Slashdot Mirror

All of these predictions are testable. If we find life that isn't related to that life we know about, we'd need to change or abandon the theory. If there are large gaps we can't explain, we would need updated theories, like punctuated equilibrium, to explain how A can evolve into B. Evolution would be hard-pressed to explain life without self-preservation instincts or without proteins. But there haven't been, to our knowledge, any contradictions like this.

All of these predictions are based on historical evidence, which is expected because most evolution happens very slowly. But there really isn't anything wrong with "reactionary" theories - they're better than cop-out answers, which I consider creationism to be. But evolution predicts other things which can be verified in the lab - the natural selection of moths, development of antibiotic resistant strains of bacteria, and even speciation in fruit flies and the like.

Most interesting to me, evolution as a process can be applied in more ways than simply to explain life on earth. You can use it to describe market economics. You can even use it to design circuits, programs, and robotics.

Sage Report

Trial Lawyers, Inc

In general it sounds as if you are unfamiliar with the topic and the methods required to solve the problems involved.

Robert

This is a very neat application of genetic algoriths! Haven't seen something so interesting since this Tron cycle game.

they've done this at Brandeis. unpatched windows xp/2000 computers are banned from the network.

Since we're already offtopic...
Hmmm, a press release by the Oregon Food Bank? Talk about a organization with a vested interest in people being hungry.
The source of their "hunger study" is Brandeis U. a small Jewish (read liberal) University. Is this the *best* the OFB can do, or is it the only research they can find that justifies their existance and continued funding?

How do they define "hunger"?
21% of households with children had to cut or skip meals
39% of those said it happened every month
so 8.2% cut or skip one or more meals per month. Once? Twice? Everyday? Without better numbers this is just a huge pile of FUD.
The OFB and those like them are delighted that Oregonians keep swallowing it (with a big swig of guilt, no doubt)

For what I think is the source of the fourfold player type thing (explore, socialize, kill, achieve), see this 1996 article by Richard Bartle, a mud pioneer.

...with the entertainment consisting in exploration and storyline rather than in player status and achievement.

Sounds like the closest thing yet to an actual online RPG, and it's not even being called a MMORPG

Bartle wrote long ago how there are four types of online gamers -- killers, achievers, socializers, and explorers. While this may be somewhat simplistic, it turns out to be pretty accurate. This game simply focuses on Exploration and Socialization. The Sims Online focuses on Socialization and Achievement (get into those top-10 lists!). AC2 focuses almost entirely on Killing.

There are lots of "actual online RPGs" out there, they just never made it very big. If you want true roleplaying, try out Underlight for example.

Besides, I don't see anything how this game is going to "enforce roleplaying" at all.

The four things that people typically enjoyed personally about MUDs were:

i) Achievement within the game context.
Players give themselves game-related goals, and vigorously set out to achieve them (...)

ii) Exploration of the game.
Players try to find out as much as they can about the virtual world. Although initially this means mapping its topology (ie. exploring the MUD's breadth), later it advances to experimentation with its physics (ie. exploring the MUD's depth).

iii) Socialising with others.
Players use the game's communicative facilities, and apply the role-playing that these engender, as a context in which to converse (and otherwise interact) with their fellow players.

iv) Imposition upon others.
Players use the tools provided by the game to cause distress to (or, in rare circumstances, to help) other players. (...)

There are any number of legitimate reasons why MMORPG players who prefer the three types other than "Achievemer" would run multiple characters. "Explorers" would want to try many different classes or races. "Socializers" would want a different character to suit different moods or hang out with different crowds. "Imposers" (player-killer types) would need plenty of backups....

Furthermore, Holocron's post made no mention of whether any reasonable pricing scheme other than forcing users to start entire new accounts (doubtless containing much redundant information) was even considered.

The statement that multiple accounts are used primarily for muling belies an overly constrained mindset about how and why people play MMORPGs. I can only conclude that cutting out three fourths of SW:G's potential market with this draconian pricing move will only have a negative impact on profits.

That first video on his site reminds me of a couple programs I've seen that do this sort of stick evolution. IIRC, the programs also have a distributed computing type function where they exchange what they evolve with other clients. If anyone's interested:

Framsticks

They're fun programs to try out. Framsticks is still under active development but the Golem Project is no more. Although I believe you can still download the last Golem client from their website.

Both sites should contain relevant articles/research on stick evolution for anybody who wants to learn more.

I'm not giving up the AOL discs I have recieved in the mail.
I have found them instead, to be very usefully in helping me balance the feng shui in my dorm room.

This simulation is on par (maybe a *little* more in-depth than) the GOLEM project that's been running out of Brandeis for quite a while now (you can download your own evolution simulation). Basic blocks which when recombined with each other "develop" into more complex things that can be rated on their ability to function in a certain role and be recombined to hopefully produce something even more efficient.

More impressive forced-evolution science is the DNA shuffling work of labs and companies like Maxygen. This is truly evolution in a tube and deals directly with the genetic sequence as opposed to higher-level vague cell-type simulations.

But the great thing about science is that there's room enough in it for pretty much any research.

It is called GOLEM and is found at http://golem03.cs-i.brandeis.edu/. It's not a bad little program and fun to watch as a screensaver (personally more fun than the SETI@Home fast Fourier transforms flying by).

This article is interesting, however, because it moves the agents into the physical world where it isn't possible to obtain the same kind of idealized environments that are possible in silico.

Here is an example of a lego robot designed by a genetic algorithm.

This how to guide gives step by step instructions (with pictures, yay!) on how to secure an Outlook client.

An online publication venue for this kind of work (and a place to go to read other related work) is the Journal of Virtual Environments (formerly Journal of Mud Research).

With these modules it would be easy to simulate those robots with a genetic algorithm. This way the compilation of modules which fits the task best is computed.
A simliar approach to that is the golem project
Imagine you set the task to build a house and the robots that fulfil it best are copied with slight mutations. Then the simulation starts again until you have the perfect housbuilder robots.

There's really no need to pay for this sort of thing. Searching for "redhat update" on Freshmeat reveals 5 GPLed update tools. I even wrote one myself to meet my university's specific needs. Download it, run it with -writeconfig, edit the config file to point to your favorite mirror, copy the script to /etc/cron.daily, and you're set for automatic update retrieval.

Beyond basic "rpm -Fvh" functionality, it can be configured to send mail to a specific address when updates arrive, ignore certain packages (with regexp support), and write a script which performs the updates when run.

It's called HURL (Hurl Updates Redhat Linux), and you can get a copy here. Drop me some mail if you like it or have suggestions.

There's a very cool application of genetic algorithms that I saw a few years back. Danny Hillis was trying to evolve sorting networks, a way of representing a sorting algorithm for a fixed number of inputs. (See volume 3 of Knuth, The Art of Computer Programming). He wanted to do it using genetic algorithms, on 16-input sorting networks. The best known one at the time used 60 comparison/swaps to sort 16 inputs.

The problem is, in order to measure the "fitness" of a sorting network, you should give it all possible sets of numbers and see how many it sorts correctly (you also give a fitness bonus to smaller networks). It turns out you just need to give it all possible sets of 0's and 1's to see if it will sort any set of numbers correctly, so Hills would have to test each network on 65,536 inputs to see how well it did.

That would take too long, so he wanted to only test the networks on a subset of possible inputs. The clever thing was he made the particular subset used also evolve, as a kind of "parasite" on the sorting networks. The parasites were "rewarded" (had higher fitness) when they broke sorting networks. That way, the system would keep around precisely those test cases which could break the current population of sorting networks, so it was always focusing the testing exactly on the trouble cases, and ignoring the ones "known" to work, and thus saving a ton of time/effort.

Hillis evolved a sorting network which used 61 comparison-swaps, just 1 away from the best man-made one known. I was at Thinking Machines (Hillis' company) for a while, and fiddled around with this myself a bit, thinking that a bit more simulation must beat the record, but I never did beat it.

Hillis had a paper, called "Co-Evolving Parasites Improve Simulated Evolution as an Optimization Procedure", published in Artificial Life II (Langton et al, editors), Addison Wesley, 1991, pages 313-324. A note in my database indicates it may also have been published in the journal Physica D, vol. 42, p. 228-234.

A search also just turned up Hugues Juille, who has apparently done some more work in this area. He evolved a 60 comparison sorting network for 16 inputs, tying the record. And he broke a (25-year-old) record for 13-input sorting networks, doing it in 45 comparison/swaps.

There's a very cool application of genetic algorithms that I saw a few years back. Danny Hillis was trying to evolve sorting networks, a way of representing a sorting algorithm for a fixed number of inputs. (See volume 3 of Knuth, The Art of Computer Programming). He wanted to do it using genetic algorithms, on 16-input sorting networks. The best known one at the time used 60 comparison/swaps to sort 16 inputs.

The problem is, in order to measure the "fitness" of a sorting network, you should give it all possible sets of numbers and see how many it sorts correctly (you also give a fitness bonus to smaller networks). It turns out you just need to give it all possible sets of 0's and 1's to see if it will sort any set of numbers correctly, so Hills would have to test each network on 65,536 inputs to see how well it did.

That would take too long, so he wanted to only test the networks on a subset of possible inputs. The clever thing was he made the particular subset used also evolve, as a kind of "parasite" on the sorting networks. The parasites were "rewarded" (had higher fitness) when they broke sorting networks. That way, the system would keep around precisely those test cases which could break the current population of sorting networks, so it was always focusing the testing exactly on the trouble cases, and ignoring the ones "known" to work, and thus saving a ton of time/effort.

Hillis evolved a sorting network which used 61 comparison-swaps, just 1 away from the best man-made one known. I was at Thinking Machines (Hillis' company) for a while, and fiddled around with this myself a bit, thinking that a bit more simulation must beat the record, but I never did beat it.

Hillis had a paper, called "Co-Evolving Parasites Improve Simulated Evolution as an Optimization Procedure", published in Artificial Life II (Langton et al, editors), Addison Wesley, 1991, pages 313-324. A note in my database indicates it may also have been published in the journal Physica D, vol. 42, p. 228-234.

A search also just turned up Hugues Juille, who has apparently done some more work in this area. He evolved a 60 comparison sorting network for 16 inputs, tying the record. And he broke a (25-year-old) record for 13-input sorting networks, doing it in 45 comparison/swaps.

Take a look at JScheme. You can embedd it into your Java spps.

Take a look at Silk, a Scheme interpreter as a jar, that you can use in your java apps. I used it for scripting in ArgoUML and liked it a lot.
http://www.cs.brandeis.edu/silk/silkweb/index.html

Biologically occurring chemical designs are REALLY very good at what they do. So efficient, that we might be tempted to flat out state that even clever mechanically engineered designs will never approach them. But we should keep in mind that in the "design space" that biological evolution works in not every possible idea is tried or tested. Its just that there is the potential for any combination to be explored.

Human (or machine) created designs can come up with "new" ways to do things. Some of those designs might be more efficient in certain applications than biological designs. A simple example might be the wheel. The last time I checked no macroscopic organisms on this planet where using wheels to get around. On flat surfaces (deserts/ice/savannas/etc.) wheels are more efficient than legs. It is not that evolution couldn't "invent" wheels, it either hasn't successfully happened before, or the design was discarded in favor of the more functionally flexible limbs we do posses. Check out the GOLEM project for more ideas.

I do not think that biological designs have the end all say on nanotechnological possibilities. But the science of biology currently has far more small scale tools and vastly greater experience (nature's experience, that is) at its disposal for the development of useful nanotechnology.

Killers can go hostile and attack anyone

Socializers have both a chat context and global in-game chatting

Achievers have titles, including hardcore, levels, and equipment to seek after

Explorers get some because the world is new every time

So, Diablo 2 clever caters to anyone with an interest in RPGs by having something for all player types -- but their system of equipment makes the achiever an absolute addict. There's ALWAYS another piece of equipment waiting to be had better than what you have. And, sadly, unlike a game of skill like, say, Quake 3, there isn't a great deal of real gain to be had from repeat play, other than equipment gathering. Admittedly, if you were a true groundbreaker, you came up with a playing style that was unique (say, the Tweaker Sorceress or the Hammerdin), but after a gentle early learning curve, most people were eq-dependant.

So, is the game actually FUN? I played it a ton, but if fun is enjoyment, then I couldn't say it was. I wanted to enjoy it but I was too frusted with all those things about D2 that have become terrible jokes:

Grotesque amounts of lag (on a T1, I should add)

Utterly stupid PvP (who hits who first?)

Totally imbalanced character classes

Lame Multiplay (95% of people want to solo in 8-player games, and 80% of them want to do it on the River of Flame, all at once, and barbarians can steal kills, etc)

On top of all this, the "realms" were subject utterly to cheating of the first order. Since I myself spent _hours_ showing support people how to 'dupe' with one of the methods (there were at least 4 different 'duping' bugs, if not more, over the course of the game), I can attest to the fact that they don't take it seriously, thus scrapping the d2 "economy", for better or for worse.

For all these reasons, and more, I've steeled myself against buying the expansion. I don't imagine the major gameplay problems have been improved. I'll buy the Throne of Bhaal expansion back for BG2 instead, and bide my time waiting on Neverwinter Nights, which may well end up becoming the most popular online RPG ever, given that PLAYERs will be able to build, run, and link their worlds, and create new content with an incredibly rich toolkit. Back to the four player classes, I can already see most of them very much enjoying it.

Enjoy D2 if that's your bag, I don't begrudge anyone their game. But D2 has gone down in my book as excellent psychology attached to a very poor game, and despite being one of those people who would spend hundreds of dollars on some of my favorite games because I enjoy them so much, I'm not going to reward Blizzard with another $35.

Well before you start, make sure you read, in addition to the Larry Niven and Issac Asimov books listed in the other two replies, this book, which also encompasses this very similar premise. When I first saw this news story I thought "urban monad." I read Silverberg's book over twenty years ago and have been generally fascinated with the concept ever since. It's a pity that as an adult I developed such a fear of heights, or I would be clamoring to live in one of these.

This story was barely funny, at best. I don't think it was worthy of front page on Slashdot. Particularly when there are other's that have been submitted and rejected that are much more interesting.

In reference to the prior comment: The Hardest Game of Tron You'll Ever See

1. Any Language. [...] I know the Java bytecode isn't tied to Java the language, but realistically, that's the way Sun as limited it.

The other purported advantages you list for C#/.NET/CLR/MSIL are similarly specious. Large-fraction-of-native-speed, cross-platform, secure distribution is already available with the Java-the-platform, and the other advantages you mention are at least as easy with Java (Java-the-language and Java-the-platform) as with .NET. Further, free Java/JVM implementations are mature and widely available (from, for example, IBM, Blackdown, TransVirtual (Kaffe), and Sun, for varying definitions of "free").

.NET is not an example of Microsoft "getting it." .NET is an example of Microsoft continuing not to get it -- reinventing the wheel, rather than building on perfectly good existing systems.

--

Instead, use bigloo, scheme48, or some other Scheme with a decent FFI. Then you can use the native Scheme calls.

An even-better alternative is to use SILK or KAWA, which are scheme systems that can interface with Java, thus using JDBC.

There's a great paper called "Java reflects easily through Scheme" which discusses SILK and extending Scheme to allow access to Java code.

Good luck -- but with Scheme, you won't need it. What a great language!


~wog

For those scared to learn latex but want a superior document processing system LyX does work and work well. I have used it for many articles and papers for work and school. And as an added incentive there is a mif to lyx filter if you happen to be one of those now stuck by Adobe's decision.

lyx website.

mif 2 lyx translator.

Neat, revolutionary stuff. Hopefully something comes of them.

there's a mirrored picture here

for people that really want to see an image of it, here it is.

Golem@Home is my favorite. Use spare cycles to design/evolve new robotic 'lifeforms'.

Entropia has several science and medical oriented research projects underway.

Popular Power is working on new influenza vaccines.

Folderol is doing Human Genome stuff.

There are dozens of others out there, but if nothing turns you on, the folks at the Cosm Project have an open source platform for building your own distributed computing project.

Done.

See DEMO.

Hey, here's a link to the software they used:

http://golem03.cs-i.bran deis.edu/download/LiveTruss120.zip

Here's a link to the 1.20 version:

http://golem03.cs-i.bran deis.edu/download/LiveTruss120.zip

This was easy enough to find, but nevertheless:

http://www.demo.cs.brandeis.edu/golem

There are pictures of the robots available there, plus videos (in MPEG, joy!) of the bots moving along a carpeted surface, and VRML models.

I might be dense, but some of these designs are actually interesting, in how the frictional physics of the carpeted surface are taken adavantage of in strange ways. Course, I'll be really impressed when the computer comes up with a top-heavy upright biped with two counter-balancing flagella.

--

The link isn't there, but a quick use of backspace will show you that there are 3 files in the download directory. One of them, magically, is the 1.19 version... which, by the way, doesn't seem to run under Win2K.

didn't check for Win2000 compatibility, but here's a link to v1.19 http://www.demo.cs .brandeis.edu/pr/golem/download/LiveTruss119.zip

Make sure you download the simulation software at http://www.demo.cs.brandeis. edu/pr/golem/download.html. It only runs under Windows, but it's curious to watch the process running. I've got it running on a couple of computers overnight. It will be interesting to see what is crawling around at the office tomorrow morning.

DRI is only for 3D support.

For video capture under Linux using The Marvel or Rainbow Runner G, go to here. The drivers are still in an unstable "alpha" state, but work well enough to use and start hacking on.
________________________________

The majority of the cards that do work are supported under the BT848 chipset, but there is work in progress for the Matrox Marvel G200 with Video 4 Linux:

http://www.cs.brandeis.edu/~eddie/mga4 linux/

I have followed the Broadcast 2000 development for a while, and I am both impressed by and thankful for their work. The drivers appear to be the missing link. Unfortunately, on the Broadcast 2000 page, they skip the issue of drivers, simply saying that any "Video4Linux" drivers will work, without so much as providing a link to a Video4Linux project page.

At least it's a good sign that they're sticking to a standard, rather than creating their own drivers. Unfortunately, the driver situation seems to be quite a mess. There's Video4Linux, Video4Linux 2, LiViD, and other projects and I'm not sure how they're related. Most of them seem to support mostly TV-in-a-window cards rather than full-motion video capture.

It looks like work on the Matrox Marvel drivers is coming along, which is good for me because I just bought an Athlon 550 and Matrox Marvel G400-TV! >geek gloatCineGX which seems to have disappeared. The idea was to create a framework for handling video, applying filters and codecs in a pipelike fashion. That would mean that you could, say, stream an AVI file from your disk, and show it as a picture within a live TV feed, showing it on the screen and spitting it into an MPEG2 file on disk all at the same time (provided you have enough processing power and/or hardware assistance). Most editing could be reduced to problems of mixing and matching streams and filters. The editing GUI would exist only to provide a project management function.

• http://dendrite.cs.brandeis.edu/proje cts.html
• http://www.cs.brandeis.edu/~hornby/lux o1.mov

• http://dendrite.cs.brandeis.edu/proje cts.html
• http://www.cs.brandeis.edu/~hornby/lux o1.mov

The Brandeis University Dynamical & Evolutionary Machine Organization (DEMO)
An overview along with AVI and VRML files are here

The Brandeis University Dynamical & Evolutionary Machine Organization (DEMO)
An overview along with AVI and VRML files are here

In the image gallery there is a picture of the actual lego bridge.

First off, while you can get to it from the links above, the impatient among you can click here for an animation (animated gif) of the bridge building process.

Second, after watching that, can someone explain to me how, in this case, GA show any advantage over boring old Newell and Simon means-ends analysis? I mean, in general, yes, I see the advantages of GA, but this looks like a case for "identify your situation", "identify your goal", "identify an operator that might reduce the distance between your situation and goal", "apply operator", "repeat." Why let randomness figure into it?