No, those are already compressed, which is why the average is closer to 2:1 because of their max compression than to 50:1 from the rest of the more compressible data.
We're comparing the uncompressed data to the compressed data.
Someday all that exercise equipment in gyms will reclaim all their expended human energy into powering their own devices. Since even top performers like Lance Armstrong produce only 500W for under 20min, maybe we can just hope that exercisers can work off their lighting bill, if not heat their showers.
Hikers with a body suit, though, might be able to cook their dinner.
Yeah, why does it matter anymore which language some binary was written in, in order to link it to another binary library? Everything uses a stack of pointers or values to pass arguments, and a symbol table to resolve references to instructions and data external to the binary package stating the reference. That seems to offer a language-neutral resolution of those references across binary packages, regardless of which language was used to generate them.
Sure, there could be some differences, like argument ordering on a stack, and even datatype conversion between them (eg. Pascal "count+data" vs C "NULL terminated" strings), even byte endianness. But those are deterministic differences that could be directly mapped to one another in a wrapper (even at the expense of some performance). And having the sourcecode for both could allow analysis that reverse engineers any incompatibility at link time to wrap or rewrite it.
Is there really no hope for reconciling the binaries for inter-language linking without extra programmer intervention?
Mobile devices have very large storage, which can be compressed to varying degrees at will, better than 50% averaged across all data types. It wouldn't be very hard to make a filesystem (or other storage type) for any of them that stores an equal amount of fake data, with a fake password, with everything compressed in the same space as an uncompressed set of real data. Such a filesystem could look just like a real filesystem in every way, including total size, but hide the real data behind fake data and fake password. If it's all encrypted, it would be very hard to tell the difference, especially in an airport screening line.
Of course, that would probably violate some law. And "only the bad guys" would do it. But if those bad guys actually have something to hide that also violates those security laws, then of course they'll break that law's "coverup" prohibitions, too.
Terrorist and other criminal orgs with enough resources to be a real threat, and carry notebooks and phones around on flights they don't just blow up, will be able to afford such a filesystem. And once there is one in the wild, anyone will get it, probably for free.
So this is yet another stupid simcurity (simulated security) measure. It's intimidation of everyone to scare us into thinking our government is "doing something severe" to terrorists, when it's just abusing our own freedom. While wasting everyone's time, eroding our trust of our government, and letting the terrorists go free.
Sounds like they're already using sophisticated decoys at DHS: fake security to hide the dangerous absence of any real security.
Any indication that this 45nm generation of Cells will actually ship with multiple dies connected with a cache-coherent bus? Any wiseguys suggesting leaving an entire wafer interconnected by the bus, for the most mammoth CPU ever produced?
Even any hope that >512MB XDR RAM will be available this year?
Or how about some sign that someone's designing a Cell into an actual PC?
Linux is not "THE OS". It's the kernel. Without the rest of, say, GNU's apps that come in, say, Ubuntu, it's not an OS, just a kernel. Which isn't enough for apps to run on - except those few that don't need an OS.
Different distros can qualify as different OS'es if they're so incompatible that apps aren't compatible among them. Not just installation packages like.deb or.rpm, but the actual apps which depend on different distro features, whether that's different directory structure, different other bundled apps (like, say, ps), or other APIs that differ and break app compatibility.
Sure, journalists and users will use the terms wrong. But what counts is what works, regardless of what anyone says. Ubuntu, for example, is more than an OS, because it includes other content, like for example, Firefox plugins, that apps don't depend on, but which enhance apps. It's really a "user environment", which includes kernel, OS, apps, Desktop, content, and other material that people use when they get Ubuntu.
And Linus is empowered to speak for only the kernel among all of those. Since Linus doesn't try to hold everything depending on his kernel hostage, he doesn't speak for anything more, even though he possibly could if he were a powermad jerk. Which he's not. If he were to try, someone would fork the kernel, even if they couldn't call it "Linux" anymore (though others certainly could, mistakenly). And with so many Linux distro brands already established, that wouldn't really cripple the "Linux" industry.
Of course Torvalds speaks for Linux - the kernel. That's all he controls, with his trademark on "Linux" and his undisputed control of what is released in the kernel. So when he speaks, he speaks for that with authority.
He doesn't speak for any distro, never did, never claimed to. But that's part of the problem with calling, say, Ubuntu "Linux". Most of Ubuntu, or Red Hat, or aN4rCHi$tOS, or any other distro, is not the kernel. It's a lot of other software that's compatible with a Linux kernel it relies on. Most of which is usually GNU software, with its own spokespeople - who often disagree fundamentally with Torvalds. The people running those distro projects speak for them. And therefore they speak for what people call "Linux" more than Torvalds does.
And when they don't speak for someone who disagrees, that person is free to make their own "Linux" and speak for it.
I know the corporate mass media can't understand that kind of community ownership and independence. But Slashdotters should be able to tell the difference.
Programmers are also writers whose work gets syndicated on the Internet without royalty payments. The difference is that TV writers have a union.
Programmers might like to think that the other difference is that there are more than enough TV writers, but not nearly enough programmers, so supply/demand keeps programmer jobs/benefits/salaries safe. But programming is getting easier for anyone to do - at least the kind of programming that most people will buy, just like the mostly crappy writing on TV. In a few years, the global supply of programmers will be a lot bigger, and programmers won't have as much safe privileges in the market.
Then programmers will want a union, and wish they'd helped TV writers secure these basic fairness rules of commerce that would protect everyone writing creatively for profit.
This exact disaster is exactly why Microsoft cannot be allowed to extend its monopoly to absorb Yahoo. When Yahoo fails the way Hotmail now has with Firefox, the way Hotmail did when MS tried to cut it over to Windows servers, the Internet will take a major hit. Even if it drives consumers to GMail, that just reinforces Google's dominance, without credible competition.
The Internet itself is a hothouse for competition. The global environment for megacorps, though, is precisely the opposite. When the business drives the apps, which it always will at that scale, the Internet's flexibility will become a hothouse for monopolies. Since the entire world depends on the Internet, that Internet monoculture must be stopped. That's why people have governments: to stop the ambitious among them from exploiting advantages that hurt everyone else.
There is every evidence that Microsoft's control of Yahoo would be a disaster, and no evidence that it would be good for anyone but Microsoft (and maybe the Yahoo shareholders they buy off). If the deal goes through, that's the proof that the people need to change our governments to actually protect us, instead of serving these monopolies.
Look, you obviously know nothing about the economics or practice of producing Sony games. You're repeating meaningless grumbling that you can't even understand about Sony and developers. The fact is that Sony's model is simply to license the games, and subsidize an extremely powerful console that runs them. A superior console that's sold more than the X-Box, which means a bigger market for games.
"No clear market", because the same game can run on PC or PS3, depending on which one you prefer (or which HW you already have)? Worried about stealing sales from PS3 when they make money from either? You know nothing about marketing if you think having extra if overlapping segments to target is worse than having only one. Chicken and egg is not a problem when there's a PS3 version to buy. Quit pretending you know anything about marketing just because you've heard some marketing buzzwords. Even real marketers are typically full of crap - poser armchair marketers are clowns.
"Graphics 6 feet / 18 inches away" doesn't even make any sense.
Sony already makes Intel PCs that are never as cheap as their direct specs competition. And they're usually higher quality, but much higher cost. There are no PCs with Blu-Ray built in, so Sony would have that market to itself for a while, and it would be a much superior machine. Which Sony prefers to sell, making lots of money off it right now.
There's some free clues. Stop wasting my time with trivially dismissable complaints that show you know nothing about marketing or even how games or PCs are made and sold.
Where has this been discussed? The Cell processor is a PowerPC with effectively 7 or so DSPs on a fast integrated onchip bus, with RAM fast enough to match. There is no reason whatsoever that the chip cannot be a perfectly good CPU for "general purpose computing", given the kind of processing that currently demands.
Most of the CPU load these days is managing multimedia streams, from audio, video, animation and telephony. The SPUs and their bus are perfect for that, while the PPC is perfect for the task of scheduling their processing with existing kernels like Linux.
The PS3 running Ubuntu right now is pretty good proof. Even without a video HW chip or any code running on SPUs yet, it's an adequate PC. With X and video playing ported to the SPUs, even in their prototype quality, it's competitive with x86 PCs for the money, even though it's got to swap due to its artificially low 512MB RAM. If the RAM were 1GB or higher, if it had a video chip, if its multimedia processing were running on the SPUs, it would be a 3.2GHz 2xhyperthread PPC with very little to do except app logic and no IO latency. That would be an excellent PC.
Yes, I think that Sony would love getting the licensing fees from the ported games. All those games pay Sony a license to play on the PS3. The contracts include licensing for ports to other platforms that aren't the PS3. More sales, more money for Sony.
Also, Sony gets royalties on every Cell sold, having helped create it (and owning many of its patents).
So yes, I am suggesting "that IBM should sell processors to a computer manufacturer who will use them to make a desktop box that plays PS3 ports". Like to Sony. Then they could make a lot more money. But even selling to competitors they'd make money. Sony loves that, so I think they'd like that idea. Thanks for suggesting it.
The PS3 is interesting because it's so much power in such a cheap box, but it's subsidized by Sony. I think Sony will be lowering prices less while reducing the subsidy more.
But where are the Cell PCs already? The PS3 is cute, but it's locked down with a Sony hypervisor, it's got no PCI or other expansion, only a single SATA connector, and a puny 512MB hardwired RAM (its Cell can rip through 512MB, peforming 64bit floating point math on it all, in under 0.0025s). Its RSX video chip is locked out from Linux, so no HW acceleration (and no addon videocard is possible).
IBM is now cranking out these chips. It lost Apple, its biggest CPU (PPC) customer, to Intel. Where's a PC built on a Cell that includes PCI-e, expandible XDR RAM, Gb-e networking, and a more open nVidia graphics card (or two)? Since the Cell is cheap due to its higher yields, a $1000 Cell PC could make a $1000 Intel PC (Mac or Windows/Linux/etc) look like a 286 with its extremely high speeds. Sony has proven it can be mass manufactured with mostly commodity parts for under $750.
Since Ubuntu already runs on Cell, a cheap Windows killer could take the Cell architecture to the top of the CPU stakes in record time from release. It would be a much easier/cheaper/faster target for porting PS3 games than Intel PCs. Apple, which supposedly dropped PPC for Intel because of heat:performance limitations, would have to look seriously at a return to PPC, especially since 45nm Cell with only a few SPUs could be a perfect fit for an iPhone successor. If not from Apple, then from someone smart enough to use Cell in the biggest market of all.
A smaller die means a smaller, cheaper package; it also means that yields will be better and that each chip will cost less overall.
The redundancy of the Cell's 8 SPUs (DSP coprocessors) is the main point of the Cell's design. Defective SPUs (nearly always from dust particles in the nearly - but not quite - perfect "clean rooms" in which they're manufactured) can be tested and turned off as they roll off the assembly line. The shut down SPUs are even physically disconnected from power by hard fuses, so they don't cost any performance in operation. The perfect Cells with 8 SPUs cost the most, in high-end IBM RS/6000 workstations (and some blade servers). 7 SPUs go into PS3s. The rest of the yield, supposedly down to a single SPU (but even 0 SPUs still have a 3.2GHz PPC and superfast IO), go into HDTVs and other consumer electronics. All of the yield gets sold, instead of a fraction in older manufacturing processes.
So smaller dies don't really affect Cell yields. Smaller dies just mean smaller parts of the wafer that would get spoiled by a single defect, which is already taken care of with the redundant SPUs.
In fact, smaller dies mean multiple defects are less likely to land on a single die. Which means that more Cells would turn into low-SPU, cheaper Cells. While larger dies would concentrate multiple defects into a single dies, by landing on a single die more often, leaving more perfect Cells getting the highest prices.
45nm does mean more Cells, at any defect rate, per wafer. Which means, for the same number of defects per wafer, more dies per wafer. So there is a yield increase, but not for the same reasons as traditional ones. And of course 45nm has so many other valuable benefits, like speed, and more transistors if they keep the same die size, that the move is very valuable overall.
The main problem with Froogle results is that accessories are included along with the main product. I wish there were a way to select a product in the results, and just see all the competing products offered at different prices. "Query by example" is already offered in Google's Web search for "similar pages" (that I never use:).
I think Froogle is just a neglected Google service overall, probably for some grand strategic reason. It disappeared entirely from Google's homepage and result pages for a while. But when some other giant falters, like eBay has, I hope Froogle swings into action with a better app.
Looks like posting right after waking up helped me get this story exactly wrong. What's incredible is that before you corrected me, several other people modded me up. Maybe the rest of my post about eBay's monopoly is "interesting", even if this new move is no demonstration of that. But still the lack of critical thinking on Slashdot is an elemental force never to underestimate.
Craigslist fulfills my eBay needs quite well. I was worried when ebay bought out 25% of Craigslist, but so far it seems to have had little or no effect.
I don't know how CL sellers might have/not changed since the eBay buyin, but I noticed that the quality of buyers went drastically downhill within about 6 months of it. It used to be that everyone I bought from on CL was honest and reliable, even a nice person it was pleasant to meet during the transaction. Now it's full of flakes who will "negotiate" (often obnoxiously) including making appointments to pickup (and halt negotiations with alternate buyers), then disappear. And just generally lowball prices. Their email habits and literacy are awful now, too.
I think it's mainly a function of CL going big, through eBay's partnership in marketing somehow. Before, CL was word of mouth starting in a community of pretty cool people. Now it represents the public at large, which on average sucks in person.
The problem is that when sellers get negative feedback, they retaliate against the buyers. So eBay's solution is to prevent negative feedback? Why doesn't eBay prevent seller retaliation? Prevent a seller from posting negative feedback against any buyer who posted negative feedback to that seller in the past month. Investigate claims from buyers of mere retaliation, and stop sellers from posting any negative feedback for a month on the first violation, stop for six months on the second, suspend their account for a month on the third, suspend for six months on the fourth, and shut them down on the fifth confirmed retaliation. Or some other aggressive policy that shows everyone that mere retaliation isn't worth it.
Instead, eBay will stop all negative feedback. Which is the only feedback that I ever look at, to see what will go wrong (things going right is the expected default, until I look at feedback). That will turn all eBay transactions into uncertainty, which is bad for the entire market.
But I guess eBay can rely on its monopoly (look it up, it means "market controller", not "sole marketer") to keep business roaring. Remember that eBay also controls PayPal, the unregulated Internet global banking monopoly, and Skype, the unregulated Internet global telco (not yet a monopoly, but gaining...). While eBay was protecting the consumer, those global market dominances in retail, banking and telephony were not such a threat. But now that they're showing the corporate bias towards secrecy to "solve" problems of abuse, they need a hard look.
Someone's got to protect the consumer, even if it means just forcing eBay to allow consumers to inform each other what sellers and eBay are working against them. It doesn't have to be a government. Something like Froogle's reviews could harness people power around the world to do it even better.
Google's Froogle shopping search filter should rate eBay sellers just like it rates other "stores". Not rate just eBay itself, but per seller. Google should allow reviewers posting reviews from their eBay account to have weighted review points, or their own rating, and discard reviews posted from someone who received a negative review from their target in the past month or so.
eBay is a market monopoly that needs balancing. If eBay is stopping its own users from being that counterbalance to its own users, then someone like Google, which exists to just connect users together with each other's content, is a great counterbalance.
How about instead of going for maglev gold, we just run rail conveyor belts along the main commuter routes, with flatbed spots and de/accelerator ramps from feeder entrances/exits? Drive on in the suburbs, park, and drive off in a highly predictable 15-20 minutes, into a parking lot building. No congestion or collisions, no gas burned (the central power generation and pollution control is much more efficient, and included in your fare). Fill the rails with cars to use the full capacity of the rails.
If a whole NYC/LA maglev that could accommodate maybe a dozen trains at a time costs only $70B, then a dozen "autoartery" car trains serving NYC or LA should cost at most 10% of that. Hell, Giuliani's lame "Sky Train" that gets a few tourists between JFK airport and Jamaica (filthy and remote) station, instead of just extending existing subways 1/4 mile, cost $8B. If we spent that on these car trains, they'd serve over 1M cars a day, which for $10B would mean $10 round trips would take under 3 years to pay back. $4:gal gas in 20MPG cars in traffic means $10 is about half what most people pay right now, not to mention normal car wear, collisions and just going crazy in traffic.
In the 1920s and 1930s, this kind of car train would have been science fiction. Why don't we catch up with "Golden Age" SF first, and then move on to the SF that came later with what we save?
My statement isn't a logical fallacy of redefining a label up to deny contradictory evidence. It's a simple statement that the US isn't living up to its basic defining characteristics, even while we base extreme actions on claims of those characteristics. It's a demand that we change, not an excuse that the country isn't guilty of those things.
Or do you think that demands to meet expectations, especially expectations that are created by insisting they're essential and to be forced on others, are never reasonable demands?
Any country making both democracy and security its highest priorities for years, even at cost of a perpetual state of emergency, suspended liberty, thousands dead and many tens of thousands wounded (multiplied in the non-American casualties), unsupportable debts, alienating allies and activating enemies, would immediately remove these untrustworthy machines and never allow their vendors or technologies into the critical path of its government again.
Such a country would never have allowed such a risk at all, either before or after such vulnerabilities were publicly exposed.
But instead, this story will become a footnote. Precisely because there's an election going on. An election that is threatened by these untrustworthy machines.
Since those priorities were set and executed by a government installed on the reports of these kinds of untrustworthy machines, I guess we've got everything we deserve.
Slashdot Mirror
No, those are already compressed, which is why the average is closer to 2:1 because of their max compression than to 50:1 from the rest of the more compressible data.
We're comparing the uncompressed data to the compressed data.
Someday all that exercise equipment in gyms will reclaim all their expended human energy into powering their own devices. Since even top performers like Lance Armstrong produce only 500W for under 20min, maybe we can just hope that exercisers can work off their lighting bill, if not heat their showers.
Hikers with a body suit, though, might be able to cook their dinner.
Yeah, why does it matter anymore which language some binary was written in, in order to link it to another binary library? Everything uses a stack of pointers or values to pass arguments, and a symbol table to resolve references to instructions and data external to the binary package stating the reference. That seems to offer a language-neutral resolution of those references across binary packages, regardless of which language was used to generate them.
Sure, there could be some differences, like argument ordering on a stack, and even datatype conversion between them (eg. Pascal "count+data" vs C "NULL terminated" strings), even byte endianness. But those are deterministic differences that could be directly mapped to one another in a wrapper (even at the expense of some performance). And having the sourcecode for both could allow analysis that reverse engineers any incompatibility at link time to wrap or rewrite it.
Is there really no hope for reconciling the binaries for inter-language linking without extra programmer intervention?
Mobile devices have very large storage, which can be compressed to varying degrees at will, better than 50% averaged across all data types. It wouldn't be very hard to make a filesystem (or other storage type) for any of them that stores an equal amount of fake data, with a fake password, with everything compressed in the same space as an uncompressed set of real data. Such a filesystem could look just like a real filesystem in every way, including total size, but hide the real data behind fake data and fake password. If it's all encrypted, it would be very hard to tell the difference, especially in an airport screening line.
Of course, that would probably violate some law. And "only the bad guys" would do it. But if those bad guys actually have something to hide that also violates those security laws, then of course they'll break that law's "coverup" prohibitions, too.
Terrorist and other criminal orgs with enough resources to be a real threat, and carry notebooks and phones around on flights they don't just blow up, will be able to afford such a filesystem. And once there is one in the wild, anyone will get it, probably for free.
So this is yet another stupid simcurity (simulated security) measure. It's intimidation of everyone to scare us into thinking our government is "doing something severe" to terrorists, when it's just abusing our own freedom. While wasting everyone's time, eroding our trust of our government, and letting the terrorists go free.
Sounds like they're already using sophisticated decoys at DHS: fake security to hide the dangerous absence of any real security.
Any indication that this 45nm generation of Cells will actually ship with multiple dies connected with a cache-coherent bus? Any wiseguys suggesting leaving an entire wafer interconnected by the bus, for the most mammoth CPU ever produced?
Even any hope that >512MB XDR RAM will be available this year?
Or how about some sign that someone's designing a Cell into an actual PC?
Linux is not "THE OS". It's the kernel. Without the rest of, say, GNU's apps that come in, say, Ubuntu, it's not an OS, just a kernel. Which isn't enough for apps to run on - except those few that don't need an OS.
.deb or .rpm, but the actual apps which depend on different distro features, whether that's different directory structure, different other bundled apps (like, say, ps), or other APIs that differ and break app compatibility.
Different distros can qualify as different OS'es if they're so incompatible that apps aren't compatible among them. Not just installation packages like
Sure, journalists and users will use the terms wrong. But what counts is what works, regardless of what anyone says. Ubuntu, for example, is more than an OS, because it includes other content, like for example, Firefox plugins, that apps don't depend on, but which enhance apps. It's really a "user environment", which includes kernel, OS, apps, Desktop, content, and other material that people use when they get Ubuntu.
And Linus is empowered to speak for only the kernel among all of those. Since Linus doesn't try to hold everything depending on his kernel hostage, he doesn't speak for anything more, even though he possibly could if he were a powermad jerk. Which he's not. If he were to try, someone would fork the kernel, even if they couldn't call it "Linux" anymore (though others certainly could, mistakenly). And with so many Linux distro brands already established, that wouldn't really cripple the "Linux" industry.
Of course Torvalds speaks for Linux - the kernel. That's all he controls, with his trademark on "Linux" and his undisputed control of what is released in the kernel. So when he speaks, he speaks for that with authority.
He doesn't speak for any distro, never did, never claimed to. But that's part of the problem with calling, say, Ubuntu "Linux". Most of Ubuntu, or Red Hat, or aN4rCHi$tOS, or any other distro, is not the kernel. It's a lot of other software that's compatible with a Linux kernel it relies on. Most of which is usually GNU software, with its own spokespeople - who often disagree fundamentally with Torvalds. The people running those distro projects speak for them. And therefore they speak for what people call "Linux" more than Torvalds does.
And when they don't speak for someone who disagrees, that person is free to make their own "Linux" and speak for it.
I know the corporate mass media can't understand that kind of community ownership and independence. But Slashdotters should be able to tell the difference.
Programmers are also writers whose work gets syndicated on the Internet without royalty payments. The difference is that TV writers have a union.
Programmers might like to think that the other difference is that there are more than enough TV writers, but not nearly enough programmers, so supply/demand keeps programmer jobs/benefits/salaries safe. But programming is getting easier for anyone to do - at least the kind of programming that most people will buy, just like the mostly crappy writing on TV. In a few years, the global supply of programmers will be a lot bigger, and programmers won't have as much safe privileges in the market.
Then programmers will want a union, and wish they'd helped TV writers secure these basic fairness rules of commerce that would protect everyone writing creatively for profit.
Duopoly is effectively monopoly. Oligarchy is the same problem as monarchy from the consumer's point of view.
This exact disaster is exactly why Microsoft cannot be allowed to extend its monopoly to absorb Yahoo. When Yahoo fails the way Hotmail now has with Firefox, the way Hotmail did when MS tried to cut it over to Windows servers, the Internet will take a major hit. Even if it drives consumers to GMail, that just reinforces Google's dominance, without credible competition.
The Internet itself is a hothouse for competition. The global environment for megacorps, though, is precisely the opposite. When the business drives the apps, which it always will at that scale, the Internet's flexibility will become a hothouse for monopolies. Since the entire world depends on the Internet, that Internet monoculture must be stopped. That's why people have governments: to stop the ambitious among them from exploiting advantages that hurt everyone else.
There is every evidence that Microsoft's control of Yahoo would be a disaster, and no evidence that it would be good for anyone but Microsoft (and maybe the Yahoo shareholders they buy off). If the deal goes through, that's the proof that the people need to change our governments to actually protect us, instead of serving these monopolies.
Look, you obviously know nothing about the economics or practice of producing Sony games. You're repeating meaningless grumbling that you can't even understand about Sony and developers. The fact is that Sony's model is simply to license the games, and subsidize an extremely powerful console that runs them. A superior console that's sold more than the X-Box, which means a bigger market for games.
"No clear market", because the same game can run on PC or PS3, depending on which one you prefer (or which HW you already have)? Worried about stealing sales from PS3 when they make money from either? You know nothing about marketing if you think having extra if overlapping segments to target is worse than having only one. Chicken and egg is not a problem when there's a PS3 version to buy. Quit pretending you know anything about marketing just because you've heard some marketing buzzwords. Even real marketers are typically full of crap - poser armchair marketers are clowns.
"Graphics 6 feet / 18 inches away" doesn't even make any sense.
Sony already makes Intel PCs that are never as cheap as their direct specs competition. And they're usually higher quality, but much higher cost. There are no PCs with Blu-Ray built in, so Sony would have that market to itself for a while, and it would be a much superior machine. Which Sony prefers to sell, making lots of money off it right now.
There's some free clues. Stop wasting my time with trivially dismissable complaints that show you know nothing about marketing or even how games or PCs are made and sold.
Where has this been discussed? The Cell processor is a PowerPC with effectively 7 or so DSPs on a fast integrated onchip bus, with RAM fast enough to match. There is no reason whatsoever that the chip cannot be a perfectly good CPU for "general purpose computing", given the kind of processing that currently demands.
Most of the CPU load these days is managing multimedia streams, from audio, video, animation and telephony. The SPUs and their bus are perfect for that, while the PPC is perfect for the task of scheduling their processing with existing kernels like Linux.
The PS3 running Ubuntu right now is pretty good proof. Even without a video HW chip or any code running on SPUs yet, it's an adequate PC. With X and video playing ported to the SPUs, even in their prototype quality, it's competitive with x86 PCs for the money, even though it's got to swap due to its artificially low 512MB RAM. If the RAM were 1GB or higher, if it had a video chip, if its multimedia processing were running on the SPUs, it would be a 3.2GHz 2xhyperthread PPC with very little to do except app logic and no IO latency. That would be an excellent PC.
Yes, I think that Sony would love getting the licensing fees from the ported games. All those games pay Sony a license to play on the PS3. The contracts include licensing for ports to other platforms that aren't the PS3. More sales, more money for Sony.
Also, Sony gets royalties on every Cell sold, having helped create it (and owning many of its patents).
So yes, I am suggesting "that IBM should sell processors to a computer manufacturer who will use them to make a desktop box that plays PS3 ports". Like to Sony. Then they could make a lot more money. But even selling to competitors they'd make money. Sony loves that, so I think they'd like that idea. Thanks for suggesting it.
The PS3 is interesting because it's so much power in such a cheap box, but it's subsidized by Sony. I think Sony will be lowering prices less while reducing the subsidy more.
But where are the Cell PCs already? The PS3 is cute, but it's locked down with a Sony hypervisor, it's got no PCI or other expansion, only a single SATA connector, and a puny 512MB hardwired RAM (its Cell can rip through 512MB, peforming 64bit floating point math on it all, in under 0.0025s). Its RSX video chip is locked out from Linux, so no HW acceleration (and no addon videocard is possible).
IBM is now cranking out these chips. It lost Apple, its biggest CPU (PPC) customer, to Intel. Where's a PC built on a Cell that includes PCI-e, expandible XDR RAM, Gb-e networking, and a more open nVidia graphics card (or two)? Since the Cell is cheap due to its higher yields, a $1000 Cell PC could make a $1000 Intel PC (Mac or Windows/Linux/etc) look like a 286 with its extremely high speeds. Sony has proven it can be mass manufactured with mostly commodity parts for under $750.
Since Ubuntu already runs on Cell, a cheap Windows killer could take the Cell architecture to the top of the CPU stakes in record time from release. It would be a much easier/cheaper/faster target for porting PS3 games than Intel PCs. Apple, which supposedly dropped PPC for Intel because of heat:performance limitations, would have to look seriously at a return to PPC, especially since 45nm Cell with only a few SPUs could be a perfect fit for an iPhone successor. If not from Apple, then from someone smart enough to use Cell in the biggest market of all.
The redundancy of the Cell's 8 SPUs (DSP coprocessors) is the main point of the Cell's design. Defective SPUs (nearly always from dust particles in the nearly - but not quite - perfect "clean rooms" in which they're manufactured) can be tested and turned off as they roll off the assembly line. The shut down SPUs are even physically disconnected from power by hard fuses, so they don't cost any performance in operation. The perfect Cells with 8 SPUs cost the most, in high-end IBM RS/6000 workstations (and some blade servers). 7 SPUs go into PS3s. The rest of the yield, supposedly down to a single SPU (but even 0 SPUs still have a 3.2GHz PPC and superfast IO), go into HDTVs and other consumer electronics. All of the yield gets sold, instead of a fraction in older manufacturing processes.
So smaller dies don't really affect Cell yields. Smaller dies just mean smaller parts of the wafer that would get spoiled by a single defect, which is already taken care of with the redundant SPUs.
In fact, smaller dies mean multiple defects are less likely to land on a single die. Which means that more Cells would turn into low-SPU, cheaper Cells. While larger dies would concentrate multiple defects into a single dies, by landing on a single die more often, leaving more perfect Cells getting the highest prices.
45nm does mean more Cells, at any defect rate, per wafer. Which means, for the same number of defects per wafer, more dies per wafer. So there is a yield increase, but not for the same reasons as traditional ones. And of course 45nm has so many other valuable benefits, like speed, and more transistors if they keep the same die size, that the move is very valuable overall.
The main problem with Froogle results is that accessories are included along with the main product. I wish there were a way to select a product in the results, and just see all the competing products offered at different prices. "Query by example" is already offered in Google's Web search for "similar pages" (that I never use :).
I think Froogle is just a neglected Google service overall, probably for some grand strategic reason. It disappeared entirely from Google's homepage and result pages for a while. But when some other giant falters, like eBay has, I hope Froogle swings into action with a better app.
You're right, I'm wrong.
Looks like posting right after waking up helped me get this story exactly wrong. What's incredible is that before you corrected me, several other people modded me up. Maybe the rest of my post about eBay's monopoly is "interesting", even if this new move is no demonstration of that. But still the lack of critical thinking on Slashdot is an elemental force never to underestimate.
I don't know how CL sellers might have/not changed since the eBay buyin, but I noticed that the quality of buyers went drastically downhill within about 6 months of it. It used to be that everyone I bought from on CL was honest and reliable, even a nice person it was pleasant to meet during the transaction. Now it's full of flakes who will "negotiate" (often obnoxiously) including making appointments to pickup (and halt negotiations with alternate buyers), then disappear. And just generally lowball prices. Their email habits and literacy are awful now, too.
I think it's mainly a function of CL going big, through eBay's partnership in marketing somehow. Before, CL was word of mouth starting in a community of pretty cool people. Now it represents the public at large, which on average sucks in person.
eBay owns 25% of CraigsList.
The problem is that when sellers get negative feedback, they retaliate against the buyers. So eBay's solution is to prevent negative feedback? Why doesn't eBay prevent seller retaliation? Prevent a seller from posting negative feedback against any buyer who posted negative feedback to that seller in the past month. Investigate claims from buyers of mere retaliation, and stop sellers from posting any negative feedback for a month on the first violation, stop for six months on the second, suspend their account for a month on the third, suspend for six months on the fourth, and shut them down on the fifth confirmed retaliation. Or some other aggressive policy that shows everyone that mere retaliation isn't worth it.
Instead, eBay will stop all negative feedback. Which is the only feedback that I ever look at, to see what will go wrong (things going right is the expected default, until I look at feedback). That will turn all eBay transactions into uncertainty, which is bad for the entire market.
But I guess eBay can rely on its monopoly (look it up, it means "market controller", not "sole marketer") to keep business roaring. Remember that eBay also controls PayPal, the unregulated Internet global banking monopoly, and Skype, the unregulated Internet global telco (not yet a monopoly, but gaining...). While eBay was protecting the consumer, those global market dominances in retail, banking and telephony were not such a threat. But now that they're showing the corporate bias towards secrecy to "solve" problems of abuse, they need a hard look.
Someone's got to protect the consumer, even if it means just forcing eBay to allow consumers to inform each other what sellers and eBay are working against them. It doesn't have to be a government. Something like Froogle's reviews could harness people power around the world to do it even better.
Google's Froogle shopping search filter should rate eBay sellers just like it rates other "stores". Not rate just eBay itself, but per seller. Google should allow reviewers posting reviews from their eBay account to have weighted review points, or their own rating, and discard reviews posted from someone who received a negative review from their target in the past month or so.
eBay is a market monopoly that needs balancing. If eBay is stopping its own users from being that counterbalance to its own users, then someone like Google, which exists to just connect users together with each other's content, is a great counterbalance.
How about instead of going for maglev gold, we just run rail conveyor belts along the main commuter routes, with flatbed spots and de/accelerator ramps from feeder entrances/exits? Drive on in the suburbs, park, and drive off in a highly predictable 15-20 minutes, into a parking lot building. No congestion or collisions, no gas burned (the central power generation and pollution control is much more efficient, and included in your fare). Fill the rails with cars to use the full capacity of the rails.
If a whole NYC/LA maglev that could accommodate maybe a dozen trains at a time costs only $70B, then a dozen "autoartery" car trains serving NYC or LA should cost at most 10% of that. Hell, Giuliani's lame "Sky Train" that gets a few tourists between JFK airport and Jamaica (filthy and remote) station, instead of just extending existing subways 1/4 mile, cost $8B. If we spent that on these car trains, they'd serve over 1M cars a day, which for $10B would mean $10 round trips would take under 3 years to pay back. $4:gal gas in 20MPG cars in traffic means $10 is about half what most people pay right now, not to mention normal car wear, collisions and just going crazy in traffic.
In the 1920s and 1930s, this kind of car train would have been science fiction. Why don't we catch up with "Golden Age" SF first, and then move on to the SF that came later with what we save?
Gore v Bush 2000 550 votes
My statement isn't a logical fallacy of redefining a label up to deny contradictory evidence. It's a simple statement that the US isn't living up to its basic defining characteristics, even while we base extreme actions on claims of those characteristics. It's a demand that we change, not an excuse that the country isn't guilty of those things.
Or do you think that demands to meet expectations, especially expectations that are created by insisting they're essential and to be forced on others, are never reasonable demands?
Any country making both democracy and security its highest priorities for years, even at cost of a perpetual state of emergency, suspended liberty, thousands dead and many tens of thousands wounded (multiplied in the non-American casualties), unsupportable debts, alienating allies and activating enemies, would immediately remove these untrustworthy machines and never allow their vendors or technologies into the critical path of its government again.
Such a country would never have allowed such a risk at all, either before or after such vulnerabilities were publicly exposed.
But instead, this story will become a footnote. Precisely because there's an election going on. An election that is threatened by these untrustworthy machines.
Since those priorities were set and executed by a government installed on the reports of these kinds of untrustworthy machines, I guess we've got everything we deserve.