... lived in the so-called "age of reason". The economic theory he created is pretty much based on the principle that people act rationally and in their best interest when doing an economic transaction.
Unfortunately, that's simply not true. Perhaps the best example is the herd behavior that's very cleverly exploited by Microsoft. Not to mention the efficacy of Microsoft-sponsored FUD. There are actually people out there who consider that Windoze is more secure and stable... no kiddin'.
It's true that gcc can generate P4-specific code, for instance, but in most cases and with generic compilation flags this will barely make a difference. I am personally not aware of a single mainstream linux app that does 5% better when optimized for P4 as opposed to generic 386 (I'm not including here apps that have hand-crafted assembly for a couple of instruction-set architectures, like, say, mplayer). I guess things might change slightly with gcc 4.* which has automatic vectorization, but in my experience automatically-vectorizable C code is very rare, unless written specifically that way.
You're absolutely right here, and I totally agree that it's ultimately up to a judge. The FAQ of the license, however, specifically identifies static and dynamic linking as derivative works; anything else (i.e. talking over a socket) falls under the "gray area".
The question, of course, is what represents derivative work in such a case. IANAL, but something that can be easily identified as a distinct part, running in a different process, etc and doesn't include a single line of code of the original codebase is very unlikely to be considered derived work (GPL has a specific paragraph on that, although it's somewhat vague). Even if the only db it works with is MySql.
Unfortunately, we don't have a legal precedent for this. Regardless, it's very similar to what nVidia is doing with its linux binaries
I really think that using a more generic interface, such as JDBC/ODBC would effectively act as a GPL barrier. Sure, it's not as fast, but it should work. Furthermore, if you built from scratch a library that talked with the mysql server, you could give it a generic license.
I'm not arguing that Oracle didn't do it with the intention to kill the competitor, just that the consequences aren't as drastic, at least not in the short term.
True, an ideal CMOS doesn't have any leakage; these days, however, the very small feature sizes translate into more and more leakage, mainly because of the tunneling effect. It's not uncommon for the leakage power to be one fifth of the entire power consumption; unfortunately, with future generations, this will only get worse.
Are you saying that regulations are bad? E.g. regulations that would inforce neutrality on the net? I'm sorry, but this seems a bit paradoxical:). I personally consider that regulations should be kept to a minimum, but some of them are absolutely necessary.
It's *very* difficult to get a compiler to exploit this kind of parallelism. Unless you're doing scientific Fortran loopy code, where it's much easier to do things like automatic vectorization/parallelization, it's basically almost impossible for the compiler (out of curiosity, try to use the automatic openmp parallelization feature within Intel C Compiler on standard C/C++ code; the results will likely underwhelm you). Unfortunately, even if you do have scientific code, the slave processing units only do simple precision (IIRC).
In my opinion, this thing will run well games, but that's about it. I've seen so far 2 presentations by IBM about the Cell processor (at (micro-)architecture conferences). Both times, the question on everybody's mind was "How do you program these things?". The answer was pretty much a hand-wavy "oh hmmm, well, blah blah blah manual"
Well, more like lack of something. Lower gravity means that it's much easier to launch something off the moon. Lack of atmosphere means that you get a lot more bang for your solar panel buck.
His rule marked the transition from the "heretic" Atenism (worshiping of the Sun god as the only true god) of his father back to the old ways of the Egyptian religion. For example, his name was originally Tutankhaten (Living Image of Aten) but he changed it to Tutankhammun (Living Image of Ammun), to show that he abandoned the religion of his father [thanks wiki]. Those were really troubled times, so it's quite interesting to know why exactly he died.
There's many things we don't know, starting with the origin of hommo sapiens. I personally consider historical non-determinism to make the subject much more interesting.
Unmodded, the temperature after a couple of hours of Quake 4 was ~150F. With watter cooling it got to ~100F. Such a dramatic change in temperature can greatly increase the lifetime of the system.
XBOX360 is an awesome gaming platform, but it's pretty clear that the first version is doing poorly power- and cooling-wise.
As a society, we have to make sure that girls aren't dissuaded/stereotyped-away from pursuing a CS career. However, if they choose not to pursue such a career, I don't think that's a real problem.
Here's my 2-cent theory: studies have shown that the average female has a vocabulary that's twice as large as the one of the average male (interestingly-enough, starting with a very early age). Of course, this is just an average, it's very easy to find people that do not match the average. IMO, this difference simply makes the average woman more inclined to pursue careers which involve communication. In IT, most of the work is just between you and the computer, there's little human interaction going around.
The instruction set architecture (ISA) defines the user-processor contract (what instructions do, but also other features such as protection, etc). The micro-architecture is how you implement the ISA. In-order vs. out-of-order is simply a micro-architectural choice.
The fact that the processor executes in-order only affects performance, and not the compatibility. For example, most x86 code runs on in-order cores (such as the 486, Pentium I), out-of-order cores (most modern x86 processors) and even VLIW cores with a translation layer (Transmeta)
The main problem with such a system is that in some cases it will actually contribute to accidents. Now, the United States significantly decrease the accuracy of the GPS system in war time (understandably so). Are we gonna see more car accidents in Canada when the US goes to war ?
While I agree that SPEC numbers are better than GHz, the "memory-bound when new" part is simply inaccurate. The current trend in architecture is that memory performance increases much slower than processor performance, and as a consequence the relative speed of memory vs. the speed of the core is getting worse. Because of this, benchmarks tend to become more and more memory bound (not the other way around, as you imply). If you don't believe me, please read "MLP yes! ILP no!" by Andy Glew (from Intel)
Slashdot Mirror
Unfortunately, that's simply not true. Perhaps the best example is the herd behavior that's very cleverly exploited by Microsoft. Not to mention the efficacy of Microsoft-sponsored FUD. There are actually people out there who consider that Windoze is more secure and stable ... no kiddin'.
... is all you're gonna get on a USB connection (at 5v) ... That's pretty punny for a regular hdd.
It's true that gcc can generate P4-specific code, for instance, but in most cases and with generic compilation flags this will barely make a difference. I am personally not aware of a single mainstream linux app that does 5% better when optimized for P4 as opposed to generic 386 (I'm not including here apps that have hand-crafted assembly for a couple of instruction-set architectures, like, say, mplayer). I guess things might change slightly with gcc 4.* which has automatic vectorization, but in my experience automatically-vectorizable C code is very rare, unless written specifically that way.
Approximately 1 American life for 100 Iraqi lives
Software != books. Sure, such analogies probably influence a judge as far as precedent is concerned, but it's still a very far stretch.
You're absolutely right here, and I totally agree that it's ultimately up to a judge. The FAQ of the license, however, specifically identifies static and dynamic linking as derivative works; anything else (i.e. talking over a socket) falls under the "gray area".
Unfortunately, we don't have a legal precedent for this. Regardless, it's very similar to what nVidia is doing with its linux binaries
I'm not arguing that Oracle didn't do it with the intention to kill the competitor, just that the consequences aren't as drastic, at least not in the short term.
True, an ideal CMOS doesn't have any leakage; these days, however, the very small feature sizes translate into more and more leakage, mainly because of the tunneling effect. It's not uncommon for the leakage power to be one fifth of the entire power consumption; unfortunately, with future generations, this will only get worse.
Why did you let them in? Let them get a warrant instead.
Are you saying that regulations are bad? E.g. regulations that would inforce neutrality on the net? I'm sorry, but this seems a bit paradoxical :). I personally consider that regulations should be kept to a minimum, but some of them are absolutely necessary.
... it ends with the USA losing its technological edge
... between trading with a dictatorship and being an instrument of censorship for that dictatorship.
In my opinion, this thing will run well games, but that's about it. I've seen so far 2 presentations by IBM about the Cell processor (at (micro-)architecture conferences). Both times, the question on everybody's mind was "How do you program these things?". The answer was pretty much a hand-wavy "oh hmmm, well, blah blah blah manual"
Well, more like lack of something. Lower gravity means that it's much easier to launch something off the moon. Lack of atmosphere means that you get a lot more bang for your solar panel buck.
Do you think that the current model will work for the next 5 years? Are you considering P2P as a way to reduce load on the servers ?
There's many things we don't know, starting with the origin of hommo sapiens. I personally consider historical non-determinism to make the subject much more interesting.
Would you buy a game knowing that its own developer doesn't believe in ?
*NOT* terms and conditions. This is the style of GNU licenses - preamble describing intent, but not legally binding, then the legally-binding terms .
XBOX360 is an awesome gaming platform, but it's pretty clear that the first version is doing poorly power- and cooling-wise.
Here's my 2-cent theory: studies have shown that the average female has a vocabulary that's twice as large as the one of the average male (interestingly-enough, starting with a very early age). Of course, this is just an average, it's very easy to find people that do not match the average. IMO, this difference simply makes the average woman more inclined to pursue careers which involve communication. In IT, most of the work is just between you and the computer, there's little human interaction going around.
They're quite effective at showing a nice facade to the west, while most of the population doesn't really have access to the facade goodies.
The fact that the processor executes in-order only affects performance, and not the compatibility. For example, most x86 code runs on in-order cores (such as the 486, Pentium I), out-of-order cores (most modern x86 processors) and even VLIW cores with a translation layer (Transmeta)
The main problem with such a system is that in some cases it will actually contribute to accidents. Now, the United States significantly decrease the accuracy of the GPS system in war time (understandably so). Are we gonna see more car accidents in Canada when the US goes to war ?
While I agree that SPEC numbers are better than GHz, the "memory-bound when new" part is simply inaccurate. The current trend in architecture is that memory performance increases much slower than processor performance, and as a consequence the relative speed of memory vs. the speed of the core is getting worse. Because of this, benchmarks tend to become more and more memory bound (not the other way around, as you imply). If you don't believe me, please read "MLP yes! ILP no!" by Andy Glew (from Intel)