Slashdot Mirror

Wasn't the Tunguska incident from a icy comet/comet fragment? So, not technically an asteroid.

You see estimating probability from prior events always results in some error - you may overestimate it or underestimate it.

If you are relatively insensitive to the direction of the error than it is usually best to use maximum likelihood estimator - which is indeed the prior rate. This can be used, for example, to analyze the distribution of craters on Mars.

However, in our case (Earth) underestimating probability is much worse than overestimating it. A typical way to deal with this is to construct a risk function that depends on existing data and your future probability estimate and then try to minimize it.

As you can expect in our case this results in the probability estimate being larger than the observed rate, so it is not impossible that the estimate is 50% in next 10 years, even though no large asteroids or comets collided with Earth in the previous 95 years (remember Tunguska ?).

All of them I approve, but what's up with Japan? Japan gets some 1,200 minor earthquakes per DAY. how in the world do they expect to overcome the seismic noise floor (pun somewhat intended)?

The extent to which communication is a bottleneck in parallel processing depends strongly on the problem at hand and the algorithm used to tackle it. Some problems are amenable to batch processing (e.g. Seti@home), others require some level of boundary-synchonisation (simple fluid codes), others require synchronisation across all nodes (e.g. more complex plasma simulations)

For batch processing tasks, there isn't an issue. For the other's the loose synchronisation may be acceptable depending on the knock-on effect. Loosening the synchronisation obviously decreases the network and infrastructural burden on the job allowing the algorithm to scale better, but the effect of this has to be carefully studied.

This is important to the application developer, but is not particularly relevent to grids per-say. Grid activity, at the moment, is mainly towards developing code at a slightly lower level than application-dependant communication. It is already building up an infrastructure in which jobs can run which tries to remove any dependancy on a central machine. This is because having a central server is a design that doesn't scale well (and also introduces a single point-of-failure). The Globus toolkit provides a basic distributed environment for batch parallel processing, including a PKI-based Grid security system: GSI.

On top of this, several projects are developing extra functionality. For example, the DataGrid project is adding may component, such as automatic target selection, fabrication management (site management, fault tolerance, ...), data management (replica selection, management and optimisation, grid-based RDBMS), network monitoring infrastructure and so on.

The basic model is currently batch-processing, but this will be extended soon to include sub-jobs (both in parallel and with a dependency tree) and an abstract information communication system which could be used for intra-job communication (R-GMA).

The applications will need to be coded carefully to fully exploit the grid, and reducing network overhead is an important part of this, but The Grid isn't quite at that stage, yet. But we're close to having the software needed for people to just submit jobs to the grid, without caring who provides the computing resource, or the geographical location they'll run.

...he could move to Siberia, which was possibly hit by an asteroid quite recently. What are the chances of that happening twice?

Other information on this project can be found here, here (Caltech), or here. This link to Princeton University seems to explain the project much better, at least to me.

that famous meteorite that fell in Siberia thousands of years ago?

That had the same power of an atomic bomd. They migh have genetically mutated.

Americas:
http://www.zdnet.com/ - Ziff-Davis, North America
http://download.cnet.com/ - CNET/Download.com, California
ftp://.beoscentral.com/pub/ - BeOS CentralJohnson City, TN
http://freebe.nerdygirls.com/ - Oak Ridge, TN
ftp://www.beforever.com/pub/beforever /freebe/ - BeForever, Omaha, NE
ftp://mirrors.rochester.rr.com/pub/be/ - RoadRunner.Com, Rochester, NY
ftp://ftp.be.com/pub/beos/ - Be, Inc. San Jose, CA

Europe/Australia:
ftp://mirror.aarnet.edu.au/pub/beos/ - AARNet, Brisbane, Australia
ftp://ftp.beeurope.com/pub/ - BeEurope, Paris, France
ftp://ftp.worldonline.fr/ - World Online, Paris, France
ftp://ftp.gigabell.net/pub/beos/ - Gigabell.Net, Frankfurt, Germany
ftp://ftp.ph-freiburg.de/pub/m irrors/ftp.be.com/beos - P.H. Freiburg, Freiburg, Germany
http://www.computerchannel.de/download /beos/ - ComputerChannel, Hamburg, Germany
ftp://ftp.xtdnet.com/pub/ - XTDNet, Karlsruhe, Germany
ftp://ftp.zdf.de/pub/ - Neues-3Sat Online, Mainz, Germany
http://pcteor1.mi.infn.it/beos/ - Univ. of Milan, Milan, Italy
ftp://dl.xs4all.nl/pub/ - XS4All, Amsterdam, Netherlands
ftp://ftp.sunet.se/pub/os/BeOS/ - SUNET, Uppsala, Sweden

Here's the abstract, and here's the full preprint paper. It's an interesting, if quite densely technical, read.

Here's the abstract, and here's the full preprint paper. It's an interesting, if quite densely technical, read.

The San Fransisco Chronicle has an article, the paper itself is located at www.lngs.infn.it.

You seem to assume that this book can't possibly have any basis in fact, because it's a book on "social sciences."

Go read Dr. Richard Feynman's commencement speach "Cargo Cult Science" and you will see why such an attitude is not unreasonable.

There has been a conflict between the hard and soft sciences for decades. Just two or three years ago, a physicist submitted an "article" to a psychological journal which expounded upon the effects of quantum gravity on behavior. It was total nonesense, a hoax from the start, but it was manufactured for a specific reason. The journal accepted it without ever investigating its validity. A hard-science journal always has a board of reviewers who frequently reject papers, explaining to the author where their work is lacking. Most of the time, those authors then fill in the gaps, and often get accepted later when they've done a better job of it.

To continue with the story, this article was published in the journal, lauded by psychiatrists. Then the hoax was revealed by the author, much to the delight of chemists, physicists, and other hard-science people. The psychologists howled in complaint, much to our amusement.

In the end, it was their own fault for not investigating the validity of that paper. That's what make science work, peer review. In the hard sciences, we try to validate or prove wrong other's work. Experiments are repeated to ensure the validity of the result. Reproducibility is where it's at.

I can say there's a great deal of work put into being sure that the measuring methods
mean something, the experiments are relevant to the question being asked, and the data is collected in a
statistically relevant way.

All of these things he's doing are important, certainly, but they are only the begining. I'm sure your psychologist is working hard, but is he reproducing favorable experiements, or does he feel that's a waste of effort? Does he put as much work into disproving his theory as he does into proving it? When he has the data, does he look for uncontrolled variables? Does he look for side-affect reasons that produced his favorable results which would undermine their support of his theory? Is he constantly refining his theory to match the experimental data, or refining his data to match his theory?

It's much like open-source software. In this case, however, we're studying nature. We're either right, or we're not, and if we're not, nature will put us right when experiments prove a bad theory is wrong. We must be carefull not to fool ourselves, though.

Oh come now, good sir. As I said, there is nothing wrong with philsophy or religion. My grievence was that unscientifically obtained and analyzed information was being presented under the banner of science. That is what psychology has always claimed to be, a science. This book does nothing to further the legitimacy of this claim, from all that has been presented of it. You would do well to read Dr. Feynman's commencement speach about Cargo Cult Science before you fire back with a rebuttle similar to the one you just did.

Further, your insinuation that people who are criticing this matter are "uncomfortable" or "unable to think [about]" these matters is more than just a poor grasp of the issues at hand, it is insulting. Part of the scientific process is the meticulous analysis of a theory and the evidence supporting it by people other than those who presented it. This is the cornerstone of science. If you take exception to people who try to do this and insult them to boot, you had best keep out of all things scientific, for you clearly then are completely lacking in the understanding necessary to participate in such discussions.

Finally, as to your remark Certainly dismissing it as "psychobabble" reveals a hostility towards the entire notion of philosophizing about the self. I would re-emphasize that this meme material is being presented as psychologial (thus a scientific) theory. Remember, she is a psychologist at a university, and books published by her of this nature is clearly going to be recieved as psychological in nature. Indeed, from all indications, she promotes her theories as such. If it is philosophy, then the author should make that disclaimer that her book is not a book of or about psychology, but philosophy. To do otherwise is disingenuous at best, pernicious at worst. One cannot have it both ways.

I would also point out to you that ad hominem attacks are always the resort of the desperate, those who have no legitimate facts or points to argue about, with respect to the matter at hand. Insulting is a poor way of proving the validity of your (or others') thoughts.