Why don't we use multicast? Simple. There are probably a total of about 12 home users on the MBone, and they probably aren't even aware they have access. Remember that multicast, according to spec, has its own address range (Class D) that "normal" internet routers will just drop as bad. This means that in order for multicasting to work as designed, every router, from source to where the multicast is "split" to clients, must grok multicasting. The MBone just doesn't have much penetration, and people aren't asking for it, although it's absolutely perfect for a broadcast of live events.
Okay, the GIMP is truly great and a credit to the Free Software community. However, the industry standard remains Photoshop. People have a lot invested into it: all their previous work is in PSD format, and they probably have all sorts of custom filters and such. Also, if you want to collaborate with others, it's much more likely that they will have and know how to use Photoshop rather than the GIMP. While it's easy to get the GIMP, it can be confusing at first (just like Photoshop) and people don't want to have to relearn everything just to work with you. In short, the GIMP is a wonderful piece of software and is gaining on Photoshop rapidly, but it just doesn't have the same position as an industry standard.
802.11x and Bluetooth target totally different markets. 802.11x is designed to link computers together; Bluetooth is designed to link peripherals to computers. This means that Bluetooth is much simpler to work with (so embedded microcontrollers can understand it) and very low-power (so it can run for a decent length of time from cheap batteries). 802.11x, on the other hand, takes some computing power on each end and is designed with the assumption that there's AC power or a long-life battery available. Bluetooth is simpler; that's the point. You'll never see an 802.11x mouse for sale, but you can buy several Bluetooth mice right now.
Take a look at the Project for a New American Century, a neocon organisation that is dedicated to bringing American "leadership" to the world and making sure that big buisiness can't be stopped by some damn upstart who happens to be better than them.
Their leadership has in the past stated that Saudi Arabia is far too volatile to be a major oil source for America and we should look for or "install" alternative leadership in the area (cite: the policy paper Rebuilding America's Defenses [PDF]). Who fits that bill? Hmm, a despot who we can tie via vauge rumors to Sept. 11, who is hated by a good deal of the world and is militarialy fairly weak? Saddam, your number's up!
Of course, Sept. 11 itself was something of a wet dream for these boys: check out page 51 (page 63 of the PDF), which states that the transformation of America from its current state into a superpower willing to smash anyone in the way of her mighty Corporations will be gradual, "absent some catastrophic and catalyzing event - like a new Pearl Harbor." A reason for massive military spending? Check! A way to tie massive military buildups to patriotism? Check! A way to declare liberals dirty traitorous bastards! Check and mate!
No, fly-by-wire is simply the replacement of direct physical linkages (via wires, hydraulics, etc) between cockpit controls and control surfaces by electronic sensors at the cockpit end and actuators (like, say, servos) at the control surface end. While it's trivial to route those signals through a computer for processing, that's not always done. In fact, you seem to think that fly-by-wire implies autopilot; your previous statement that a FBW aircraft should return to straight, level flight if no control input is given sort of assumes that an autopilot is present.
As for aero parts being "kludged on" later, you seem to have no understanding of how parts are attached to the airframe. Assuming that this problem actually needs aerodynamic modifications, (it might be solvable by simply shifting internal weight) those modifications will be attached directly to the frame. They won't just be pop-riveted to the skin or something. In any aircraft, if parts start falling off, conditions are already so bad that you probably are screaming for dear life or are already dead. While modifications to the original design might not be as efficient as a completely new design, they will still be air- and spaceworthy. Scaled Composites has invested millions into developing a working space vehicle; they're not about to throw all that away by slapping poorly thought-out modifications onto Spaceship 1 and calling it good.
I take it you were there and able to see everything that went on; otherwise, of course, you would stop making such uninformed statements about the nature of aviation.
First, the "lack of understanding" about drag and the gear door is perfectly normal. If things operated identically in real life and in computer simulations, there would be no need for physical testing at all. Anomalies come out in tests and are fixed. It might well be that the door design is not at fault; perhaps the finish is rough (something that can cause trouble at very high speed). Perhaps it didn't seal properly. Whatever the cause, it will be found out and fixed. That's what the tests are there for.
As for your second point, you have absolutely no grounding in reality when you say that any modern fly-by-wire aircraft should revert to level flight when the pilot is giving zero input and some aero parts are broken. First of all, Spaceship One is not fly-by-wire, it's manually controlled for a good part of its flight. Second, with aero parts missing, no aircraft will be able to fly just as well as it can under normal conditions. While it might be possible to fly a modern aircraft under extreme conditions (like missing control surfaces) it would be extremely difficult and not simply handled by the computer.
Burt Rutan is famous for building safe and reliable aircraft; I doubt that he will take shortcuts when it comes to safety.
While this is almost certainly smaller than a tiny rare-earth magnet, it still won't cause compass needles all over the Earth to point in funny directions; just those around the building it's installed in.
Sorry, I meant to say that the big magnet is bigger than a little rare-earth like you might find in your hard drive.
No, this is the largest magnetic field ever created (that we know about), although it does cover a fairly small area. The Earth's magnetic field is weak but very, very large; magnetic fields drop off rapidly (Inverse cube of distance, if I recall correctly) as you move away from the source. Similarly, speaker magnets are fairly weak (on the order of a few hundred to a thousand gauss) but since they are large, they affect a greater area than a small rare-earth magnet that has a 1 Tesla (10,000 Gauss) field strength - a few meters away, the magnetic field is lost in the "background noise" of the Earth's magnetic field. While this is almost certainly smaller than a tiny rare-earth magnet, it still won't cause compass needles all over the Earth to point in funny directions; just those around the building it's installed in. In summary: powerful magnet != big field.
VNC has surprisingly good performance, especially compared to X, but still lags behind MS/Citrix Terminal Services in many cases due to fundamental design differences. VNC's design uses a framebuffer streamed over the network to the client, while for the most part MS Terminal Server can just tell the client to render basic Win32 objects at specific locations.
Needless to say, this means that MS Terminal Server wins in the performance department as long as you are using standard Win32 apps. However, it is also much less flexible - if you want to run an app that doesn't use standard controls (like, say, Winamp) you are back on VNC's territory, and VNC is designed around streaming bitmaps, which gives it a slight advantage.
Of course, the MS approach doesn't work too well on Linux, because of the large variety in toolkits; it's easier to simply stream bitmaps than to create a tool that recognizes and works well with 10 or more different toolkits and even then leaves many apps out in the cold.
For your application, however, I would recommend VNC: since you are simply streaming a bitmap (I doubt MS Terminal Server groks X), VNC has a number of different compression options and tradeoffs that can help improve performance a great deal over what X offers. While it may take some tweaking to get really good performance, you will almost certainly get some gain out of going to VNC in your application.
When asked about SCO, Linus likens it to a paternity case and has this to say:
Even though SCO has refused to undergo the technical equivalent of DNA testing, and
even though my (and other people's) DNA is probably all over Linux [emphasis mine].
However, video signals are analog and high-bandwidth. To run one over IP, you would need to have an ADC at the source end, plus compression hardware if you ever want to run more than TV resolutions in realtime. The point of this device is to put a computer in one room and a console in another, not to allow use of a computer from around the world.
If you need access to your US-based computer from China, you'd be much better off using VNC, X, MS Terminal Services, or another remote-console app, combined with a PC Weasel and serial terminal server if you need console access from boot.
The point of this system is to provide a console a few hundred feet away that is functionally equivalent to a console directly connected to the computer. It seems to perform that function adequately. It is not designed to provide halfway-around-the-world access; the products I mentioned are far better for that. Furthermore, it would be a bad idea to include IP functionality in this product: it would mean a significant increase in cost, while simply duplicating features that are mostly available through Free software running on the host machine. Sure, IP encapsulation on a device like this would be great. However, it is simply impossible from an engineering standpoint to make it work via IP without shooting the price into the stratosphere or requiring an insanely high-bandwidth link.
Why encumber a product with IP when it doesn't need it? This is designed for end-to-end transmissions over a single wire; IP is ridiculous overkill incorporating useless overhead for things like routability and the ability to communicate with multiple devices.
In fact, the designers of this product seem to have even dispensed with Ethernet (note that it only says it'll work with Cat5 and contains its own gain control system - something that to me tells me that it's using some sort of broadband signaling, rather than the baseband shared by almost all forms of Ethernet.
As for "physical network requirements," everything that works over a network can't avoid using some form of physical communication; all the fancy protocols in the world won't do you any good if you can't get two systems to exchange bits. While it's possible to avoid permenantly tying yourself to one particular medium (via modular interfaces), it's impossible to have a network without physical communication. You just can't have IP between two locations without something to communicate packets - there has to be some form of communication. There's a reason why the physical layer is at the bottom of the OSI and TCP/IP models. It's because if the physical layer fails, there is no network.
A long time. While 802.11b and the GBA wireless adaptor work on a similar frequency, that doesn't mean that they're using the same underlying protocols. In fact, this is almost certainly closer to Bluetooth than anything, although I haven't seen any tech details other than those provided in the article.
While GBA emulators can certainly communicate with each other (via any sort of TCP/IP network, including one over 802.11x), it's going to take some time and some brilliant hacking and RF geekery to get GBAs and PCs talking wirelessly (without special hardware from Nintendo intended for developers).
In the mean time, it would almost certainly be cheaper and easier to interface to the GBA's wired link port - hacking up a USB interface, while not exactly easy, would be child's play compared to getting a 2.4 GHz wireless link with a proprietary protocol working between a GBA and a PC.
How about 64-bit Apache, My/postgre SQL and all the other opensource server goodies? Not to mention a cheap way to accelerate science and math work for universities using in-house software. Oh, and then there's grep, which you even mentioned; its performance could probably be improved significantly by recompilation to take advantage of the new registers in x86-64, if nothing else.
Cheap 64-bit server/workstation hardware has been way too long in coming; it looks like AMD is finally going to introduce some low-end competition into the market and force Intel to either slash Itanic prices drastically, put out its own lower-end 64-bit server/workstation chip, or simply cede a major server market to AMD (yeah, right).
An older CRPG called Dungeon Master used a system somewhat similar to the one you described; you had to experiment and combine different "runes" for spells, or find notes describing how to cast each spell in the dungeon. Dungeon Master's system for spellcasting was rather cumbersome, due to the fact you had to input each rune individually whenever you cast a spell, but interface problems could easily be sidestepped by setting up a "frequently used" spell list. Also, while the system was static and thus vulnerable to people posting spell lists online, it certainly wouldn't be too hard to implement a random generator for spells at the beginning of the game.
This might even be an interesting way to discriminate between magic users who use talent as opposed to study (like sorcerers and wizards in D&D 3E) in MMORPGs; sorcerers may have more power, but can't share spells, while wizards might have less casting power but can share spells with one another.
If you're interested in a modern game that uses a similar system, check out Morrowind. The whole game is incredibly open-ended. While it doesn't have any experimentation requirements for spells, you are allowed to create your own and buy them in standard fashion from wizards, as well as use lots of predefined spells. Also, its alchemy system comes a lot closer to your ideal: you can actually pick flowers and berries from plants (which have some small effects), then use a mortar and pestle to mix potions and other tools to modify the effects. The lists of possible potions on Gamefaqs run for many pages; it's really an incredibly versatile system.
Bargain systems never left integrated video behind; SiS has made good money by cranking out integrated-video chipsets for systems from Pentiums to Athlons. For manufacturers like Emachines who operate on razor-thin margins, even $15 or $20 more for a separate video card means the difference between a decent profit and a loss-leader.
Not only that, the trend is swinging back towards integrated video on midrange systems; memory is finally fast enough that it's possible to get decent video performance using shared system RAM. The Nforce series has quite decent integrated graphics; they are roughly on par with a Geforce2 with slower memory. While I haven't built any systems for customers using ATI's new integrated graphics, I've heard good things about them as well. For the average user, who wants decent 2d performance but doesn't care about getting 890 FPS at 1600x1200 with the detail sliders maxxed out in the latest games, the new integrated-graphics chipsets provide an excellent compromise and offer some 3d ability as well - certainly enough to play older games, and usually the power to play newer games as well, although at lower resolutions and detail levels than could be managed by a $450 graphics board.
Although I'm not the original poster, I can tell you the problem with a dialup service: it costs money. When companies are cutting their IT departments' budgets to the bone, a $20/mo connection to a dialup ISP isn't a good idea. Also, if you need to report that someone has deployed a fiber-seeking backhoe and taken a major line down, there's a good chance your backup ISP is offline too. TAP, on the other hand, works as long as you can get an analog dialtone, which will often work even if higher-bandwidth connections are down.
As for solutions to the problem, I'm going to have to cast another vote for a voice modem and text-to-speech system; this allows you to configure error messages as much as you like, and costs you money only for setup and in an actual emergency. Also, you can get basic delivery confirmation based on who picks up and who doesn't. It might not be the best idea for a system that requires multicast over a large group or contact over a noisy link (voice synths can be hard to understand).
The flexibility, though, is the real strong point: with a bit of perl, you could probably get the system to repeat back error data. Even better, with a bit more work you could probably set up basic commands via touch tones (press one for hard reboot, press two for soft reboot, press three to restart apache, etc.) and perhaps not have to come in at all.
So someone would go through every item in your office, trying to find possible alphanumeric strings that might be a password, and type it in? Using a password like "CD" or "book" is a very bad idea, but using the password "0441328008-sand" (the ISBN of my copy of Heretics of Dune, which I just picked at random out of my 1000+ books, plus a random word relating to the book), isn't something that's easily guessable.
Furthermore, until it gets firmly implanted in my tactile memory, I just have to remember "Heretics of Dune" rather than a long ugly string of numbers. Things aren't nearly as easy for an attacker, though. Any attacker looking to get my password would have to first know that it is a book they're looking for, then go through every single book I own, typing in likely numbers (not only the ISBN, but also the barcode, and any other likely numbers; for example, I might work the price in there somehow).
Also, an attacker would have to have physical access to my home for a good long time to even know what books, CDs and other things I own. The set of all possible passwords, although restricted compared to a truly random string, is still incredibly massive and would take a long time to crack with a dictionary attack. Assuming I change the password every 2 to 3 months, the attacker would be better off looking for exploits to bypass the password mechanism entirely.
Start off by asking them about computer/console games. Even among kindergarteners, a significant number will probably be familiar with Playstation or computer edutainment games. If possible, show a video clip (keep it non-violent, think Mario rather than Soldier of Fortune 2 or UT2003).
From there, go into something like, "Well, how does the computer know to make Mario jump when I press this button? How does it know when the bad guy gets jumped on and is squished? It follows instructions, and it's my job to give computers those instructions." Explain that you have to figure out what people want to do, then tell the computer how to follow instructions other people give it. Keep it very simple, and make it as visual and interactive as possible. Remember that the kids have an attention span of maybe five minutes, even with all the pretty visual aids you can conjure, so keep it short and sweet.
Finally, a web page for the younger kids probably isn't that great an idea. Most probably aren't familiar with the Internet, and a web page is much more static than a game. Add to that the fact that most probably can't read very well, and you have a recipe for failure. Remember to keep things as visually oriented as possible.
Actually, consoles do wear out. When you consider that there are quite a few moving parts (DVD drive spindle & eject motors are the big ones, but fans can fail as well), and the system may well be placed where it can't get enough cooling, there's plenty to wear out (not to mention the fact that the electronics can degrade on their own).
If you factor in Sony's infamous customer support, it becomes cheaper and easier to just buy a new Playstation 2 rather than send it in, wait 3 months, and find out it's been lost in a Cambodian rice field.
While your appliances may not fail, a significant number of toasters and microwaves do so every day; this means a statistically significant number of people will be buying replacement units. The concept is the same for the PS2: while it's fairly reliable, the massive installed base means that many fail every day and must be replaced.
Sorry, but process engineering is a dead field right now. Many companies are cutting "improvement" departments like process engineering and IT to the bone because they don't directly produce end products; middle and upper management often look down on divisions that don't do "real work." While companies doing this are essentially shooting themselves in the foot, most of manufacturing is cutting back or eliminating entirely divisions that serve to improve productivity. Since they're not directly related to product output, they get cut first and the company doesn't suffer immediately.
Also, what makes you think that an out-of-work sysadmin or programmer would be qualified? All the process engineers I know have spent many years working on the shop floor in their industries, and know the processes involved like the back of their hand. If they can't find work, what hope is there for someone who walks in from what is essentially a completely different industry?
Oh yeah, tube amps sound way better than solid-state.
I hope you're being sarcastic; the only area where there's even a difference between the output of tubes and transistors is when they're overdriven. That's a concern in recording, where overdriving the preamp can create interesting effects; guitar players know all about this and many prefer tube amps for the pleasing distortion. It is not a concern for playback, where the goal is to reproduce the original recording exactly rather than to overdrive the amp and create different sounds.
As for tube amps in high-power situations, that's still the norm. The reason tubes fell to discrete transistors was mainly due to the fact that tubes have to be heated to work right. While several tube heaters in a small radio mean serious inefficiency, a 200W tube heater coil in a 200 KW radio transmitter means that all of 0.1% of your broadcast power is used for the tube heater - no big deal. Add to that the fact that large transistors are very expensive and the difficulty of moving heat away from the junctions in something that large, and tubes are still the natural choice for really high-power applications.
The problem isn't that movies use bad physics as plot devices, the problem is that movies use bad physics for no reason at all. Instead of running from an explosion (seen in any action movie), why not have Arnie dive for cover, only to see a jet of flame shoot past half a second later and annihilate the bad guys? Gunfight scenes would arguably be more tense without "Hollywood magazines" in the guns; now the hero and the villian have to worry about limited ammunition and getting a good shot without being shot themselves. Instead, directors choose to magically expand Uzi and AK-47 magazines to hold about 5000 bullets each.
While I enjoy good sci-fi movies that suspend physics, they should have a reason to do so. Why can Neo, Morpheus and Trinity do impossible things? Okay, they're able to warp the Matrix with their minds. I can accept that, at least during the movie, because it doesn't destroy my suspension of disbelief. Diving from the explosion, being knocked down by the hot gases, and getting up without anything more than a few scrapes? No, because that cannot happen in the real world. Suspension of disbelief is like duct tape: it works to fix a lot of things, but you can only do so much with it before things just don't work anymore.
If it's social satire, where's the element of truth? Good satire needs to have a kernel of truth "stretched" to the extreme; think of Charlie Chaplin in Modern Times, being force-fed by a machine while he works on an assembly line. While factory owners never actually attempted to make their workers work and eat similtaneously, Chaplin still made a point: that factory owners were willing to sacrifice workers' basic decency for higher rates of production and more profit.
I don't see a corresponding "basic point" here. Is the author trying to tell us that games can take over our lives? Maybe, but that doesn't seem to be a major social problem in today's society; for every one gamer who allows games to take over his or her life, there are many who treat gaming as a fun and relaxing hobby. While some people do treat this as a problem, the identification of this writer as the P-I's videogames columnist seems to rule that possiblity out. Still, the article seems to be written by a person who is somewhat afraid of computers, but wants to write something on the "gaming culture." There's no point that it tries to make; the only humor is weak and solely derived from hyperbole. Social satire should be funny and have a point; this posesses neither of those characteristics.
As for the analogy with Bonsai Kitten, there's very little similarity. BK is a troll, designed to shock people into an emotional response. I see no evidence of this being similar; if it were going for shock value, it would probably describe gamers as somewhat similar to this. If this is a troll, it's a fairly weak one: it doesn't appeal to any emotions, but rather simply makes you want to point out the author's errors and move on.
While one does get the sense that the author is trying to be funny and perhaps a bit controversial, there's no content that makes me laugh or think. Instead, I get the same feeling you might get from a comedian whose jokes aren't getting a response. I think, "what a moron!" and move on; not the response that either a humorist or troll would look for. It doesn't matter whether you rate the article on humor value, shock value, or informative content: it fails on all three counts.
Slashdot Mirror
Why don't we use multicast? Simple. There are probably a total of about 12 home users on the MBone, and they probably aren't even aware they have access. Remember that multicast, according to spec, has its own address range (Class D) that "normal" internet routers will just drop as bad. This means that in order for multicasting to work as designed, every router, from source to where the multicast is "split" to clients, must grok multicasting. The MBone just doesn't have much penetration, and people aren't asking for it, although it's absolutely perfect for a broadcast of live events.
(taps sword to shoulders)
Rise, Sir GIMP Pimp!
Okay, the GIMP is truly great and a credit to the Free Software community. However, the industry standard remains Photoshop. People have a lot invested into it: all their previous work is in PSD format, and they probably have all sorts of custom filters and such. Also, if you want to collaborate with others, it's much more likely that they will have and know how to use Photoshop rather than the GIMP. While it's easy to get the GIMP, it can be confusing at first (just like Photoshop) and people don't want to have to relearn everything just to work with you. In short, the GIMP is a wonderful piece of software and is gaining on Photoshop rapidly, but it just doesn't have the same position as an industry standard.
Please have 1.38 x 10^7 blank, formatted floppy disks ready.
Insert Disk 1 of 13,800,000 and press [ENTER] to continue...
802.11x and Bluetooth target totally different markets. 802.11x is designed to link computers together; Bluetooth is designed to link peripherals to computers. This means that Bluetooth is much simpler to work with (so embedded microcontrollers can understand it) and very low-power (so it can run for a decent length of time from cheap batteries). 802.11x, on the other hand, takes some computing power on each end and is designed with the assumption that there's AC power or a long-life battery available. Bluetooth is simpler; that's the point. You'll never see an 802.11x mouse for sale, but you can buy several Bluetooth mice right now.
Their leadership has in the past stated that Saudi Arabia is far too volatile to be a major oil source for America and we should look for or "install" alternative leadership in the area (cite: the policy paper Rebuilding America's Defenses [PDF]). Who fits that bill? Hmm, a despot who we can tie via vauge rumors to Sept. 11, who is hated by a good deal of the world and is militarialy fairly weak? Saddam, your number's up!
Of course, Sept. 11 itself was something of a wet dream for these boys: check out page 51 (page 63 of the PDF), which states that the transformation of America from its current state into a superpower willing to smash anyone in the way of her mighty Corporations will be gradual, "absent some catastrophic and catalyzing event - like a new Pearl Harbor." A reason for massive military spending? Check! A way to tie massive military buildups to patriotism? Check! A way to declare liberals dirty traitorous bastards! Check and mate!
As for aero parts being "kludged on" later, you seem to have no understanding of how parts are attached to the airframe. Assuming that this problem actually needs aerodynamic modifications, (it might be solvable by simply shifting internal weight) those modifications will be attached directly to the frame. They won't just be pop-riveted to the skin or something. In any aircraft, if parts start falling off, conditions are already so bad that you probably are screaming for dear life or are already dead. While modifications to the original design might not be as efficient as a completely new design, they will still be air- and spaceworthy. Scaled Composites has invested millions into developing a working space vehicle; they're not about to throw all that away by slapping poorly thought-out modifications onto Spaceship 1 and calling it good.
First, the "lack of understanding" about drag and the gear door is perfectly normal. If things operated identically in real life and in computer simulations, there would be no need for physical testing at all. Anomalies come out in tests and are fixed. It might well be that the door design is not at fault; perhaps the finish is rough (something that can cause trouble at very high speed). Perhaps it didn't seal properly. Whatever the cause, it will be found out and fixed. That's what the tests are there for.
As for your second point, you have absolutely no grounding in reality when you say that any modern fly-by-wire aircraft should revert to level flight when the pilot is giving zero input and some aero parts are broken. First of all, Spaceship One is not fly-by-wire, it's manually controlled for a good part of its flight. Second, with aero parts missing, no aircraft will be able to fly just as well as it can under normal conditions. While it might be possible to fly a modern aircraft under extreme conditions (like missing control surfaces) it would be extremely difficult and not simply handled by the computer.
Burt Rutan is famous for building safe and reliable aircraft; I doubt that he will take shortcuts when it comes to safety.
No, this is the largest magnetic field ever created (that we know about), although it does cover a fairly small area. The Earth's magnetic field is weak but very, very large; magnetic fields drop off rapidly (Inverse cube of distance, if I recall correctly) as you move away from the source. Similarly, speaker magnets are fairly weak (on the order of a few hundred to a thousand gauss) but since they are large, they affect a greater area than a small rare-earth magnet that has a 1 Tesla (10,000 Gauss) field strength - a few meters away, the magnetic field is lost in the "background noise" of the Earth's magnetic field. While this is almost certainly smaller than a tiny rare-earth magnet, it still won't cause compass needles all over the Earth to point in funny directions; just those around the building it's installed in. In summary: powerful magnet != big field.
Needless to say, this means that MS Terminal Server wins in the performance department as long as you are using standard Win32 apps. However, it is also much less flexible - if you want to run an app that doesn't use standard controls (like, say, Winamp) you are back on VNC's territory, and VNC is designed around streaming bitmaps, which gives it a slight advantage.
Of course, the MS approach doesn't work too well on Linux, because of the large variety in toolkits; it's easier to simply stream bitmaps than to create a tool that recognizes and works well with 10 or more different toolkits and even then leaves many apps out in the cold.
For your application, however, I would recommend VNC: since you are simply streaming a bitmap (I doubt MS Terminal Server groks X), VNC has a number of different compression options and tradeoffs that can help improve performance a great deal over what X offers. While it may take some tweaking to get really good performance, you will almost certainly get some gain out of going to VNC in your application.
If you need access to your US-based computer from China, you'd be much better off using VNC, X, MS Terminal Services, or another remote-console app, combined with a PC Weasel and serial terminal server if you need console access from boot.
The point of this system is to provide a console a few hundred feet away that is functionally equivalent to a console directly connected to the computer. It seems to perform that function adequately. It is not designed to provide halfway-around-the-world access; the products I mentioned are far better for that. Furthermore, it would be a bad idea to include IP functionality in this product: it would mean a significant increase in cost, while simply duplicating features that are mostly available through Free software running on the host machine. Sure, IP encapsulation on a device like this would be great. However, it is simply impossible from an engineering standpoint to make it work via IP without shooting the price into the stratosphere or requiring an insanely high-bandwidth link.
In fact, the designers of this product seem to have even dispensed with Ethernet (note that it only says it'll work with Cat5 and contains its own gain control system - something that to me tells me that it's using some sort of broadband signaling, rather than the baseband shared by almost all forms of Ethernet.
As for "physical network requirements," everything that works over a network can't avoid using some form of physical communication; all the fancy protocols in the world won't do you any good if you can't get two systems to exchange bits. While it's possible to avoid permenantly tying yourself to one particular medium (via modular interfaces), it's impossible to have a network without physical communication. You just can't have IP between two locations without something to communicate packets - there has to be some form of communication. There's a reason why the physical layer is at the bottom of the OSI and TCP/IP models. It's because if the physical layer fails, there is no network.
While GBA emulators can certainly communicate with each other (via any sort of TCP/IP network, including one over 802.11x), it's going to take some time and some brilliant hacking and RF geekery to get GBAs and PCs talking wirelessly (without special hardware from Nintendo intended for developers).
In the mean time, it would almost certainly be cheaper and easier to interface to the GBA's wired link port - hacking up a USB interface, while not exactly easy, would be child's play compared to getting a 2.4 GHz wireless link with a proprietary protocol working between a GBA and a PC.
Cheap 64-bit server/workstation hardware has been way too long in coming; it looks like AMD is finally going to introduce some low-end competition into the market and force Intel to either slash Itanic prices drastically, put out its own lower-end 64-bit server/workstation chip, or simply cede a major server market to AMD (yeah, right).
This might even be an interesting way to discriminate between magic users who use talent as opposed to study (like sorcerers and wizards in D&D 3E) in MMORPGs; sorcerers may have more power, but can't share spells, while wizards might have less casting power but can share spells with one another.
If you're interested in a modern game that uses a similar system, check out Morrowind. The whole game is incredibly open-ended. While it doesn't have any experimentation requirements for spells, you are allowed to create your own and buy them in standard fashion from wizards, as well as use lots of predefined spells. Also, its alchemy system comes a lot closer to your ideal: you can actually pick flowers and berries from plants (which have some small effects), then use a mortar and pestle to mix potions and other tools to modify the effects. The lists of possible potions on Gamefaqs run for many pages; it's really an incredibly versatile system.
Not only that, the trend is swinging back towards integrated video on midrange systems; memory is finally fast enough that it's possible to get decent video performance using shared system RAM. The Nforce series has quite decent integrated graphics; they are roughly on par with a Geforce2 with slower memory. While I haven't built any systems for customers using ATI's new integrated graphics, I've heard good things about them as well. For the average user, who wants decent 2d performance but doesn't care about getting 890 FPS at 1600x1200 with the detail sliders maxxed out in the latest games, the new integrated-graphics chipsets provide an excellent compromise and offer some 3d ability as well - certainly enough to play older games, and usually the power to play newer games as well, although at lower resolutions and detail levels than could be managed by a $450 graphics board.
As for solutions to the problem, I'm going to have to cast another vote for a voice modem and text-to-speech system; this allows you to configure error messages as much as you like, and costs you money only for setup and in an actual emergency. Also, you can get basic delivery confirmation based on who picks up and who doesn't. It might not be the best idea for a system that requires multicast over a large group or contact over a noisy link (voice synths can be hard to understand).
The flexibility, though, is the real strong point: with a bit of perl, you could probably get the system to repeat back error data. Even better, with a bit more work you could probably set up basic commands via touch tones (press one for hard reboot, press two for soft reboot, press three to restart apache, etc.) and perhaps not have to come in at all.
Furthermore, until it gets firmly implanted in my tactile memory, I just have to remember "Heretics of Dune" rather than a long ugly string of numbers. Things aren't nearly as easy for an attacker, though. Any attacker looking to get my password would have to first know that it is a book they're looking for, then go through every single book I own, typing in likely numbers (not only the ISBN, but also the barcode, and any other likely numbers; for example, I might work the price in there somehow).
Also, an attacker would have to have physical access to my home for a good long time to even know what books, CDs and other things I own. The set of all possible passwords, although restricted compared to a truly random string, is still incredibly massive and would take a long time to crack with a dictionary attack. Assuming I change the password every 2 to 3 months, the attacker would be better off looking for exploits to bypass the password mechanism entirely.
From there, go into something like, "Well, how does the computer know to make Mario jump when I press this button? How does it know when the bad guy gets jumped on and is squished? It follows instructions, and it's my job to give computers those instructions." Explain that you have to figure out what people want to do, then tell the computer how to follow instructions other people give it. Keep it very simple, and make it as visual and interactive as possible. Remember that the kids have an attention span of maybe five minutes, even with all the pretty visual aids you can conjure, so keep it short and sweet.
Finally, a web page for the younger kids probably isn't that great an idea. Most probably aren't familiar with the Internet, and a web page is much more static than a game. Add to that the fact that most probably can't read very well, and you have a recipe for failure. Remember to keep things as visually oriented as possible.
If you factor in Sony's infamous customer support, it becomes cheaper and easier to just buy a new Playstation 2 rather than send it in, wait 3 months, and find out it's been lost in a Cambodian rice field.
While your appliances may not fail, a significant number of toasters and microwaves do so every day; this means a statistically significant number of people will be buying replacement units. The concept is the same for the PS2: while it's fairly reliable, the massive installed base means that many fail every day and must be replaced.
Also, what makes you think that an out-of-work sysadmin or programmer would be qualified? All the process engineers I know have spent many years working on the shop floor in their industries, and know the processes involved like the back of their hand. If they can't find work, what hope is there for someone who walks in from what is essentially a completely different industry?
As for tube amps in high-power situations, that's still the norm. The reason tubes fell to discrete transistors was mainly due to the fact that tubes have to be heated to work right. While several tube heaters in a small radio mean serious inefficiency, a 200W tube heater coil in a 200 KW radio transmitter means that all of 0.1% of your broadcast power is used for the tube heater - no big deal. Add to that the fact that large transistors are very expensive and the difficulty of moving heat away from the junctions in something that large, and tubes are still the natural choice for really high-power applications.
While I enjoy good sci-fi movies that suspend physics, they should have a reason to do so. Why can Neo, Morpheus and Trinity do impossible things? Okay, they're able to warp the Matrix with their minds. I can accept that, at least during the movie, because it doesn't destroy my suspension of disbelief. Diving from the explosion, being knocked down by the hot gases, and getting up without anything more than a few scrapes? No, because that cannot happen in the real world. Suspension of disbelief is like duct tape: it works to fix a lot of things, but you can only do so much with it before things just don't work anymore.
I don't see a corresponding "basic point" here. Is the author trying to tell us that games can take over our lives? Maybe, but that doesn't seem to be a major social problem in today's society; for every one gamer who allows games to take over his or her life, there are many who treat gaming as a fun and relaxing hobby. While some people do treat this as a problem, the identification of this writer as the P-I's videogames columnist seems to rule that possiblity out. Still, the article seems to be written by a person who is somewhat afraid of computers, but wants to write something on the "gaming culture." There's no point that it tries to make; the only humor is weak and solely derived from hyperbole. Social satire should be funny and have a point; this posesses neither of those characteristics.
As for the analogy with Bonsai Kitten, there's very little similarity. BK is a troll, designed to shock people into an emotional response. I see no evidence of this being similar; if it were going for shock value, it would probably describe gamers as somewhat similar to this. If this is a troll, it's a fairly weak one: it doesn't appeal to any emotions, but rather simply makes you want to point out the author's errors and move on.
While one does get the sense that the author is trying to be funny and perhaps a bit controversial, there's no content that makes me laugh or think. Instead, I get the same feeling you might get from a comedian whose jokes aren't getting a response. I think, "what a moron!" and move on; not the response that either a humorist or troll would look for. It doesn't matter whether you rate the article on humor value, shock value, or informative content: it fails on all three counts.