed their records from the computer system and one man handled their accounts manually. Much cheaper that way.
Probably the same here. Something like a database with a nine digit field for "donations to charity". Much cheaper to handle Bill Gates' tax by hand then to make a software change just for one person.
'' I think the "pay $4-5 to make it all legal" idea would only work if all record labels participated, and all ISPs participated. You'd have to basically force every ISP to add this "music-download tax," and implement it across the board... otherwise customers are going to be flocking to the competition that doesn't include this tax, and continue downloading things for free. ''
Instead of making it a "download-tax", why not make it a license to posess music, no matter where it comes from? My model would be something like iTunes, where a song gets marked to a specific owner, but if you pay the license tax, your computer plays music, no matter who the owner is. You would be allowed to copy anyone's music without restriction, and your computer plays it as long as the license tax is paid. If you stop paying, your computer stops playing. Still, nobody could sue you because the music is not playable.
P2P download qualities would improve, because instead of clueless kids, adults with respect for quality and either respect of the law or fear for consequences of breaking the law, would then start uploading.
'' However what happens to the tracking if the artists themselves decide to boost their income by having bots download their songs as often as possible? ''
That should be no problem. Lets say everybody pays x dollars a month, and lets say three dollars of everyone's payments is to be distributed to the artist. If all you download in one month is one Britney Spears song, she gets your three dollars. If you download her song tenthousand times, she gets three dollars. If you download ten different songs, everyone gets 30 cents, if you download 1000 different songs, everyone gets 0.3 cents.
The bot can only produce three dollars of income to an artist, but it needs an ISP address where more than three dollars are paid, so it is a net loss.
What would be dangerous is a virus that gets copied on many machines of paying consumers and downloads stuff they don't want.
'' My big fear is that the Disney brass learned their lessons from the Daimler-Chrysler scam: perform a "merger", let the new "partners" have their head for a while, but retain the real power. A year passes, and then there's a purge. I don't think there will be any Pixar people left at Disney after 1 year, and Jobs will leave in 18 months: substantially richer, but disgusted. ''
That would be rather stupid; paying over $7bn for essentially nothing. Not that it is impossible, but it would be incredibly stupid.
'' Actually I was discussing the UK situation. Here's a BBC article about how intelligent design appears well thought of in the UK (17% of population with 22% for creationism and 48% for evolution). For me, as a scientist, it was quite surprising to here this. AFAIK, the study was reasonable and the findings valid. ''
I found this very strange, because I doubt that anyone here in Britain has ever heard of "Intelligent Design", except for those who have heard of it in the context of brain-damaged religious twats who are even worse than Creationists.
Most probably thought of BMW or Apple Computer when they were asked about "Intelligent Design".
'' I've actually heard people from Mensa say that you can't study for the test. What bullshit. You can study the problem types etc. in the same way you can study or learn about anything. ''
People who take intelligence tests repeatedly will know how to maximise their scores; mostly by reducing the time spent at each question, thus increasing the percentage of wrong answers, but maximising the number of correct answers. Using a good strategy to use your time will already improve your scores.
''... I'm marking a pile of work handed in by 2nd year university students studying mathematics. Instead of giving numerical grades and seeing how good or bad the little chavlets have done we are required to avoid disheartening them by grading them on the following scale: E - Excelent - Tried every question and got most of them mostly right.
G - Good - Tried most of the questions and got them mostly right.
S - Satisfactory - Tried some of the questions, didn't get everything wrong.
U - Unsatisfactory - Didn't bother to hand in, or got absolutely everything wrong
Yeah. This is at one of the better Universities around in the UK (though not one of the best). If schools are as bad (and they are) is it any wonder? ''
That was exactly the scale that was used when I studied maths 25 years ago. We had exactly one guy who managed to achieve the (E), and that guy was a pure genius. But then, trying all questions and getting them mostly right would have required four hours of writing at high speed without any break, and that alone made it difficult.
'' The Cell will be nothing special for typical integer workloads...in fact it will probably perform inferior to offerings from Intel and AMD. In fact any double-precision arithmatic is 10 times slower!! ''
Actually, Cell will be quite powerful on many double precision floating point problems. Throughput per processor is not very high, but latency is only twice the throughput, memory access is incredibly fast, all the supporting operations are incredibly fast, so it is very very easy to get performance that is equal to the theoretical limit of the chip.
You have to do some macro changes to split a task to run on more than one SPE, but that may be easier than trying to get peak performance out of a Pentium or G5 chip. Especially on a G5, you need massive parallelism at the instruction level to get peak performance.
'' As our tech increases we need less and less knowledge about it. My mom knew how to wire a fuse. I know how to screw in one. My kid knows how to throw a circuit breaker. Wich one of us would be more likely to be able to get a car moving when there is no replacement fuse available?
The kid. While both of you are looking for the fuse the kid realises that the car's got an onboard computer (black box) which has crashed and needs resetting, finds the reset switch (possibly a hole needing a paperclip), and holds it down for 3 seconds. ''
In my experience, the kid will be too stupid to realise anything. Second, he won't know that reset switches are hidden so they can't be pressed by accident. Third, he will be too stupid to find that hole. Fourth, even if he finds that hole, he will be too stupid to improvise something that can be used to press a reset button hidden in that hole.
''.. its the source code to Office that we want. If it were possible to buy a competing office suite for Windows, Linux, Mac OS X, etc, that was 99% compatible with all office file formats, then we would have true competition in an open and level market. ''
The source code to Microsoft Office would not help you one bit. Anyone using that source code to build Office compatible applications would be sued for copyright violations, and rightfully so. Source code doesn't help to create compatible applications at all. Source code helps to find bugs, or to find vulnerabilities. We don't want to find bugs in the Microsoft Office source code, that is Microsoft's job.
What you need is a specification. Like the Open Document format; anybody can implement an application that is compatible with the Open Document format, no problem at all. Or all the specificiations for internet protocols, anybody can implement them. Or take the H.264 spec; that is a few hundred pages of highly technical documentation, and there are dozens of encoders and decoders available. Source code is useless.
'' Rather than worry about "opening up Windows, or decoupling IE from Windows" isn't it time that people start innovating and competing with Microsoft? ''
Yes, that is why Microsoft has to produce the documentation how to interact with their software, so that others can start innovating and competing. Without that documentation, the Microsoft monopoly prevents competition.
Now it seems that Microsoft is uncapable to provide documentation, so as a publicity stunt they propose to let others read their source code instead of documentation. But as any programmer knows (are there any here on Slashdot? ), SOURCE CODE IS NOT DOCUMENTATION.
'' I can guarantee you for the price you paid for that MacIntel (Intelitosh?), you'd be much better off spending half of the cash and buying from another retailer ''
So where exactly can you buy a notebook with Core Duo processor for half the price of an Intel MacBook Pro?
'' It's like porting Windows to MAC. Why bother? Most, if not all, MAC people don't want or require it. Granted there are a few people with exceptions but for the 99.9% why is so important? ''
The problem with your argument is that the number "99.9%" is not correct. It is more like 80% or 90%. And you forget that there are considerable numbers of Linux users and Windows users who would just like to buy a computer that is completely contained in the LCD screen (without having to run MacOS X).
'' This is a good argument for the need for open source election software. Another factor besides transparency is the need for peer review. Well designed software should be able to withstand some level of scrutiny from outside developers, and no set of code should be more well designed than that which is responsible for tabulating the record concerning the orderly transfer of power. ''
It is a good argument for publicly available source code, with a license to everybody and their dog to duplicate and compile the source code in order to examine its operation and its correctness.
Anybody, including Diebold's competitors, should be able and allowed to find and point out any flaws in their software, but there is no reason why competitors should be allowed to use the source code in their own products.
No matter how difficult it is for someone outside Apple to make Windows XP boot, I would say the following to Apple:
1. If Apple were to sell Macintosh hardware with Windows XP preinstalled instead of MacOS X, then a considerable number of people would buy these machines. Not "considerable" as in "Dell goes out of business" but "considerable" as in a few percent of Apple revenues.
2. If Apple were to sell Macintosh hardware that can dual boot into MacOS X and Windows XP without any problems, a much greater number would buy those machines. Dual boot = run one, reboot, run the other, reboot...
3. If Apple were to sell Macintosh hardware that can run MacOS X and Windows XP simultaneously, they could sell tons of those. Even if "simultaneously" means that one of the OSes is in sleep mode while the other is running, with some form of communication so that cut&paste works.
'' I just wonder (assuming it works correctly only in FireFox for a clear example) how many IE users and Opera users and Safari users etc wouldn't just go to a different site as opposed to downloading, installing and then using another browser for *one site*. ''
You might also ask yourself (or your marketing people) how many people will get pissed off if your webpage doesn't work with their favorite browser, so that having a webpage that doesn't support browser X might be worse than having no webpage at all!
And what _absolutely_ pisses people off is if your webpage actively disables their browser. That is worse than a webpage that doesn't work properly. A webpage that doesn't work shows the developers were incompetent (they might just be under time pressure, but the user will assume incompetence); that is forgivable. A webpage that shuts me out means an active decision was made that the company doesn't want my customs. That is unforgivable.
'' On the Apple website right over the pic of the Macbook is a line reading "What's an Intel chip doing in a Mac? A lot more than it' ever done in a PC". That's pretty funny. I wonder what they'll say in 10 years when AMD has killed Intel (praying that it happens, of course)? I'd love a Mac with an Athlon and Win (games), Linux, Mac OS, and BSD all on the same machine. ''
I think you make the assumption that Intel is going to roll over and die. I think quite the opposite is going to happen.
Intel was an enormous company. They could build a chip based on the most crappiest design possible (the Itanium) and make it through sheer will power. Except that it doesn't work and it cost them dearly. They could build another chip based on the second most crappiest design possible (the Pentium 4) and make it through sheer will power. Except that it doesn't work and now they are falling behind.
But it seems that Intel has started realising that they are not invincible (at about the same time the AMD fanboys started thinking that Intel is dead); and for the first time in years they have an actual roadmap with promising products. And I think in Apple Computer they have found the best customer they could wish for: One that won't let them get away with crap like Itanium and P4.
'' How long will it now take before we start seeing iTunes Payola. Networks eager to get their shows on people's ipods will certainly be tempted to try to strike special deals to make their shows more easily obtainable. They may try to offer Apple cash, try to reduce the cost to consumers, or try to find ways to get their links on the front page. ''
It might happen, but it won't be as bad as on radio. On the radio, if they pay, you can hear it, if they don't pay, you can't hear it. On iTunes, what would happen: If they pay, it is advertised big on page 1. If they don't pay, it is still there on page 3 or 4.
Slashdot Mirror
'' The car also had built in GPS with maps, and seemed to be able to drive itself anywhere, though not from the exact beginning to the exact end (it relied on humans for such things as parking.) ''
Which is weird, because I would have thought that parking would be one thing an automated system should be able to do better than any human. And it is much easier to make it failsafe than most parts of driving: When in doubt, hit the brakes.
'' The moment you get a human interacting with computer driven cars, all chaos will result. No computer system, radar system, and automated response system can anticipate a drunk human driver swerving across 6 lanes of traffic at 100 mph in order to make an exit. ''
It doesn't have to anticipate anything. A good radar system should be able to keep track of the location, speed and direction of that car with higher precision than you, calculate its course, and take evasive action faster and more precisely than you can, while taking into consideration where other cars are as well. Assuming that some other cars are automated as well and some aren't, and the automated ones know who is automated and who is not, all those cars could act coordinated as well (like everyone not directly involved making space for those closer to the action).
'' This would be a great advancement in driving for the elderly. Just imagine, a car that plows through a farmers market automatically! They should beta test it in Florida. ''
I've seen this attitude being posted here several times. Just yesterday, four children (four, eight, and two fifteen your old) died in a car accident about four miles away from my home. The oldest was driving. They hit the car of an elderly couple frontally; the kids were on the wrong side of the road. The people in the other car had light injuries only, probably because they were wearing seatbelts.
During the year-2000 craze, one insurance company in Britain reported that their computer software had always been able to handle customers who were born in the century before the current century (they had to, because the system was started around 1960 and there were plenty of pensioners born in the previous century at that time), but they figured out that to change it to handle customers two centuries before the current one would be a significant, expensive, and dangerous change.
On the other hand, they only had half a dozen customers born in the 19th century and still alive in the year 2000, so they removed their records from the computer system and one man handled their accounts manually. Much cheaper that way.
Probably the same here. Something like a database with a nine digit field for "donations to charity". Much cheaper to handle Bill Gates' tax by hand then to make a software change just for one person.
'' I think the "pay $4-5 to make it all legal" idea would only work if all record labels participated, and all ISPs participated. You'd have to basically force every ISP to add this "music-download tax," and implement it across the board... otherwise customers are going to be flocking to the competition that doesn't include this tax, and continue downloading things for free. ''
Instead of making it a "download-tax", why not make it a license to posess music, no matter where it comes from? My model would be something like iTunes, where a song gets marked to a specific owner, but if you pay the license tax, your computer plays music, no matter who the owner is. You would be allowed to copy anyone's music without restriction, and your computer plays it as long as the license tax is paid. If you stop paying, your computer stops playing. Still, nobody could sue you because the music is not playable.
P2P download qualities would improve, because instead of clueless kids, adults with respect for quality and either respect of the law or fear for consequences of breaking the law, would then start uploading.
'' However what happens to the tracking if the artists themselves decide to boost their income by having bots download their songs as often as possible? ''
That should be no problem. Lets say everybody pays x dollars a month, and lets say three dollars of everyone's payments is to be distributed to the artist. If all you download in one month is one Britney Spears song, she gets your three dollars. If you download her song tenthousand times, she gets three dollars. If you download ten different songs, everyone gets 30 cents, if you download 1000 different songs, everyone gets 0.3 cents.
The bot can only produce three dollars of income to an artist, but it needs an ISP address where more than three dollars are paid, so it is a net loss.
What would be dangerous is a virus that gets copied on many machines of paying consumers and downloads stuff they don't want.
'' My big fear is that the Disney brass learned their lessons from the Daimler-Chrysler scam: perform a "merger", let the new "partners" have their head for a while, but retain the real power. A year passes, and then there's a purge. I don't think there will be any Pixar people left at Disney after 1 year, and Jobs will leave in 18 months: substantially richer, but disgusted. ''
That would be rather stupid; paying over $7bn for essentially nothing. Not that it is impossible, but it would be incredibly stupid.
'' Actually I was discussing the UK situation. Here's a BBC article about how intelligent design appears well thought of in the UK (17% of population with 22% for creationism and 48% for evolution). For me, as a scientist, it was quite surprising to here this. AFAIK, the study was reasonable and the findings valid. ''
I found this very strange, because I doubt that anyone here in Britain has ever heard of "Intelligent Design", except for those who have heard of it in the context of brain-damaged religious twats who are even worse than Creationists.
Most probably thought of BMW or Apple Computer when they were asked about "Intelligent Design".
'' I've actually heard people from Mensa say that you can't study for the test. What bullshit. You can study the problem types etc. in the same way you can study or learn about anything. ''
People who take intelligence tests repeatedly will know how to maximise their scores; mostly by reducing the time spent at each question, thus increasing the percentage of wrong answers, but maximising the number of correct answers. Using a good strategy to use your time will already improve your scores.
'' ... I'm marking a pile of work handed in by 2nd year university students studying mathematics. Instead of giving numerical grades and seeing how good or bad the little chavlets have done we are required to avoid disheartening them by grading them on the following scale:
E - Excelent - Tried every question and got most of them mostly right.
G - Good - Tried most of the questions and got them mostly right.
S - Satisfactory - Tried some of the questions, didn't get everything wrong.
U - Unsatisfactory - Didn't bother to hand in, or got absolutely everything wrong
Yeah. This is at one of the better Universities around in the UK (though not one of the best). If schools are as bad (and they are) is it any wonder? ''
That was exactly the scale that was used when I studied maths 25 years ago. We had exactly one guy who managed to achieve the (E), and that guy was a pure genius. But then, trying all questions and getting them mostly right would have required four hours of writing at high speed without any break, and that alone made it difficult.
'' The Cell will be nothing special for typical integer workloads...in fact it will probably perform inferior to offerings from Intel and AMD. In fact any double-precision arithmatic is 10 times slower!! ''
Actually, Cell will be quite powerful on many double precision floating point problems. Throughput per processor is not very high, but latency is only twice the throughput, memory access is incredibly fast, all the supporting operations are incredibly fast, so it is very very easy to get performance that is equal to the theoretical limit of the chip.
You have to do some macro changes to split a task to run on more than one SPE, but that may be easier than trying to get peak performance out of a Pentium or G5 chip. Especially on a G5, you need massive parallelism at the instruction level to get peak performance.
'' As our tech increases we need less and less knowledge about it. My mom knew how to wire a fuse. I know how to screw in one. My kid knows how to throw a circuit breaker. Wich one of us would be more likely to be able to get a car moving when there is no replacement fuse available?
The kid. While both of you are looking for the fuse the kid realises that the car's got an onboard computer (black box) which has crashed and needs resetting, finds the reset switch (possibly a hole needing a paperclip), and holds it down for 3 seconds. ''
In my experience, the kid will be too stupid to realise anything. Second, he won't know that reset switches are hidden so they can't be pressed by accident. Third, he will be too stupid to find that hole. Fourth, even if he finds that hole, he will be too stupid to improvise something that can be used to press a reset button hidden in that hole.
'' .. its the source code to Office that we want. If it were possible to buy a competing office suite for Windows, Linux, Mac OS X, etc, that was 99% compatible with all office file formats, then we would have true competition in an open and level market. ''
The source code to Microsoft Office would not help you one bit. Anyone using that source code to build Office compatible applications would be sued for copyright violations, and rightfully so. Source code doesn't help to create compatible applications at all. Source code helps to find bugs, or to find vulnerabilities. We don't want to find bugs in the Microsoft Office source code, that is Microsoft's job.
What you need is a specification. Like the Open Document format; anybody can implement an application that is compatible with the Open Document format, no problem at all. Or all the specificiations for internet protocols, anybody can implement them. Or take the H.264 spec; that is a few hundred pages of highly technical documentation, and there are dozens of encoders and decoders available. Source code is useless.
'' Rather than worry about "opening up Windows, or decoupling IE from Windows" isn't it time that people start innovating and competing with Microsoft? ''
Yes, that is why Microsoft has to produce the documentation how to interact with their software, so that others can start innovating and competing. Without that documentation, the Microsoft monopoly prevents competition.
Now it seems that Microsoft is uncapable to provide documentation, so as a publicity stunt they propose to let others read their source code instead of documentation. But as any programmer knows (are there any here on Slashdot? ), SOURCE CODE IS NOT DOCUMENTATION.
'' I can guarantee you for the price you paid for that MacIntel (Intelitosh?), you'd be much better off spending half of the cash and buying from another retailer ''
So where exactly can you buy a notebook with Core Duo processor for half the price of an Intel MacBook Pro?
'' It's like porting Windows to MAC. Why bother? Most, if not all, MAC people don't want or require it. Granted there are a few people with exceptions but for the 99.9% why is so important? ''
The problem with your argument is that the number "99.9%" is not correct. It is more like 80% or 90%. And you forget that there are considerable numbers of Linux users and Windows users who would just like to buy a computer that is completely contained in the LCD screen (without having to run MacOS X).
'' This is a good argument for the need for open source election software. Another factor besides transparency is the need for peer review. Well designed software should be able to withstand some level of scrutiny from outside developers, and no set of code should be more well designed than that which is responsible for tabulating the record concerning the orderly transfer of power. ''
It is a good argument for publicly available source code, with a license to everybody and their dog to duplicate and compile the source code in order to examine its operation and its correctness.
Anybody, including Diebold's competitors, should be able and allowed to find and point out any flaws in their software, but there is no reason why competitors should be allowed to use the source code in their own products.
'' Not really. IF the person bought the songs off of ITunes and deleted the copy on his computer why wouldn't this be legit? ''
Do you believe in the tooth fairy?
Does anyone know how I can mod the article down as "stupid"?
No matter how difficult it is for someone outside Apple to make Windows XP boot, I would say the following to Apple:
1. If Apple were to sell Macintosh hardware with Windows XP preinstalled instead of MacOS X, then a considerable number of people would buy these machines. Not "considerable" as in "Dell goes out of business" but "considerable" as in a few percent of Apple revenues.
2. If Apple were to sell Macintosh hardware that can dual boot into MacOS X and Windows XP without any problems, a much greater number would buy those machines. Dual boot = run one, reboot, run the other, reboot...
3. If Apple were to sell Macintosh hardware that can run MacOS X and Windows XP simultaneously, they could sell tons of those. Even if "simultaneously" means that one of the OSes is in sleep mode while the other is running, with some form of communication so that cut&paste works.
'' I just wonder (assuming it works correctly only in FireFox for a clear example) how many IE users and Opera users and Safari users etc wouldn't just go to a different site as opposed to downloading, installing and then using another browser for *one site*. ''
You might also ask yourself (or your marketing people) how many people will get pissed off if your webpage doesn't work with their favorite browser, so that having a webpage that doesn't support browser X might be worse than having no webpage at all!
And what _absolutely_ pisses people off is if your webpage actively disables their browser. That is worse than a webpage that doesn't work properly. A webpage that doesn't work shows the developers were incompetent (they might just be under time pressure, but the user will assume incompetence); that is forgivable. A webpage that shuts me out means an active decision was made that the company doesn't want my customs. That is unforgivable.
Market caps of Pixar is $6.95bn. There will be no $7bn takeover. Maybe a $10bn takeover, but not $7bn.
'' Another reason Apple's alliance with Intel wasn't such a good idea. Should have gone AMD, Steve. ''
Any idea what the performance and price characteristics of ZRAM are, or are you just a clueless AMD fanboy?
'' On the Apple website right over the pic of the Macbook is a line reading "What's an Intel chip doing in a Mac? A lot more than it' ever done in a PC". That's pretty funny. I wonder what they'll say in 10 years when AMD has killed Intel (praying that it happens, of course)? I'd love a Mac with an Athlon and Win (games), Linux, Mac OS, and BSD all on the same machine. ''
I think you make the assumption that Intel is going to roll over and die. I think quite the opposite is going to happen.
Intel was an enormous company. They could build a chip based on the most crappiest design possible (the Itanium) and make it through sheer will power. Except that it doesn't work and it cost them dearly. They could build another chip based on the second most crappiest design possible (the Pentium 4) and make it through sheer will power. Except that it doesn't work and now they are falling behind.
But it seems that Intel has started realising that they are not invincible (at about the same time the AMD fanboys started thinking that Intel is dead); and for the first time in years they have an actual roadmap with promising products. And I think in Apple Computer they have found the best customer they could wish for: One that won't let them get away with crap like Itanium and P4.
'' How long will it now take before we start seeing iTunes Payola. Networks eager to get their shows on people's ipods will certainly be tempted to try to strike special deals to make their shows more easily obtainable. They may try to offer Apple cash, try to reduce the cost to consumers, or try to find ways to get their links on the front page. ''
It might happen, but it won't be as bad as on radio. On the radio, if they pay, you can hear it, if they don't pay, you can't hear it. On iTunes, what would happen: If they pay, it is advertised big on page 1. If they don't pay, it is still there on page 3 or 4.