The goal of this project isn't to recreate SETI@Home but to give astronomers all over the place access to data collected by instruments in places where they aren't. We've got thousands of instruments gulping down data but most of it doesn't ever get processed, just stored for later. Like the article says, anybody can have access to massive amounts of raw data. A grad student in the UK can download data gathered from telescopes in Hawai'i and write his or her own program to process them looking for the data they want. A group of amateur astronomers could request a bunch of wide field images and scavange through them looking for comets or asteroids.
If we ban cloning does that mean monocellular life will be against the law? Cloning is something readily done in nature, your entire body is constructed of cloned cells. They've all got your DNA and act just like the other versions of themselves.
What cloning won't do:
1. Allow you to make a clone of someone and replace them in society with an exact replica. A clone of me made tomorrow would still take a normal amount of time to grow up and may or may not be anything like me. Genetically we'd be identical but unless he traveled back in time to live my life for me he probably wouldn't end up anything like me.
2. Allow me to create an army of super clone warriors to take over the world. Said soldiers would have to be gestated and raised like a normal army of soldiers.
What cloning embryos WOULD allow:
1. Do gene mapping and stem cell research with a very large subject base with little genetic discrepency. Every wonder why fruit flies and a few simple plants have been used for the past whatever years for biological experimentation? There's little genetic diversity and they're plentiful.
2. Figure out how to regenerate cells by cloning them so you can repair almost any part of the body damaged by just about anything. There's not a whole lot of a chance for rejection when you're your own oragan donor.
Cloning research doesn't require an embryo to be gestated. Then of course there are those holding to the notion that life begins as an embryo and all that jazz. That is just picking at straws because you don't have enough understanding of the process to make a logical argument against it. If you want to save a baby stop jacking off and ovulating but don't harrass somebody trying to make you and your kids have a better life.
Re:Minor catch in your plans . . .
on
Mining On The Moon
·
· Score: 3, Informative
I think that's far too conservative of an estimate for fusion powered spacecraft (technologically, politically it could be never when a fusion powered spacecraft is sent to another planet). The most energy efficient AFAIK fusion reactors we're researching are colliding beam fusion systems (Tokamak). These work really well fusing Deuterium-Tritium but can be used with Deuterium-Helium and Hydrogen-Boron fuels as well. The problem with Deuterium-Helium fusion is the near complete lack of He-3 on the Earth's surface. We're much too hot with too little gravity to hold onto Helium for very long in any form and our atmosphere blocks most of the cosmic rays that form it. The Moon however would be a good source of it since there's nothing protecting the regolith from cosmic ray bombardment. With CBF reactors once you get stable fusion with one fuel source most of your work is done in getting to use other fuel sources. I think cutting your timetable in half would be a little better of an estimate.
Re:The Real Treasure Of The Moon...
on
Mining On The Moon
·
· Score: 3, Interesting
The catch-22 with a lunar launched mission is the cost of getting equipment from the moon. Even if you merely lauch modular facilities to process regolith to construct a spaceship with that still counts as cost for a Mars mission. Mining ice on the Moon for a chemical rocket is dumb anyhow, Helium-3 is much energetic of a fuel and will get a craft to Mars in a much shorter timeframe. Furthermore you need not use the fucking space shuttle to build a Mars ship, that would be absolutely ludicrous, the SST is one of the most expensive fucking launch systems in use currently. It'd be cheaper to build a magnetic linear accelerator up the side of a mountain and shoot stuff into orbit.
I thought Protector started off really well until it slowly devolved into "super genius can do absolutely anything at all" and saves the day for the human race. Waiting to find out what happens next kept me reading until the end but in retrospect I think I enjoyed Ringworld a bit better. I'm reading Ringworld Throne right now hoping not to be disappointed with the series.
Re:I have experience with both...
on
GTK-- vs. QT
·
· Score: 2
I thought Havoc Pennington's book on GTK+ was pretty good documentation. It's called GTK+/Gnome Application Development and published by New Riders IIRC. I picked it up for 14.95$ and have found is very useful.
I don't even think the schools really need to install RH on the systems because it can be done by whoever MS buys the boxes from (Dell or Compaq most likely) so it can be as easily done as writing an image onto the drive. That also works for maintaining the systems. The school's can be given disks with live images on them which are just dd'ed onto the hard drives. As long as there is a modicum of password security so people don't get to run around as root it'd be hard for anyone to really fuck up an installation anyways.
Well..for those of us who can read. Redhat is offering to stick Linux on these systems and provide unlimited support for the hardware MS buys. KDE and GNOME look alot like Windows and in many cases act alot like it. As long as the installation is kept pretty compact it shouldn't confuse anybody. You click the button at the bottom to get to your applications. Nautilus is a pretty workable GUI file manager desktop icons pretty easily get people to the media devices they want to use. It isn't like the schools have to admin the boxes themselves anyways. They call up RH if there is ever a problem and if the boxes have even modest internet connectivity an RH trouble shooter can run stuff remotely to fix the system.
Actually Harrison Ford and co DO get paid for their likenesses to be used on toys and whatnot. Part of the contract they sign says that they get a mad fat check every time some shit comes out with their likeness on it. Same goes for most actors whom appear as toys. Thats why you don't see huge blockbuster movies on the big three networks very often. The network has to pay the studio who has to pay the actors and whomever else was contracted.
I was just wondering the other day how feasible it would be to use a laser as a communication device between two birds. You can get really nice range with little EM interference with only a wee bit or output power. The one obstacle I kept running into whilst pondering a laserlink was keeping the beam aimed at another bird in a different orbit. Well hot damn and way to go. I guess I was hit in the face with the same muse as the dudes at the ESA just a little bit late and without any satellites under my control to play with...so far.
Anand wrote a very good article with plenty of meat and potatos to satisfy but I think he is sort of missing a very important point. He mentions in the article how the PS2 falls behind the XBox in some way because it has less fancy graphics and sound capabilities. SONY HAS SOLD NEARLY 20 MILLION OF THE FUCKING THINGS AND HAS HUNDREDS OF GAMES AVAILABLE FOR IT. I think Anand needs to reassess his position on the PS2. It is still selling for 300$ because people are still eating it up at this price. The XBox may be able to do a bajillion polygons per second but it still doesn't have the game franchises that make bank on console systems. Nintendo's also in a good position because they are destined to get the little kids who want to play Pokemon until their eyes fall out. Who cares if Pikachu's only rendered with half a million fucking surfaces, people want to play the games not write a master's thesis on the theoretical graphic capabilities of a computer system. I bought a GC on my way home from Louisiana and since I've been home my brother's been playing Rogue Squadron almst non-stop. He hasn't yet complained about the lack of theoretical polygons the GC can render yet and I sort of doubt he will. The XBox will only truely contend with Nintendo and Sony when it has games in high demand. I thought it's launch titles were pretty crappy compared to the GC's though a little better than what the PS2 originally offered. Besides that I got a GC and two games for the price of either the PS2 or XBox.
Isn't strange how fucking stupid the slashdot crowd is sometimes, especially something involving Microsoft. One genius posed the question "Do you think Gates and Balmer can envision great games?" Do you REALLY FUCKING THINK BILL GATES DESIGNS GAMES FOR THE XBOX? Other quite intelligent people are telling everyone they ought to buy XBoxes because Microsoft HAS to be losing money on each one you buy. The factories spitting out the XBox make just about every component in the console with most components being contained on a single board. It is very cheap to manufacture SBCs. The 300$ price tag doesn't leave much room for profit but they aren't losing anym oney on the damn boxes. Stop using pricewatch.com to find the prices of computer hardware and then trying to extrapolate the price of a console system. Others claim the XBox is little more than commodity PC hardware in a little black box. The memory subsystem isn't exactly an i815 chipset or something. Buy one if you want or don't buy one if you don't want to play the games. I'm waiting for Sunday to pick up a Gamecube. I can get a console and a couple games for the price of a PS2 or XBox console. Plus I get something for my GBA to talk to.
One option you might look into rather than USB cameras would be to use video capture boards and bullet cameras as your capture device. Most small cameras will get you from 360 to 380 lines of resolution in colour and usually 420 lines in B/W. They're fairly cheap and really rugged. In fact almost any CCTV camera ought to work fine as most output a composite video signal. Most run on a 12v DC power supply so you can even run them off batteries if you really wanted or needed to. You can bet small CCTV cameras will give you a bit better of a video signal than USB cameras and most capture boards allow for uncompressed video that you can pipe into whatever streaming codec you are hip to use. A board and camera combo might cost you from 150 to 200$. Here's a link to compatible capture boards.
The choice seems clear if you lack the ability to process information in a logical manner. USB is a host-based bus, it needs a host controller in order for any of the devices to talk to one another. All data has to pass through a central controller (on your PC) in order for any of the devices to even see one another. FireWire on the otherhand is a host independent system as each FW device has a FW controller as one of its logic conponents. This essentially makes all FW devices their own hosts. They can connect directly to one another and intercommunicate or in a daisy chain configuration each device can talk to any other device without the intervention of some contral controller. A FW camera and HD both plugged into a computer you can tell the camera to send video to the HD and it will with no further intervention of the computer. A USB configuration like that would require all data to pass through the computer's USB controller and then into the hard drive. There's FW drives that plug directly into DV cameras and can offload video without a computer even involved. The next FW spec increases the throughput to a couple gigabits per second, in some cases current FW components with a driver update would be compatible with the new FW spec.
Woohoo InfiniBand. Finally small companies can build clusters that can compete with multi-million dollar SGI super systems. Beowulf clusters using Ethernet interconnects are nothing compared to a system with memory bridge level hardware interconnects. Current Linux clusters are often based around Ethernet which is very slow and has a high overhead when you want to put it to the use of letting processes on different machines talk to one another. Big Beowulf clusters have been shown to work well on embarrassingly parallel computational problems (key cracking, some types of CFD, and packet analisys like SETI@Home) but aren't really up to snuff with truely high end super computers without much faster interconnects. Some clusters (I think Microway and a few others) are built using Myrinet interconnects which are faster and actually provide the sort of connections big clusters need. Even using Myrinet you end up with alot of scalability problems because you only get your max throughput when transfering very large blocks of data to the same host. Having a Northbridge based interconnect (rather than something like a Myrinet card which is connecting to the PCI bus which in turn hooks to the Southbridge) is a very good way to speed up clustered systems. IBM, Sun, and SGI learned this a long time ago which is why they use high speed interconnects between individual boxes rather than trying to overextend network protocols.
The other guy caught Bluetooth vs. 802.11b so I'll pick up USB vs. FireWire. What the fuck are you smoking? USB is a host based connection protocol and FireWire is a host independant protocol. That makes them not really even in the same ballpark. Just because they are both serial protocols don't mean jack shit. USB brokers all connections through a host controller, this host controller also acts as the switch for all communication on the bus. Everything has to pass through the host controller for any devices to talk to one another. FireWire is host independant which basically means every device on the bus is a peer of every other device (all devices are hosts). A FireWire DV camera can talk directly to a FW hard drive or DVD-RAM with no computer involved whatsoever. A USB webcam isn't going to be writing video to a USB hard drive any time soon because neither of them has a host controller. FireWire devices can also talk to each other directly rather than through a central host. Just because USB 2 is a high bandwidth version of USB doesn't mean it truely competes with FW. USB and FW are going to both be included in systems for a while since they both serve different purposes. You're not going to be seeing USB2 DV cameras for sale while FW cries in its corner.
Latency is a prime concern of the I2 network architects. One of the cooler aspects of the whole I2 thing in hardware is the use of packet switching schemes that resemble circuit switching schemes. Instead of a packet just flying around the network trying to get to the other end of it by whatever path a router deems viable, I2 routers route all of the data from two sources along the same path. Packetswitching was designed to be redundant for lossy networks, not fast for high availability networks. Routes can easily be recalculated but having data all go the same way lowers latency dramatically. I remember reading a bit ago about a realtime video conference between a professor and his class in the US and a professor and his class in Japan. Uncompressed realtime video doesn't work real well on the more traditional packet switched networks of the current internet. An interesting aspect of the circuit switching like packet switching is that specific pathways between a destination and origin can be planned out ahead of time and even sold. Instead of merely selling X megabits (or maybe gigebits) of bandwidth a company could sell a direct path or high priority path between two network nodes. Junior sending a video of his skateboard competition to grandma wouldn't interfere with the realtime traffic of a big Fortune50 paying beaucoup dollars for their network while Junior and Grandma are paying 50$ for their cable modems.
The 802.11 protocol isn't robust enough to handle a huge network cloud like you're thinking of. It uses many techniques of the wired 802.x schemes to detect nodes on the network, these work fine for wired networks where it's as simple as detecting a completed circuit but for wireless networks alot of overhead is added. Besides the problems in 802.11 for large numbers of nodes, you'd also need to come up with an efficient dynamic routing scheme. Packets can't just propogate over the entire network with no points that sort of direct their flow. If the wired internet worked like that data would never reach its destination. Wireless clouds are inefficient and messy and not very scalable at all. The bandwidth of 802.11 can't ever exceed B log2 (1 + SNR). Wired networks have the same restriction but can just add wires instead of refining their encoding schemes.
Re:Why use a PC like architecture???
on
RLX Gets Denser
·
· Score: 3, Interesting
A reason which is blantantly obvious to those of us who use non-x86 processors in our computers is that software originally written for x86 processor takes a bit of work to port to other processors and then the ports aren't always as functional as the original piece of software. RLX boxes run a slightly customized Linux installation but for the most part are using off the shelf GNU software. Were a company to come out with boxes using they'd have to invest alot of money into making sure their software was up to snuff on their hardware. Transmeta and those using Transmeta chips rounded the corners by making their product compatible with a shitload of software already existing.
Media producer and gamer are broad catagories. A couple years ago only a speed freak gamer would think of owning a 500 MHz machine, now my grandparents own a 500 MHz Gateway and a 700 Mhz HP and all they really do is type up greeting cards and use the internet. It is alot easier to do any sort of graphics work on a 700 MHz Celeron then it is trying to do it on a 66 MHz 486. On that point now all "media producers" are professionals. Lots of people are make their own audio CDs, touching up old family photo albums, and getting creative with home movies on their computers. More and faster are keys to these sorts of applications. I used to do video editing work on old PowerMac 8500AVs and it took forever to do even the simplest things with small video clips, you wouldn't think of doing a major project without access to a machine with a several thousand dollar hardware video card. With a 799$ iMac people can do in an afternoon what it would take me a week to do.
For the bitchers, please remember than the GNU-Darwin project no matter how inaptly named is not supported by Apple and thus anything mentioning the GNU-Darwin project may fuck up an Apple supplied OSX installation. Besides that I think this project is pretty nifty, it's one of the reasons I was pretty excited about getting OSX on my Powerbook. Not only would I get the cool features of OSX I also get to fire up Terminal.app and use a huge number of FreeBSD ports. The article from the other day where both OS 7.6 and WindowsXP were running on an iBook 466 I found pretty interesting. There was little content but the demonstration was pretty cool.
This is not an upgrade for an Apple supplied OSX installation, this is an installation for a Darwin (not supplied or supported by Apple) installation. If you're using OSX and running this script you'll probably fuck something up seriously because OSX is a bit different from the Darwin distro. You can run the GNU-Darwin Ports on OSX but it would be pretty silly for me to run this script on my 10.1 installation on my Powerbook.
Slashdot Mirror
All of your GPS are belong to us!
The goal of this project isn't to recreate SETI@Home but to give astronomers all over the place access to data collected by instruments in places where they aren't. We've got thousands of instruments gulping down data but most of it doesn't ever get processed, just stored for later. Like the article says, anybody can have access to massive amounts of raw data. A grad student in the UK can download data gathered from telescopes in Hawai'i and write his or her own program to process them looking for the data they want. A group of amateur astronomers could request a bunch of wide field images and scavange through them looking for comets or asteroids.
If we ban cloning does that mean monocellular life will be against the law? Cloning is something readily done in nature, your entire body is constructed of cloned cells. They've all got your DNA and act just like the other versions of themselves.
What cloning won't do:
1. Allow you to make a clone of someone and replace them in society with an exact replica. A clone of me made tomorrow would still take a normal amount of time to grow up and may or may not be anything like me. Genetically we'd be identical but unless he traveled back in time to live my life for me he probably wouldn't end up anything like me.
2. Allow me to create an army of super clone warriors to take over the world. Said soldiers would have to be gestated and raised like a normal army of soldiers.
What cloning embryos WOULD allow:
1. Do gene mapping and stem cell research with a very large subject base with little genetic discrepency. Every wonder why fruit flies and a few simple plants have been used for the past whatever years for biological experimentation? There's little genetic diversity and they're plentiful.
2. Figure out how to regenerate cells by cloning them so you can repair almost any part of the body damaged by just about anything. There's not a whole lot of a chance for rejection when you're your own oragan donor.
Cloning research doesn't require an embryo to be gestated. Then of course there are those holding to the notion that life begins as an embryo and all that jazz. That is just picking at straws because you don't have enough understanding of the process to make a logical argument against it. If you want to save a baby stop jacking off and ovulating but don't harrass somebody trying to make you and your kids have a better life.
I think that's far too conservative of an estimate for fusion powered spacecraft (technologically, politically it could be never when a fusion powered spacecraft is sent to another planet). The most energy efficient AFAIK fusion reactors we're researching are colliding beam fusion systems (Tokamak). These work really well fusing Deuterium-Tritium but can be used with Deuterium-Helium and Hydrogen-Boron fuels as well. The problem with Deuterium-Helium fusion is the near complete lack of He-3 on the Earth's surface. We're much too hot with too little gravity to hold onto Helium for very long in any form and our atmosphere blocks most of the cosmic rays that form it. The Moon however would be a good source of it since there's nothing protecting the regolith from cosmic ray bombardment. With CBF reactors once you get stable fusion with one fuel source most of your work is done in getting to use other fuel sources. I think cutting your timetable in half would be a little better of an estimate.
The catch-22 with a lunar launched mission is the cost of getting equipment from the moon. Even if you merely lauch modular facilities to process regolith to construct a spaceship with that still counts as cost for a Mars mission. Mining ice on the Moon for a chemical rocket is dumb anyhow, Helium-3 is much energetic of a fuel and will get a craft to Mars in a much shorter timeframe. Furthermore you need not use the fucking space shuttle to build a Mars ship, that would be absolutely ludicrous, the SST is one of the most expensive fucking launch systems in use currently. It'd be cheaper to build a magnetic linear accelerator up the side of a mountain and shoot stuff into orbit.
That's so fucking funny. If I only had mod points...
I thought Protector started off really well until it slowly devolved into "super genius can do absolutely anything at all" and saves the day for the human race. Waiting to find out what happens next kept me reading until the end but in retrospect I think I enjoyed Ringworld a bit better. I'm reading Ringworld Throne right now hoping not to be disappointed with the series.
I thought Havoc Pennington's book on GTK+ was pretty good documentation. It's called GTK+/Gnome Application Development and published by New Riders IIRC. I picked it up for 14.95$ and have found is very useful.
I don't even think the schools really need to install RH on the systems because it can be done by whoever MS buys the boxes from (Dell or Compaq most likely) so it can be as easily done as writing an image onto the drive. That also works for maintaining the systems. The school's can be given disks with live images on them which are just dd'ed onto the hard drives. As long as there is a modicum of password security so people don't get to run around as root it'd be hard for anyone to really fuck up an installation anyways.
I was thinking more along the lines of African swallows. Weird. 20 damned seconds to reply. Has it been 20 seconds yet?
Well..for those of us who can read. Redhat is offering to stick Linux on these systems and provide unlimited support for the hardware MS buys. KDE and GNOME look alot like Windows and in many cases act alot like it. As long as the installation is kept pretty compact it shouldn't confuse anybody. You click the button at the bottom to get to your applications. Nautilus is a pretty workable GUI file manager desktop icons pretty easily get people to the media devices they want to use. It isn't like the schools have to admin the boxes themselves anyways. They call up RH if there is ever a problem and if the boxes have even modest internet connectivity an RH trouble shooter can run stuff remotely to fix the system.
Actually Harrison Ford and co DO get paid for their likenesses to be used on toys and whatnot. Part of the contract they sign says that they get a mad fat check every time some shit comes out with their likeness on it. Same goes for most actors whom appear as toys. Thats why you don't see huge blockbuster movies on the big three networks very often. The network has to pay the studio who has to pay the actors and whomever else was contracted.
I was just wondering the other day how feasible it would be to use a laser as a communication device between two birds. You can get really nice range with little EM interference with only a wee bit or output power. The one obstacle I kept running into whilst pondering a laserlink was keeping the beam aimed at another bird in a different orbit. Well hot damn and way to go. I guess I was hit in the face with the same muse as the dudes at the ESA just a little bit late and without any satellites under my control to play with...so far.
Anand wrote a very good article with plenty of meat and potatos to satisfy but I think he is sort of missing a very important point. He mentions in the article how the PS2 falls behind the XBox in some way because it has less fancy graphics and sound capabilities. SONY HAS SOLD NEARLY 20 MILLION OF THE FUCKING THINGS AND HAS HUNDREDS OF GAMES AVAILABLE FOR IT. I think Anand needs to reassess his position on the PS2. It is still selling for 300$ because people are still eating it up at this price. The XBox may be able to do a bajillion polygons per second but it still doesn't have the game franchises that make bank on console systems. Nintendo's also in a good position because they are destined to get the little kids who want to play Pokemon until their eyes fall out. Who cares if Pikachu's only rendered with half a million fucking surfaces, people want to play the games not write a master's thesis on the theoretical graphic capabilities of a computer system. I bought a GC on my way home from Louisiana and since I've been home my brother's been playing Rogue Squadron almst non-stop. He hasn't yet complained about the lack of theoretical polygons the GC can render yet and I sort of doubt he will. The XBox will only truely contend with Nintendo and Sony when it has games in high demand. I thought it's launch titles were pretty crappy compared to the GC's though a little better than what the PS2 originally offered. Besides that I got a GC and two games for the price of either the PS2 or XBox.
Isn't strange how fucking stupid the slashdot crowd is sometimes, especially something involving Microsoft. One genius posed the question "Do you think Gates and Balmer can envision great games?" Do you REALLY FUCKING THINK BILL GATES DESIGNS GAMES FOR THE XBOX? Other quite intelligent people are telling everyone they ought to buy XBoxes because Microsoft HAS to be losing money on each one you buy. The factories spitting out the XBox make just about every component in the console with most components being contained on a single board. It is very cheap to manufacture SBCs. The 300$ price tag doesn't leave much room for profit but they aren't losing anym oney on the damn boxes. Stop using pricewatch.com to find the prices of computer hardware and then trying to extrapolate the price of a console system. Others claim the XBox is little more than commodity PC hardware in a little black box. The memory subsystem isn't exactly an i815 chipset or something. Buy one if you want or don't buy one if you don't want to play the games. I'm waiting for Sunday to pick up a Gamecube. I can get a console and a couple games for the price of a PS2 or XBox console. Plus I get something for my GBA to talk to.
One option you might look into rather than USB cameras would be to use video capture boards and bullet cameras as your capture device. Most small cameras will get you from 360 to 380 lines of resolution in colour and usually 420 lines in B/W. They're fairly cheap and really rugged. In fact almost any CCTV camera ought to work fine as most output a composite video signal. Most run on a 12v DC power supply so you can even run them off batteries if you really wanted or needed to. You can bet small CCTV cameras will give you a bit better of a video signal than USB cameras and most capture boards allow for uncompressed video that you can pipe into whatever streaming codec you are hip to use. A board and camera combo might cost you from 150 to 200$. Here's a link to compatible capture boards.
The choice seems clear if you lack the ability to process information in a logical manner. USB is a host-based bus, it needs a host controller in order for any of the devices to talk to one another. All data has to pass through a central controller (on your PC) in order for any of the devices to even see one another. FireWire on the otherhand is a host independent system as each FW device has a FW controller as one of its logic conponents. This essentially makes all FW devices their own hosts. They can connect directly to one another and intercommunicate or in a daisy chain configuration each device can talk to any other device without the intervention of some contral controller. A FW camera and HD both plugged into a computer you can tell the camera to send video to the HD and it will with no further intervention of the computer. A USB configuration like that would require all data to pass through the computer's USB controller and then into the hard drive. There's FW drives that plug directly into DV cameras and can offload video without a computer even involved. The next FW spec increases the throughput to a couple gigabits per second, in some cases current FW components with a driver update would be compatible with the new FW spec.
Woohoo InfiniBand. Finally small companies can build clusters that can compete with multi-million dollar SGI super systems. Beowulf clusters using Ethernet interconnects are nothing compared to a system with memory bridge level hardware interconnects. Current Linux clusters are often based around Ethernet which is very slow and has a high overhead when you want to put it to the use of letting processes on different machines talk to one another. Big Beowulf clusters have been shown to work well on embarrassingly parallel computational problems (key cracking, some types of CFD, and packet analisys like SETI@Home) but aren't really up to snuff with truely high end super computers without much faster interconnects. Some clusters (I think Microway and a few others) are built using Myrinet interconnects which are faster and actually provide the sort of connections big clusters need. Even using Myrinet you end up with alot of scalability problems because you only get your max throughput when transfering very large blocks of data to the same host. Having a Northbridge based interconnect (rather than something like a Myrinet card which is connecting to the PCI bus which in turn hooks to the Southbridge) is a very good way to speed up clustered systems. IBM, Sun, and SGI learned this a long time ago which is why they use high speed interconnects between individual boxes rather than trying to overextend network protocols.
The other guy caught Bluetooth vs. 802.11b so I'll pick up USB vs. FireWire. What the fuck are you smoking? USB is a host based connection protocol and FireWire is a host independant protocol. That makes them not really even in the same ballpark. Just because they are both serial protocols don't mean jack shit. USB brokers all connections through a host controller, this host controller also acts as the switch for all communication on the bus. Everything has to pass through the host controller for any devices to talk to one another. FireWire is host independant which basically means every device on the bus is a peer of every other device (all devices are hosts). A FireWire DV camera can talk directly to a FW hard drive or DVD-RAM with no computer involved whatsoever. A USB webcam isn't going to be writing video to a USB hard drive any time soon because neither of them has a host controller. FireWire devices can also talk to each other directly rather than through a central host. Just because USB 2 is a high bandwidth version of USB doesn't mean it truely competes with FW. USB and FW are going to both be included in systems for a while since they both serve different purposes. You're not going to be seeing USB2 DV cameras for sale while FW cries in its corner.
Latency is a prime concern of the I2 network architects. One of the cooler aspects of the whole I2 thing in hardware is the use of packet switching schemes that resemble circuit switching schemes. Instead of a packet just flying around the network trying to get to the other end of it by whatever path a router deems viable, I2 routers route all of the data from two sources along the same path. Packetswitching was designed to be redundant for lossy networks, not fast for high availability networks. Routes can easily be recalculated but having data all go the same way lowers latency dramatically. I remember reading a bit ago about a realtime video conference between a professor and his class in the US and a professor and his class in Japan. Uncompressed realtime video doesn't work real well on the more traditional packet switched networks of the current internet. An interesting aspect of the circuit switching like packet switching is that specific pathways between a destination and origin can be planned out ahead of time and even sold. Instead of merely selling X megabits (or maybe gigebits) of bandwidth a company could sell a direct path or high priority path between two network nodes. Junior sending a video of his skateboard competition to grandma wouldn't interfere with the realtime traffic of a big Fortune50 paying beaucoup dollars for their network while Junior and Grandma are paying 50$ for their cable modems.
The 802.11 protocol isn't robust enough to handle a huge network cloud like you're thinking of. It uses many techniques of the wired 802.x schemes to detect nodes on the network, these work fine for wired networks where it's as simple as detecting a completed circuit but for wireless networks alot of overhead is added. Besides the problems in 802.11 for large numbers of nodes, you'd also need to come up with an efficient dynamic routing scheme. Packets can't just propogate over the entire network with no points that sort of direct their flow. If the wired internet worked like that data would never reach its destination. Wireless clouds are inefficient and messy and not very scalable at all. The bandwidth of 802.11 can't ever exceed B log2 (1 + SNR). Wired networks have the same restriction but can just add wires instead of refining their encoding schemes.
A reason which is blantantly obvious to those of us who use non-x86 processors in our computers is that software originally written for x86 processor takes a bit of work to port to other processors and then the ports aren't always as functional as the original piece of software. RLX boxes run a slightly customized Linux installation but for the most part are using off the shelf GNU software. Were a company to come out with boxes using they'd have to invest alot of money into making sure their software was up to snuff on their hardware. Transmeta and those using Transmeta chips rounded the corners by making their product compatible with a shitload of software already existing.
Media producer and gamer are broad catagories. A couple years ago only a speed freak gamer would think of owning a 500 MHz machine, now my grandparents own a 500 MHz Gateway and a 700 Mhz HP and all they really do is type up greeting cards and use the internet. It is alot easier to do any sort of graphics work on a 700 MHz Celeron then it is trying to do it on a 66 MHz 486. On that point now all "media producers" are professionals. Lots of people are make their own audio CDs, touching up old family photo albums, and getting creative with home movies on their computers. More and faster are keys to these sorts of applications. I used to do video editing work on old PowerMac 8500AVs and it took forever to do even the simplest things with small video clips, you wouldn't think of doing a major project without access to a machine with a several thousand dollar hardware video card. With a 799$ iMac people can do in an afternoon what it would take me a week to do.
For the bitchers, please remember than the GNU-Darwin project no matter how inaptly named is not supported by Apple and thus anything mentioning the GNU-Darwin project may fuck up an Apple supplied OSX installation. Besides that I think this project is pretty nifty, it's one of the reasons I was pretty excited about getting OSX on my Powerbook. Not only would I get the cool features of OSX I also get to fire up Terminal.app and use a huge number of FreeBSD ports. The article from the other day where both OS 7.6 and WindowsXP were running on an iBook 466 I found pretty interesting. There was little content but the demonstration was pretty cool.
This is not an upgrade for an Apple supplied OSX installation, this is an installation for a Darwin (not supplied or supported by Apple) installation. If you're using OSX and running this script you'll probably fuck something up seriously because OSX is a bit different from the Darwin distro. You can run the GNU-Darwin Ports on OSX but it would be pretty silly for me to run this script on my 10.1 installation on my Powerbook.