Also, Linux/Open source cannot sue!! How can you show anti-competitive behaviour on an entity that has no legal existence? Especially one which makes no profit?
Linux cannot sue. However RedHat, VA and other CAN sue. They probably won't, but they still can. For instance, if Microsoft uses illegal tactics (how suprising!) to get a contract instead of RedHat, they could be sued and it could be possible to prove that some damage has been done.
I think the web has a much stronger effect than digital photos. Whenever a web site changes, you can't get the previous version anymore. Even if it is archived, it will probably end up getting lost within 20 years. In a couple years, there'll be no trace left of what the web looks like today and I'm not sure we could recover more than 10% of what was there just 5 years ago.
Even worse, there's something 1984-esque about the web: it allows to "modify history" at will. You can change the content of your web site without anybody really being able to "prove" that you did.
Re:Good for the average joe
on
Netscape 6.2
·
· Score: 2
I'm using Netscape 6.2 (Linux) right now, and it looks pretty good. There are some advantages to Netscape vs. Mozilla. The most important one is QA. I'm betting NS 6.2 has less really annoying bugs than you'd expect to find in a nightly build, or even a Mozilla milestone. Also, so far NS 6.2 is the first mozilla-based browser for which Java worked out of the box on my Linux workstation.
I remember playing with bottle rockets a while ago. I don't remember the pressure, but we were using a bike pump I think. Unlike what they're showing, our nozzled was actually just the end of the bottle. When we'd let it go, it emptied almost instantly while the bottle would rise above 10 meters faster than we could track it.
Wonder what it would look like when fired horizontally. I bet you'd do the 20 meters in about one second. I'm not sure you could make it stay on the ground though. And, well there might be safety issues...
I wonder whether most people would agree with me or not. Unlike some people that say "this action should be considered terrorism", for me terrorism has nothing to do with the actions and everything to do with the intent.
To me, if Mr. X puts a bomb in a plane to kill his wife, that's first degree murder (though not terrorism). However if Mr. Y does the same thing for a political cause, it is terrorism, although the action is exactly the same. The same way, for me a serial killer is not a terrorist, though I think none is "better" of "worse" than the other,
Does that make any sense. Surely at some point it could be hard to determine the intent in a trial, but for me it's important to make the distinction. Otherwise you just end up with all crimes being labeled as "terrorism act" and the word doesn't mean anything anymore.
...Or maybe it's just the province of Quebec. We used to get calls from auto-dialers and it stopped about 5-10 years ago when they passed the law. I haven't received any such calls since then.
In Canada, it is now illegal to do "automatic telemarketing", that is it has to be a real person calling. That decreases a lot the amount of telemarketing calls we have.
You forgot "solving systems of 50,000 equations." People always bring that one up, as unrealistic as it.
(Note: I'm not talkig about home use here) Actually, 50,000 equations is a rather small system. Any idea what weather prediction looks like? Something like 10 equations per grid point, with a grid that's something like 200x200x50=2,000,000. So you end up with a 20 million equation system. Also, many CAD software (eg finite element simulations) also need to solve *huge* systems. The faster the computer, the more precise the simulation (because you can afford more grid points).
I hold the copyright for a couple (mostly GPL) programs. Am I allowed to hack into the RIAA machines so I can verify they're not violating the GPL? That would be interesting. For some time there's been rumors of MS using GPL'ed code in Windows, I guess we can verify that now!
Thinking about it, I'm not sure even the RIAA wants that kind of laws...
I think the main difference between SCSI and IDE drives is not the interface, but the general quality of the components. If it were only the interface, then the price difference would be constant across the (capacity) range. What happens is that SCSI is targeted at servers that have to be reliable, while IDE drives just minimize capacity/$. Haven't looked at the numbers, but I'd expect that MTBF for SCSI to be more than twice what it is for IDE.
Of course I'm often wrong, but I don't say things like "I'll attack any other nation that harbors terrorists"... That's the difference. At least admit your mistakes and don't try to play "good guy vs. bad guy" with people basically doing the same stuff you do.
That one always makes me laugh. After all the terrorist groups the US helped (via CIA,...), including Bin Laden's group, I guess it's time the US follows the next logical step. "US declares war to itself and bombs Washington after it refused to turn over to itself ex-CIA directors."
Then would the US have extradited the pilots involved in the "accident" between the american plane and the chinese fighter a while ago (well, if they hadn't land in china in the first place)?
Of course, it's not exactly the same case and we could spend a lot of time arguing about that. But the thing is: this is probably the first time war is declare to a country for *one* person.
What's hypocritical is the way they're trying to justify with this "poor afghan people" thing, almost like it's a humanitary thing. BTW, don't get me wrong, I'll be pretty happy when (if?) the Taliban are gone.
By the way, speaking of Give up Bin Ladin or give up power, what would bw the american opinion if some american guy does something bad in, say China, and the chinese government says to the US: "Give us this guy or we bomb you"?
Just because the U.S. isn't being hypocritical dealing with the Taliban...
Then could you explain to me why the US never really cared too much about them (unlike Iraq, which they bomb everytime the persident has an affair) until sept. 11? They don't care about the Taliban and the people in Afghanistan, they just want bin Laden and that's all.
Can't comment for other OS, but Linux tends to be pretty good at using all the RAM you've got to cache disk data. Even though I rarely use more than 256 MB, upgrading to 768 MB made a significant performance improvement for me, as Linux quickly fill the remaining 512 MB with disk cache, without me bothering with setting up a ramdrive.
Re:Check out the Preemptible Kernel patches...
on
Kernel 2.4.11 Released
·
· Score: 5, Informative
These sound real good. Is there a reason that these patches are not the default behavior? Is there a downside to having a premptible kernel?
AFAIK, there are two reasons why these patches aren't in default kernel. First, I understand that decreases latency at the price of slightly decreasing throughput. The second is that though the patch is small, its effects can be complex and nobody's too sure it doesn't have any bad side effects (crash, oops,...), especially on SMP systems.
I've never tried optimizing for SSE, but someone in lab did once. He reported higher performance when doing his computations element-at-a-time than vector-at-a-time. His conclusion, for his particular application, was that memory latency was killing SSE.
Maybe his problem was really special, but most likely he didn't know how to write SSE code. First, if you write your code correctly, the worse you can do is a bit better than the x87, because you can use SSE with scalar and take advantage of the linear register model (as opposed to stack).
The only time I've converted some code for SSE, I got a 2-3x improvement in speed. There is one thing you really have to be careful when writing SSE code: ALIGNE EVERYTHING TO 16 BYTES. That's very important. The "movaps" (move aligned packed single) is more than twice faster than the "movups" (unaligned) when data is unaligned (movups on aligned is not too bad). That makes all the difference. Also, sometimes you need to change the way you do your computations.
For the case I have (optimizing a speech recognition engine), just by changing the order of of the loops (inner vs. outter), we got a 2-3x improvement (still with x87) because of increased L1 hit rate. Then when switching to SSE, it ended up to a 5x improvement over the original code. Had I just re-written the code in SSE (with cache optimization), the gain would have been around 25%, because memory would still be the bottleneck.
As for libraries not being optimized, just look at FFTW (www.fftw.org) or ATLAS (www.netlib.org/atlas) and you might change your mind.
On the other hand, for a weather simulation, I would bet on the cluster.
No way. Weather sinulations involves lots of linear algebra on huge matrices. When you parallelize that, you need a lot of communication between the nodes. With a cluster of P100, communication will kill you right there (10x of so penalty). It's not that much the network bandwidth, but the latency. Weather simulation is one of the hardest problems to parallelize and that's why until recently, SMP was prefered to MPP (and of course clusters of small workstations).
As for the memory bandwidth, depending on the problem, sometimes the L1 is really effective.
I'm seeing P4s do a fair bit less than the P3 per cycle, and the P3, P2, and PPro cores didn't seem *that* much faster per clock than the original Pentiums.
For "unoptimized" applications, that may be approximatly true. However, if you're going to build a cluster, you're also going to optimize your code for it. What kills the P4 is branch misprediction. However, by carefully writing your code, you can avoid most of these problems. Also, most of the big clusters are for numerical code, for which branch predictions does well (plus you can do lots of loop unrolling).
Another thing that the P4 (and PIII) has is SSE. On a P4 2 GHz, you can do theoratically 8 gflops. In practice, if you write good code, you'll bet between 1 and 2 gflops. On a plain pentium, the FPU is not pipelined, so a P100 (I'm guessing here) probably has a theoretical maximum of ~25 mflops with a performance for real code around ~10 mflops. That means the P4 is probably 100-200 times faster at floating that a P100.
Of course, you're right in saying that other architectures are probably faster that P4....and by the way I'm not saying that the P4 is great... but if you're doing numerical stuff and using SSE, it's VERY fast (in my experience, 3DNow! has been faster than SSE at the same clock rate, but 2GHz is too much higher than the fastest Athlon).
Actually, the best MIPS/Wh is probably with the slower versions of the current laptop chips. Maybe portable G3/G4?
Also, I don't think you'd get much useful stuff done with early Pentiums and 486. Consider that a P4 2 GHz has 20 times the clock speed and probably does twice as much per cycle, so it's ~40X faster. Now, if you connect 40 P100 together, unless your problem is completly parallel (like breaking keys, as opposed to most linear algebra), you're going to lose at least a factor of 2 there. This means that in order to equal 1 P4 @ 2 GHz, you'll need almost 100 Pentium 100 Mhz. This means that 10 P4 would be like a thousand Pentiums. At these numbers, it's going to cost so much in networking and power...
I'd say (pure opinion here) the slower you'd want to have today is something like a Duron 1 GHz and the best MIPS/$ is probably with a bunch of dual Athlon 1.4 GHz (A dual is not cheaper that 2 single, but you get more done because of parallelism issues).
I think there's a middle between the two extremes. I think you should not be liable *except* in the few cases where it's pretty obvious that the intent was to directly cause damage. If someone writes an editor that has a bug that ends up wiping your drive, it's an accident. However if the same guy advertises an editor that startes by system("/bin/rm -rf $HOME"), I think it should be illegal. Where to draw the line, I don't know.
But advocating that you're not responsible if you spread worms/virus is plain dangerous. Why don't you extend that to crasking into systems... after all, you're just sending bits to random ports on a system...
Slashdot Mirror
Also, Linux/Open source cannot sue!! How can you show anti-competitive behaviour on an entity that has no legal existence? Especially one which makes no profit?
Linux cannot sue. However RedHat, VA and other CAN sue. They probably won't, but they still can. For instance, if Microsoft uses illegal tactics (how suprising!) to get a contract instead of RedHat, they could be sued and it could be possible to prove that some damage has been done.
I think the web has a much stronger effect than digital photos. Whenever a web site changes, you can't get the previous version anymore. Even if it is archived, it will probably end up getting lost within 20 years. In a couple years, there'll be no trace left of what the web looks like today and I'm not sure we could recover more than 10% of what was there just 5 years ago.
Even worse, there's something 1984-esque about the web: it allows to "modify history" at will. You can change the content of your web site without anybody really being able to "prove" that you did.
I'm using Netscape 6.2 (Linux) right now, and it looks pretty good. There are some advantages to Netscape vs. Mozilla. The most important one is QA. I'm betting NS 6.2 has less really annoying bugs than you'd expect to find in a nightly build, or even a Mozilla milestone. Also, so far NS 6.2 is the first mozilla-based browser for which Java worked out of the box on my Linux workstation.
I remember playing with bottle rockets a while ago. I don't remember the pressure, but we were using a bike pump I think. Unlike what they're showing, our nozzled was actually just the end of the bottle. When we'd let it go, it emptied almost instantly while the bottle would rise above 10 meters faster than we could track it.
Wonder what it would look like when fired horizontally. I bet you'd do the 20 meters in about one second. I'm not sure you could make it stay on the ground though. And, well there might be safety issues...
I wonder whether most people would agree with me or not. Unlike some people that say "this action should be considered terrorism", for me terrorism has nothing to do with the actions and everything to do with the intent.
To me, if Mr. X puts a bomb in a plane to kill his wife, that's first degree murder (though not terrorism). However if Mr. Y does the same thing for a political cause, it is terrorism, although the action is exactly the same. The same way, for me a serial killer is not a terrorist, though I think none is "better" of "worse" than the other,
Does that make any sense. Surely at some point it could be hard to determine the intent in a trial, but for me it's important to make the distinction. Otherwise you just end up with all crimes being labeled as "terrorism act" and the word doesn't mean anything anymore.
In Quebec, a surprising proportion of the population insist they can't speak English.
Yes, I do that often. It's very convenient, a bit like saying you can't read when called by the newspaper telemarketers.
...and by the way, a lot (probably 30%-50%) of people in Quebec can't speak Englih.
Sorry, as I said in the other thread, it was probably Quebec only. I haven't had an auto-dialer call for quite a long time.
...Or maybe it's just the province of Quebec. We used to get calls from auto-dialers and it stopped about 5-10 years ago when they passed the law. I haven't received any such calls since then.
In Canada, it is now illegal to do "automatic telemarketing", that is it has to be a real person calling. That decreases a lot the amount of telemarketing calls we have.
You forgot "solving systems of 50,000 equations." People always bring that one up, as unrealistic as it.
(Note: I'm not talkig about home use here) Actually, 50,000 equations is a rather small system. Any idea what weather prediction looks like? Something like 10 equations per grid point, with a grid that's something like 200x200x50=2,000,000. So you end up with a 20 million equation system. Also, many CAD software (eg finite element simulations) also need to solve *huge* systems. The faster the computer, the more precise the simulation (because you can afford more grid points).
I hold the copyright for a couple (mostly GPL) programs. Am I allowed to hack into the RIAA machines so I can verify they're not violating the GPL? That would be interesting. For some time there's been rumors of MS using GPL'ed code in Windows, I guess we can verify that now!
Thinking about it, I'm not sure even the RIAA wants that kind of laws...
I think the main difference between SCSI and IDE drives is not the interface, but the general quality of the components. If it were only the interface, then the price difference would be constant across the (capacity) range. What happens is that SCSI is targeted at servers that have to be reliable, while IDE drives just minimize capacity/$. Haven't looked at the numbers, but I'd expect that MTBF for SCSI to be more than twice what it is for IDE.
Of course I'm often wrong, but I don't say things like "I'll attack any other nation that harbors terrorists"... That's the difference. At least admit your mistakes and don't try to play "good guy vs. bad guy" with people basically doing the same stuff you do.
any other nation that harbors terrorists
...), including Bin Laden's group, I guess it's time the US follows the next logical step. "US declares war to itself and bombs Washington after it refused to turn over to itself ex-CIA directors."
That one always makes me laugh. After all the terrorist groups the US helped (via CIA,
Then would the US have extradited the pilots involved in the "accident" between the american plane and the chinese fighter a while ago (well, if they hadn't land in china in the first place)?
Of course, it's not exactly the same case and we could spend a lot of time arguing about that. But the thing is: this is probably the first time war is declare to a country for *one* person.
How is that being hypocritical ?
What's hypocritical is the way they're trying to justify with this "poor afghan people" thing, almost like it's a humanitary thing. BTW, don't get me wrong, I'll be pretty happy when (if?) the Taliban are gone.
By the way, speaking of Give up Bin Ladin or give up power, what would bw the american opinion if some american guy does something bad in, say China, and the chinese government says to the US: "Give us this guy or we bomb you"?
Just because the U.S. isn't being hypocritical dealing with the Taliban...
Then could you explain to me why the US never really cared too much about them (unlike Iraq, which they bomb everytime the persident has an affair) until sept. 11? They don't care about the Taliban and the people in Afghanistan, they just want bin Laden and that's all.
Can't comment for other OS, but Linux tends to be pretty good at using all the RAM you've got to cache disk data. Even though I rarely use more than 256 MB, upgrading to 768 MB made a significant performance improvement for me, as Linux quickly fill the remaining 512 MB with disk cache, without me bothering with setting up a ramdrive.
These sound real good. Is there a reason that these patches are not the default behavior? Is there a downside to having a premptible kernel?
...), especially on SMP systems.
AFAIK, there are two reasons why these patches aren't in default kernel. First, I understand that decreases latency at the price of slightly decreasing throughput. The second is that though the patch is small, its effects can be complex and nobody's too sure it doesn't have any bad side effects (crash, oops,
I've never tried optimizing for SSE, but someone in lab did once. He reported higher performance when doing his computations element-at-a-time than vector-at-a-time. His conclusion, for his particular application, was that memory latency was killing SSE.
Maybe his problem was really special, but most likely he didn't know how to write SSE code. First, if you write your code correctly, the worse you can do is a bit better than the x87, because you can use SSE with scalar and take advantage of the linear register model (as opposed to stack).
The only time I've converted some code for SSE, I got a 2-3x improvement in speed. There is one thing you really have to be careful when writing SSE code: ALIGNE EVERYTHING TO 16 BYTES. That's very important. The "movaps" (move aligned packed single) is more than twice faster than the "movups" (unaligned) when data is unaligned (movups on aligned is not too bad). That makes all the difference. Also, sometimes you need to change the way you do your computations.
For the case I have (optimizing a speech recognition engine), just by changing the order of of the loops (inner vs. outter), we got a 2-3x improvement (still with x87) because of increased L1 hit rate. Then when switching to SSE, it ended up to a 5x improvement over the original code. Had I just re-written the code in SSE (with cache optimization), the gain would have been around 25%, because memory would still be the bottleneck.
As for libraries not being optimized, just look at FFTW (www.fftw.org) or ATLAS (www.netlib.org/atlas) and you might change your mind.
On the other hand, for a weather simulation, I would bet on the cluster.
No way. Weather sinulations involves lots of linear algebra on huge matrices. When you parallelize that, you need a lot of communication between the nodes. With a cluster of P100, communication will kill you right there (10x of so penalty). It's not that much the network bandwidth, but the latency. Weather simulation is one of the hardest problems to parallelize and that's why until recently, SMP was prefered to MPP (and of course clusters of small workstations).
As for the memory bandwidth, depending on the problem, sometimes the L1 is really effective.
I'm seeing P4s do a fair bit less than the P3 per cycle, and the P3, P2, and PPro cores didn't seem *that* much faster per clock than the original Pentiums.
...and by the way I'm not saying that the P4 is great... but if you're doing numerical stuff and using SSE, it's VERY fast (in my experience, 3DNow! has been faster than SSE at the same clock rate, but 2GHz is too much higher than the fastest Athlon).
For "unoptimized" applications, that may be approximatly true. However, if you're going to build a cluster, you're also going to optimize your code for it. What kills the P4 is branch misprediction. However, by carefully writing your code, you can avoid most of these problems. Also, most of the big clusters are for numerical code, for which branch predictions does well (plus you can do lots of loop unrolling).
Another thing that the P4 (and PIII) has is SSE. On a P4 2 GHz, you can do theoratically 8 gflops. In practice, if you write good code, you'll bet between 1 and 2 gflops. On a plain pentium, the FPU is not pipelined, so a P100 (I'm guessing here) probably has a theoretical maximum of ~25 mflops with a performance for real code around ~10 mflops. That means the P4 is probably 100-200 times faster at floating that a P100.
Of course, you're right in saying that other architectures are probably faster that P4.
Actually, the best MIPS/Wh is probably with the slower versions of the current laptop chips. Maybe portable G3/G4?
Also, I don't think you'd get much useful stuff done with early Pentiums and 486. Consider that a P4 2 GHz has 20 times the clock speed and probably does twice as much per cycle, so it's ~40X faster. Now, if you connect 40 P100 together, unless your problem is completly parallel (like breaking keys, as opposed to most linear algebra), you're going to lose at least a factor of 2 there. This means that in order to equal 1 P4 @ 2 GHz, you'll need almost 100 Pentium 100 Mhz. This means that 10 P4 would be like a thousand Pentiums. At these numbers, it's going to cost so much in networking and power...
I'd say (pure opinion here) the slower you'd want to have today is something like a Duron 1 GHz and the best MIPS/$ is probably with a bunch of dual Athlon 1.4 GHz (A dual is not cheaper that 2 single, but you get more done because of parallelism issues).
If a million monkeys type out the source code to MS Office.
Isn't that how it was written in the first place anyway?
I think there's a middle between the two extremes. I think you should not be liable *except* in the few cases where it's pretty obvious that the intent was to directly cause damage. If someone writes an editor that has a bug that ends up wiping your drive, it's an accident. However if the same guy advertises an editor that startes by system("/bin/rm -rf $HOME"), I think it should be illegal. Where to draw the line, I don't know.
But advocating that you're not responsible if you spread worms/virus is plain dangerous. Why don't you extend that to crasking into systems... after all, you're just sending bits to random ports on a system...