The normal way to operate a cluster is to have a shared (NFS) file system across all the systems, thereby solving the data distribution problem (please note though that this prevents you from doing too much file base IO because it's too slow, you might want to make a local/scratch directory on each node)
Besides the NFS share you'll need some kind of parallel programming library like MPI or pvm, and a job scheduler of some sorts. The libraries you can find on the web (maybe in precompiler RPMS, look for the mpich MPI implementation for a start), and will provide you with a programming framework for doing all the networking and setting up the topology. The scheduler can be as simple as the one provided with MPI/pvm (ie. you name a few hosts and your job gets run on those), or, if there's a number of people accessing the cluster at the same time, you might want to try a real queuer (like gridware).
The parallellization is something you'll have to do yourself and it's the hardest part of clustering.
Or you could try pslatex: works like a charm, esp. if you use
\documentstype[times]
Re:"Pedophilia is *good*!" - Greeks
on
Roasting Sacred Cows
·
· Score: 3, Insightful
>So we should wonder: what was the point inthe
>history of Western culture where pedophilia
>switched from white to black? And why?
Easy: With the introduction of christianity....
The largest part of the 'mass' (energy) of the neutrino still comes from its kinetic energy. What has been found in the experiment is that it also has a *rest* mass (ie. a mass at zero velocity).
To get light to interfere with itself (which is basically what they're trying to do in the Keck telescope) you'd have to align the various mirrors and other optical components so that the light from the two telescopes comes in at the same relative path length over the entire image. A path length deviation of several tens of nanometers is already too much.
In practice it means that your mirrors have to be aligned so precisely, that if you scale the whole telescope array up to the size of the US, the positioning would have to be flat and constant up to a scale of meters (or feet).
Research is, in fact, one of the most open sourced processes going on.
Researchers get their status and thus their grant money from the number of publications and the number of times their publications are quoted by other publications, so it is in their best interest to make publications as easily obtainable as possible
For example, check out the arXiv preprint archive, which contains a large proportion of all publications in physics (at least for my field, soft condensed matter physics), in postscript and LaTeX source (!).
Of course, commercial journal publisher (like Elsevier) use this to obtain full copyright on articles, and publishing them in journals which cost 10s of thousands of dollars for a subscription. A good university or institute can't do without them, because they have a natural monopoly on what's published.
The only thing i use pen and paper for still is for calculations; having a recognition program for that to mathml/latex/mathematica would be really useful..
Imagine writing down your integral and have it calculated by mathematica...
AFAIK the incas lived pretty close to the equator, so, unless the stars are pretty close to each other and the pole (which increases the error margin), the trick would not be very useful for them
Probably not, because this trick only works from well north of the equator. South of the equator at least one of the stars is under the horizon if they're aligned vertically
If i understand the watermarking correctly, it is based on the fact that some parts (frequencies?) of the audio signal are inaudible to human listeners, so a watermark signal there wouldn't degrade the sound. Isn't it quite trivial to generate some noise/random signals at the *modulated* frequencies *of* the watermark signal, or just read the watermark signal, invert it and put it back? What am i missing?
Alladin sells the latest version of their Ghostscript postscript interpreter, but open-sources the older versions. Of course the feasibility of this depends on the nature of the software, but it's still a fairly 'nice' way of making money and open-sourcing.
The 'euphemism' "Stockpile Stewardship and Managment program" actually aims for a way of not having to make any more nuclear weapons by maintaining and conserving the current ones.
If it turns out that the expiration dates of the weapons is far enough in the future, no new weapons will have to be made for a while, BUT to verify the durability, some experiments will have to be made, and since real-world nuclear tests politically sensitive (how's this for a euphemism?:-)); doing the experiments virtually, would be a boon.
Personally i wouldn't mind seeing all of those weapons passing their expiration dates, but spontaneous detonation would be rather nasty.... (though highly improbable i guess)
Now THAT'S going to piss off your neighbours!
on
"Invisible" Speakers
·
· Score: 1
Living in a small apartment in Amsterdam, I can imagine the joy of watching my neighbours in fury at the house resonating with the prodigy on a late drinking night with my friends....
The alpha architecture seems to be conspicuous by it's absence! Too bad, because tuning stuff (compiler flags etc) gets you lots of gains on these machines, and it's really picky about the types of memory accesses etc.
These cases are UGLY! They're a perfect example of a distinctly American type of kitsch, the 'decorations' (a good-looking box doesn't NEED decoration) on the box look like the decorations on cowboy boots and I can hear the country music coming out of those ugly boxes....
(not to say that Europeans or Asians don't have their own kitsch)
Slashdot Mirror
???
Fat is *less* dense than water (it floats).
Maybe the right kind of cold fat sinks (as opposed to the water which expands upon freezing)?
Anyway, sinking ice cream is strange.
Is there a way to use a new(er) kernel with XFree86 4.0.3 DRI? It seems my mb agpgart (VIA KT266) is not supported by 2.4.7 but it may be by 2.4.14.
The WWW, developed at CERN by Tim Berners-Lee springs to mind..
The normal way to operate a cluster is to have a shared (NFS) file system across all the systems, thereby solving the data distribution problem (please note though that this prevents you from doing too much file base IO because it's too slow, you might want to make a local /scratch directory on each node)
Besides the NFS share you'll need some kind of parallel programming library like MPI or pvm, and a job scheduler of some sorts. The libraries you can find on the web (maybe in precompiler RPMS, look for the mpich MPI implementation for a start), and will provide you with a programming framework for doing all the networking and setting up the topology. The scheduler can be as simple as the one provided with MPI/pvm (ie. you name a few hosts and your job gets run on those), or, if there's a number of people accessing the cluster at the same time, you might want to try a real queuer (like gridware).
The parallellization is something you'll have to do yourself and it's the hardest part of clustering.
Hope this helps :-)
err. i mean pdflatex..
\documentstype[times]
>So we should wonder: what was the point inthe
>history of Western culture where pedophilia
>switched from white to black? And why?
Easy: With the introduction of christianity....
The largest part of the 'mass' (energy) of the neutrino still comes from its kinetic energy. What has been found in the experiment is that it also has a *rest* mass (ie. a mass at zero velocity).
In practice it means that your mirrors have to be aligned so precisely, that if you scale the whole telescope array up to the size of the US, the positioning would have to be flat and constant up to a scale of meters (or feet).
YYYY-MM-DD HH:MM:SS would be the format,
2001-04-01 13:13:13
would be fitting..
Research is, in fact, one of the most open sourced processes going on.
Researchers get their status and thus their grant money from the number of publications and the number of times their publications are quoted by other publications, so it is in their best interest to make publications as easily obtainable as possible
For example, check out the arXiv preprint archive, which contains a large proportion of all publications in physics (at least for my field, soft condensed matter physics), in postscript and LaTeX source (!).
Of course, commercial journal publisher (like Elsevier) use this to obtain full copyright on articles, and publishing them in journals which cost 10s of thousands of dollars for a subscription. A good university or institute can't do without them, because they have a natural monopoly on what's published.
Sander
The only thing i use pen and paper for still is for calculations; having a recognition program for that to mathml/latex/mathematica would be really useful..
Imagine writing down your integral and have it calculated by mathematica...
oh well.
AFAIK the incas lived pretty close to the equator, so, unless the stars are pretty close to each other and the pole (which increases the error margin), the trick would not be very useful for them
Probably not, because this trick only works from well north of the equator. South of the equator at least one of the stars is under the horizon if they're aligned vertically
If i understand the watermarking correctly, it is based on the fact that some parts (frequencies?) of the audio signal are inaudible to human listeners, so a watermark signal there wouldn't degrade the sound. Isn't it quite trivial to generate some noise/random signals at the *modulated* frequencies *of* the watermark signal, or just read the watermark signal, invert it and put it back? What am i missing?
Alladin sells the latest version of their Ghostscript postscript interpreter, but open-sources the older versions. Of course the feasibility of this depends on the nature of the software, but it's still a fairly 'nice' way of making money and open-sourcing.
The 'euphemism' "Stockpile Stewardship and Managment program" actually aims for a way of not having to make any more nuclear weapons by maintaining and conserving the current ones.
:-)); doing the experiments virtually, would be a boon.
If it turns out that the expiration dates of the weapons is far enough in the future, no new weapons will have to be made for a while, BUT to verify the durability, some experiments will have to be made, and since real-world nuclear tests politically sensitive (how's this for a euphemism?
Personally i wouldn't mind seeing all of those weapons passing their expiration dates, but spontaneous detonation would be rather nasty.... (though highly improbable i guess)
Sander
Why would any company want to do anything in Belgrade, let alone starting a R&D center?
was just a little too lazy to read... :-)
As it seems to be the limiting factor...
Is the /etc/passwd (or equivalent) file world-readable in QNX? (While i'm at it, what exactly is QNX and why is it so special?)
Sander
:-)
Living in a small apartment in Amsterdam, I can imagine the joy of watching my neighbours in fury at the house resonating with the prodigy on a late drinking night with my friends....
Sander
The alpha architecture seems to be conspicuous by it's absence!
Too bad, because tuning stuff (compiler flags etc) gets you lots of gains on these machines, and it's really picky about the types of memory accesses etc.
These cases are UGLY!
They're a perfect example of a distinctly American
type of kitsch, the 'decorations' (a good-looking
box doesn't NEED decoration) on the box look
like the decorations on cowboy boots and I can
hear the country music coming out of those ugly
boxes....
(not to say that Europeans or Asians don't have
their own kitsch)
Sander