Slashdot Mirror

catalyst for almost CO free process
this have achieved with Brookhaven National Laboratory

I don't want to sound all negative here, but... I don't have a choice, do I? A visible light spectrophotometer will not "detect toxins", no matter how much you try to make it open-source or crowd-sourced. The very concept of identifying compounds by visible light absorbance is very much flawed. Thing is, *most* molecules will not absorb visible or near UV light in a way that is specific enough. Real Chemists (TM) traditionally use the so-called IR fingerprint region for this purpose. This region is from approx. 700 to 1500 cm-1 (about 6 to 20 uM - that is 6000 to 20000 nanometers). A special detector is needed for these wavelengths. The one we have in our lab is cooled with LN2 and costs south of $15K. We also have a UV-Vis spectrophotometer, which has its own purpose. That purpose is not "identifying toxins", or analyzing any unknowns. Now, on to my next point. Identifying molecules is challenging, because they are very, very, very mindbogglingly small. Chemists have been grappling with this challenge for a long time. There are many spectrometric methods out there, including IR and UV-Vis (briefly discussed above), near-IR (900 to ~1800 nm, useful for *some* fingerprinting), and NMR (60-1000 MHz, very informative, bit needs a BIG magnet). Spectral data for many molecules of interest has been compiled into readily accessible databases, and is easily accessible. Some of the databases are proprietary/pay-per-view: https://ftirsearch.com/features/libraries/sea407.htm Some are semi-public: http://riodb01.ibase.aist.go.jp/sdbs/cgi-bin/direct_frame_top.cgi And some are government/public: http://webbook.nist.gov/chemistry/ The people who started this project do not seem to grasp of the very basic concepts of chemistry, nor did they do any research on the subject. Reading a Wikipedia article on UV-Vis would have been a good start. What is even more disconcerting is that the fundraising effort behind this cardboard spectroscope has been a success. One just has to hope that nobody buys this to screen their food for "toxins", or to teach their kids chemistry.

You mean like th HRP-4C: http://www.aist.go.jp/aist_e/latest_research/2009/20090513/20090513.html

except Google Earth had a network sites for monitoring set up throughout Japan within a few days of the Fukushima news - government sites, university sites, private companies with monitoring, individuals with geiger counters. This site is near my house: http://www.aist.go.jp/aist_e/taisaku/en/measurement/index.html

The radiation in Tokyo is less than the radiation in New York, so many places have stopped monitoring continuously now. According to most of the press, we should have been dead by now...

The hack consisted of accessing wireless POS terminals from the car park

By cracking WEP, BTW. Any other real-world incident that involved WEP cracking you have encountered? BTW, I found this paper on "IVs to Skip for Immunizing WEP against FMS Attack" from 2008, which seems to be a better attempt at skipping weak IVs than before. Of course it is still better to use WPA if you can.

Meh. That hologram you linked to on engadget looks impractical and not very scalable because it requires a mirror rotating at high speeds. Its nothing like this.
Yes, it is primitive but its the closest thing to what we've seen in science fiction yet. Hopefully they will expand on this and improve on its resolution.

Screw 2D screens, we should be pouring funding into these technologies:

http://www.aist.go.jp/aist_e/latest_research/2006/20060210/20060210.html (TRUE 3D)
http://www.wired.com/gadgetlab/2008/10/philips-3d-hdtv/ (Stereoscopic w/o glasses)

... Especially since metals behave differently in microgravity, possibly leading to new alloys and manufacturing processes not possible or practical on Earth.

First, the real links. I don't know why the blogger didnt't include them, and I don't think this should have gone on the front page without them. Oh well, there's always the comments...

Novel Ferroelectric NAND Flash Memory Cell Demonstrates 10000 Times More Program and Erase cycles than Conventional Memory Cells (AIST press release, surprisingly science-dense).
Highly Scalable Fe(Ferroelectric)-NAND Cell - contribution to the Non-Volatile Semiconductor Memory Workshop, 2008 (you may have access to only the abstract).

This is NOT flash ram, it's ferroelectric RAM. This doesn't matter much to the consumer who can use it much the same way, but it's a different principle. Apparently they've (semi-)tested 100 million r/w cycles, and expect that it can hold data for 10 years (extrapolated from some curve). Besides, it uses a lower voltage than flash, and they expect it to scale down further. Nice. It even looks like it might work. SSDs for teh win :-)

A bit of research from the original AIST site bring quite a lot of info.

The from the original tech report:

Shigeki Sakai (Leader) et al. of the Novel Electron Devices Group, the Nanoelectronics Research Institute (Director: Seigo Kanemaru) of the National Institute of Advanced Industrial Science and Technology (AIST) (President: Hiroyuki Yoshikawa) in collaboration with Ken Takeuchi, Associate Professor of the Graduate School of Engineering, the University of Tokyo (Univ. Tokyo) have demonstrated that the use of ferroelectric gate field-effect transistors (FeFETs) as memory cells dramatically improves the performance of NAND flash memory. The FeFET, the newly developed memory cell, can be programmed and erased as many times as 100 million or more and with programming voltage of less than 6 V, whereas the conventional NAND flash memory cells have ten thousand program/erase endurance cycles with approximately 20 V programming voltage. It has been assumed that conventional NAND flash memory can be downsized to 30 nm at the minimum, whereas this novel memory cell will meet the needs of the next 20-nm and 10-nm technology generations. And thus, this memory cell is expected to be used in a next-generation, high-density, high-capacity nonvolatile memory.

Results of the research was reported at the 23rd Nonvolatile Semiconductor Memory Workshop (the 23rd IEEE NVSMW / the 3rd ICMTD 2008) held in France, May 18â"22, 2008.

Or perhaps just more research into atmospheric plasma like this: Three Dimensional Images in the Air, video.

Give them time. They're working on that...and getting pretty amazing results: http://www.aist.go.jp/aist_e/latest_research/2006/ 20060210/20060210.html

says the date of the research

http://www.aist.go.jp/aist_e/aist_today/2006_21/ho t_line/hot_line_13_2.html

The Group responsible for the device
http://unit.aist.go.jp/photonics/english/group/bun shi_TF_e.html

says the date of the research

http://www.aist.go.jp/aist_e/aist_today/2006_21/ho t_line/hot_line_13_2.html

The Group responsible for the device
http://unit.aist.go.jp/photonics/english/group/bun shi_TF_e.html

Here's the press release it links to. Sadly both the article and the press release are from February 2006...

Actually if you follow the link at the bottom to the "previous robot with coordination problems" you eventually get to this page for MTRAN http://unit.aist.go.jp/is/dsysd/mtran/English/expe rimentE.htm where you can see a somewhat similar robot changing shape. It's interesting. A light flashes when it disables a magnet or something, and then can rotate the modules away from their previously locked position.

I'm afraid the Japanese lump looks much better it even demonstrates the units changing shape which is what its about I think :)

http://unit.aist.go.jp/is/dsysd/mtran/English/expe rimentE.htm

Diamond as semiconductor has been studied for some time. There are some progress recently. But it is not easy to make big single crystal diamond yet, not to mention single crystal diamond wafer. Just imagine how difficult it could be to polish the diamond wafer without introducing many defects, and how hard could it be to characterize how many defects you get in the wafer.

Phosphorus http://www.aist.go.jp/aist_e/latest_research/2005/ 20050615/20050615.html

HTTP-FUSE KNOPPIX is a true (What is true?) Web Based OS.

http://unit.aist.go.jp/itri/knoppix/http-fuse/inde x-en.html

It is a 6MB CD bootable Linux and gets the disk image from Web Servers.

Plan9 runs on Xenopix DVD. Xenoppix runs Plan9 on anonymous PC.

http://unit.aist.go.jp/itri/knoppix/xen/index-en.h tml

Xenoppix is a combination of Virtual Machine Monitor "Xen" and 1CD/DVD "KNOPPIX".
It runs Plan9 and NetBSD on Xen-DomU(GuestOS) and KNOPPIX on XenDom0(HostOS).

I mean, seriously - it can't be the servos, sensors, mechanical linkages, or plastic "body parts". Lynxmotion's Robonova only costs $1000.00 for a kit, and that includes digital feedback servos. I know for a fact that the accellerometers, force sensors, and electronic gyros don't cost that much, we could probably safely say another $1000.00 (and that would likely be a high estimate, although I haven't been able to find pricing on miniature 3-axis force sensors, but they are nothing more than fancy small strain gauges). If you wanted vision, you could add on a cheap CMUCam system from Parallax - something which doesn't seem to be a part of the HRP-2m. For an SBC of the size and power quoted (240MHz SH-4 processor, 32MB of RAM, business card sized, linux capable), I was able to find a few examples after a bit of googling for dev kits (of similar spec) for around $1200.00 (personally, I would stick with a cheapo desktop communicating with the 'bot via bluetooth or a similar wireless link, coupled to a servo controller, and maybe a Parallax Stamp or Propeller for management). Software is "open" (though not really "open source" - you need to license it), and free to download (for non-commercial use), so there isn't any cost there (unless, as another poster pointed out, there is a firmware issue - which I didn't research).

I just don't see how there could be justification for $7000.00, let alone $14000.00! As I attempted to describe, a similar robot with similar capabilities can be picked up almost "off-the-shelf" (albeit, from a few different vendors instead of just one) for almost half the price of the "low" estimate, while being generous with my estimated pricing. Likely, you could do even better - for example, using Lynxmotion's servo brackets plus lower-cost non-digital servos, you might able to build something like the Robonova for about 2/3's or so of the cost. If you made your own brackets from cheap alluminum, you could knock even more off the cost (though you will still pay for it in time to machine those brackets).

You ultimately "get what you pay for", so I wouldn't go too cheap - the Lynxmotion servo brackets and cheaper non-digital hi-tec servos would probably be the lowest cost (in terms of money, time, and aggravation) that I would be willing to spend, personally. However, there comes a point where you are just spending money to spend money - and for something like this, that point is about $3000.00 (give or take a few hundred), unless your biped is a fair amount taller than 14 inches or so. So, where is the price increase coming from for this machine? The answer is right in the article:

Much of the technology behind Choromet was originally developed for the HRP-2 Promet (pictured at right), a life-sized research robot marketed by Kawada, and costing tens of millions of yen (upwards of $170,000).

Can you say "recovery of invested funds" (by overcharging buyers of a 'toy')?

If you download the choromet.wmv file you can play it with MPlayer. Works for me: http://www.aist.go.jp/aist_j/press_release/pr2006/ pr20060526/choromet.wmv

"Add to that they have to re buy all their dvd's that they are still pissed about rebuying from vhs" Hah! Just wait soon as they're done reselling us all our DVDs on HD-DVD they'll sell em to us AGAIN on HD-DVD Superbit! And, of course, five years from now we'll be buying them all over on our new holographic media players... http://www.inphase-technologies.com/ Or in a new format designed for 3d? http://www.aist.go.jp/aist_e/latest_research/2006/ 20060210/20060210.html Personally, I can't wait for the MPAA to start selling licenses to content... then, maybe each time it comes out in new media we'll get it for free....

The answer is dynamic reconfiguration. With smart cells you can have much greater flexibility. Consider M-Tran, a self-reconfigurable modular robot. These designs are (potentially) much better than centralised systems, because you can reconfigure them any way you want. Want to add a new antenna on your spaceship. Ask some cells to prepare for holding it, passing over their current functions to some neighbouring cells.

Yes, you can have a central database, but then you need to waste a lot of system resources on communications. Waste, because for many functions/decisions you don't need to consult the central authority. This isn't a problem when you only have 10-100 smart units (sensors or actuators), but what when you have 10000-100000 of them? Design of such system is a choice between computing and communications. When the number of units is small, you choose a centralised design, when it's huge, you distribute the processing.

Think again what they are designing. They are designing skin. Do skin cells in your body consult the brain (or the spinal cord) to decide whether and how to heal a scratch? Would that make sense? Does that make sense for a spacecraft?

cool advances in diamond semiconductors continue By the way, gem quality carbon crystals are already a reality. Funny DeBeers and other diamond cartels are funding development of machines that can distinguish natural diamond from synthetically created carbon crystals (not hard, natural diamonds have "inclusions"), and propoganda that of course your loved one would much rather have a "real" diamond and man-made carbon crystals aren't diamond. Oh, yeah, what would you like dear, a half-carat natural diamond or a 10 carat pure diamond ^h^h^h^h^h^h^h^h carbon crystal with no defects as the natural ones have?

You are not entirely correct, even though you say some interesting things. For example, it is a well-known fact that there are some complex chemicals that humans can't build, we must find these "building blocks of life" in nature and consume them. One notable example is Vitamin C. Just like these buzzing robots need more cubic blocks we need more chemicals. Just like our cells, proteins, etc. have binding spots that can connect to a particular type of molecules, so do these robots.

Yes, there is a quantitative difference. The robots in the video are small, very simple and the program they use requires the building blocks to be manually placed in a predefined locations.

But combine it with something like M-Tran, add slightly better programming, make thousands of these blocks and add a powerful computer (or parallel computers inside the blocks) and you suddenly can have much more impressive things. Yes, these 4-block thingies from Cornell can only bend and buzz. But the Japanese 15-block thingies can run on four legs and can evolve new movement modes in virtual space.

Soon someone will build a 1000-block thingy and it will be much more impressive. Then, if you like this particular path, you can refine the technology, shrink the building blocks, make much more of them and you can have something approaching life in complexity. Move to the atomic level and you can leave the biological life behind.

I think that Japanese M-Tran (videos galore) is more impressive. It is a similar technology for modular robots. Each module has a motor, a battery and some magnets to connet with other modules.

The blocks can connect with each other in different forms, from a snake-like robot that crawls to a four-legged robot that runs. Even more interestingly, the programs for movement are generated using genetic algorithms in a virtual environment first. An M-Tran robot can be any odd configuration of the blocks, then after some thinking (done on an external computer right now, I guess) it can use that particular shape to move around and (in the future) do something else.

You can rather easily modify M-Tran robots so that they could copy themselves. Just add a magnetic sensor that would feel other blocks and program it to wander randomly around to stumble into blocks (fancy sensors like eyes that can see the unique markers on the blocks can make it more efficient).

Of course, self-replication is just one example of self-construction and not a particularly interesting or important one. We are so excited about it only because evolution works this way. You have heredity and it works by having stuff make copies of itself. There is no reason why our future robots or nanobots must self-replicate (but it makes for a simplier brute force explanation of what they can do) - they can just design a good shape and form it (for example, by using genetic programming). In a typical scenario you would not have the blocks form a hundred of similar robots (or a few types of robots) - instead the blocks would form a single organism (not necessarily fully connected) that would do exactly what is needed and would reconfigure itself on the fly.

Of course, replication is easier and may prove useful to some extent, I am just saying that there are ultimately better ways to tackle problems than by throwing hundreds of identical robots at them.

Here's a similar project, each part is made of two blocks and a hinge instead of a single cube:

http://unit.aist.go.jp/is/dsysd/mtran/English/inde x.html

There are some truely amaizing videos of M-TRAN II in action. Doing things like actually walking on 4 legs, deforming itself into a snake shape and inchworming itself under an arch and then reforming into 4 legs and walking away. They even make the same clicking sound as they walk like the Replicators. Check out the videos here:

http://unit.aist.go.jp/is/dsysd/mtran/English/expe rimentE.htm

Here's a similar project, each part is made of two blocks and a hinge instead of a single cube:

http://unit.aist.go.jp/is/dsysd/mtran/English/inde x.html

There are some truely amaizing videos of M-TRAN II in action. Doing things like actually walking on 4 legs, deforming itself into a snake shape and inchworming itself under an arch and then reforming into 4 legs and walking away. They even make the same clicking sound as they walk like the Replicators. Check out the videos here:

http://unit.aist.go.jp/is/dsysd/mtran/English/expe rimentE.htm

I probably missed the karma boat on this one but c'est la vie. These are NOT self-replication robots, what they are, are self re-configurable robots, and quite honestly, not that impressive, if you want to see something with blow you away check out MURATA Satoshi and his bots. They are rediculous, just watch the video (asf sorry)http://complexity.vub.ac.be/~comdig/ALife9/M urata.asf
http://unit.aist.go.jp/is/dsysd/mtran/English/inde x.html

Scientific American (warning: loaded with ads etc)

Not for the light-hearted, a thorough review in Reviews of Modern Physics (subscription required, if you cannot access the article, drop me an email at karvind@NOSPAM.gmail.com)

On Ferroelectric spintronics from Colossal Storage.

Spintronics and Quantum Dots. Discussion about one possible implementation.

Another good introduction.

Hope it helps.

Here is coLinux+KNOPPIX

http://unit.aist.go.jp/itri/knoppix/colinux/index- en.html

QEMU with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/qemu/index-en. html

coLinux with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/colinux/index- en.html

Here is an installer to NTFS of Windows2000/XP.

Install2win

http://unit.aist.go.jp/itri/knoppix/win/index-en.h tml

QEMU with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/qemu/index-en. html

coLinux with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/colinux/index- en.html

Here is an installer to NTFS of Windows2000/XP.

Install2win

http://unit.aist.go.jp/itri/knoppix/win/index-en.h tml

QEMU with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/qemu/index-en. html

coLinux with KNOPPIX

http://unit.aist.go.jp/itri/knoppix/colinux/index- en.html

Here is an installer to NTFS of Windows2000/XP.

Install2win

http://unit.aist.go.jp/itri/knoppix/win/index-en.h tml

from the page

http://www.aist.go.jp/aist_j/press_release/pr2005/ pr20050304/pr20050304.html

about half way down

from the page

http://www.aist.go.jp/aist_j/press_release/pr2005/ pr20050304/pr20050304.html

about half way down

Shaving off those 3 minutes might help.

knoppix with "readahead" and "bootchart" makes quick boot from CD and visualize the result.

http://unit.aist.go.jp/itri/knoppix/readahead/inde x-en.html

"readahead" is a tool to populate the page cache with data from files so that subsequent reads from these files will not block on disk I/O.
"bootchart" is a tool for performance analysis and visualization of the GNU/Linux boot process.

DVD Knoppix is good ( http://iso-top.de ) . UNESCO Linux is good (Look for freeduc-cd ). SuSE Live 9.2 is good. Better if you glue in a 'qemu' and make it able to autoplay-boot virtually under Windows. Puppy Linux on a USB stick is also good. Who needs disks ?

Also figuring out biology seems to be a lot harder than figuring out networking, at least there are all kinds of nefarious things but also serendipitous things found. Like one presentation I just heard had a U.S. scientist who announced that they had discovered an entire signalling network in human cells that was like the one found in yeast cells. And apparently more proteins can be encoded than the number of genes, because of alternate orderings (counting from different displacements in the gene, I think, ask a real bioinformatics expert). One talk I heard a year ago that stuck with me was a scientist who had devised a way to find signalling pathways in cells quickly; by forcing the cell to die if certain requirements were not met, he created a parallel computer that allowed him to discover a whole swath at once. There is also a lot of math and statistics, as well as a lot of biological knowledge behind it, it is not strange to see various statistical tests, references to different computer programs they used for analysis, or a mention of simulated annealing (well maybe that one not so often, came up yesterday though).

One interesting thing is that they (the H-Invitational people / Japan Bioinformatics Consortium) have I believe twice held what they call annotation jamborees, much like a hackfest! In 2002 they had 120 scientists gather (mostly Japan but from all over the world) in a big room with a computer per person. They locked them in for 10 days, and annotated IIRC over 20,000 genes, basically doing a figure some man years of work in a week, inputting data so it can be searched, analyzed, and crossreferenced.

They do have a comparison between mouse and human genome there, I wonder if something similar could be done in open source in terms of annotating and indexing a libary of open source code in different languages, really all in one pseudo language would be more useful perhaps. Anyway biologists are learning from computer scientists learning from mathematicians, and someone famous has said that in the future, all science will be computer science.

Bioinformatics people are doing text mining and data mining, but also there are many flavors and types of analysis programs designed to penetrate and match up information as encoded by tiny molecules, folded proteins, genes, and so on. Here are some links to get started. Also note the perl for bioinformatics books, and there was a big oreilly bioinformatics conference archived from 2003 and other links too (see bio.oreilly.org link below).

I cannot speak for everyone, but I can convey what I have heard, that there have long been communication gaps that have held back some of this, actually cultural differences. For example physicists like pure math and biologists deal in dirty, wet things.. when people successfully combine different perspectives in this area [more] discoveries start getting made. In Japan at least they are trying to figure out how to grow more bioinformaticists, since students tend to go only towards either biology or towards computer science (why study twice as hard). But there seems to be a lot of interesting stuff in there for both sides.

PLoS Bio article
some clusty
faq

The reproduction part of creating robot might be achievable some day when the robots can produce their own control logic. One thing attractive about this robot is the simplicity, op amp , resistor, capactors, that makes reproducing robots by robots themselves much earier than the digital conterpart.
Low-Drive-Voltage Printable Organic Field Effect Transistor

Check out http://plyojump.com and its accompanying blog for very good summaries on Japanese humanoid robot developments. ( the site seems to be down at the moment, but google cache helps )
QRIO was presented already back in august, at Robodex2003. QRIO is a direct followup, "production release" for previous development codenamed SDR-4XII.

There were other bots presented at Robodex, that were able to perform jumps and even somersaults.

The most interesting two IMO, are not megacorps entertainment bots ASIMO, AIBO etc, but humanoids that are of practical use or very low-budget, like HRP-II that is able to drive a backhoe, remotely assisted

And other one, SILF developed by a single person ( student ? ) on obviosly quite a low budget. Still, the bot is able to perform jumps.

Here

Also note that it's actually based on the ideas initially developed by HTCPCP protocol, which just turned 5 years.

Whoever said that computers can't handle superposition has never heard of convolutional neural networks.

Really, comparing human intelligence to computer intelligence doesn't seem like a good idea unless we're going to define what kind of computer intelligence it is.
Neural computing really screws the comparison up - the kinds of computing that normal computers are good for are quite different from the kinds of computing that neural nets are well suited to. Furthermore, different neural net architectures make for different capabilities - the tasks a feedforward network are best suited to are very different from the tasks a bayesian network are best suited to.

Take a look at this page for a good run-though of the different kinds of nets.

• ftp://gd.tuwien.ac.at/infosys/browsers/mozilla/sou rces/
• http://gd.tuwien.ac.at/infosys/browsers/mozilla/so urces/

• ftp://mozilla.mirror.pacific.net.au/mozilla/
• http://mozilla.mirror.pacific.net.au/
• ftp://ftp.planetmirror.com.au/pub/mozilla/
• http://planetmirror.com.au/pub/mozilla/

• ftp://ftp.belnet.be/packages/mozilla/

• ftp://ftp.ntrl.net/pub/mirror/mozilla/

• ftp://sunsite.ualberta.ca/pub/Mirror/mozilla/

• ftp://sunsite.cnlab-switch.ch/mirror/mozilla/

• ftp://ftp.ibiblio.org/pub/packages/infosystems/WWW /clients/mozilla/
• http://www.ibiblio.org/pub/packages/infosystems/WW W/clients/mozilla/
• ftp://ftp.tux.org/pub/net/mozilla/
• http://www.cise.ufl.edu/ftp/mirrors/mozilla/
• ftp://ftp.yggdrasil.com/mirrors/site/ftp.mozilla.o rg/pub/
• ftp://sunsite.utk.edu/pub/netscape-source/
• ftp://archive.progeny.com/mozilla/
• http://archive.progeny.com/mozilla/
• rsync://archive.progeny.com/mozilla/
• http://mirrors.xmission.com/mozilla/
• ftp://mozilla.teleglobe.net/ftp.mozilla.org/pub/

• ftp://ftp.cvut.cz/mozilla/
• http://ftp.cvut.cz/mozilla/

• ftp://ftp-stud.fht-esslingen.de/pub/Mirrors/ftp.mo zilla.org/pub/mozilla/
• ftp://ftp.fh-wolfenbuettel.de/pub/www/mozilla/
• ftp://ftp.uni-bayreuth.de/pub/packages/netscape/mo zilla/
• ftp://sunsite.informatik.rwth-aachen.de/pub/mirror /ftp.mozilla.org/pub/
• ftp://ftp.leo.org/pub/comp/general/infosys/www/bro wsers/mozilla/
• ftp://ftp.rhein-zeitung.de/mirrors/mozilla.org/
• ftp://ftp.uni-erlangen.de/pub/mirrors/mozilla/
• http://ftp.uni-erlangen.de/pub/mirrors/mozilla/

• http://mirrors.sunsite.dk/mozilla/
• ftp://mirrors.sunsite.dk/mozilla/
• rsync://mirrors.sunsite.dk/mozilla/

• ftp://ftp2.mbp.ee/pub/mozilla/

• ftp://ftp.rediris.es/mirror/mozilla/
• http://ftp.rediris.es/mirror/mozilla/
• ftp://ftp.etsimo.uniovi.es/pub/mozilla/
• http://www.etsimo.uniovi.es/pub/mozilla/

• ftp://ftp.funet.fi/pub/mirrors/ftp.mozilla.org/pub /

• ftp://ftp.univ-lille1.fr/pub/mozilla/
• ftp://ftp.oleane.net/pub/mozilla/
• http://ftp.oleane.net/pub/mozilla/
• ftp://ftp.free.fr/pub/Networking/www/Mozilla
• ftp://fr2.rpmfind.net/linux/mozilla/
• http://fr2.rpmfind.net/linux/mozilla/

• ftp://ftp.ntua.gr/pub/www/Mozilla/ ftp://ftp.duth.gr/pub/mozilla -->

• ftp://ftp.comp.hkbu.edu.hk/pub/mozilla/
• ftp://sunsite.ust.hk/pub/mozilla/

• ftp://ftp.c3.hu/pub/mirrors/mozilla/
• http://mozilla.szentimre.hu/

• ftp://ftp.EUnet.ie/mirrors/ftp.mozilla.org/pub/

• http://mozilla.fresh.co.il/

• ftp://ftp.cin.nihon-u.ac.jp/pub/net/www/mozilla ftp://his.ktarn.or.jp/pub/mirrors/mozilla/ --->
• ftp://ring.aist.go.jp/pub/net/www/mozilla/
• ftp://ring.crl.go.jp/pub/net/www/mozilla/
• ftp://ring.etl.go.jp/pub/net/www/mozilla/
• ftp://ring.exp.fujixerox.co.jp/pub/net/www/mozilla /
• ftp://ring.nacsis.ac.jp/pub/net/www/mozilla/
• ftp://ring.so-net.ne.jp/pub/net/www/mozilla/
• ftp://ftp.jaist.ac.jp/pub/Mozilla/
• ftp://ftp.lab.kdd.co.jp/Mozilla/
• ftp://ftp.kddlabs.co.jp/Mozilla/
• http://mirror.nucba.ac.jp/mirror/mozilla/
• ftp://mirror.nucba.ac.jp/mirror/mozilla

• ftp://ftp.hpcnet.ne.kr/pub/Mozilla/
• ftp://linux.sarang.net/mirror/network/client/web/m ozilla/

• ftp://ftp.uninett.no/pub/network/www/mozilla/

• ftp://sunsite.icm.edu.pl/pub/mozilla/
• http://sunsite.icm.edu.pl/pub/mozilla/
• ftp://ftp.task.gda.pl/pub/mozilla/

• ftp://ftp.global-one.pt/pub/Packages/Mozilla/

• ftp://ftp.chg.ru/pub/WWW/mozilla/

• ftp://ftp.sunsite.kth.se/mirrors/ftp.mozilla.org/
• ftp://ftp.fredan.org/mozilla/

• http://mirror.averse.net/pub/mozilla/
• ftp://mirror.averse.net/pub/mozilla/

• http://www.mozilla.sk/pub/

• ftp://ftp2.sinica.edu.tw/pub3/www/mozilla/
• ftp://ftp.nctu.edu.tw/WWW/mozilla/
• rsync://ftp.nctu.edu.tw/ftp/WWW/mozilla

• ftp.mirror.ac.uk/sites/ftp.mozilla.org/pub/ ftp://ftp.gbnet.net/pub/mozilla/
• http://www.gbnet.net/public/mozilla/ -->
• ftp://mozilla.teleglobe.net/ftp.mozilla.org/pub/

Here's my guess: The funneling effect of the arrowhead allows more randomly-moving particles flow in that direction rather than the rather small reverse opening.

That's a GUESS, allright! Did you view the videos? No, I guess not.

Clearly the ones going the "wrong" way turn around inside the "arrows" (triangles, really). The microtubules appear to stick to the walls of the track, thus they go down the angled side of the triangle, hit the corner at the base and turn, follow the base and fail to negotiate the 90 degree turn to continue down the track the way they were headed (wrong-way), hit the next corner and turn again, following the side of the triangle toward the center of the track, which they join now going in the desired direction. It doesn't take too long for the approx. half going the wrong way to get turned around within the first triangle they encounter.

The only real mystery is why they fail to make the 90 degree turn from the base to the track. This picture clearly shows the true shape of the "arrowheads", and from the looks of it the microtubules should stick to the track wall and turn down the track, continuing the wrong way -- the radius is certainly greater than the triangle corners, which they negotiate just fine. Why do they "unstick" from the track wall and cross the "upstream" track opening? Are the microtubules longer than the track is wide (longer than the opening in the base)? How flexible are the microtubules? Can they bend around corners?

It would be interesting to see their behaviour traversing various geometric shapes. I'm sure they tried more than just triangles, but they're only showing us the "useful" one.

Here's a better SRP link:
http://berlin.arcot.com/srp/

Here's a python implementation of the protocol:
http://www.aist.go.jp/NIBH/~tomh/sr psocket.html

http://www.aist.go.jp/NIBH/~b0616/ Lab/Links.html has a list of many interesting A-life applets.

Langston's Ant is listed on the page, and is located at:
http://www.home.fh-karl sruhe.de/~gran0011/java/langton.html