One line on perl typically does a lot more than one line of C code (even without absurd "golf" tricks). The same is true of other high level languages. So even leaving out issues of programmer quality, what does this really mean?
Tha standard argument is that, regardless of which language you're using, for a sufficiently large project, the number of lines of code a programmer can produce is relatively constant. Some languages may require less lines to get the same ammount of work done, but the work to produce a KLOC is approximately constant.
I forget the source (Mythical Man Month, maybe?) but there was some study that showed that programmers writing ASM produced the same LOC as those using some early HLL (PL/1?). I've heard it said that this number remains relatively constant as more powerful languages get introduced.
The problem with the test is that they obscure the links. To me, the big test of a scam v. a real email is where the links point to rather than the content and the test uses javascript to obscure where they're going.
Having a single system image is a -massive- bennefit to people who want to run simulations without having to actually learn the complexities of clustering. Considering that this will be used by a lot of scientists & engineers that are engaged in non-related research, I could see this cluster being used, most of the time, as simply 20 separate 512 processor systems.
We've got a high-performance computing center here on campus and their largest cluster has 512 processors (in 256 nodes). It's an older setup, only running P3-730s. Each node of this NASA cluster is going to be like 5x faster than our entire cluster.
The problem is that 2.6 might not stay stable. In a stable branch, if x.y.z is stable, x.y.z+1 should be stable; if it's not, somebody made a mistake.
The article is saying that x.y.z+1 might make changes that they -know- are going to break things so that they can work on new features, something that previously was only done in development series kernels.
It seems kinda stupid to have the clients basing their updates on clock time. Doing an update on client startup and then every 60min after that would be just as easy as doing it on the clock time & would basically eliminate the whole DDOSesque thing.
Of course they're not really telling it straight with the CPU requirements; the Willamette core P4s were pretty sorry chips. An Athlon at 1.5GHz would smoke a P4 1.5GHz; saying they're equivalent is being political.
The thing you're overlooking is that media has become an integral part of our culture. Look at the works of the past; great music, art & literature; they're freely available. What effect would perpetual copyright have on future generations?
SGI's had the Altix, which goes up to 256 procs & runs Linux, in their production lineup for a while now. This is a pretty big jump, but in realistic terms, if a 256-way machine isn't doing it, why is a 1024-way machine?
What's the point of porting Rogue dirrectly to the Dreamcast when you could simply tie it onto the back of a Free Unix system? About the only real thing gained is the controller support which could probably be done with a hacked keyboard driver a lot easier.
If anything, it'll boost sales of AMD chips, they have a much larger following among gamers 'cuz of the price/performance ratio. For a random data point, over at [H]ard|OCP's forums, there's like five times the traffic in the AMD board than there is on the Intel board.
Grr. Slashcode's eating my hyperlinks again.
[H] = http://www.hardocp.com
By August, you should probably be able to get a card that'll tear that game up for $150-60; the low-end cards on the list are already only $50-60. I'd be suprised if my Radeon 9600 can't play this one acceptably, and that's a $100 card.
When I was working in the university library as a student employee, trying to code up web apps on a 120MHz Pentium while 2GHz P4s were in public areas used for little more than searching the catalog, I was told that it was because of some sort of funding issue. Easier to get grants for things students touch or somesuch.
It's not like there's not backgammon addicts running around IRL, playing for $5-10 (or more) per point (potential of 3*x^2 points per game, depending on cube action) running around bars & coffee houses.
Just because backgammon's not as popular as something like poker does not mean that it's any less of a gambling game (and, IMHO, it's the perfect blend of skill & random chance, making it a supreme gambling game). Gambling is a recognized addiction; doing it online doesn't change a thing.
While going to some tiny little town in Iowa has definate problems associated with their lack of,well... everything, there are plenty of cities that aren't NYC/LA that can easily accomodate a few thousand visitors.
Pretty much any city with more than a few hundred thousand people would do the job & have conference facilities that would fit the bill and cost a lot less than being in one of the top-10 biggest cities.
The thing is that DDR2 probably will never get timings down to the same speeds as DDR. You may have noticed that DDR timings haven't changed much since day one; speeds and capacity have gone up but the range of available timings hasn't changed much. The impression that I've gotten in the past is that slower timings are inherrent in the design of DDR2.
In Intel's long tradition of pushing a new technology before it actually surpassed the previous generation (P2, P4, P4-Prescott) DDR2 is, at best, on par with systems already in place.
If you look at some benchmarks of DDR2 performance, you have to wonder why anyone would even consider buying it right now.
"Expect DDR2 memory at 533MHz to be comparable to DDR1 at 400MHz, but don't expect to see any "noticeable" memory bandwidth performance gains till DDR2 667 and above with low timings!"
I'm suprised that nobody's come out and suggested to just get cable; if sattelite TV was that big of a priority for you, you should've made sure you could use it before you moved.
Slashdot Mirror
One line on perl typically does a lot more than one line of C code (even without absurd "golf" tricks). The same is true of other high level languages. So even leaving out issues of programmer quality, what does this really mean?
Tha standard argument is that, regardless of which language you're using, for a sufficiently large project, the number of lines of code a programmer can produce is relatively constant. Some languages may require less lines to get the same ammount of work done, but the work to produce a KLOC is approximately constant.
I forget the source (Mythical Man Month, maybe?) but there was some study that showed that programmers writing ASM produced the same LOC as those using some early HLL (PL/1?). I've heard it said that this number remains relatively constant as more powerful languages get introduced.
I'd take Graham over John Katz any day...
Why do you differentiate between sysadmin and janitor?
The two jobs have a lot in common. I keep pushing for having my job title changed to "IT Janitor".
How about a compromise like SGI's NUMA systems? Seems like a reasonable compromise between fast memory & lots of CPUs when memory access is important.
The problem with the test is that they obscure the links. To me, the big test of a scam v. a real email is where the links point to rather than the content and the test uses javascript to obscure where they're going.
Don't forget the 600MHz parts... they were newly upgraded to the Coppermine core (more cache) and consistantly OCed to 900MHz (66Mhz FSB to 100MHz).
Having a single system image is a -massive- bennefit to people who want to run simulations without having to actually learn the complexities of clustering. Considering that this will be used by a lot of scientists & engineers that are engaged in non-related research, I could see this cluster being used, most of the time, as simply 20 separate 512 processor systems.
We've got a high-performance computing center here on campus and their largest cluster has 512 processors (in 256 nodes). It's an older setup, only running P3-730s. Each node of this NASA cluster is going to be like 5x faster than our entire cluster.
What's wrong with a 2.14.x brancH?
The problem is that 2.6 might not stay stable. In a stable branch, if x.y.z is stable, x.y.z+1 should be stable; if it's not, somebody made a mistake.
The article is saying that x.y.z+1 might make changes that they -know- are going to break things so that they can work on new features, something that previously was only done in development series kernels.
It seems kinda stupid to have the clients basing their updates on clock time. Doing an update on client startup and then every 60min after that would be just as easy as doing it on the clock time & would basically eliminate the whole DDOSesque thing.
Of course they're not really telling it straight with the CPU requirements; the Willamette core P4s were pretty sorry chips. An Athlon at 1.5GHz would smoke a P4 1.5GHz; saying they're equivalent is being political.
The thing you're overlooking is that media has become an integral part of our culture. Look at the works of the past; great music, art & literature; they're freely available. What effect would perpetual copyright have on future generations?
SGI's had the Altix, which goes up to 256 procs & runs Linux, in their production lineup for a while now. This is a pretty big jump, but in realistic terms, if a 256-way machine isn't doing it, why is a 1024-way machine?
...and they're not that fast.
Sun's biggest unit is only 100 procs.
Girls are the consumer model of those creatures you see in your porn.
What's the point of porting Rogue dirrectly to the Dreamcast when you could simply tie it onto the back of a Free Unix system? About the only real thing gained is the controller support which could probably be done with a hacked keyboard driver a lot easier.
If anything, it'll boost sales of AMD chips, they have a much larger following among gamers 'cuz of the price/performance ratio. For a random data point, over at [H]ard|OCP's forums, there's like five times the traffic in the AMD board than there is on the Intel board.
Grr. Slashcode's eating my hyperlinks again. [H] = http://www.hardocp.com
By August, you should probably be able to get a card that'll tear that game up for $150-60; the low-end cards on the list are already only $50-60. I'd be suprised if my Radeon 9600 can't play this one acceptably, and that's a $100 card.
When I was working in the university library as a student employee, trying to code up web apps on a 120MHz Pentium while 2GHz P4s were in public areas used for little more than searching the catalog, I was told that it was because of some sort of funding issue. Easier to get grants for things students touch or somesuch.
Why is this even worth writing an article about?
It's not like there's not backgammon addicts running around IRL, playing for $5-10 (or more) per point (potential of 3*x^2 points per game, depending on cube action) running around bars & coffee houses.
Just because backgammon's not as popular as something like poker does not mean that it's any less of a gambling game (and, IMHO, it's the perfect blend of skill & random chance, making it a supreme gambling game). Gambling is a recognized addiction; doing it online doesn't change a thing.
Juan_Juarez@ODay-Couriers.com
Meh. $70 is a big jump from "free with any hardware that's remotely capable of playing DVDs"...
While going to some tiny little town in Iowa has definate problems associated with their lack of ,well... everything, there are plenty of cities that aren't NYC/LA that can easily accomodate a few thousand visitors.
Pretty much any city with more than a few hundred thousand people would do the job & have conference facilities that would fit the bill and cost a lot less than being in one of the top-10 biggest cities.
The thing is that DDR2 probably will never get timings down to the same speeds as DDR. You may have noticed that DDR timings haven't changed much since day one; speeds and capacity have gone up but the range of available timings hasn't changed much. The impression that I've gotten in the past is that slower timings are inherrent in the design of DDR2.
In Intel's long tradition of pushing a new technology before it actually surpassed the previous generation (P2, P4, P4-Prescott) DDR2 is, at best, on par with systems already in place.
If you look at some benchmarks of DDR2 performance, you have to wonder why anyone would even consider buying it right now.
"Expect DDR2 memory at 533MHz to be comparable to DDR1 at 400MHz, but don't expect to see any "noticeable" memory bandwidth performance gains till DDR2 667 and above with low timings!"
I'm suprised that nobody's come out and suggested to just get cable; if sattelite TV was that big of a priority for you, you should've made sure you could use it before you moved.