Slashdot Mirror

(L. alumen, alum) The ancient Greeks and Romans used alum as an astringent and as a mordant in dyeing. In 1761 de Morveau proposed the name alumine for the base in alum, and Lavoisier, in 1787, thought this to be the oxide of a still undiscovered metal.

Wohler is generally credited with having isolated the metal in 1827, although an impure form was prepared by Oersted two years earlier. In 1807, Davy proposed the name aluminum for the metal, undiscovered at that time, and later agreed to change it to aluminum. Shortly thereafter, the name aluminum was adopted to conform with the "ium" ending of most elements, and this spelling is now in use elsewhere [this being a US site, they don't feel obliged to say "everywhere else"] in the world.

Aluminium was also the accepted spelling in the U.S. until 1925, at which time the American Chemical Society officially decided to use the name aluminum thereafter in their publications.

There is a classical entanglement between regular EM waves, which is formally identical to the quantum state entanglement for photons (as appearing in Bell's inequalities). There is nothing mysterious about that meaning of "entanglement," everything is perfectly local, computers made using them have no special powers (beyond what an analog computer would have, if one could build a usable one). Dutch physicist Robert Spreeuw (well known name in the optics of atoms) had worked out quite far the mathematical correspondence between the two formally identical phenomena, including the extensions to the so-called "quantum computing." So, in that sense, the "entangled" state has been produced many times, in quantum or classical systems.

What separates the "quantum entanglement" is the presumed ability of quantum entangled state to collapse instantly the remote subsystem state when a measurement is done on the other subsystem. To test that key property, on which the Quantum Computing contact with reality vitally rests, Bell inequality (or its more recent variations) is used. And that is where no test to date has come even close to being "loophole free." While papers and preprints may not emphasize or even mention the loophole, the fatal loophole is present in every single experiment. Merely showing that the state is consistent with the entangled state (which is how far the actual, raw, non-adjusted data ever goes) doesn't prove that it has the key property of remote subsystem collapse.

To prove that, the raw data would have to be used, before any assumptions about the missing data (inferred to exist from the triggers of some, but not all necessary photo-detectors, and which is routinely pressumed to obey the "fair sampling" hypothesis) are used to adjust the raw data (raw detector counts). Additional data adjustments are also done, specific to the experimental setup (especially common are subtractions of background counts, again under unproven and unprovable assumptions about unmeasurables).

Only after all the adjustments, the wishfully massaged data does violate (amazingly enough) the Bell inequalities (which is the objective if one wishes to prove the existence of quantum entanglement).

Having been once a theoretical physicist (and having written masters thesis in this area), I had corresponded over years, via mail and email with most experimenters who had published the "almost" proofs of the quantum entaglement. And when pushed, yes every one will tell you that indeed, there is a loophole of one kind or another, the hand put assumption about the missing data which makes the adjusted data violate the Bell inequalities. Without it, on the raw data, none violates the Bell inequalities.

To put it in plain terms for the non-physicists here, one could for example, having no data on where you were 1 minute ago, assume that you were 1000 miles away from your current place. And, say, now we can all find you on your current place reading this message. We're missing the data on where you were 1 minute ago (and there are no outside witnesses to say one way or another) but with our 1000 mile assumption, that seems pretty amazing how you managed to travel that far in so short time. Now we take for granted that you did travel 1000 miles in 1 munute, and then we start constructing ever more marvelous hypothetical technology based on that enormous speed for an earthly vehicle. Sure, you could come up with some pretty nifty transportation systems if you build upon that 1000 miles/minute vehicle you supposedly must have used. And if someone is going to pay us to do research on these possibilities, we're not going to pursue seriously the issue whether our key unverified assumption (your location 1 minute ago) might be wrong.

That is what the issue in quantum entanglement comes down to. Without making the unverifiable assumptions of certain particular kind (like that you were 1000 miles away 1 minute ago) about the missing data, nothing unusual or amazing can be deduced, the entanglement measured on actual data is no different than the classical one (which of course, can't collapse instantly the remote subsystem state).

There've been some experiments in Italy (so no ref) that have been running over the past couple of years (not "the 1980's") that have "proved" entanglement over very large distances (kilometres)

You probably mean Geneva experiments by Tittel et al from 1998-99; check on the Los Alamos preprint archive for their 10 km exeperiments. There is the group's email address at the top of the abstract, go ask them if the experiments were loophole free (they were not). Now, of course, they'll tell you stories and do the usual handwaving dance to convince you (if you appear in-the-know enough to be worthy their reply) why their unprovable assumptions were "natural" and why the "loophole" isn't likely to change their conclusions, should it ever be possible to do experiment without it. Yeah, sure.

And that is the bottom line. Just like with my "assumption" about your location 1 minute ago, I could dance and handwave all I want, why it is "natural" to assume, provided there are no witnesses to contradict me (i.e. data missing for good), that you were 1000 miles away, unless one can show without such assumption existence of the 1000 miles a minute vehicle, why should anyone believe it. Or fund it with their tax dollars.

When you peel off the layers of technical jargon and euphemisms (such as "loophole") protecting the dirty little secret and its priesthood's well being from the outsiders, that's the vital presumed "fact" on which the whole marvelous technology of Quantum Computing rests on -- the wishful assumption about the unmeasured, unmeasurable and irretrievably lost data points.

More on Quantum Cryptography at:

http://qso.lanl.gov/qc/

For those of us who are Paranoically Inclined (tm)

I want the future now!

http://www.lanl. gov /worldview/news/releases/archive/00-146.html

This one works.

One place this could come in handy is in the Linux BIOS project. This project has Linux booting straight out of the Mothersboard's NVRAM.

The U.S. National Labs offer lots of internships/positions for computer scientists involving Linux or Unix. Here is an incomplete set links for the curious:

• Lawrence Livermore National Laboratory
• Sandia National Laboratory
• Sandia's California site
• Los Alamos National Laboratory
• Argonne National Laboratory

Here is the technical article from the Los Alamos preprint server.

Here is the technical article from the Los Alamos preprint server.

(LINK)

sig you!!

You might wish to check out the Linux BIOS project at http://www.acl.lanl.gov/linuxbios/. It's not exactly the same, but remarkably similar.

Yeah - even without a complete model framework for quantum gravity, if you take into account the Holographic Principle, there is a theoretical limit on the amount of information stored in a particular location: 1 quantum bit per Planck area.

Numerically, if you have some super high density quantum ram dimm with a surface area of 10cm, it could contain theoretically as much as 10^{67} bits of information (that's about a million trillion trillion terabytes of storage - plenty of mp3's...).

Of course, this is just in theory... practical limits will be hit loooooong before that.

Here is a link to Seth Lloyd's paper:

http://xxx.lanl.gov/abs/quant-ph/9908043

Abstract:

Computers are physical systems: what they can and cannot do is dictated by the laws of physics. In particular, the speed with which a physical device can process information is limited by its energy and the amount of information that it can process is limited by the number of degrees of freedom it possesses. This paper explores the physical limits of computation as determined by the speed of light $c$, the quantum scale $\hbar$ and the gravitational constant $G$. As an example, quantitative bounds are put to the computational power of an `ultimate laptop' with a mass of one kilogram confined to a volume of one liter.

... all these "environmentalists" were, even then, claiming that all the pollutantants that we're spewing into the atmosphere was going to drastically affect the global climate. Tricky bit was, the doomsayers were NOT talking about global warming... they were saying that industrial soceity was bringing about global COOLING!!! ...

Well, come on guys... which is it?

You can't have it both ways.

If environmentalists were predicting a climate change 50 years ago that's actually pretty insightful. That the direction of the change was wrong is hardly surprising, given the massive complexity to predicting weather. We can't even reliably predict the weather for the next day! What would have been clear 50 years ago was that we were pumping massive amounts of waste into the atmosphere (and rivers etc.), so much that we were changing its composition. It's good to know that some people back then had the foresight to realise that this may have effects on the climate.

As for all the comments that doubt that global warming is going on, you only have to do a simple web search to find some statistics on how the mean temperature has risen over the past 100 years.

Here's a few links I just found:
This graph of the mean world temperature over the last 150 years
From this report
Global Temperatures
A paper from the Proceedings of National Academy of Science

And there is heaps and heaps more evidence if you care to look.

The fact that the world leaders had a conference in Kyoto to discuss global warming means that the evidence is clear enough to worry them. It should probably worry everyone else too.

... all these "environmentalists" were, even then, claiming that all the pollutantants that we're spewing into the atmosphere was going to drastically affect the global climate. Tricky bit was, the doomsayers were NOT talking about global warming... they were saying that industrial soceity was bringing about global COOLING!!! ...

Well, come on guys... which is it?

You can't have it both ways.

If environmentalists were predicting a climate change 50 years ago that's actually pretty insightful. That the direction of the change was wrong is hardly surprising, given the massive complexity to predicting weather. We can't even reliably predict the weather for the next day! What would have been clear 50 years ago was that we were pumping massive amounts of waste into the atmosphere (and rivers etc.), so much that we were changing its composition. It's good to know that some people back then had the foresight to realise that this may have effects on the climate.

As for all the comments that doubt that global warming is going on, you only have to do a simple web search to find some statistics on how the mean temperature has risen over the past 100 years.

Here's a few links I just found:
This graph of the mean world temperature over the last 150 years
From this report
Global Temperatures
A paper from the Proceedings of National Academy of Science

And there is heaps and heaps more evidence if you care to look.

The fact that the world leaders had a conference in Kyoto to discuss global warming means that the evidence is clear enough to worry them. It should probably worry everyone else too.

Of course, we're now using Linux for the usual SAMBA and webserver stuff as well.

This link to the Avalon system at Los Alamos may be reachable from the outside. I'm inside the firewall right now, so I can't verify if the link will work for the rest of you (it works for me). There is another relatively new Beowulf cluster at Los Alamos, 128 dual processor P-III's, IIRC, named Rockhopper. Sorry, don't have any links for that machine. Some of the folks who put that machine together now work for the TurboLabs division of TurboLinux.

Garbage in, garbage out. Make a model and then dig up the evidence the model fits the evidence. How is it that Slashdot escapes this paradox--do 99% of Slashdot users read instead of posting? I don't think so.

Commonsense ought to tell anybody that new popular sites get more links than older, static sites. But Huberman has to publish a paper in Nature to prove that. He also contorts his brain to try to prove the power law distribution and page count of sites.

There's nothing new to these ideas--they are just trotting out the old "Pareto principle" that hardly anybody tries to explain. You find self-help books now in the Business section on how to "apply" the 80/20 rule. But the "law" that Pareto discovered is too universal in nature to provide an answer to inequality in the social realm, and too abstract for anybody to say that it is greed that motivates Gnutella sharing.

Even if one admits that music sharing is a market (not clear what the costs are) or that some market principles can help in distribution (it's too early to tell, the Net is an experiment, bandwidth increases to meet demand, it is still almost too cheap to meter, it's becoming easier for everybody to publish their own works, people have other motivations to publish than to make money, and so on) Huberman still don't have the right answers--and they can't derive them simply from the data they produce.

Instead, look at the work of Jean-Philippe Bouchaud and Marc Mezard http://xxx.lanl.gov/abs/cond-mat/0002374. As explained ably by Mark Buchanan in New Scientist 19 August 2000, pp22-26, the natural laws of physics can explain the inequality in the power law distribution much better than economists in the past have attempted to do.

If one needs to remedy inequality then there are several programs one can attempt. One is taxes--tax the rich and give the money to the poor. But taxes often are captured back by the rich and the poor stay poor. With music sharing, if the receivers paid a small tax to the system, with the money going to the producers, one might or might not see the receivers producing more files on their own. Maybe they are motivated by fame instead of money. Or the big collectors might not increase their collections if they were given a small amount of money.

Another remedy these physicists suggest trying is simply raising the "temperature" of the system. By increasing the transactions in the system, through free trade, fair rules, more exchange, and some competition, then there is less volatility for individuals--less chance for somebody to go down and stay down.

Libertarians such as Huberman might be interested in this idea. Certainly they worship free trade. But they ought to be careful of bringing stale models to the new Internet. It might not work the same way they predict in other systems.

I for one don't really see the Gnutella "discrepancies" as a problem. There is certainly not yet any "tragedy of the commons." Because "tragedy" implies that there are limited resources. For example, on my cable modem segment I am no doubt considered a "hog" because my web server gives out big text files to people all over the world. But even if I increased my bandwidth usage tremendously, I would still be far from saturating the system. All that would happen is that some people might be delayed a few seconds by collisions. If the system ever got saturated, I would have to take my server off anyway, and then the system can work toward a new stability.

Since I don't think Huberman's economic, market model is very good at the moment, I think that we ought to explore other alternatives. For example, Frances Cairncross, editor of The Economist, in her excellent book The End of Distance, refers to at least three other ways markets could succeed in music publishing: the advertising model (Cairncross attaches Ester Dyson's name to that), the performance model (John Perry Barlow), or the sort of subscription model (Wall Street Journal formally, Stephen King less formally, ideas that date back to de Sola Pool's idea of the Internet as a model for freedom).

Some pieces that free software could do much better:

• dynamic optimization rather than a simple JIT (see also Self),
• layers of intermediate representation carrying more semantic information, giving a big boost to projects with code available,
• a more powerful security model,
• using the same deduction framework to assist with optimizations and parsing,
• going even further and integrating the language into the programming environment,
• completely kick-ass garbage collection,
• and so on...

Re-inventing OLE / DCOM is silly. There's so much else to do with much greater pay-offs...

The web site has links to previous projects to give you some ideas about the kind of work that some of the teams have done, but overall I will say that the work is of remarkably high quality. This is a school-year-long event, so many of the projects will need to be shortened for semester-long use. They may also need to be made more simple.

As a side note, I should mention that although all of the projects can benefit from supercomputer time, the Challenge is over 10 years old as a program now. As a result, most of the projects run find on mid-range desktops (but are, neverthless, computationally intensive tasks).

Go have a look at the SETI@Home Downloads page. Tell me there are no processors in there that might have vector-optimised maths units.

Go have a look at the Crusoe Technology page, or the VMWare site. Tell me that it's impossible to use hardware or software to emulate or translate from one instruction set to another. Besides, what's stopping the KrosnoConv "surplus military" stock being military-spec MIPS or SPARC clones? I didn't read the "about the company" bit, so I expected this was just an American company picking up after the excesses of the United States war machine.

Heck, check any Pentium III and tell me that it's impossible to execute another device's instruction set (8088) natively.

The only points that concerned me about the KrosnoConv boards were the Linux-in-Flash claim, and 32Mb of RAM per processor for less than $US200 (either very slow memory, or only 32Mbit perhaps, co-packaged RAM from the old 8086 days). There are projects out there to put Linux in a PC BIOS, or even an LS120. You can get CompactFlash cards, which behave like very small hard drives (either Flash memory pretending to be an IDE drives, or IBM microdrives really being IDE drives). But they're not cheap.

x86 is not the only architecture that SETI@Home supports. Why shouldn't someone produce an add-in card that uses your existing infrastructure? I would still be interested in getting a (cheap!) board full of heavy-maths processors to do hardware encryption for Virtual Private Networks, or even just a heavy duty key server.

I've already got the expensive bits like hard drives, network cards, monitors, cases and memory. I'd actually love to have a "parasitic" processor running its own OS, where I can download software to it, and have it process data that I store for it in my real RAM. Kinda like a multi-processor machine, where one or more processors are especially suited to encryption math.

Don't just spout "that's not how things work". Because with the introduction of technology like that used in the Crusoe, or even older technology like that used in VMWare, or any C64 emulator, you know this is how things work now.

wmSpaceWeather shows this kind of info on a widget.

I can't find a web page, but tarballs are available here. rpms and debs are also available from other sources.

This is what I heard too. Actually when I read it they said something to the effect of "If the government would put more money into research, fuel cell engines for cars could be produced within 3-5 years."

I just had to give a short speech for a speech class last week. I did it on fuel cells. There are a lot of interesting developments going on with that technology. Apparently the only commercially ready fuel cells are phosphoric acid fuel cells (originally developed by NASA about 30 years ago I believe). There are several other types that should be ready within the next 5 years though.

They're also coming up with some interesting ideas for fuel sources. Powerball Technologies had one of the more interesting ideas.

Here's a few more links to check out:

Fuel Cells 2000

Commercializing Fuel Cells

Fuel Cells - Green Power. This is a 36 page PDF file from Los Alamos National Lab

Here's a bunch more.

The Linux BIOS Project was in this /. article just a week ago. It's not ready for prime time yet, but does all the BIOS step in about 0.1 seconds! Combine this with the right Linux distribution (debian has a fast (binary) service startup program - do others?), and a carful choice of which services you run, and Linux would boot REALLY fast! :)

Well, BeOS is fast, that's for sure. You might also want to look at the Linux BIOS page - it promises VERY fast linux loading... when the work is done.

If all you want is a GUI of some type, DOS + Win 3.1 will load very fast, if you arrange it nicely. There's a DOS version of the GNU compiler environment here, so you can do C/C++ development under DOS, but I don't know about a GUI IDE. I recall there are several text IDEs.

DOS won't take up too much space, either, so you might be able to keep Win98 on there for other boot times.

Did I just read the phrase "zero-ohm resistor" or do my EE eyes deceive me?
(from http://www.acl.lanl.gov/linuxbios/pictures.htm )

For all I know, they've deemed the Los Alamos cafeteria schedule classified because it might be used to help poison some foolhardy scientist in ten years.

Nope. The Los Alamos cafeteria schedule is in fact available on the web for the entire world to see at
http://www.lanl.gov/labview/services/CafeteriaMe nu/menu.htm.

I heard about Grover's algorithm two years ago, and there have been papers about it on the LANL archive from as far back 1996! It apparently allows unordered searches in a list of N elements in O(sqrt(N)) time, whereas the best a classical machine can do is O(N). See quant-ph/9605043: "A fast quantum mechanical algorithm for database search", by Lov K. Grover (Bell Labs, Murray Hill NJ). Where has everyone been? This is old, old news. Now all that we need is someone to implement a quantum computer...

I heard about Grover's algorithm two years ago, and there have been papers about it on the LANL archive from as far back 1996! It apparently allows unordered searches in a list of N elements in O(sqrt(N)) time, whereas the best a classical machine can do is O(N). See quant-ph/9605043: "A fast quantum mechanical algorithm for database search", by Lov K. Grover (Bell Labs, Murray Hill NJ). Where has everyone been? This is old, old news. Now all that we need is someone to implement a quantum computer...

[...] This yields new applications - an algorithm is presented that can create an arbitrarily specified quantum superposition on a space of size N in O(sqrt(N)) steps.[...]

So while providing a significant speedup over the classic O(N/2) steps, this search algorithms does not overcome the barrier from exponential to polynomial search times (like e.g. Shor's quantum algorithm for prime factorisation); i.e. you would still require O(2^(k/2)) steps to find some k-bit string.

If you are interested in quantum computation, check out QCL, my quantum computation language and QC simulator (for Linux, you guessed it ;-)). An implementation of Grover's original search algorithm is included in the tarball.

At the 1998 BEAM Workshop, I had the opportunity to meet Mark Tilden, inventor of BEAM and employee of Los Alamos National Laboratories. He demonstrated something along these lines, but more in line with the BEAM philosophy. I don't know what happened with it, though. If you're interested in more about BEAM, check out Solarbotics, a really great Canadian company with tons of info, kits, and general resources.
Phyrkrakr
"God doesn't play dice"-Einstein

Our taxes also fund the local fire department, but that doesn't mean they have to give us rides in their big red trucks for free.

I take it by "Free software ethos" you really mean you want journals to be free? Part of the reason academic journals cost so much is because (a) they have a very limited audience, (b) they generally don't sell their space for advertising, and (c) their target audiences, research institutions and universities, can (usually) afford these prices. Subscription rates for individuals, while expensive, are not outrageously so in my opinion for most journals.

If you've ever tried purchasing an esoteric book in the science or mathematics fields, you've probably experienced something similar: a 150-page book may retail for 150$, when the local grocery store hawks pulp fiction by the metric ton. As you identified, it results from their business model: if you are only going to sell a few thousand of something, then a high markup is required in order to make even a modest profit on your work. While I agree that academic books and journals could be cheaper, and they should be so when the distribution costs are lowered due to electronic publishing, I doubt that they could be made completely free without sacrificing quality in the process. Many journals that publish electronically (for example, the Physical Review Letters) offer lower subscription rates for the electronic version of their journals than the paper version.

Incidentally, free electronic journal services do exist, e.g. the Los Alamos e-Print archive at xxx.lanl.gov. One thing you will probably notice is that while many of the articles are outstanding, just as many are "I wiped my nose this morning and decided what I saw on the tissue was publishable so here it is" quality. It's hit-or-miss with these articles sometimes. Standard practice among many disciplines is to archive an early draft of the work on the ePrint archive and then publish the refereed, edited, corrected version in a journal such as the Physical Review....

In this enlightened scenario, there would be very much greater dissemination of the knowledge produced, to the benefit of a very much wider set of users.

...which brings me to my point: High quality, refereed journals that cost money are, in many ways, superior to unfiltered electronic archives precisely because they charge for their services and then in turn use a portion of that money to perform quality control. Part of what one pays for is the process of having experts in the field (hopefully) perusing each article closely to catch mistakes made by the authors or elucidate points the authors left unclear. Editors coordinate the refereeing process, and publishers maintain an infrastructure for ensuring this process happens in a timely manner. As long as people are willing to pay for quality control, then a market will exist for these journals. Electronic publishing can do away with many of the costs of publication and distribution of the information, but I don't see how it can reduce the cost to zero without asking publishers to simply get out of the publishing business altogether.

My thanks to everyone who commentd on the first-round results in the Software Carpentry design competition. I've replied to a few points directly below. Please note that many of these issues are addressed in the project's FAQ as well.

forcing everyone who wants to participate in this project to use Python seems unfair.

Allowing multiple languages moves the burden of learning new syntax and execution models from developers to users. (Essentially, it allows a minority to be lazy, while creating work for the majority.) Feedback from the 140 students who took the software engineering courses I co-taught at Los Alamos National Laboratory in 1998-2000 was unequivocal: as long as we keep making the tradeoff in favor of developers, people will continue to be frustrated and confused by the tools we build.

As an aside, one entrant from Europe said that he'd be a lot more impressed with programmers' sincerity if they put as much effort into creating documentation in multiple (human) languages as they do into arguing over what (machine) language should be used for coding...

...since winners are chosen on functionality/interface and not on implementation details why does the language matter?

We did ask people to keep their designs as language-neutral as possible, and I for one would be very excited to see implementations of (for example) the winning bug-tracking entry in Perl, Java, and other languages as well as Python. (If nothing else, it would permit a real "compare and contrast" study of the languages.) However, since we had chosen an implementation language, we felt it would be dishonest not to say so.

One posting to this thread said that the discussion list had "wandered off" onto the topic of representing programming languages in XML. This actually isn't off-topic: XML means that we finally have a standard language-neutral way of representing hierarchical data, such as the static dependencies in a makefile. However, we still don't have an equivalent way to represent procedural information (loops, conditionals, etc.). As a result, any non-trivial makefile (or configuration file for any other tool) must be bound to a particular language: most makefiles, for example, are tied to the syntax of the Bourne shell. If anyone's looking for a way to change the world, they could start by fixing this...

The whole idea of free software is just that -- that it's free and is being made to produce good software and not because someone is pursuing a cash prize. A certain project being developed with the Software Carpentry Project competition in mind might be rushed through development just so it can meet the contest deadline. End result: A buggier product with fewer features.
...coding solely to earn money was exactly the concept that free software was supposed to eliminate.

No: it's "free software" as in "free speech", not as in "free beer". As Richard Stallman said:

Having a design competition does not go against the principles of free software. The idealism of the Free Software Movement is not that "there should be no competition". It is that we should work for our freedom as computer users, by writing free software to do all the jobs that matter, and thus eliminating non-free software from our lives.

We are awarding prizes because we think people should be rewarded for hard work. The only regret is that we can't award more...

I'd rather they had specified a means by which scripting tools could plug into the system, so that any language (possibly with a small modification) could be used. Of course, the question then becomes how to plug in...

Amen. As the FAQ says:

Component software is clearly the next big thing in software development... However, while COM, CORBA, and (Enterprise) JavaBeans are becoming more popular, they are still new to most programmers, and still do not have robust cross-platform Open Source implementations.

Less formally, COM is proprietary, CORBA is bloated, and EJB is immature and lacks stable bindings for other languages. Creating a lightweight, usable, cross-platform, multi-lingual component system would be a good way to address the issues that Rob Pike raises in: http://cm.bell-labs.com/who/rob/utah200 0.ps

Another thing that my scan of the site didn't turn up was interfacing to source code management...

Amen again. Everyone who has suggested categories for next year's competition has mentioned version control, and no-one seems happy with the tools we have today. However, we had to start somewhere...

Greg Wilson
Software Carpentry Project Coordinator

I am talking, of course, Larry McVoy's thoughts on scalability and SMP clusters. Here is a link on the problems with SMP, and here are the slides without explanation.

The theory goes like this. In an SMP system all of the CPUs have to be made to pay attention when any of the CPUs wants to do something where races would be bad. To do that you need good latency, which means that you need to fine-tune what is locked where and for how long. This introduces a lot of overhead.

Instead what Larry wants is to have a machine with a lot of CPUs turn itself internally into a cluster of Linux machines that just happen to network Really Fast. There are good theoretical reasons why this should scale Really Well.

One of the key items in this vision is the ability to run virtual machines within Linux. Guess what User Mode Linux is? :-) The other piece of the puzzle is making a cluster work like one machine, and Ron Minnich has been doing some work there.

In 2 years, care for a 1000 CPU multi-threaded database server? With failover? :-) :-) :-)

Cheers,
Ben

Well, could this be made into an app like SETI@Home? A nice distributed app that runs on all sorts of computers with some pretty screensaver (maybe of the current pics being processed) might be something people really like. Even just a catalogue would be pretty extensive. But if a whole lot of people each proccess one picture, it might be worth it.

There are possibly some applications that could be automated, such as building a complete two-point correlation function for the clustering of the objects in the field, or maybe trying to categorize all the objects by colour, redshift and position into groupings in space and colour. However, most of these tasks are doable in a reasonable amount of computing time - say two-weeks computation on an UltraSparc machine (although the two-point correlation function is an O(n^2) problem, that requires 10^16 comparisons at a rough estimate, with maybe 10^8 comparisons a second, that would require ... umm ... err ... about 3 years of CPU time). So yes - possibly an automated tool might well be worth it. I strongly suspect that few astronomers would bother to do the correlation function for the whole field at all scales, and would settle for looking at the function for scales up to around 4 degrees separation on the sky (that's much bigger than the largest known cluster of galaxies).

However, looking at the automatically processed picture strips, I see all sorts of problems with background level correction (the background appears to be wavey in these pictures so there is definitely room for improvement). Modern astronomical analysis often requires significant time spent on looking at a particular frame of interest - I spent over a year examining and refining an image of a pair of Quasars as part of my thesis - so my feeling is that there is much to be gained by picking an object which interests you, possibly from a Radio or X-ray survey, and following it up with the SDSS survey here. With this much data I think you can be assured that the Astronomy community will get to grips with the important statistical analysis on it's own. What it won't be able to do is follow up every field, every interesting quasar or galaxy and really really work on it. It may be possible to see gravitational lensing (although it won't be very clear since the point spread function will be around an arcsec) or do some funky image processing to try and deconvolve the images to recover more detail. In fact, there are lots of things to play with which are unlikely to ever get done on every part of this image data, so grab yourself a copy of IRAF or Source Extractor and go play.

Cheers,

Toby Haynes

Well, could this be made into an app like SETI@Home? A nice distributed app that runs on all sorts of computers with some pretty screensaver (maybe of the current pics being processed) might be something people really like. Even just a catalogue would be pretty extensive. But if a whole lot of people each proccess one picture, it might be worth it.

There are possibly some applications that could be automated, such as building a complete two-point correlation function for the clustering of the objects in the field, or maybe trying to categorize all the objects by colour, redshift and position into groupings in space and colour. However, most of these tasks are doable in a reasonable amount of computing time - say two-weeks computation on an UltraSparc machine (although the two-point correlation function is an O(n^2) problem, that requires 10^16 comparisons at a rough estimate, with maybe 10^8 comparisons a second, that would require ... umm ... err ... about 3 years of CPU time). So yes - possibly an automated tool might well be worth it. I strongly suspect that few astronomers would bother to do the correlation function for the whole field at all scales, and would settle for looking at the function for scales up to around 4 degrees separation on the sky (that's much bigger than the largest known cluster of galaxies).

However, looking at the automatically processed picture strips, I see all sorts of problems with background level correction (the background appears to be wavey in these pictures so there is definitely room for improvement). Modern astronomical analysis often requires significant time spent on looking at a particular frame of interest - I spent over a year examining and refining an image of a pair of Quasars as part of my thesis - so my feeling is that there is much to be gained by picking an object which interests you, possibly from a Radio or X-ray survey, and following it up with the SDSS survey here. With this much data I think you can be assured that the Astronomy community will get to grips with the important statistical analysis on it's own. What it won't be able to do is follow up every field, every interesting quasar or galaxy and really really work on it. It may be possible to see gravitational lensing (although it won't be very clear since the point spread function will be around an arcsec) or do some funky image processing to try and deconvolve the images to recover more detail. In fact, there are lots of things to play with which are unlikely to ever get done on every part of this image data, so grab yourself a copy of IRAF or Source Extractor and go play.

Cheers,

Toby Haynes

If the cooler could be scaled up, it would take 3 kilowatts to maintain the temperature of a 30 watt processor.

Not so (I was surprised as well). At Los Alamos they have used optical refigeration to cool a lump of Ytterbium (4mm x 4mm x 7mm - so it is a visible lump) by 0.3K (not to, by). Details here. Only one laser required too. But yes, the story was clueless.

The original New Scientist article is online, as is the full paper which has much more content.

This is interesting, but even if it turns out that they can be found (or built), there may be problems. If they can be moved, you can turn one into a time machine (giving causality the finger) by accelerating one end to relativistic speeds and taking it on a trip, as noted in the actual paper (but ignored by both the New Scientist and BBC articles).

A reasonable SF treatment of this particular idea is in Robert Forward's Timemaster. The characters make cardboard look 3D, and the prose isn't the most beautiful, but the main hook is the physics speculation--and Forward does that quite well.

I don't think many astronomers at all were convinced by the data when it first came out, and many (including the serious planet hunters I know) were convinced that it was not a planet.

It was, I think, far to early for a press release to have been made, and I think the press too readily accepts the judgement of only one or two scientists. Even when they do include the opinions of skeptics in their articles, they are often strongly biased in favor of the "new discovery."

I think it would be interesting to have a web site that is a combination of the e-print server and slashdot. Not only would this help discussion within the comminuty, a lurking press might get an insight into the state of different ideas.

As far as i can tell the aol webpage is done by some antigravity fanboy, Jean Louis Naudin. The person who actually came up with all this crap is Fran De Aquino from Maranhao State University, Brazil. He has his own webpage at:
http://www.elo.com.br/~deaquino/
He apparently has presented stuff in journals, at least according to the good folks at los alamos:
http://eprints.lanl.gov/cgi-bin/w3vdkhgw?qryRDAD2_ .g_;phys-9904018
The references cite that he's been in the Electric Spacecraft Journal, i have no idea what sort of a publication this is though, it could be a trashy magazine about UFO abductions or what not parading as a journal, i'm not too sure.

I found a better copy of Podkletnov's paper on the Los Alamos National Laboratory's e-print archive server. It's available in a bunch of formats, including PostScript, PDF, ASCII, and DVI. The previous link I posted didn't have the diagrams included with his paper.

It's better to actually read the paper and draw your own conclusions than to simply listen to what other people think about it and accept their views.

---
Epitaph

You can read some more information about the work of the Los Alamos scientists at http://www.lanl.gov/w orldview/news/releases/archive/00-041.html. Curiously, Moore's Law seems to hold for quantum computers as well, since it was nearly 18 months since the same researchers intoduced the first 3 qubit quantum computer (using nuclear magnetic resonance and a trichloroethylene molecule). To quote the article: Of course, if Moore's Law is at work here," Laflamme added, "then we could have a 30-qubit quantum computer in less than five years." A 30 qubit machine could perform certain tasks (such as Shor's algorithm or a variant for factoring large primes) many times faster than even the most powerful present-day supercomputers.

You can only trap small nobel gas and metal atoms. Typically, covalent elements will react with the fullerene. Its been tried with hydrogen, e.g., look here, here, and here.

You can only trap small nobel gas and metal atoms. Typically, covalent elements will react with the fullerene. Its been tried with hydrogen, e.g., look here, here, and here.

You can only trap small nobel gas and metal atoms. Typically, covalent elements will react with the fullerene. Its been tried with hydrogen, e.g., look here, here, and here.

Here's further information on the this theory. I think it's quite good, myself. Note that New scientist barely scratches the surface of it:

www.physics.adelaide.ed u.au/ASGRG/ACGRG1/papers/cahill.ps

By the way, if one is after wild and wacky theories, as well as pretty damn good ones, you can do worse than check out the pre-print server on xxx.lanl.gov (Uk mirror at xxx.soton.ac.uk) This is one of the oldest sites on the net.

Avalon (You can tell I don't do web work, heh... forgot the http)

- Rei

Check out Mark Tilden from LANL, and his robots built from a handful of discrete components
Here

Actually, I think some hybrid combination of this with digital management and nonvolatile storage might eventually win out.

"We have identified nearly 3000 faint galaxies, of which nearly 1000 galaxies are of redshift z > 2 and more than 50 galaxies are of redshift z > 5 (ranging up to and beyond z = 10)."

Now, I should caution you that these are photometric redshifts, somewhat more speculative than those derived from matching spectral lines (as was done with the quasar atz=5.5). But in principle with large enough telescopes we can go back and do the spectroscopy and verify these redshifts, they won't all be correct but most of them will.

Another interesting possibility that you should look at if you are interested in having a clever answer to the question 'what's the furtherst thing in the universe' are gamma ray bursts. Though redshifts for these are hard to get it is possible to make speculative arguments about their redshift distribution based on the idea that some of them may appear to 'last longer' due to cosmological time dialation.

Finally, as several others have already pointed out the Cosmic Microwave Background estimated to be at z of about 1500, is about as far away as we are going be able to see. Farther off, you are looking back into the universe when it was so hot and dense that it was 'opaque'. The CMB represents the point in the evolution of the universe when things cooled enough for neutral atoms to form. It turns out that electrons running around without a proton make it really hard for photons to get anywhere in a straight line. At z=1500 those free electrons got used up to make neutral hydrogen and the universe suddenly became 'transparent'.

Incidentally, the universe had to become transparent for the gas to ever cool and form galaxies, stars, planets, and people. In this way and many others the cosmic microwave background represents the beginning of all of the structure that we see around us.

If anyone maintains a list of blocked sites other than those mentioned in the article or on the article's links, please post.

--