I've got a Beatles CD that says (C) 1963 right on the disc. Either the CD format was in beta longer than we thought, or more likely, the copyright applies to the audio recording independent of the medium it's fixed in.
Wrong. I've been through the patent application process quite a few times. In many cases our corporate lawyers had to drum into the engineers' heads that the patent office has a much lower standard of "novelty" and "obviousness" than what any reasonable person would think. Few if any of the applications that I was involved with were rejected.
Don't forget that there's a difference between the poorly functioning USPTO and the more stringent tests applied in infringements.
You seem to admit that large number of invalid patents are currently being issued. Like I said, they don't have to be inventive to be issued.
Maybe the courts are somewhat more prudent, but most parties won't be able to afford getting to that point. The market barrier works just fine through intimidation without the patent ever getting tested in court. In particular, Microsoft knows that their current top worry, Linux, will have a particularly hard time dealing with patents, valid or not. It's pretty easy to see that the large number of patents they are filing are intended to create a minefield to ensnare anybody trying to compete with them using distributed development model.
Maybe they've developed a new inventive approach. If so, fair enough. If not, then the patent will not be enforceable and they've just burned cash on it.
To get a patent, it doesn't have to be better than anything else, or even "inventive", it just has to be different and "useful" (i.e., it kind of works). That's enough to generate vendor lockin and competitor lockout.
Let's say they invent a new image format called "MSPEG", which is just like JPEG, but they add some twist to address some minor limitation. Now let's say that they make MSPEG the default image format for a large number of Windows applications. What becomes the most valuable aspect of the MSPEG IP? Is it the marginal new feature that almost nobody is taking advantage of, or is it the fact that everybody else is now using the format? Of course, it's the second item.
IMO, IP law should be changed to discount the vendor lockin portion of the IP value. Royalties and license fees should only apply to the extent that a work is instrinically better than other alternatives. Nobody should be able to collect fees just because they are a gatekeeper to compatibility with something. For example, the only value of the GIF format over the last few years has been compatibility with other GIF enabled software. All of its other functionality has long ago been superceded by other formats. Therefore, Unisys should not have been given the privilege to demand fees from people who have to use GIFs to be compatible with applications that require GIFs.
I'm sure a single IBM mainframe could do the same amount of work in half the amount of time and cost a fraction of what that Linux cluster cost.
Mainframes are optimized for batch processing. Interactive queries do not take full advantage of their vaunted I/O capacity.
Moreover, while a mainframe may be a good way to host a single copy of a database that must remain internally consistent, that's not the problem Google is solving. It's trivial for them to run their search service off of thousands of replicated copies of the Internet index. Even the largest mainframe's storage I/O would be orders of magnitude smaller than the massively parallel I/O operations done by these thousands of PCs. Google has no reason to funnel all of the independent search queries into a single machine, so they shouln't buy a system architecture designed to do that.
anything more than four cylinders and a coffee can muffler is a waste.
Actually, I saw a documentary on dragsters a while back. They briefly covered a group of people who were trying to create a new class of riceboy-inspired nitro fueled funny cars that use 4-cylinder engines instead of V-8s. Unless I'm mistaken, even half of a modern dragster engine would have well over the 1000 HP of that W-16. Of course, the dragsters would have no muffler of any kind.
What kind of a retard goes into debt $20,000 to produce free software? It's not like you'll ever make a profit off of it......
He can put down on his resume that he created one of the top 10 most popular Linux distros, and that he supervised quite a few interesting technical innovations unique to that distro.
This experience would help qualify for a job with a salary quite a bit more than $20,000 (not to mention more influence and responsibility) over that of a random code monkey.
That fish looks just a little too happy to be held by a seagull.
I don't know. The first thing that I thought was that that seagull looks totally stoned. It's got bloodshot eyes, messed up hair, a zoned out expression, and its fingers look just like they're holding a joint. It's pretty easy to mentally add an enormous smoldering doobie to the image.
The fish has a huge smile on his face and is staring directly at the seagull's virtual joint. He may be anticipating that the seagull is going to be kind enough to give him a puff or two.
The "diff," obviously, is a matter of degrees. Society can stand a little bit of unlawful activity.
It is a matter of degrees, but you're wrong about which degrees. Regardless of your misinformed opinion, it is actually legal for you to tape a show off HBO, and it is legal for you to lend that tape to a friend. The degree part comes in because the law is fuzzy and ill-defined as to the point where increased volume of copying and sharing actually becomes illegal.
To repeat: most people are not currently breaking the law with their VCRs. Your assumption that the vast majority of the population are petty thieves is simply wrong. What is being tested by this matter of degrees is not how much illegality is being tolerated; it is what activities are legal and what are actually illegal.
Now, P2P is probably beyond the fuzzy line defining illegality, but that's a different matter altogether.
Yeah, those hundreds of PhDs they have working there will *never* figure that out. I hear they started with a 16 bit signed integer for their primary key and only after months of hard work upgraded it to 32 bit. Time to close down shop, it's impossible to fix.
Actually, they already have the fix implemented, and it's currently in the process of being rolled out. The upgraded system makes use of a split primary key which comprised of a "selector" subkey and a "segment" subkey. The selector key is shifted left by four bits and then arithmetically added to the segment key. This clever scheme expands the index by a factor of 16; Google will soon be able to host over 64 billion pages!
Maybe the retard is the one who makes assumptions that weren't in the original statement.
So, simply put your claims are retarded. It takes an intense amount of effort to make a nuclear reactor go critical, theft is not a problem,
I never claimed that people were going to steal nuclear waste to get a critical mass for making bombs. A dirty bomb may not be the worst scenario, but it would release more radioactivity than quite a bit of coal, which was the original point.
However, I did mention diversion of nuclear materials into an illicit weapons program by a 2-bit country. Don't say it can't hapen; it has already happened several times, and our president is currently flummoxed trying to defuse the latest case of that in North Korea. If there were no nuclear power generation here or abroad, then these countries would have a far harder time hiding or justifying the kind of activities that go into developing nuclear weapons.
You also ignored the threat of sabotage or commando-style attack. The specific worst case is attacking the spent fuel storage pond, which can hold more radioactivity than the reactor itself, and which is seldom shielded, and which may spontaneously ignite if its cooling system is disabled.
Just like how coal plants release more radition and heavy metals into the environment that nuclear plants.
If and only if nothing ever goes wrong at even the worst managed of hundreds of independent sites around the world.
(And spare me the standard "modern designs can't fail" line. Sabotage, theft and diversion into illicit nuclear weapons programs are already far larger risks than technical mishaps, and they aren't addressed by reactor design.)
Probably because the airlines can still get a piece of the action with this scheme. They can do a deal with the phone companies to split an outrageously priced "roaming" fee for these calls. People would be much more likely to use their own phones than those crappy phones in the backs of seats. (I can't remember ever seeing anyone actually use one of those things.)
Even if they only charged half or less of the unbelievable current air phone price, the total call volume would go way up, and so would the airlines' revenues from phone services.
CC was the most liberal of them all, it would complile and run your email.
This highlights once again how Windows is a more flexible and modern development platform than Un*x. With Windows, email can be run automatically and remotely, without the need for a separate compilation step.
I'm no expert but if something's on fire, getting it wet is the least of your worries.
However, sometimes nothing is on fire. A while back they were doing a video in one of the labs near where I worked, and a hot light was set up too close to a sprinkler head and set it off. The entire lab was doused with the stinky stale water from the fire control system, ruining much of the equipment.
"If you were plowing a field, which would you rather use? Two strong oxen or 1024 chickens?"
When Seymour Cray made that statement, he was probably pointing out the difference between his he-man vector processors vs. clusters of the wimpy microprocessors of old.
After reading the article, it seems that this new Cray is powered by a bunch of the exact same AMD microprocessors that a cluster of Linux boxes would use. So what they have now is more like an ox-shaped sack stuffed with chickens.
If Java was truely Open Source, then Microsoft could have forked it to allowed J++ to exist on Windows and blow a hole in the "write once, run everywhere" theory.
Well, since Microsoft couldn't do that, they just switched to plan B. They used 5 years of hindsight to write a new language like Java, but with some nicer new features, then they applied this new language to their OS monopoly to get instant market penetration with little effort.
Meanwhile, Microsoft froze their support for Java until it was hopelessly obsolete; this passive-aggressive move blew a hole in "write once, run anywhere" all by itself. Microsoft's moves seem to have succeeded in taking much of the steam out of Sun's goal of taking over the world with Java.
It seems to me that this course of events was a big factor in Sun's recent "surrender". I don't see how things have come out any better for Sun than if they had set Java loose.
Maybe they do have more multi-purpose registers (I don't believe you 100%), but if you cannot access them via assembly code, then the compiler can't see them either.
Well, the way CPU performance works today isn't really intuitive. A modern CPU can slice each opcode into several independent primitive operations and run each of those independently. In fact, it can reorder the suboperations from a variety of opcodes and do the work as it can be done, delaying for later the primitives that depend on long-latency things like data from cache. It can also execute many operations after a branch before it knows that the branch will be taken, and throw away all of those results if the branch gets mispredicted. The CPU may be simultaneosly working on dozens of opcodes at any given point in time.
To support this craziness, the CPU uses "register renaming", which allots dynamic assignments for the user-visible registers from a bank of generic hardware registers. At any one time, you may have several versions of "EAX" simultaneously exist in the CPU; these represent the value of EAX at different logical points in the program code (some of the values may later be found to be useless because of speculative execution).
So what the programmer thinks of as a bottleneck of loading and storing EAX a couple of times in succession may turn out not to be a bottleneck at all if the values are logically independent. The instructions may be reordered so that both loads of EAX exist at the same time, regardless of what one would assume from looking at the linear opcode sequence. In this case you get to simultaneously use more registers than what you can see.
While its hard for assembly programmers to keep this straight, compiler writers can emit code that is aware of the CPUs behavior to take advantage of these features as much as possible. The X86 instruction set is a kind of bytecode abstraction; the compiler and CPU can mutually understand that there are ways to transcend the apparent limitations of that visible architecture.
The bottom line is that register pressure is an issue, but register renaming in the X86 helps to mitigate it. Moreover, AMD's 64-bit extension adds lots of new programmer-visible registers, further reducing the problem. The real challenges going forward with current CPU designs today are improving branch prediction with ever-deepening pipelines, increasing cache size as the CPU speed continues to outstrip DRAM speed, and managing power consumption as gate leakage and transistor count increase. All CPU architectures need to deal with these issues, X86 and RISC included.
(Itanium was meant to be a new approach to the branch-prediction issue, pushing the intelligence to the software compiler; it hasn't been a resounding success. It also really pushed the cache size by including monster caches, and this has been the main reason for its reputation as an expensive power guzzler. The CPU core really isn't that big or complex.)
It truly is a pain to do any assembly programming on the x86.
So?
The 99.9% of people writing apps in any langauge as abstract as C or higher don't have to worry about the CPU architecture. If it compiles and runs these languages at a price/performance ratio favorable to other CPUs, then nobody sould have a problem with it.
The true runtime architecture of an X86 CPU (and most RISC chips as well) has been mostly unfathomable to humans since the Pentium Pro came out. The X86 instruction set is just a backwards-compatible abstraction that is used to logically specify what needs to be done. The chip transforms these instructions to something completely different at runtime. For example, X86 chips already do have dozens of the "multi-purpose" registers you're pining for; you just don't see them at the visible instruction set level. When you do "assembly programming" on a modern CPU, you're not much closer to the real hardware than you are writing in C.
When you have the cash they do, you don't go away.
Well, they're supposed to have about $40B in cash. In 2002, AOL/TW posted a loss of $99B. If Microsoft were to start making mistakes (possible if, for example, current top management retires, gets bored, or tries to rise above its level of competence), then losses like that could put a major dent in their pile of cash in a relatively short time.
Of course the name would never go away completely. It seems that old tech companies never die, they just participate in questionable mergers with troubled peers.
Maybe sun doesn't need to actually sell the systems to Wal-mart. I recently bought a "GE" electric griddle at Wal-mart. I found it interesting when I looked at the fine print on the box and the grill itself. It said something along the lines of "designed exclusively for Wal-mart blah blah by so-and-so (not GE); made in China."
I quickly realized that General Electric had little if anything to do with this griddle. Wal-mart probably just paid GE a few cents to license their logo and slap it on the griddle and the box; otherwise Wal-mart probably had this product designed and imported all on their own. GE would have very little to lose from this arrangement. They have a very nice looking logo, and they get money for letting people print it with no additional work or risk on GE's part. The GE brand name also gets wider exposure this way.
If sun were smart they'd set up a similar scheme. Wal-mart would stock computers directly from from their current cut-rate suppliers, and they would pay Sun a couple of bucks to slap on the Sun logo and load the Sun software. Sun wouldn't have to worry about the actual price of the hardware as they would never actually touch it.
Slashdot Mirror
The full-install ISOs are available from SuSE; they just happen to be pressed onto CDs.
I've got a Beatles CD that says (C) 1963 right on the disc. Either the CD format was in beta longer than we thought, or more likely, the copyright applies to the audio recording independent of the medium it's fixed in.
Wrong. I've been through the patent application process quite a few times. In many cases our corporate lawyers had to drum into the engineers' heads that the patent office has a much lower standard of "novelty" and "obviousness" than what any reasonable person would think. Few if any of the applications that I was involved with were rejected.
Don't forget that there's a difference between the poorly functioning USPTO and the more stringent tests applied in infringements.
You seem to admit that large number of invalid patents are currently being issued. Like I said, they don't have to be inventive to be issued.
Maybe the courts are somewhat more prudent, but most parties won't be able to afford getting to that point. The market barrier works just fine through intimidation without the patent ever getting tested in court. In particular, Microsoft knows that their current top worry, Linux, will have a particularly hard time dealing with patents, valid or not. It's pretty easy to see that the large number of patents they are filing are intended to create a minefield to ensnare anybody trying to compete with them using distributed development model.
To get a patent, it doesn't have to be better than anything else, or even "inventive", it just has to be different and "useful" (i.e., it kind of works). That's enough to generate vendor lockin and competitor lockout.
Let's say they invent a new image format called "MSPEG", which is just like JPEG, but they add some twist to address some minor limitation. Now let's say that they make MSPEG the default image format for a large number of Windows applications. What becomes the most valuable aspect of the MSPEG IP? Is it the marginal new feature that almost nobody is taking advantage of, or is it the fact that everybody else is now using the format? Of course, it's the second item.
IMO, IP law should be changed to discount the vendor lockin portion of the IP value. Royalties and license fees should only apply to the extent that a work is instrinically better than other alternatives. Nobody should be able to collect fees just because they are a gatekeeper to compatibility with something. For example, the only value of the GIF format over the last few years has been compatibility with other GIF enabled software. All of its other functionality has long ago been superceded by other formats. Therefore, Unisys should not have been given the privilege to demand fees from people who have to use GIFs to be compatible with applications that require GIFs.
Mainframes are optimized for batch processing. Interactive queries do not take full advantage of their vaunted I/O capacity.
Moreover, while a mainframe may be a good way to host a single copy of a database that must remain internally consistent, that's not the problem Google is solving. It's trivial for them to run their search service off of thousands of replicated copies of the Internet index. Even the largest mainframe's storage I/O would be orders of magnitude smaller than the massively parallel I/O operations done by these thousands of PCs. Google has no reason to funnel all of the independent search queries into a single machine, so they shouln't buy a system architecture designed to do that.
As opposed to the original AT&T UNIX license which thankfully prevented companies like Sun, HP, IBM, SCO, SGI, DEC, BSDi, etc. from forking...
What terms of use? What agreement?
There is no contractual agreement between a DVD purchaser and the merchant.
Actually, I saw a documentary on dragsters a while back. They briefly covered a group of people who were trying to create a new class of riceboy-inspired nitro fueled funny cars that use 4-cylinder engines instead of V-8s. Unless I'm mistaken, even half of a modern dragster engine would have well over the 1000 HP of that W-16. Of course, the dragsters would have no muffler of any kind.
I wouldn't want to take any artifacts from the ark either, given that most of it would probably consist of thousands of different kinds of coprolites.
He can put down on his resume that he created one of the top 10 most popular Linux distros, and that he supervised quite a few interesting technical innovations unique to that distro.
This experience would help qualify for a job with a salary quite a bit more than $20,000 (not to mention more influence and responsibility) over that of a random code monkey.
I don't know. The first thing that I thought was that that seagull looks totally stoned. It's got bloodshot eyes, messed up hair, a zoned out expression, and its fingers look just like they're holding a joint. It's pretty easy to mentally add an enormous smoldering doobie to the image.
The fish has a huge smile on his face and is staring directly at the seagull's virtual joint. He may be anticipating that the seagull is going to be kind enough to give him a puff or two.
Hopefully the seagull doesn't have the munchies.
It is a matter of degrees, but you're wrong about which degrees. Regardless of your misinformed opinion, it is actually legal for you to tape a show off HBO, and it is legal for you to lend that tape to a friend. The degree part comes in because the law is fuzzy and ill-defined as to the point where increased volume of copying and sharing actually becomes illegal.
To repeat: most people are not currently breaking the law with their VCRs. Your assumption that the vast majority of the population are petty thieves is simply wrong. What is being tested by this matter of degrees is not how much illegality is being tolerated; it is what activities are legal and what are actually illegal.
Now, P2P is probably beyond the fuzzy line defining illegality, but that's a different matter altogether.
Actually, they already have the fix implemented, and it's currently in the process of being rolled out. The upgraded system makes use of a split primary key which comprised of a "selector" subkey and a "segment" subkey. The selector key is shifted left by four bits and then arithmetically added to the segment key. This clever scheme expands the index by a factor of 16; Google will soon be able to host over 64 billion pages!
So, simply put your claims are retarded. It takes an intense amount of effort to make a nuclear reactor go critical, theft is not a problem,
I never claimed that people were going to steal nuclear waste to get a critical mass for making bombs. A dirty bomb may not be the worst scenario, but it would release more radioactivity than quite a bit of coal, which was the original point.
However, I did mention diversion of nuclear materials into an illicit weapons program by a 2-bit country. Don't say it can't hapen; it has already happened several times, and our president is currently flummoxed trying to defuse the latest case of that in North Korea. If there were no nuclear power generation here or abroad, then these countries would have a far harder time hiding or justifying the kind of activities that go into developing nuclear weapons.
You also ignored the threat of sabotage or commando-style attack. The specific worst case is attacking the spent fuel storage pond, which can hold more radioactivity than the reactor itself, and which is seldom shielded, and which may spontaneously ignite if its cooling system is disabled.
If and only if nothing ever goes wrong at even the worst managed of hundreds of independent sites around the world.
(And spare me the standard "modern designs can't fail" line. Sabotage, theft and diversion into illicit nuclear weapons programs are already far larger risks than technical mishaps, and they aren't addressed by reactor design.)
Probably because the airlines can still get a piece of the action with this scheme. They can do a deal with the phone companies to split an outrageously priced "roaming" fee for these calls. People would be much more likely to use their own phones than those crappy phones in the backs of seats. (I can't remember ever seeing anyone actually use one of those things.)
Even if they only charged half or less of the unbelievable current air phone price, the total call volume would go way up, and so would the airlines' revenues from phone services.
This highlights once again how Windows is a more flexible and modern development platform than Un*x. With Windows, email can be run automatically and remotely, without the need for a separate compilation step.
Google: "Searching 4,285,199,774 web pages" That's quite a big difference.
At least this Gigablast name is closer to the truth. They are only exaggerating their page count by a factor of 3.7 : 1.
By my math, Google comes up short by 2.3x10^90 : 1.
However, sometimes nothing is on fire. A while back they were doing a video in one of the labs near where I worked, and a hot light was set up too close to a sprinkler head and set it off. The entire lab was doused with the stinky stale water from the fire control system, ruining much of the equipment.
When Seymour Cray made that statement, he was probably pointing out the difference between his he-man vector processors vs. clusters of the wimpy microprocessors of old.
After reading the article, it seems that this new Cray is powered by a bunch of the exact same AMD microprocessors that a cluster of Linux boxes would use. So what they have now is more like an ox-shaped sack stuffed with chickens.
Well, since Microsoft couldn't do that, they just switched to plan B. They used 5 years of hindsight to write a new language like Java, but with some nicer new features, then they applied this new language to their OS monopoly to get instant market penetration with little effort.
Meanwhile, Microsoft froze their support for Java until it was hopelessly obsolete; this passive-aggressive move blew a hole in "write once, run anywhere" all by itself. Microsoft's moves seem to have succeeded in taking much of the steam out of Sun's goal of taking over the world with Java.
It seems to me that this course of events was a big factor in Sun's recent "surrender". I don't see how things have come out any better for Sun than if they had set Java loose.
Well, the way CPU performance works today isn't really intuitive. A modern CPU can slice each opcode into several independent primitive operations and run each of those independently. In fact, it can reorder the suboperations from a variety of opcodes and do the work as it can be done, delaying for later the primitives that depend on long-latency things like data from cache. It can also execute many operations after a branch before it knows that the branch will be taken, and throw away all of those results if the branch gets mispredicted. The CPU may be simultaneosly working on dozens of opcodes at any given point in time.
To support this craziness, the CPU uses "register renaming", which allots dynamic assignments for the user-visible registers from a bank of generic hardware registers. At any one time, you may have several versions of "EAX" simultaneously exist in the CPU; these represent the value of EAX at different logical points in the program code (some of the values may later be found to be useless because of speculative execution).
So what the programmer thinks of as a bottleneck of loading and storing EAX a couple of times in succession may turn out not to be a bottleneck at all if the values are logically independent. The instructions may be reordered so that both loads of EAX exist at the same time, regardless of what one would assume from looking at the linear opcode sequence. In this case you get to simultaneously use more registers than what you can see.
While its hard for assembly programmers to keep this straight, compiler writers can emit code that is aware of the CPUs behavior to take advantage of these features as much as possible. The X86 instruction set is a kind of bytecode abstraction; the compiler and CPU can mutually understand that there are ways to transcend the apparent limitations of that visible architecture.
The bottom line is that register pressure is an issue, but register renaming in the X86 helps to mitigate it. Moreover, AMD's 64-bit extension adds lots of new programmer-visible registers, further reducing the problem. The real challenges going forward with current CPU designs today are improving branch prediction with ever-deepening pipelines, increasing cache size as the CPU speed continues to outstrip DRAM speed, and managing power consumption as gate leakage and transistor count increase. All CPU architectures need to deal with these issues, X86 and RISC included.
(Itanium was meant to be a new approach to the branch-prediction issue, pushing the intelligence to the software compiler; it hasn't been a resounding success. It also really pushed the cache size by including monster caches, and this has been the main reason for its reputation as an expensive power guzzler. The CPU core really isn't that big or complex.)
So?
The 99.9% of people writing apps in any langauge as abstract as C or higher don't have to worry about the CPU architecture. If it compiles and runs these languages at a price/performance ratio favorable to other CPUs, then nobody sould have a problem with it.
The true runtime architecture of an X86 CPU (and most RISC chips as well) has been mostly unfathomable to humans since the Pentium Pro came out. The X86 instruction set is just a backwards-compatible abstraction that is used to logically specify what needs to be done. The chip transforms these instructions to something completely different at runtime. For example, X86 chips already do have dozens of the "multi-purpose" registers you're pining for; you just don't see them at the visible instruction set level. When you do "assembly programming" on a modern CPU, you're not much closer to the real hardware than you are writing in C.
Well, they're supposed to have about $40B in cash. In 2002, AOL/TW posted a loss of $99B. If Microsoft were to start making mistakes (possible if, for example, current top management retires, gets bored, or tries to rise above its level of competence), then losses like that could put a major dent in their pile of cash in a relatively short time.
Of course the name would never go away completely. It seems that old tech companies never die, they just participate in questionable mergers with troubled peers.
I quickly realized that General Electric had little if anything to do with this griddle. Wal-mart probably just paid GE a few cents to license their logo and slap it on the griddle and the box; otherwise Wal-mart probably had this product designed and imported all on their own. GE would have very little to lose from this arrangement. They have a very nice looking logo, and they get money for letting people print it with no additional work or risk on GE's part. The GE brand name also gets wider exposure this way.
If sun were smart they'd set up a similar scheme. Wal-mart would stock computers directly from from their current cut-rate suppliers, and they would pay Sun a couple of bucks to slap on the Sun logo and load the Sun software. Sun wouldn't have to worry about the actual price of the hardware as they would never actually touch it.