Anyone able to point a finger at the legislation that enables them to do this? Or is there none, and they are just overstepping the mark?
There is none. However, the government won't prosecute police for criminal actions unless there is political pressure to do so. In the US there is currently no pressure for 'abuse of powers' issues, although there sometimes is for more conventional crimes (murder, assault, etc). People may not like Bush, but they're quite happy with their police state. And it's notoriously difficult to get a civil suit to stick for these things.
Ultimately, once the police have forcibly confiscated your camera and 'lost' the film, the pictures are gone and nobody's going to do anything about it. You can get a court to order them to give back what they still have, but stuff goes missing all the time and nobody is held accountable for it, so that's not particularly productive.
At any rate, most people speak at over 100 words a minute (my teenage daughters: 200 wpm), even a really good typist does, what? 40, 50 wpm?
Only in the conversational mode, with its sloppy, incomplete grammar and total absence of punctuation. Spoken English is not like written English. Speaking the written form is very slow. You either need a secretary to interpret, or you have to think ahead and plan your sentences. Typing's faster, largely because it offloads work from the part of your brain that handles speech to the part that handles motor control.
Voice recognition is not a viable dictation method.
Yes, I really mean that. Typing is faster than talking, for anybody who's any good at it. And not by a little - by a lot. If you want to get a document into a computer quickly, use a keyboard.
There are two practical applications for voice recognition that currently exist in the real world. The first is scenarios where it's not practical to reach a keyboard. Home automation, portable devices, stuff like that. Accuracy is important here, but it's also easy - there will be a small set of allowed phrases. That's much easier than full-English input. The second is probably the most important one - input for people who cannot use a keyboard, due to some form of disability. Accuracy here is considered a luxury. Nice to have but not really necessary.
So, it's not really a big deal. But you'll note that nothing here suggests that non-disabled people will have any use for voice input to Windows.
Realistically, this is the new minesweeper for Vista. It's a toy. Each new version of Windows has a few of these, because these toys are what sell non-OEM copies of the thing. People who go out and buy the new upgrade in a store, for however many thousands of dollars, are mostly doing it to get the new toys. Microsoft is aware of this and puts a lot of effort into making each successive version have new shiny toys; the rest is largely unimportant.
So Apple should be allowed to bundle voice recognition (not to mention handwriting recognition, calendering, mail, chat, photo management, video authoring, web development, and audio editing) but Microsoft shouldn't?
Correct, that is what antitrust law is all about. If you don't like it, feel free to campaign for antitrust law to be revoked. I'm sure a lot of megacorps would like that.
Whether or not it's currently a part of their long-term plans is obviously unknown, but in the past Gates has talked about how he would like the hardware to be free, and the user to pay a monthly rental charge for the software.
The frame rate of the average human eye is somewhere around 40 fps
Human eyes operate asynchronously, they don't have a fixed flicker rate like that. As such, they can spot things that happen much faster than 40fps, but only if those things are large enough - it's all dependant on the number of receptors that can detect the change.
For example: if you had a 50fps video camera, and you flashed an image at it for 1/50th of a second, then the camera probably wouldn't detect anything at all - the image most likely will not intersect with the camera's measuring interval. However, if you show this image to the human eye, then you will probably see something - but only a few of your receptors will be firing at the right time, so you'll see the image with limited resolution.
To further complicate things, there is the persistence of vision effect - that's why your brain translates a 25fps flickering image into motion, even though you're actually capable of perceiving the flicker. The details of how, when and why this operates are still not fully understood, so it is very difficult to pin down exactly what the limits of human perception are in this area. You can easily spot a lamp that flashes on for 1/50th of a second, but you cannot perceive the individual frames in a movie. Current psychologists and biologists have a range of pretty good approximations to describe the effects, but the precise details of what happens remain elusive.
I'm pretty sure the "bandwidth" between my eyes and brain is a little faster than even the best ethernet connection
The limiting factor is how fast your brain can interpret the signals to produce understanding. Most people seem to have a capacity for maybe a dozen bits of new information per lifetime.
And you DO NOT have such resolution as your monitor.
Actually you do. However, you don't normally press your eyeball against the glass. At a distance of about.5m, which is normal, your eyes don't have sufficient resolution any more. If you move your head up close to the glass, you should be able to perceive the individual pixels. It's important to remember that there are three dimensions here. The expected viewing distance determines the necessary dpi of the viewing device.
People would use commercial closed source software on Linux and Free OSS on Windows. I mean, wow. There really are people that will choose to use the best tool for the job.
I've seen people choose to use commercial software, volunteer-driven software, closed source, open source, Linux, Windows, and all kinds of other stuff, but I cannot recall the last time I saw somebody choose to use the best tool for the job.
Most people will choose the first match against these criteria, in order:
The thing they used last time
What they were instructed to use
What all the people near them are using (and other variations on peer pressure)
What they think is 'cool'
The one with the colour scheme they like best
Each of these might not have an answer, in which case they move on to the next item. Also if the answer is unavailable or otherwise cannot be used, they move on to the next item (so if the thing they used last time isn't installed here, or all the people near them are using different things, they'll skip those ones). If they fall off the bottom of the list, they just choose the cheapest (or pick randomly).
Some people will swap those last two. Some people will swap the first two (but only after repeatedly being told to). But I have never encountered a user who had "the one which gets the job done most effectively" on their list. Nobody seriously goes out there, looks at the available tools, and picks the one which will do the best job. Ever. (Reading reviews is not the same thing as evaluating a product, and it's probably an instance of either #3 or #4). The closest that I've seen is people evaluating all the options, discarding all the ones that don't actually work, and then picking the cheapest (which is so not the same thing).
No, the printed graphics don't contain such source code. They're just a bitmap image. However, the postscript file that has not yet been printed may well be under the GPL, even though, when executed, it generates an image that is not required to be GPLed.
They have *NO* way of knowing that there even *IS* an inner filesystem in the unused part of the outer filesystem.
Doesn't really matter. They can ensure that there is no hidden filesystem by overwriting all the unused parts of the disk with random data. Truecrypt may give you plausible deniability in court, but it's no good for smuggling data.
Should graphics drawn with The GIMP be GPL because they contain circles made by The GIMP's copyrighted code? If not, then what's the fundamental difference between The GIMP and phpBB that should restrict the output of one and not the other?
The difference is that the output of phpBB contains a large amount of data provided by phpBB (HTML gunk, etcetera). You can find copies of that data in the phpBB source tree. An image drawn with the GIMP, on the other hand, is only going to contain data supplied by the user (modulo file format overheads, which aren't copyrightable anyway). If I draw an image with the GIMP, you will not find any similarities between the image and the GIMP source tree.
The essential part is that the 'output' you're looking at contains data created by the author of the program, not by the user of the program. That means it's not really 'output' at all, but a copy of part of the program. If that is sufficiently complex and creative to be copyrightable, then the author of the program has some control over what you can do with it. Whether or not they choose to exercise that control is up to them.
For example, compilers such as gcc usually put non-trivial amounts of copyrightable data into their output - much of this can be found in libgcc.a. These aren't parts of the compiled program, they're parts of the implementation of the C language. The FSF could use this to exercise control over what you do with programs compiled with gcc, but they choose not to; the gcc license explicitly exempts this sort of thing from the terms of the GPL.
They'll just strip your car down to the frame, then reject your entry to the country and leave you standing there with a pile of mechanical parts. Customs is fucked up and has no proper appeals process or oversight of these things.
There is one thing that seems constant: The mix of successful, marginally successful, and just plain failed projects feels the same as ever, even though I'm positively sure that our knowledge of how to create software is much greater than it was.
Indeed. My take on all this is that people just don't care all that much about whether a given project is successful, or whether the product is any good. Sure, they'd like it to happen, but it's not all that important to them and they aren't going to go out of their way to make it happen. The industry seems to encourage this attitude.
After all, there will always be another project after it, so you'll still get paid. And if the product sucks, you can simply delude yourself that it didn't, by revising your objectives to match what was produced. Finishing the project is the only thing that counts, not how well it works or how much better it could have worked.
We must reduce the size of government if we're to avoid the slide into tyranny.
I've often thought that the US federal system is backwards anyway. It collects power in the hands of a few at the top who have little or no connection to the people that their actions affect. It would make far more sense if the federal government were the least powerful and local government were the superior power (over their constituency). The federal government would then adopt a role more like the UN, brokering treaties and agreements between the states instead of handing down badly formed decrees. If they made a stupid ignorant decision that wasn't appropriate for your area, you could then correct it at the state or county level.
This addiction to unnecessary consistancy is getting out of hand.
A lot of the newer mobile GPU (like GeForce Go) are capable of greatly reducing their overall consumption when their total demand is low. They ramp up when needed.
Mobile? The desktop ATI cards, including the twin-slot monoliths, have done this for years.
Do the math, it's not actually costing a person much more per month to go from 600 to a 1000 watt PSU
A 1000 watt PSU does not use 10/6 times as much power as a 600 watt PSU. There are two reasons for this:
The PSU only draws power proprotional to the load on it. The rating is the maximum draw, not the minimum. Two PSUs of equal efficiency but different ratings, supplying the same load, should draw the same amount of power.
The ratings are lies anyway. The manufacturers add up the numbers in ways that make NO SENSE AT ALL in order to get the largest numbers they can. They do this because it sells PSUs.
The single figure rating on a PSU tells you nothing at all about its capacity, efficiency, power drain, or suitability for your purpose. You need the per-rail ratings (usually printed on a sticker on the PSU) to get useful information, it's about a dozen numbers. This would be true even if the single figure rating was not a lie.
And furthermore, civil trials are much more shades of gray. If a person does something that is technically illegal but in practice causes no harm, the verdict will be 'liable' but the judge will apply the minimum sentence (a few dollars and an instruction not to do it again). No judge in the world is going to make you pay the maximum fine for downloading one movie that you've already paid for. They'll probably be pretty irritated at the MPAA for filling their docket with this nonsense; judges are very busy people and get short-tempered with cases that don't accomplish anything.
I don't know about the MPAA ones, but a few cases like these (RIAA perhaps? I forget) have gone to trial and the persecutor lost. Those were all absurd cases though (like the defendant didn't own a computer or similar) so it was just *stupid* to take them to court in the first place.
This whole extortion racket is an insulting joke, and really shouldn't be legal.
My understanding is that a British ISP basically told the European equivalent of the MPAA/RIAA to GTFO unless they could provide convincing evidence of the accused users downloading specific files at specific times.
Here in the UK we have some nice laws that require the ISPs to do this. If the ISP had handed over the information, then not only could their customers have sued them for it (and won), but it would also have been a criminal act. You cannot pass on personal details about your customers to somebody else without your customer's permission or a court order. There are no exceptions that will let them get out of this one. The company is responsible for obtaining the necessary permission, and they can't just hide it in the small print - they must offer you the choice, they cannot require you to grant it. (You'll never get this in the US because it would basically put Choicepoint out of business).
you would think that someone by now would have developed a decent stand-alone box that could transfer video from your computer to your TV over your network in a variety of formats REALIABLY. So far they all feature either piss-poor performance or are VERY picky and flaky about the video formats they'll play.
This is a BIG PROBLEM. People never quite realise how big a problem it is. The problem being that video is very very big.
The video bandwidth to your display device is a minimum of several hundred Mb/sec. That's far more than you can fit on a 100Mbit network link (which manages about 10Mb/sec at best). It's equivalent to something like three or four Gbit network links running at maximum efficiency. The kind of hardware that can haul that much data is expensive - it's not something you're going to be able to cram into a small stand-alone box with today's hardware, you're looking at a full server-class system with multiple PCI busses.
So, we can either reduce the video quality (giving you that piss-poor performance - this is a common approach), or we can use lossy compression (also giving piss-poor performance because the sending box does not have the CPU power to do very good video encoding in real time, and the developers don't have the software to do it very well either - most good encoding is done much slower than real time). Some combination of these two effects is what all those first bunch of devices you're complaining about do.
The alternative approach is to say that the box doesn't compress the video stream itself - instead, it receives a video stream from the user, pre-compressed. But of course, the receiving box has to be able to decompress that. You can either do it in software or hardware. Software can decompress just about anything with xine or vlc, but needs a 'real' CPU to do it, which is quite a problem. Hardware is limited to one codec per chip you can cram in there, so you're only going to get one or two choices. That's the problem with the other set.
About the only way this is ever going to work is if hardware improves faster than people's demand for increases in video quality, to the point where it's practical to do the decompressing in software. Since bandwidth demands normally increase to consume all available capacity, this is unlikely. Otherwise, you're going to be looking at solutions that involve putting a real computer with a TV-out device (like mythtv) next to your TV system.
I suggest you consider an alternative approach: get a good projector that's designed to take DVI input, and paint one of your walls white (or hang a screen). It's more likely to work than faffing around with a TV set.
All the idiots who would use this while driving are currently driving around looking at a map which they're holding on one hand, instead of watching the road and controlling their vehicle. The number of people who get ticketed for that every day is horrifying. Changing the thing they're holding isn't going to make a lot of difference.
Smart people give the map, or the hardware, to their passenger. Or they pull over to consult it if they don't have a passenger.
The only thing that seems worthy of notice in Google labs is google sets, which has that 'next gen AI search' feeling to it.
Any postgrad student in machine learning should know basically how to build one of those. It's not 'next gen', it's 'last gen'. It's in the damned textbook. Figuring out new methods for doing it smarter is a subject of current research.
Traditionally, IBM and Motorola/Freescale only announced a G3/G4/G5 processor whenever Apple was ready to introduce a new model using it - since Apple was the largest PPC system maker, they had some clout in that area.
Well, that's wrong somehow... I think you either meant to say "Apple was the largest G3/G4/G5 system maker" or "Apple was the largest user of the PPC in desktop computers". The largest PPC system maker would be hard to pin down, but my bet would be either one of the car or printer manufacturers. PPC is all about embedded systems, Apple's use of them was just convinient fallout.
Obviously IBM/Motorola only bothered to announce the G3/G4/G5 chips when Apple was ready to introduce a new model using them, because those are names for variations on existing PPC chip designs that were designed and produced on contract with Apple explicitly for their use. While they do refer to unique chips, they were all fairly minor variations (mostly just increased specs) on chips that the relevant maker had already created. The G5 is just a variation on IBM's established POWER4 line, for example. Apple chips have always been evolutionary, not revolutionary (even back in the 680x0 days).
The only thing that's changed here is that you happen to be reading the press in which Intel chip announcements are published, while I'm betting that you never heard about all the developments in PPC chips over the past 20 years or so, except for the ones published by Apple.
There's a big list of some of the stuff that uses PPC over here: http://en.wikipedia.org/wiki/Powerpc#Implementatio ns. It's probably a toss-up between PPC and ARM as to which is the most common microprocessor architecture on the planet (this is notoriously difficult to estimate with any precision and there's the perennial question about whether or not you should count the PICs - but if you don't, most people will agree that it's one of those two, although they'll often argue about which). Apple has never been as important as their fanboys would like people to believe.
Slashdot Mirror
There is none. However, the government won't prosecute police for criminal actions unless there is political pressure to do so. In the US there is currently no pressure for 'abuse of powers' issues, although there sometimes is for more conventional crimes (murder, assault, etc). People may not like Bush, but they're quite happy with their police state. And it's notoriously difficult to get a civil suit to stick for these things.
Ultimately, once the police have forcibly confiscated your camera and 'lost' the film, the pictures are gone and nobody's going to do anything about it. You can get a court to order them to give back what they still have, but stuff goes missing all the time and nobody is held accountable for it, so that's not particularly productive.
Only in the conversational mode, with its sloppy, incomplete grammar and total absence of punctuation. Spoken English is not like written English. Speaking the written form is very slow. You either need a secretary to interpret, or you have to think ahead and plan your sentences. Typing's faster, largely because it offloads work from the part of your brain that handles speech to the part that handles motor control.
Voice recognition is not a viable dictation method.
Yes, I really mean that. Typing is faster than talking, for anybody who's any good at it. And not by a little - by a lot. If you want to get a document into a computer quickly, use a keyboard.
There are two practical applications for voice recognition that currently exist in the real world. The first is scenarios where it's not practical to reach a keyboard. Home automation, portable devices, stuff like that. Accuracy is important here, but it's also easy - there will be a small set of allowed phrases. That's much easier than full-English input. The second is probably the most important one - input for people who cannot use a keyboard, due to some form of disability. Accuracy here is considered a luxury. Nice to have but not really necessary.
So, it's not really a big deal. But you'll note that nothing here suggests that non-disabled people will have any use for voice input to Windows.
Realistically, this is the new minesweeper for Vista. It's a toy. Each new version of Windows has a few of these, because these toys are what sell non-OEM copies of the thing. People who go out and buy the new upgrade in a store, for however many thousands of dollars, are mostly doing it to get the new toys. Microsoft is aware of this and puts a lot of effort into making each successive version have new shiny toys; the rest is largely unimportant.
Correct, that is what antitrust law is all about. If you don't like it, feel free to campaign for antitrust law to be revoked. I'm sure a lot of megacorps would like that.
Whether or not it's currently a part of their long-term plans is obviously unknown, but in the past Gates has talked about how he would like the hardware to be free, and the user to pay a monthly rental charge for the software.
Human eyes operate asynchronously, they don't have a fixed flicker rate like that. As such, they can spot things that happen much faster than 40fps, but only if those things are large enough - it's all dependant on the number of receptors that can detect the change.
For example: if you had a 50fps video camera, and you flashed an image at it for 1/50th of a second, then the camera probably wouldn't detect anything at all - the image most likely will not intersect with the camera's measuring interval. However, if you show this image to the human eye, then you will probably see something - but only a few of your receptors will be firing at the right time, so you'll see the image with limited resolution.
To further complicate things, there is the persistence of vision effect - that's why your brain translates a 25fps flickering image into motion, even though you're actually capable of perceiving the flicker. The details of how, when and why this operates are still not fully understood, so it is very difficult to pin down exactly what the limits of human perception are in this area. You can easily spot a lamp that flashes on for 1/50th of a second, but you cannot perceive the individual frames in a movie. Current psychologists and biologists have a range of pretty good approximations to describe the effects, but the precise details of what happens remain elusive.
The limiting factor is how fast your brain can interpret the signals to produce understanding. Most people seem to have a capacity for maybe a dozen bits of new information per lifetime.
Actually you do. However, you don't normally press your eyeball against the glass. At a distance of about
I've seen people choose to use commercial software, volunteer-driven software, closed source, open source, Linux, Windows, and all kinds of other stuff, but I cannot recall the last time I saw somebody choose to use the best tool for the job.
Most people will choose the first match against these criteria, in order:
Each of these might not have an answer, in which case they move on to the next item. Also if the answer is unavailable or otherwise cannot be used, they move on to the next item (so if the thing they used last time isn't installed here, or all the people near them are using different things, they'll skip those ones). If they fall off the bottom of the list, they just choose the cheapest (or pick randomly).
Some people will swap those last two. Some people will swap the first two (but only after repeatedly being told to). But I have never encountered a user who had "the one which gets the job done most effectively" on their list. Nobody seriously goes out there, looks at the available tools, and picks the one which will do the best job. Ever. (Reading reviews is not the same thing as evaluating a product, and it's probably an instance of either #3 or #4). The closest that I've seen is people evaluating all the options, discarding all the ones that don't actually work, and then picking the cheapest (which is so not the same thing).
If Windows was free then it wouldn't be such an unmitigated pile of agony. The worst of the crap would have been kicked out of it years ago.
No, the printed graphics don't contain such source code. They're just a bitmap image. However, the postscript file that has not yet been printed may well be under the GPL, even though, when executed, it generates an image that is not required to be GPLed.
That does not mean that they won't simply require the TPM to validate the system as running Windows.
Doesn't really matter. They can ensure that there is no hidden filesystem by overwriting all the unused parts of the disk with random data. Truecrypt may give you plausible deniability in court, but it's no good for smuggling data.
The difference is that the output of phpBB contains a large amount of data provided by phpBB (HTML gunk, etcetera). You can find copies of that data in the phpBB source tree. An image drawn with the GIMP, on the other hand, is only going to contain data supplied by the user (modulo file format overheads, which aren't copyrightable anyway). If I draw an image with the GIMP, you will not find any similarities between the image and the GIMP source tree.
The essential part is that the 'output' you're looking at contains data created by the author of the program, not by the user of the program. That means it's not really 'output' at all, but a copy of part of the program. If that is sufficiently complex and creative to be copyrightable, then the author of the program has some control over what you can do with it. Whether or not they choose to exercise that control is up to them.
For example, compilers such as gcc usually put non-trivial amounts of copyrightable data into their output - much of this can be found in libgcc.a. These aren't parts of the compiled program, they're parts of the implementation of the C language. The FSF could use this to exercise control over what you do with programs compiled with gcc, but they choose not to; the gcc license explicitly exempts this sort of thing from the terms of the GPL.
They'll just strip your car down to the frame, then reject your entry to the country and leave you standing there with a pile of mechanical parts. Customs is fucked up and has no proper appeals process or oversight of these things.
Indeed. My take on all this is that people just don't care all that much about whether a given project is successful, or whether the product is any good. Sure, they'd like it to happen, but it's not all that important to them and they aren't going to go out of their way to make it happen. The industry seems to encourage this attitude.
After all, there will always be another project after it, so you'll still get paid. And if the product sucks, you can simply delude yourself that it didn't, by revising your objectives to match what was produced. Finishing the project is the only thing that counts, not how well it works or how much better it could have worked.
I've often thought that the US federal system is backwards anyway. It collects power in the hands of a few at the top who have little or no connection to the people that their actions affect. It would make far more sense if the federal government were the least powerful and local government were the superior power (over their constituency). The federal government would then adopt a role more like the UN, brokering treaties and agreements between the states instead of handing down badly formed decrees. If they made a stupid ignorant decision that wasn't appropriate for your area, you could then correct it at the state or county level.
This addiction to unnecessary consistancy is getting out of hand.
Mobile? The desktop ATI cards, including the twin-slot monoliths, have done this for years.
A 1000 watt PSU does not use 10/6 times as much power as a 600 watt PSU. There are two reasons for this:
And furthermore, civil trials are much more shades of gray. If a person does something that is technically illegal but in practice causes no harm, the verdict will be 'liable' but the judge will apply the minimum sentence (a few dollars and an instruction not to do it again). No judge in the world is going to make you pay the maximum fine for downloading one movie that you've already paid for. They'll probably be pretty irritated at the MPAA for filling their docket with this nonsense; judges are very busy people and get short-tempered with cases that don't accomplish anything.
I don't know about the MPAA ones, but a few cases like these (RIAA perhaps? I forget) have gone to trial and the persecutor lost. Those were all absurd cases though (like the defendant didn't own a computer or similar) so it was just *stupid* to take them to court in the first place.
This whole extortion racket is an insulting joke, and really shouldn't be legal.
Here in the UK we have some nice laws that require the ISPs to do this. If the ISP had handed over the information, then not only could their customers have sued them for it (and won), but it would also have been a criminal act. You cannot pass on personal details about your customers to somebody else without your customer's permission or a court order. There are no exceptions that will let them get out of this one. The company is responsible for obtaining the necessary permission, and they can't just hide it in the small print - they must offer you the choice, they cannot require you to grant it. (You'll never get this in the US because it would basically put Choicepoint out of business).
This is a BIG PROBLEM. People never quite realise how big a problem it is. The problem being that video is very very big.
The video bandwidth to your display device is a minimum of several hundred Mb/sec. That's far more than you can fit on a 100Mbit network link (which manages about 10Mb/sec at best). It's equivalent to something like three or four Gbit network links running at maximum efficiency. The kind of hardware that can haul that much data is expensive - it's not something you're going to be able to cram into a small stand-alone box with today's hardware, you're looking at a full server-class system with multiple PCI busses.
So, we can either reduce the video quality (giving you that piss-poor performance - this is a common approach), or we can use lossy compression (also giving piss-poor performance because the sending box does not have the CPU power to do very good video encoding in real time, and the developers don't have the software to do it very well either - most good encoding is done much slower than real time). Some combination of these two effects is what all those first bunch of devices you're complaining about do.
The alternative approach is to say that the box doesn't compress the video stream itself - instead, it receives a video stream from the user, pre-compressed. But of course, the receiving box has to be able to decompress that. You can either do it in software or hardware. Software can decompress just about anything with xine or vlc, but needs a 'real' CPU to do it, which is quite a problem. Hardware is limited to one codec per chip you can cram in there, so you're only going to get one or two choices. That's the problem with the other set.
About the only way this is ever going to work is if hardware improves faster than people's demand for increases in video quality, to the point where it's practical to do the decompressing in software. Since bandwidth demands normally increase to consume all available capacity, this is unlikely. Otherwise, you're going to be looking at solutions that involve putting a real computer with a TV-out device (like mythtv) next to your TV system.
I suggest you consider an alternative approach: get a good projector that's designed to take DVI input, and paint one of your walls white (or hang a screen). It's more likely to work than faffing around with a TV set.
All the idiots who would use this while driving are currently driving around looking at a map which they're holding on one hand, instead of watching the road and controlling their vehicle. The number of people who get ticketed for that every day is horrifying. Changing the thing they're holding isn't going to make a lot of difference.
Smart people give the map, or the hardware, to their passenger. Or they pull over to consult it if they don't have a passenger.
And anybody who tries to look it up goes on the terrorist watch lists.
Any postgrad student in machine learning should know basically how to build one of those. It's not 'next gen', it's 'last gen'. It's in the damned textbook. Figuring out new methods for doing it smarter is a subject of current research.
Well, that's wrong somehow... I think you either meant to say "Apple was the largest G3/G4/G5 system maker" or "Apple was the largest user of the PPC in desktop computers". The largest PPC system maker would be hard to pin down, but my bet would be either one of the car or printer manufacturers. PPC is all about embedded systems, Apple's use of them was just convinient fallout.
Obviously IBM/Motorola only bothered to announce the G3/G4/G5 chips when Apple was ready to introduce a new model using them, because those are names for variations on existing PPC chip designs that were designed and produced on contract with Apple explicitly for their use. While they do refer to unique chips, they were all fairly minor variations (mostly just increased specs) on chips that the relevant maker had already created. The G5 is just a variation on IBM's established POWER4 line, for example. Apple chips have always been evolutionary, not revolutionary (even back in the 680x0 days).
The only thing that's changed here is that you happen to be reading the press in which Intel chip announcements are published, while I'm betting that you never heard about all the developments in PPC chips over the past 20 years or so, except for the ones published by Apple.
There's a big list of some of the stuff that uses PPC over here: http://en.wikipedia.org/wiki/Powerpc#Implementati