I know people who grew up in Eastern Europe. They had home computers, mostly C64s and the like imported at vast expense from the West. Apparently in Russia a few people had cloned machines, mostly from designs from Sinclair in the UK.
On the other hand I met someone who worked in a chip factory in East Germany. Everyone knew what they were doing was very far behind the west. In fact there was a joke that the first 1Mhz processor in the Eastern Bloc would fly in on a cruise missile.
This page reckons that the Soviet Block was 10-12 years behind the West at chip production.
Bootstrapping was the first major problem. How do you start a new computer and debug its OS if don't have an OS on the computer? From earlier systems I already had a small monitor program - directly burned into an EPROM - able to load binaries through a serial line. Getting the MMU (74ls610) was the second problem, because it was on the CoCom list, and it was not allowed to export to eastern countries. (Although I don't live in an eastern country, this posed some difficulties...)
So if you were an Eastern Block engineer you'd have to get someone to buy this MMU on the black market somehow which cost precious hard currency. Or you could get some local factory to make a clone. Obviously either are harder than buying it from a mailorder shop.
My guess is that the Cubans set up a front company and buy PCs somewhere in the West and then probably pirate the software.
The worst thing is that lots of awful journalists hide behind the doctrine of protecting their sources and by karma whoring to a rabidly partisan audience.
E.g. if I write an article now accusing Obama of eating babies based on anonymous sources, right wingers will back me up. A few months ago I could have done the same with Bush. Though I'd probably need a right wing and a left wing pen name to make it work.
The poor assumption is that a signal synthesized by the organic combination of some set S of frequencies is going to contain only frequencies between 0 and max(S).
That's not a poor assumption - there is an antialiasing filter that cuts off sharply before the Nyqvist frequency.
The corner cases of extremely complex waveforms-- for example, a saxophone-- that are essentially controlled by higher-frequency fluctuations that to the crude observer affect overall amplitude at a lower frequency also have some audible drift, but not noticeable by an untrained ear. Live saxophones sound far more crisp than well-recorded saxophones
Hey, I like live music much more than recorded music. But that's nothing to do with Nyqvist or signal to noise ratios or any technical detail. If I listen to live music I'm out boozing and carousing, and that is fun.
The point of DRM is not that you force people to use it. I can still watch avi files and listen to mp3s on my Vista machine. I can rip DVDs too, and download torrents. Microsoft don't care about this.
What Microsoft and Apple are trying to do is to convince the content industry that their schemes are safe enough to trust over DVDs and CDs. That's doesn't mean foolproof by any means, just fiddly to crack but cheaper and more convenient than going out and physically buying a DVD or CD or even downloading a torrent. If they can convince the content industry, the users will follow.
That's what Apple did with iTunes. It won't be open of course, that defeats the whole purpose.
Yeehaw. I done trolled the lake and caught me an audiophile.
All Nyqvist's theorem says is the if you sample a frequency F, any frequencies upto F/2 will be preserved. You make F big enough so that F/2 is outside the range of frequencies people can hear and you're good to go.
I dunno about you, but I'm pretty sure I don't want my hi fi emitting frequencies above 22-24khz which is N/2 for CDs and computer generated sounds.
But actually look what's happening here. The generations of flash I've used have gone like this
NOR. No ECC. Low density SLC NAND. Hamming code ECC. Higher density (2xNOR). MLC NAND. BCH or Reed Solomon. BCH seems to be the most common. Highest density (2xSLC)
It seems like the trend is to have less reliability in built reliability in return for smaller cells and add smarter error correction in the controller. Which was my point - you can live with any bit error rate that you can correct.
I bet the cost of silicon for the controller is negligable compared to the NAND array anyway.
Of course AMD made documented the AMD64 instruction set. They wanted Intel to adopt it, because that would make it ubiquitous. AMD probably has patents for it, but that does affect Intel because AMD and Intel have a patent cross license. That doesn't mean that NVidia could implement an x64 chip though, both Intel and AMD have been patenting things as fast as they possibly could.
True unless your talking about the X86. The X86 is register starved. X64 fixes that problem so on the mess that is the X86 yes 64 bit code can be faster.
[Benchmarks needed]
Register renaming probably probably hides lack of registers from what I've seen. Certainly doubling the number of registers in x64 hasn't led to drastic performance increases when you switch a chip from x86 to x64 mode.
You could detect and correct single bit errors, and detect multi bit errors and replace the cell with one of the ones in the spare pool. Which is how NAND ecc works.
24 bits of colour data, i.e eight bits for Red, eight bits for Green and eight bits for Blue for images is probably quite close to the limit of what your eyes can actually see. Just like 16 bits of audio is probably quite close to the perceptionlimit for audio. Audiophiles may disagree on the last bit, but then they disagree with the Nyqvist theorem and double blind tests too.
Mind you during processing there might be an argument for processing at a higher bit depth and then downconverting. E.g. you could process audio at 32 bits, make sure the errors stay in the bottom 16 bits and then when you downconvert they will disappear.
Finally 32 bits of colour data might be faster due to alignment issues. Still it's worth pointing out that 32 bit pixel formats often use 8 bits for alpha level, i.e. only for mixing images together. Probably most video is processed at more than 8 bits per component too, it's only the final output which is 8 bit. With video there are good dithering algorithms too like Floyd Steinberg. They preserve some of the information when you down convert.
Slashdot Mirror
I know people who grew up in Eastern Europe. They had home computers, mostly C64s and the like imported at vast expense from the West. Apparently in Russia a few people had cloned machines, mostly from designs from Sinclair in the UK.
On the other hand I met someone who worked in a chip factory in East Germany. Everyone knew what they were doing was very far behind the west. In fact there was a joke that the first 1Mhz processor in the Eastern Bloc would fly in on a cruise missile.
This page reckons that the Soviet Block was 10-12 years behind the West at chip production.
http://www.cpushack.net/soviet-cpus.html
There were also CoCom restrictions on selling technology. You can see how this worked with this example
http://www.canberra.edu.au/~scott/C=Hacking/C-Hacking13/os.html
Bootstrapping was the first major problem. How do you start a new computer and debug its OS if don't have an OS on the computer? From earlier systems I already had a small monitor program - directly burned into an EPROM - able to load binaries through a serial line. Getting the MMU (74ls610) was the second problem, because it was on the CoCom list, and it was not allowed to export to eastern countries. (Although I don't live in an eastern country, this posed some difficulties...)
So if you were an Eastern Block engineer you'd have to get someone to buy this MMU on the black market somehow which cost precious hard currency. Or you could get some local factory to make a clone. Obviously either are harder than buying it from a mailorder shop.
My guess is that the Cubans set up a front company and buy PCs somewhere in the West and then probably pirate the software.
Well done. This permalink should be even better. http://en.wikipedia.org/w/index.php?title=Slashdot&oldid=269993818#Truth
That link is truly a thing of beauty.
This is Gnome on top, GNU in between and Linux underneath. This is Ubuntu. Welcome to the future.
I prefer to call it Gnome/GNU/Linux.
$current_year+1 will be the year of Linux.
"Daddy! What's a GIMP?"
The worst thing is that lots of awful journalists hide behind the doctrine of protecting their sources and by karma whoring to a rabidly partisan audience.
E.g. if I write an article now accusing Obama of eating babies based on anonymous sources, right wingers will back me up. A few months ago I could have done the same with Bush. Though I'd probably need a right wing and a left wing pen name to make it work.
Enjoy your 'reality'.
The poor assumption is that a signal synthesized by the organic combination of some set S of frequencies is going to contain only frequencies between 0 and max(S).
That's not a poor assumption - there is an antialiasing filter that cuts off sharply before the Nyqvist frequency.
The corner cases of extremely complex waveforms-- for example, a saxophone-- that are essentially controlled by higher-frequency fluctuations that to the crude observer affect overall amplitude at a lower frequency also have some audible drift, but not noticeable by an untrained ear. Live saxophones sound far more crisp than well-recorded saxophones
Hey, I like live music much more than recorded music. But that's nothing to do with Nyqvist or signal to noise ratios or any technical detail. If I listen to live music I'm out boozing and carousing, and that is fun.
The point of DRM is not that you force people to use it. I can still watch avi files and listen to mp3s on my Vista machine. I can rip DVDs too, and download torrents. Microsoft don't care about this.
What Microsoft and Apple are trying to do is to convince the content industry that their schemes are safe enough to trust over DVDs and CDs. That's doesn't mean foolproof by any means, just fiddly to crack but cheaper and more convenient than going out and physically buying a DVD or CD or even downloading a torrent. If they can convince the content industry, the users will follow.
That's what Apple did with iTunes. It won't be open of course, that defeats the whole purpose.
Maybe he just resents your obvious superiority complex.
Only another few hundred In Soviet Russia jokes to scroll through before you start to get to original content.
Why can't you give your presentation in Vi?
About half an aswey.
It would have been funnier if she'd reprogrammed it to spell out DIVORCE.
Yeehaw. I done trolled the lake and caught me an audiophile.
All Nyqvist's theorem says is the if you sample a frequency F, any frequencies upto F/2 will be preserved. You make F big enough so that F/2 is outside the range of frequencies people can hear and you're good to go.
I dunno about you, but I'm pretty sure I don't want my hi fi emitting frequencies above 22-24khz which is N/2 for CDs and computer generated sounds.
I can't hear them - actually if you go here ... you'll find you can't hear them either.
http://www.jimmyr.com/blog/hearingloss.html
Having spent some time in the same room as an ultrasonic PCB cleaner, I know I don't want to be around ultrasound.
Imagine how embarassed you'd be if they beat you with a wrench to make you reveal your password and all they found afterwards was some goat porn.
Coming from an Org that encrypts everything
Tom Cruise? Is that you?
Hmm, I didn't know MLC used a stronger ECC. It seems you're right
http://www.spansion.com/application_notes/Types_of_ECC_Used_on_Flash_AN_01_e.pdf
But actually look what's happening here. The generations of flash I've used have gone like this
NOR. No ECC. Low density
SLC NAND. Hamming code ECC. Higher density (2xNOR).
MLC NAND. BCH or Reed Solomon. BCH seems to be the most common. Highest density (2xSLC)
It seems like the trend is to have less reliability in built reliability in return for smaller cells and add smarter error correction in the controller. Which was my point - you can live with any bit error rate that you can correct.
I bet the cost of silicon for the controller is negligable compared to the NAND array anyway.
MSI has to be the worst quality part maker on the market. I've had terrible experience with them.
If I was betting on it, I'd say ASUS would have the most profitable year.
I like Porter Industries. Even though they are a bit pricy, the quality of their kit is really high.
Of course AMD made documented the AMD64 instruction set. They wanted Intel to adopt it, because that would make it ubiquitous. AMD probably has patents for it, but that does affect Intel because AMD and Intel have a patent cross license. That doesn't mean that NVidia could implement an x64 chip though, both Intel and AMD have been patenting things as fast as they possibly could.
True unless your talking about the X86. The X86 is register starved. X64 fixes that problem so on the mess that is the X86 yes 64 bit code can be faster.
[Benchmarks needed]
Register renaming probably probably hides lack of registers from what I've seen. Certainly doubling the number of registers in x64 hasn't led to drastic performance increases when you switch a chip from x86 to x64 mode.
You don't need to correct multibit errors.
You could detect and correct single bit errors, and detect multi bit errors and replace the cell with one of the ones in the spare pool. Which is how NAND ecc works.
24 bits of colour data, i.e eight bits for Red, eight bits for Green and eight bits for Blue for images is probably quite close to the limit of what your eyes can actually see. Just like 16 bits of audio is probably quite close to the perceptionlimit for audio. Audiophiles may disagree on the last bit, but then they disagree with the Nyqvist theorem and double blind tests too.
Mind you during processing there might be an argument for processing at a higher bit depth and then downconverting. E.g. you could process audio at 32 bits, make sure the errors stay in the bottom 16 bits and then when you downconvert they will disappear.
Finally 32 bits of colour data might be faster due to alignment issues. Still it's worth pointing out that 32 bit pixel formats often use 8 bits for alpha level, i.e. only for mixing images together. Probably most video is processed at more than 8 bits per component too, it's only the final output which is 8 bit. With video there are good dithering algorithms too like Floyd Steinberg. They preserve some of the information when you down convert.
What about error correction?
NAND flash chips are dense enough that they have errors, it's up to the user to error correct those errors away.
I think there's some merit in scaling things to the point where they become unreliable and then using error correction to fix the unreliability.
It will jump forward 20 years to 2008?
You missed an opportunity there.
You realise that for once "FRIST PS0T!!11" would have been ontopic, Funny and Insightful.