So you're saying that when you're using, say, Winamp to play an Ogg file that you downloaded from the net, the CPU will be able to identify the song as "copyrighted" and will crash your system? Right...
Let me guess, the CPU includes a database with all copyrighted songs on Earth (plus all movies, all software, etc., and is able to compare all data that runs through it with this huge database)...? Then I guess we won't even need to rip our CDs anymore, we just have to find the song's address in this internal database...
For any piece of hardware to automagically identify a file as "copyrighted", it would need to have, hardcoded in it, intimate knowledge of the file format and decoding algorithms and it would need the file itself to have some detail that identified it as being copyrighted.
If you do it in software (ie, in the OS) the first part becomes easier. But the second part is still relevant. For a file to be tagged as "copyrighted", that tag must be added at some stage. In the specific case of music files, it would have to be added by the encoder. Do you think Microsoft would ever manage to convince, say, Xiph.org or Xing to support that kind of initiative?
I won't bother with the fact that this piece of Slashdot "news" has already been denied by AMD. I just ask you to think about it for a second. Do you think AMD would make a CPU that would refuse to run all software except Microsoft's? Especially a CPU that's aimed at small servers (hint: small servers don't normally run Windows)? And at a time when Asia (China particularly) is the fastest growing market for them (hint: China doesn't like Microsoft)?
And who cares about the RIAA or even american regulations? The world is a big place, you know? AMD's fabs are in Europe and Asia, and those two are their main markets (nearly 40% share, against only 15% in the USA).
Stop seeing conspiracies everywhere and start thinking about things for a change.
Do you think the CPU has any idea if it's rendering a 3D scene or playing an MP3 or decompressing a JPEG or spell-checking a document? Let alone know if the files are copyrighted or not.
That's up to the OS and individual applications to (try to) determine and enforce.
The only thing that changes in a "secure" CPU is the fact that programs and (especially) the operating system will be able to identify that CPU uniquely (by a serial number), similar to what the Pentium III already does (but you can turn it off on the PIII, and I think also on the P4). Then some programs will probably refuse to play certain files if they're not tagged with that CPU id. Ex., if you buy a "secure" song on-line, or if you rip one of your CD's, it probably won't play on your friend's computer (or on yours if you change the CPU, and that's why MS needs to work with CPU makers, to make sure the CPU id can be managed by the OS).
The rest is just a lot of marketing hype to get money out of the RIAA and similar associations. "See, we are working on this 'secure' hardware that won't play copyrighted music, but it's very expensive to develop and we really don't have enough money, what with this recession and everything, so if you could fork over a couple of million, we'd appreciate it..."
It's a potential gold mine for (some) IT companies, just like the Y2K bug.
Yes, but it's much slower than loading the values into the CPU. Current x86 FPUs can handle 64- and 80-bit values directly ("double" and "long double", in C), and that's enough for most situations.
I mentioned 4 terabytes because the current limit is 4 gigabytes, and because the processor that's likely to become the first 64-bit "home" CPU (AMD's Sledgehammer) will start with a 40-bit limit (1 terabyte), and later be expanded to 48-bit addressing (roughly 280 terabytes). And I'm willing to bet "home" versions of Windows will be locked at something lower (to force you to buy the "server" version if you want to use all your memory).
If you want the exact number that full 64-bit addressing can give you (2^64) just see my first message (it's 18446744073709551615 bytes).
And anyway, this doesn't apply to drives or files (that depends on the drive interface and the file system), only to memory. Current ATA-133 drives can go up to about 144 petabytes.
If one of your pr0n pictures is over 4 TB, I suspect you need some "compress your penis" pills.
The jump from 640 KB (which was an OS limit, not a limit on the memory controller itself) to the amount of memory found on most computers today (128MB to 1 GB) was about an 800-fold increase. And that took well over ten years.
Going from 32-bit to 64-bit addressing raises the memory limit four thousand million times (that's four billion if you're american). That's eighteen petabytes (although early models of AMD's "Sledgehammer" will be limited to 40- and 48-bit addressing, which is still a lot, especially when you consider that, on SMP systems, each CPU can have its own addressing space).
It's highly unlikely home computers will even need to go beyond 4 GB in the next couple of years. I think it's pretty safe to say they won't need more than 1 terabyte (40-bit addressing) within the next ten.
I mean, there's a limit even to how bloated MS Windows can get...
I think you've been playing with consoles a bit too much. 64-bit here refers to the size of general-purpose registers and memory addressing.
With 64-bit you can address over 4 terabytes. Do you feel the need for more than that?
You can also work with integers up to 18.446.744.073.709.551.615, and floating-point numbers up to 1.7976931348623158 E+308. Feel the need for more than that?
There are wider registers in the CPU (such as the dedicated SSE2 or Altivec registers), but for normal operation I think 64-bit should keep us going for quite a few years.
That could also have been Jedi hell. There were flames involved. But I agree that joining Yoda, Vader and Obi-Wan in that scene in RotJ sort of implies they're all on the same side (which is odd in itself, since Vader killed half the galaxy). Unless, of course, that was just a vision that Luke had (ie, "his mentors", whatever).
But there's still no way Yoda can be both wise and good (why else would he keep making the wrong decisions and deceiving people?) and no way Vader can be both powerful and evil (how else could he not have spotted that Leia was his daughter?). But also no way that Vader could be a good guy in disguise (or he wouldn't have killed all those people). The characters are just gridlocked.
I still think my Dallas ("it was all a - bad - dream") theory could bring some consistency to the SW plot.
I was referring to Vader. The idea was that Vader (and possibly Obi-Wan) were working against Yoda and Palpatine (the Emperor). That would explain why Yoda keeps lying, why Obi-Wan doesn't seem to remember the droids, why Vader doesn't seem to notice that Leia is a Jedi, why Luke was "hidden" in plain sight, and so on.
It is not a "terrific plot", and it's not terribly original, either (I've been hearing similar theories since Episode 1 was released). As Brin himself points out, it can't work for one simple reason: Darth Vader killed far too many people.
I mean, maybe he could have been forced to kill a couple of people just to convince the emperor he really was on the dark side. A tough moral choice but in the end he decided the sacrifice of a few was necessary to save the majority, sort of thing. But when the undercover agent kills more people than the godfather, then there's no way you can make him turn out to be a good guy after all.
No, I think Episodes 2 and 1 really did more damage than Episode 3 can possibly fix. And what's worse, if you see them in the new order, the original movies (which are excellent) will now seem to be full of contradictions. Lucas might as well have made a completely separate story (same universe, perhaps, but not directly connected to the original movies).
After I saw Episode 1, a friend of mine asked me "so, how did you find the plot?" and I said "I didn't".
The only way to fix things is if, in Episode 3, J.R. wakes up and all this turns out to have been a dream.
I'll skip the silly sensationalism. The drive won't "implode", some parts just risk melting because the laser isn't properly calibrated in the old firmware.
But the interesting question is: does this mean that, with this new 4x media, my DVR-103 (which is a 2x DVD-R recorder) will record at 4x? If so, that's very nice to know.
I already have the new firmware, by the way. Unlike the old updater (that required you to connect the drive as the primary IDE slave), the new one will automatically detect and update all your drives, no matter how they're connected (even via IEEE-1394).
Now if only HP and Sony could do the same to make their DVD+RW recorders work with DVD+R as they promised...;-)
Earth is about to be hit by a huge asteroid and the only way to save mankind it is to use something with a big enough mass to push the asteroid off its course.
That's when CowboyNeal springs into action (powered by the world's largest catapult, naturally).
In a dramatic sequence, he misses the asteroid completely. Humanity prepares to be obliterated (shots of Arafat and Sharon hugging, Bush in bed with Osama, etc.). But the scientists notice that, without CowboyNeal's mass, Earth's orbit has changed slightly, and the asteroid narrowly misses.
The film ends with a shot fo CowboyNeal landing on the moon and eating it.
Normal CCD pixels are only sensitive to one light component (usually R, G and B). So when you have, say, 3 megapixels, you only have 1 million red samples, 1 million green samples and 1 million blue samples (actually this isn't exactly right, because green is usually given priority, so on most 3 megapixel cameras you have 1.5 million green samples, 750k red samples and 750k blue samples).
Basically what this means is that if you want a clean-looking image, with no coloured halos in high-contrast areas, you should have at least 3 times as many CCD pixels as you want to have in the final image.
Also, because CCDs can be quite "noisy" under low light, it helps to have more samples so that when you average them you get a more stable colour.
As a rule of thumb, with mid-range CCDs, you'll be able to get pretty good pictures at 1/4 of the camera's resolution (ie, if the camera shoots at 2048x1536, the picture will look better - without red/blue outlines and with a lot less noise - if you resample it to 1024x768).
Some cameras will do this automatically when you select a lower resolution (ie, shoot at maximum, then resample), others will capture the image directly at the lowest resolution, using only half the CCD's pixels. If you're not sure how your camera works, you should always shoot at the maximum resolution and then downsample in an image editing program if you want to get rid of halos / noise.
RMN ~~~
Re:3500 year old technology
on
Awari Solved
·
· Score: 2
They also built the pyramids, calculated the radius of the Earth, and so on.
The human brain hasn't changed that much in 3500 years; we were a pretty smart bunch back then. And of course, we didn't have computers so we actually used it now and then.
All image capture is analog. It can be electronic (CCDs) or chemical (film), but there's always an element that "charges up" as it's hit by photons.
Compared to film, CCDs are extremely low-res (top quality 35 mm film has resolution equivalent to a 50 megapixel CCD) but, more importantly, they're slow. At very short exposure times, CCDs have so much noise that the final result is useless. The problem isn't the transfer rate, it's the time the CCDs take to "charge up" to meaningful values.
There is one alternative: use very large CCDs. The larger the CCD, the more light hits it, and the faster it can charge. But larger CCDs are more expensive and require special lenses.
Recording directly to digital does have one big advantage: you don't have to pay for the film. But the CCDs simply aren't up to film quality yet (and probably won't be for another 5 years or so). So the solution is simple: shoot on film, then digitise it.
That has to be the strangest looking Hammer I've ever seen. Doesn't even have a handle. On the plus side, it does seem to come with a lot of built-in nails.
Slashdot Mirror
...if the CD is free?
RMN
~~~
Yes, and computers will also read our lips and lock us out of our spaceships.
RMN
~~~
So you're saying that when you're using, say, Winamp to play an Ogg file that you downloaded from the net, the CPU will be able to identify the song as "copyrighted" and will crash your system? Right...
Let me guess, the CPU includes a database with all copyrighted songs on Earth (plus all movies, all software, etc., and is able to compare all data that runs through it with this huge database)...? Then I guess we won't even need to rip our CDs anymore, we just have to find the song's address in this internal database...
For any piece of hardware to automagically identify a file as "copyrighted", it would need to have, hardcoded in it, intimate knowledge of the file format and decoding algorithms and it would need the file itself to have some detail that identified it as being copyrighted.
If you do it in software (ie, in the OS) the first part becomes easier. But the second part is still relevant. For a file to be tagged as "copyrighted", that tag must be added at some stage. In the specific case of music files, it would have to be added by the encoder. Do you think Microsoft would ever manage to convince, say, Xiph.org or Xing to support that kind of initiative?
I won't bother with the fact that this piece of Slashdot "news" has already been denied by AMD. I just ask you to think about it for a second. Do you think AMD would make a CPU that would refuse to run all software except Microsoft's? Especially a CPU that's aimed at small servers (hint: small servers don't normally run Windows)? And at a time when Asia (China particularly) is the fastest growing market for them (hint: China doesn't like Microsoft)?
And who cares about the RIAA or even american regulations? The world is a big place, you know? AMD's fabs are in Europe and Asia, and those two are their main markets (nearly 40% share, against only 15% in the USA).
Stop seeing conspiracies everywhere and start thinking about things for a change.
RMN
~~~
Do you think the CPU has any idea if it's rendering a 3D scene or playing an MP3 or decompressing a JPEG or spell-checking a document? Let alone know if the files are copyrighted or not.
That's up to the OS and individual applications to (try to) determine and enforce.
The only thing that changes in a "secure" CPU is the fact that programs and (especially) the operating system will be able to identify that CPU uniquely (by a serial number), similar to what the Pentium III already does (but you can turn it off on the PIII, and I think also on the P4). Then some programs will probably refuse to play certain files if they're not tagged with that CPU id. Ex., if you buy a "secure" song on-line, or if you rip one of your CD's, it probably won't play on your friend's computer (or on yours if you change the CPU, and that's why MS needs to work with CPU makers, to make sure the CPU id can be managed by the OS).
The rest is just a lot of marketing hype to get money out of the RIAA and similar associations. "See, we are working on this 'secure' hardware that won't play copyrighted music, but it's very expensive to develop and we really don't have enough money, what with this recession and everything, so if you could fork over a couple of million, we'd appreciate it..."
It's a potential gold mine for (some) IT companies, just like the Y2K bug.
RMN
~~~
Yes, but it's much slower than loading the values into the CPU. Current x86 FPUs can handle 64- and 80-bit values directly ("double" and "long double", in C), and that's enough for most situations.
RMN
~~~
I mentioned 4 terabytes because the current limit is 4 gigabytes, and because the processor that's likely to become the first 64-bit "home" CPU (AMD's Sledgehammer) will start with a 40-bit limit (1 terabyte), and later be expanded to 48-bit addressing (roughly 280 terabytes). And I'm willing to bet "home" versions of Windows will be locked at something lower (to force you to buy the "server" version if you want to use all your memory).
If you want the exact number that full 64-bit addressing can give you (2^64) just see my first message (it's 18446744073709551615 bytes).
And anyway, this doesn't apply to drives or files (that depends on the drive interface and the file system), only to memory. Current ATA-133 drives can go up to about 144 petabytes.
If one of your pr0n pictures is over 4 TB, I suspect you need some "compress your penis" pills.
RMN
~~~
The jump from 640 KB (which was an OS limit, not a limit on the memory controller itself) to the amount of memory found on most computers today (128MB to 1 GB) was about an 800-fold increase. And that took well over ten years.
Going from 32-bit to 64-bit addressing raises the memory limit four thousand million times (that's four billion if you're american). That's eighteen petabytes (although early models of AMD's "Sledgehammer" will be limited to 40- and 48-bit addressing, which is still a lot, especially when you consider that, on SMP systems, each CPU can have its own addressing space).
It's highly unlikely home computers will even need to go beyond 4 GB in the next couple of years. I think it's pretty safe to say they won't need more than 1 terabyte (40-bit addressing) within the next ten.
I mean, there's a limit even to how bloated MS Windows can get...
RMN
~~~
I think you've been playing with consoles a bit too much. 64-bit here refers to the size of general-purpose registers and memory addressing.
With 64-bit you can address over 4 terabytes. Do you feel the need for more than that?
You can also work with integers up to 18.446.744.073.709.551.615, and floating-point numbers up to 1.7976931348623158 E+308. Feel the need for more than that?
There are wider registers in the CPU (such as the dedicated SSE2 or Altivec registers), but for normal operation I think 64-bit should keep us going for quite a few years.
RMN
~~~
That could also have been Jedi hell. There were flames involved. But I agree that joining Yoda, Vader and Obi-Wan in that scene in RotJ sort of implies they're all on the same side (which is odd in itself, since Vader killed half the galaxy). Unless, of course, that was just a vision that Luke had (ie, "his mentors", whatever).
But there's still no way Yoda can be both wise and good (why else would he keep making the wrong decisions and deceiving people?) and no way Vader can be both powerful and evil (how else could he not have spotted that Leia was his daughter?). But also no way that Vader could be a good guy in disguise (or he wouldn't have killed all those people). The characters are just gridlocked.
I still think my Dallas ("it was all a - bad - dream") theory could bring some consistency to the SW plot.
RMN
~~~
I was referring to Vader. The idea was that Vader (and possibly Obi-Wan) were working against Yoda and Palpatine (the Emperor). That would explain why Yoda keeps lying, why Obi-Wan doesn't seem to remember the droids, why Vader doesn't seem to notice that Leia is a Jedi, why Luke was "hidden" in plain sight, and so on.
RMN
~~~
It is not a "terrific plot", and it's not terribly original, either (I've been hearing similar theories since Episode 1 was released). As Brin himself points out, it can't work for one simple reason: Darth Vader killed far too many people.
I mean, maybe he could have been forced to kill a couple of people just to convince the emperor he really was on the dark side. A tough moral choice but in the end he decided the sacrifice of a few was necessary to save the majority, sort of thing. But when the undercover agent kills more people than the godfather, then there's no way you can make him turn out to be a good guy after all.
No, I think Episodes 2 and 1 really did more damage than Episode 3 can possibly fix. And what's worse, if you see them in the new order, the original movies (which are excellent) will now seem to be full of contradictions. Lucas might as well have made a completely separate story (same universe, perhaps, but not directly connected to the original movies).
After I saw Episode 1, a friend of mine asked me "so, how did you find the plot?" and I said "I didn't".
The only way to fix things is if, in Episode 3, J.R. wakes up and all this turns out to have been a dream.
RMN
~~~
I'll skip the silly sensationalism. The drive won't "implode", some parts just risk melting because the laser isn't properly calibrated in the old firmware.
;-)
But the interesting question is: does this mean that, with this new 4x media, my DVR-103 (which is a 2x DVD-R recorder) will record at 4x? If so, that's very nice to know.
I already have the new firmware, by the way. Unlike the old updater (that required you to connect the drive as the primary IDE slave), the new one will automatically detect and update all your drives, no matter how they're connected (even via IEEE-1394).
Now if only HP and Sony could do the same to make their DVD+RW recorders work with DVD+R as they promised...
RMN
~~~
it seems a shame that Hollywood had to wait until his death before they took him seriously...
You're not supposed to take him seriously; he is (was) a comedy writer.
RMN
~~~
...have never read any of Stephen King's books...
RMN
~~~
How about for the people playing the game...?
RMN
~~~
You've been reading Dave Barry's articles, haven't you...? ;-)
RMN
~~~
...I thought it was about Steve Ballmer.
RMN
~~~
Earth is about to be hit by a huge asteroid and the only way to save mankind it is to use something with a big enough mass to push the asteroid off its course.
That's when CowboyNeal springs into action (powered by the world's largest catapult, naturally).
In a dramatic sequence, he misses the asteroid completely. Humanity prepares to be obliterated (shots of Arafat and Sharon hugging, Bush in bed with Osama, etc.). But the scientists notice that, without CowboyNeal's mass, Earth's orbit has changed slightly, and the asteroid narrowly misses.
The film ends with a shot fo CowboyNeal landing on the moon and eating it.
RMN
~~~
Normal CCD pixels are only sensitive to one light component (usually R, G and B). So when you have, say, 3 megapixels, you only have 1 million red samples, 1 million green samples and 1 million blue samples (actually this isn't exactly right, because green is usually given priority, so on most 3 megapixel cameras you have 1.5 million green samples, 750k red samples and 750k blue samples).
Basically what this means is that if you want a clean-looking image, with no coloured halos in high-contrast areas, you should have at least 3 times as many CCD pixels as you want to have in the final image.
Also, because CCDs can be quite "noisy" under low light, it helps to have more samples so that when you average them you get a more stable colour.
As a rule of thumb, with mid-range CCDs, you'll be able to get pretty good pictures at 1/4 of the camera's resolution (ie, if the camera shoots at 2048x1536, the picture will look better - without red/blue outlines and with a lot less noise - if you resample it to 1024x768).
Some cameras will do this automatically when you select a lower resolution (ie, shoot at maximum, then resample), others will capture the image directly at the lowest resolution, using only half the CCD's pixels. If you're not sure how your camera works, you should always shoot at the maximum resolution and then downsample in an image editing program if you want to get rid of halos / noise.
RMN
~~~
They also built the pyramids, calculated the radius of the Earth, and so on.
The human brain hasn't changed that much in 3500 years; we were a pretty smart bunch back then. And of course, we didn't have computers so we actually used it now and then.
RMN
~~~
Would you buy a phone that was smarter than you...? :-)
RMN
~~~
Apparently this guy is already using them in his shoes.
RMN
~~~
Porn helped drive the printing and video industries. Gutenberg's second book, after the Bible, was erotic stories.
So he moved from hardcore to softcore, huh?
RMN
~~~
All image capture is analog. It can be electronic (CCDs) or chemical (film), but there's always an element that "charges up" as it's hit by photons.
Compared to film, CCDs are extremely low-res (top quality 35 mm film has resolution equivalent to a 50 megapixel CCD) but, more importantly, they're slow. At very short exposure times, CCDs have so much noise that the final result is useless. The problem isn't the transfer rate, it's the time the CCDs take to "charge up" to meaningful values.
There is one alternative: use very large CCDs. The larger the CCD, the more light hits it, and the faster it can charge. But larger CCDs are more expensive and require special lenses.
Recording directly to digital does have one big advantage: you don't have to pay for the film. But the CCDs simply aren't up to film quality yet (and probably won't be for another 5 years or so). So the solution is simple: shoot on film, then digitise it.
RMN
~~~
That has to be the strangest looking Hammer I've ever seen. Doesn't even have a handle. On the plus side, it does seem to come with a lot of built-in nails.
RMN
~~~