So how exactly do people with pie in the sky Wi-Fi plans intend to overcome 802.11b's inherent scalability problems? How many people can one AP REALLY serve?
From my experience I'd say that answer to that question is not very many. Having more than a couple people on a single AP is a recipe for pain and suffering. As the number of users on an AP increases so does the chances of packet collisions. As collisions increase the viability of the network decreases and you eventually reach a collapsing point where the network is unusable. A corollary to that rule then would be the larger your coverage area the higher a chance of collisions and thus a higher chance of the network collapsing.
You run into a similar problem with 802.3 which is solved by switching the network. With a wireless network you don't have the ability to add a switch in the middle of the network to keep the number of collisions down to a minimum. You're only got a bunch of nodes waiting their turn to talk. Switching channels isn't an option because APs can only serve particular channels.
With a coverage area of four miles then, the number of potential collisions on a channel is pretty high because your entire customer base could be in that four mile coverage area. Sweeping a broadbast between different nodes doesn't do much good on their end where all the static from other connections is an issue. On current networks you've got a small number of users because your coverage area is pretty small so problems aren't evident. You don't have problems on a wired network with only a 5 port hub either.
I think it was widely known the ISS would never come to scientific fruition in the bowels of appropriation committee meetings and planning commissions. The ISS has become and I think was intended to be-politically-a means to grab taxpayer dollars and stick them in the pockets of Congress people and their wealthy constituants.
A pork barrel is a project that puts federal dollars in the hands of Congress people in charge of the projects or appropriation committees for said projects. The best pork barrels are projects you can trick a lot of people into thinking are useful for the greater good so they don't ask any probing questions. An example would be a Represenative from a district in Vermont appropriating money for a project in that district some friends of his run a business in. For making them rich they cut said Congressman in on the fat of the "pork" for buying midgets to do battle or whatever it is rich people do. Sometimes pork barrels can be good for the people at large, a project could bring a bunch of jobs to a job poor district and then those people can eat and the country at large benefits from said federal project.
The ISS is starting to look more and more like this every day. The billions of dollars spent on the thing are going somewhere. It isn't like the solid rocket boosters of the Shuttle are lined with five dollar bills, not literally anyways. Before we had our ever impotent "War on Terror" to provide a means for getting public money into private hands the ISS was a perfect project to pork. It had a tenuous scientific basis, it would do JUST enough hard science for data to trickle in so it didn't look like a waste. As an added bonus the EU, Russia, and Japan could get in on the act and make it look to everyone like it was a giant shiny peace symbol in the sky. It's also a project that certain states *cough*California, Texas, and Floria*cough* would have a major hand in both developing and manufacturing. Billions of dollars means lots of cushy raises for government contractors. A pie in the sky science project that may or may not actually work as intended provides sweet CYA material for hearings later on.
You may or may not ask why was the ISS funded when we coulds have gotten more hard science out of smaller space projects and still bilked money out of them in particular Congressional districts. The answer is publicity. You can't go outside and take fricken pictures of the Mars Rover with a high powered zoom lens. You can take a picture of that megabright collection of aluminum cans flying around the planet. Also unlike probes launched from disposable rockets the ISS is something that needs to be maintained. Ron Popiel doesn't have a MagicStation where you set it and forget it. The ISS is a pork barrel that could have lasted for a decade or more had it been viable to do so. That's more than ten years of government contractors selling a $500 space toilet to NASA for $500,000.
Whatever dreams the ISS was supposed to fill for geeks and engineers don't matter to politicians, only the beaucoup cash that comes from those dreams matters. The ISS/Freedom/Alpha may have started as a cool science mission with attainable and useful goals but once it got into the grubby hands of Congress it turned into one giant government contract after another. As I said, now that we've got a "war" against nobody and maybe even a real war with remote control bombs and lasers on 747s the ISS isn't much needed anymore by the government. Why milk NASA's measly 14 billion when you can milk the DOD's uberbillions?
The ISS's failure is the fault of Congress and the people looking to make megabucks off taxpayer dollars, not Lance Bass. You can still despise him resoundly and wish he we eaten by wild battling midgets or whatever you want done to him but his inability to generate investor interest is not dooming the ISS.
I think some of your points are pretty ridiculous in this context. A game demo is often times free, I was playing the Jedi Knight 2 demo the other night I downloaded while I was in the shower, it cost me the electricity to leave my computer and cable modem turned on. That happens to be a lot less money and a lot less hassle than going to see a specific movie showtime or renting a movie.
Online gaming has few constraints besides your network speed. I don't NEED to play a team game in order to play the game. I can fire up UT2003 or CS and be fragging away whenever I want. I can't do that so easily with a movie.
There's nothing to really compare an extended edition DVD to a game mod. Day of Defeat and CS completely change the gameplay of Half-Life, that is no way compares to some fan edit of the Phantom Menace or a bonus DVD with some deleted scenes.
I'm not a video game fanatic either, I'm much more of a movie fanatic and have spent more time in the last month watching movies than playing video games. I'm not poking holes in your argument to defend gamers everywhere. I'm doing it because your comparisons suck.
The magnetic launcher is sort of doable but not on a scale large enough to send any decent sized payload into orbit unfortunately. NASA's Marshal Space Flight Center was doing some research on linear accelerators but I think a lot of their funding has been cut under the O'Keefe regime. Getting trains and scale models working on accelerators is one thing but an orbital payload is quite another. I think if someone were to invest in research most of their funding would end up doing to R&D rather than actually getting payloads into space. For a satellite launching startup it would likely be much better to get a bird in the sky and get your investment back plus a little profit to reinvest rather than blowing the whole kit and kaboodle on research and then be forced to sell out to someone that is just going to use your patents and research materials to make themselves money.
I'd addend to the flashlight a red filter as to better preserve your night vision. Novelty red Saran Wrap works pretty well to cover most flashlights, all you need in some wrap and tape. I prefer duct tape to get a better seal against dew as you'll get some condensation on the wrap because of cold air and a hot bulb. Sealing the wrap directly onto the flashlight lens I've found works well. If you've got a military surplus angle head flashlight you can probably pick up a red filter for it from the same place you bought it.
Also if you're going out in the middle of noplace and decide not to stay until dawn grab some reflective tape to attach strips to stuff you're taking with you like the Thermor or binocs. The full moon will give just enough light for the strips to be seen in the dark so you don't lose them when you set them down in the grass or something. Don't forget strips on your flashlights too. This seems counter intuitive until you set your flash down in the dark grass when its off and can't find it against until you sit on it.
IIRC the reaction used on these lasers only lasts at most about thirty seconds so you'd only be able to keep the laser on the target that amount of time at most. I wish I could find where I read that so I could pass it on or just check the figure to see if I am indeed remembering correctly.
Most of the chips used in handheld MP3 players are lightweight integer only processors or DSPs that handle the sort of integer ops MP3 and WMA files need to decode. With the more general purpose chips you need to write a decoder that can fit in the device's ROM and then not use signifigantly more processing power than the other decoders. Vorbis files need more math ops performed on them than MP3 files do. MP3 audio was designed to be run on systems whose processors performed fewer MIPS than a potato chip. Ogg Vorbis is from the era of Athlons and Intel space heaters. it simply requires more processing power period.
Squeezing an Ogg decoder into the same space as an MP3 decoder is hard enough but getting it to not drain the batteries in Planck time adds a whole level of complexity on top of it. The more work a chip has to do the more power it consumes. I wouldn't buy an iPod that only played 3 hours worth of music before needing a nuclear generator to recharge it.
For the hardware vendors though it is a question of space. Can an Ogg codec fit into the same ROM space as an MP3 codec and only use the same resources as said MP3 codec? If not they will not use Ogg codecs. Nor will they use Ogg codecs if it halves the battery life of the device, if the Ogg needs so much processing muscle it uses twice the wattage as the MP3 encoder they can't really sell that to people. Who cares if the device holds twice as many songs if the battery life is only half of what it would be otherwise. If playing an Ogg made my iPod only last 5 hours there's no way in hell I'd ever use them better quality or not. I routinely run my iPod for 8-10 hour stretches any period of time less than that is unacceptable for me personally.
Work on Ogg is going to continue and some intepid soul or souls are going to make a super cool Ogg decoder that can run on a paper clip taped to a Dorito but until then MP3 and WMP are going to dominate because they fit on the existing hardware.
Re:Whats the big problem with putting ogg everywhe
on
Ogg Support For iTunes
·
· Score: 3, Informative
If it's so trivial why haven't you done it already? Integer only MP3 decoders are all over the places and MP3 decoding using only integer math is well understood. MP3 is also standardized such that anyone with the specs can write a decoding algorithm for them if they desire to. The Tremor codec has just recently been released which means there's still a bit of development time before you see it adapted to handhelds like the Rio and iPod. Even if you've got a strong processor, which most MP3 handhelds don't have, you need to get your decoder on a MIPS diet so your chip isn't running full bore and sucking power out of the batteries like an electricty vampire. Integer only MPEG decoding is a well understood practice while Ogg is still relatively new even though it shares many concepts. Decoding algorithms are one thing, decoding algorithms that don't require 100+ MIPS are another.
Despite evidense to the contrary, California and Arizona are not different countries. Neither are New York and Pennsylvania. Ergo a call from one state to the next state is merely going over an abundance of fiber optic lines between the states. The difference between France and Germany is they are indeed different countries. Any call between the two countries has to switch between different national telephone systems. It isn't the amount of area you roam in but the different networks you roam in. Try roaming around northern Louisiana or central Texas making calls with impunity and see how high your phone bill is. Off network roaming no matter what continent you're on is very very expensive.
Technology may have advanced but it still costs as much money to get something into orbit as it did in the Apollo days. Despite technological progress there's been no developments in getting big things high up cheaply. L1 is a long way from Floria.
Besides Chris and Ferdinan weren't sailing for the sake of sailing and discovering, they were sailing to be rich when they got home. They were heavily backed by various crowned rulers to find trade routes to China so the rest of Europe wouldn't have to deal exclusively with the Venicians and Turks when they wanted to buy silks and spices that had come off the Silk Road.
Thanks for the info, I wasn't sure if they were just keen X terminals or framebuffers like you describe.
I wonder if it'd be possible to have a small cluster of application hosts all offering their services via X. Then another cluster of boxes to just run X clients and store the framebuffers, whenever a particular program were used the X client systems would just forward the command to the app servers who'd then export their display instructions to the clients and the terminals could just run the VNC-ish client. The processing load of the hosted apps could then be separate from the X client load. A couple boxes would be enough to serve a decent sized network with framebuffers and X protocol forwarding. The app server cluster could dynamically change their hosted X app to meet demand, i.e. a bunch of people have Mozilla or Konq running but only a few have xv or asclock. I wonder if you could even get something like that working if it would be cost effective.
Because Mozilla crashed before I submitted my original response to you this one will be shorter.
Sun already does this with their Sun Ray terminals. Your X session is attached to an ID card you stick in the terminal and when you move to another machine your session state moves with you. When you plug your card in your screen pops up where you left it and how you left it.
I don't have an Enterprise 45 and some Sun Rays lying around but I'll make a guess that in order to do this they just did a little transparent tweaking on top of X. The session could be attached not to a network address but instead to an alias. The terminals could send out an alias release packet to the alias manager on the server with the X client whe na card is removed and an alias set packet when the card is inserted. All the server would then need to do is reassign the alias wherever you moved to and your screen would show up on the next screen update.
X is already maintaining your state on the machine hosting the client. The tricky part is just attaching the session to something other than a specific network address or physical port. Resolution and colour depth can be translated by the server running on your terminal. I've never seen an open source implementation of whatever Sun does, maybe someone else can correct me if I was wrong about that or how it works. Just because I haven't seen it doesn't mean it doesn't exist.
These aren't their only rockets. The HOPE project was originally intended to be up and running now to be able to lend a hand with the ISS. Unfortunately Japan's economic climate makes it a bit difficult for the government to finish funding the project. It is indeed inefficient to send up manned crews to place a satellite into orbit which is exactly why most birds are launched using conventional rockets. In the cases of projects like the Hubble or ISS (which is a waste in itself) you need a reusable and manned craft.
Re:But I *like* those functions...
on
Phoenix 0.3 Is Out
·
· Score: 5, Insightful
You'd have more of a point if Mozilla wasn't already a huge framework of code. The parts of the Mozilla that make it a browser, mail client, or IRC client are very small compared to the rest of the Mozilla system. If you want just a browser load up Opera or Athena. Complaining about Mozilla being bloated is silly. It is an entire application framework, not merely a web browser app with a mail client.
Linux is not a very good way to design an operating system. Even as robust as Linux has become, Andrew said it best way back when to Linus when he said if Linux had turned in Linux as a project he would have failed the OS design and implementation class. That doesn't make Minix superior to Linux it merely makes it a better operating system to look at the code and learn the concepts from. It was written to teach people how to build a Unix system. A modern Linux system is not that simple anymore. Minix is and thus a much better choice to teach OS design neophytes.
Who is talking short term besides you? Planet hunters are well aware they're not likely to set foot on a planet they discover orbiting a distant star. They're not thinking they are going to find Earth 2 and fly there. They don't look because of that. They look because they just want to see if we're really as unique as we think we are. We've gone from thinking the Earth was flat and at the center of the universe to knowing it is shaped sort of like a pear and is housed inside of a normal galaxy which turned out to not even look like we originally thought it did. Now we're seeing that not only are we not at the center which the universe revolves but there are planets orbiting stars other than our own. The next step is to find out that we're not the only intelligent group of amino acids and sugar molecules putzing around our galaxy. Hell, we're probably not the only group of self organizing amino acids and sugars putzing around our solar system.
A moon colony has nothing to do with looking for extra solar planets. Compared to the cost of sending a bunch of stuff into space to crash into the moon in order to build stuff on it building stuff on Earth to look up at light coming from the sky is much more likely, possible, and useful in the near future.
Icelandic and Greek can both be represented in the ASCII character set and Korean was defined in the KSC 5601 specification so Hangul could be typed on ASCII based keyboards. I don't know anything about any languages spkoen or written in India but how easy would it be for any of them to be represented in ASCII, even with a shift methlod like RFC1557 for Korean? I know there's a character encoding for Hindi but not knowing the language it doesn't mean much to me. If your language isn't going to work on commodity computer components it is going to miss out on the enourmous production base of both hardware and software that already exists. Programming in many languages requires some knowlege or at least awareness of Latin based semantics if not outright knowlege of English. If you're asking people to program in C, Pascal, Java, or Ruby you pretty much need to teach them about the Roman character set. Write a Hindi front end for a compiler and more people who speak Hindi will program.
No it doesn't because a PV cell has a limited lifetime with diminishing returns as it ages. A 30 year old PV cell is practically useless for generating power for your household, by the time its thatold you might as well just rig it to your Palm Pilot and it MIGHT be able to charge its batteries. They're only being at most 15% or so efficient so it takes them a long long time to pay for themselves ad by the time they do their effiency has already dropped signifigantly. Calling it a zero sum process is a bit pessimistic on my part but the net gain is not as high as some would like you to believe. If you consider semiconductor manufacturers and industrial chemical suppliers dumping toxic waste where no one will find it (anytime soon) clean then thats your perogotive. I call it being duped by clean energy activists into thinking there is a magical solution to generating electricity cleanly.
Well a quick googling found this rather nifty site that might be of some use to you trying to shirk out of paying a measily $250.
Though I'm making an assumption that you just want something cheap (i.e. free as in beer) to respond to other posters I can perfectly understand why a Linux based plotter control system would be a good thing. Even running Windows the plotter/computer combination is a bit of a black box. If you can control that system easily and remotely it makes you workflow all that more efficient. I too have worked for a vinyl sign shop. We had a bunch of cheap Windows boxes built from crap found at computer fairs. I rigged up a thinnet network between them all so we could send finished files to the plotter box and print from there. It could be a pain sometimes working with Windows and its flakiness.
If I were setting up now and found find decent control programs for the plotters we used in Linux I'd strap a little Linux box to the plotter's stand and stick an Ethernet cable in it. From there you just need to mount a SMB or AFP mount from your design stations (running Windows or MacOS respectively) and then have a script on the Linux system chroned to scan the shared directory and if it finds any HPGL files to print them out. Most print shops have their proofing RIPs set up in this fashion. You just save to a network share and the RIP scans that folder looking for files to print out. It makes it pretty easy to use any design program and any RIP software because there is no need to low level communications, all they need to understand is PS. A design station/plotter server would be the same with HPGL instead of PS.
Yes, to made crystalline PV cells requires the same sort of chemicals and plant processes used in making semiconductor chips. It isn't necessarily dirty but very power intensive. Amorphous silicon PVs are also dirty to produce because of the amount of power needed and the chemicals used. Even if the chemicals are handled responsibly by the manufacturer there is no guarantee that the chemical's manufacturer handled the chemicals safely.
Calling PV power generation clean is an absurd falsehood by those promoting it, not to insult you but instead to point out the people who convinced you PV was the clean wave of the future. To generate power you need to spend power, on the whole it is a zero sum process, you don't get moreo ut of what was put in.
The reason oil is cheap easy and popular is because the energy it contains has been put there over the course of millions of years by microbes decomposing organic matter. The energy required to tap fossil fuels is much less than all of the energy contained in fossil fuels. The same goes for fisson power, the energy in the uranium was put there by a supernova billions of years ago. All we have to do is spend a little energy to tap that. Water, wind, and solar power sources are clean on the level they don't produce emissions themselves but the processes constructing them sure as hell do.
PV is clean in the same way electric cars are clean. Sure the eletric car doesn't produce emissions itself but it did take quite a bit of power to construct. There there is the fact that 55% of the nation's power comes from coal power plants, so for every kilowatt an eletric car uses you need to chalk up the fossil fuel emissions that generated that kilowatt. ULEV cars are cleaner overall than electric ones.
Hydroelectric and geothermic power generation is typically the cleanest IIRC all things considered. They are both just redirecting energy being emitted naturally and require a minimum amount of dirty processing to construct. They also last much longer than PV or wind generators and produce most power.
The only real way to clean up power usage is to make things more efficient and work with what you already have. PV cells require too much material alteration to be long term efficient. Lower power electronics, higher efficiency lighting, better industrial resource planning, solar heating, and efficient building design are all measures that can clean up power generation simply because less power is required. PVs can help lighten loads of the power grid by they are far from being a clean power source or an effective alternative to fossil fuels.
PV cells are clean? It sure would be nice if everyone could ignore the costs of manufacture. Just because it generates energy from sunlight you are already getting doesn't make it clean.
You obviously do not get the point if you're still discussing it. The original comment made the FSF out to be a bunch of guys who never did anything who want a say in the overall direction of a system they have contributed a great deal of time and effort working on. This is contrary to the fact the FSF has provided as important a contribution to Linux as the kernel developers. Without userland code a kernel is pretty much useless. You don't use a kernel you use userland code. You can't even compile the damn thing without the FSF's userland code. Without even GCC the Linux kernel is entirely irrelevant to the world at large. Linux then relies on everyone else's work to operate, saying those people don't have a say in the direction of the system as a whole is ludicrous and arrogant.
Slashdot Mirror
So how exactly do people with pie in the sky Wi-Fi plans intend to overcome 802.11b's inherent scalability problems? How many people can one AP REALLY serve?
From my experience I'd say that answer to that question is not very many. Having more than a couple people on a single AP is a recipe for pain and suffering. As the number of users on an AP increases so does the chances of packet collisions. As collisions increase the viability of the network decreases and you eventually reach a collapsing point where the network is unusable. A corollary to that rule then would be the larger your coverage area the higher a chance of collisions and thus a higher chance of the network collapsing.
You run into a similar problem with 802.3 which is solved by switching the network. With a wireless network you don't have the ability to add a switch in the middle of the network to keep the number of collisions down to a minimum. You're only got a bunch of nodes waiting their turn to talk. Switching channels isn't an option because APs can only serve particular channels.
With a coverage area of four miles then, the number of potential collisions on a channel is pretty high because your entire customer base could be in that four mile coverage area. Sweeping a broadbast between different nodes doesn't do much good on their end where all the static from other connections is an issue. On current networks you've got a small number of users because your coverage area is pretty small so problems aren't evident. You don't have problems on a wired network with only a 5 port hub either.
I think it was widely known the ISS would never come to scientific fruition in the bowels of appropriation committee meetings and planning commissions. The ISS has become and I think was intended to be-politically-a means to grab taxpayer dollars and stick them in the pockets of Congress people and their wealthy constituants.
A pork barrel is a project that puts federal dollars in the hands of Congress people in charge of the projects or appropriation committees for said projects. The best pork barrels are projects you can trick a lot of people into thinking are useful for the greater good so they don't ask any probing questions. An example would be a Represenative from a district in Vermont appropriating money for a project in that district some friends of his run a business in. For making them rich they cut said Congressman in on the fat of the "pork" for buying midgets to do battle or whatever it is rich people do. Sometimes pork barrels can be good for the people at large, a project could bring a bunch of jobs to a job poor district and then those people can eat and the country at large benefits from said federal project.
The ISS is starting to look more and more like this every day. The billions of dollars spent on the thing are going somewhere. It isn't like the solid rocket boosters of the Shuttle are lined with five dollar bills, not literally anyways. Before we had our ever impotent "War on Terror" to provide a means for getting public money into private hands the ISS was a perfect project to pork. It had a tenuous scientific basis, it would do JUST enough hard science for data to trickle in so it didn't look like a waste. As an added bonus the EU, Russia, and Japan could get in on the act and make it look to everyone like it was a giant shiny peace symbol in the sky. It's also a project that certain states *cough*California, Texas, and Floria*cough* would have a major hand in both developing and manufacturing. Billions of dollars means lots of cushy raises for government contractors. A pie in the sky science project that may or may not actually work as intended provides sweet CYA material for hearings later on.
You may or may not ask why was the ISS funded when we coulds have gotten more hard science out of smaller space projects and still bilked money out of them in particular Congressional districts. The answer is publicity. You can't go outside and take fricken pictures of the Mars Rover with a high powered zoom lens. You can take a picture of that megabright collection of aluminum cans flying around the planet. Also unlike probes launched from disposable rockets the ISS is something that needs to be maintained. Ron Popiel doesn't have a MagicStation where you set it and forget it. The ISS is a pork barrel that could have lasted for a decade or more had it been viable to do so. That's more than ten years of government contractors selling a $500 space toilet to NASA for $500,000.
Whatever dreams the ISS was supposed to fill for geeks and engineers don't matter to politicians, only the beaucoup cash that comes from those dreams matters. The ISS/Freedom/Alpha may have started as a cool science mission with attainable and useful goals but once it got into the grubby hands of Congress it turned into one giant government contract after another. As I said, now that we've got a "war" against nobody and maybe even a real war with remote control bombs and lasers on 747s the ISS isn't much needed anymore by the government. Why milk NASA's measly 14 billion when you can milk the DOD's uberbillions?
The ISS's failure is the fault of Congress and the people looking to make megabucks off taxpayer dollars, not Lance Bass. You can still despise him resoundly and wish he we eaten by wild battling midgets or whatever you want done to him but his inability to generate investor interest is not dooming the ISS.
Try BootCD from CharlesSoft.
For Jaguar
For Puma
I think some of your points are pretty ridiculous in this context. A game demo is often times free, I was playing the Jedi Knight 2 demo the other night I downloaded while I was in the shower, it cost me the electricity to leave my computer and cable modem turned on. That happens to be a lot less money and a lot less hassle than going to see a specific movie showtime or renting a movie.
Online gaming has few constraints besides your network speed. I don't NEED to play a team game in order to play the game. I can fire up UT2003 or CS and be fragging away whenever I want. I can't do that so easily with a movie.
There's nothing to really compare an extended edition DVD to a game mod. Day of Defeat and CS completely change the gameplay of Half-Life, that is no way compares to some fan edit of the Phantom Menace or a bonus DVD with some deleted scenes.
I'm not a video game fanatic either, I'm much more of a movie fanatic and have spent more time in the last month watching movies than playing video games. I'm not poking holes in your argument to defend gamers everywhere. I'm doing it because your comparisons suck.
The magnetic launcher is sort of doable but not on a scale large enough to send any decent sized payload into orbit unfortunately. NASA's Marshal Space Flight Center was doing some research on linear accelerators but I think a lot of their funding has been cut under the O'Keefe regime. Getting trains and scale models working on accelerators is one thing but an orbital payload is quite another. I think if someone were to invest in research most of their funding would end up doing to R&D rather than actually getting payloads into space. For a satellite launching startup it would likely be much better to get a bird in the sky and get your investment back plus a little profit to reinvest rather than blowing the whole kit and kaboodle on research and then be forced to sell out to someone that is just going to use your patents and research materials to make themselves money.
I'd addend to the flashlight a red filter as to better preserve your night vision. Novelty red Saran Wrap works pretty well to cover most flashlights, all you need in some wrap and tape. I prefer duct tape to get a better seal against dew as you'll get some condensation on the wrap because of cold air and a hot bulb. Sealing the wrap directly onto the flashlight lens I've found works well. If you've got a military surplus angle head flashlight you can probably pick up a red filter for it from the same place you bought it.
Also if you're going out in the middle of noplace and decide not to stay until dawn grab some reflective tape to attach strips to stuff you're taking with you like the Thermor or binocs. The full moon will give just enough light for the strips to be seen in the dark so you don't lose them when you set them down in the grass or something. Don't forget strips on your flashlights too. This seems counter intuitive until you set your flash down in the dark grass when its off and can't find it against until you sit on it.
IIRC the reaction used on these lasers only lasts at most about thirty seconds so you'd only be able to keep the laser on the target that amount of time at most. I wish I could find where I read that so I could pass it on or just check the figure to see if I am indeed remembering correctly.
Most of the chips used in handheld MP3 players are lightweight integer only processors or DSPs that handle the sort of integer ops MP3 and WMA files need to decode. With the more general purpose chips you need to write a decoder that can fit in the device's ROM and then not use signifigantly more processing power than the other decoders. Vorbis files need more math ops performed on them than MP3 files do. MP3 audio was designed to be run on systems whose processors performed fewer MIPS than a potato chip. Ogg Vorbis is from the era of Athlons and Intel space heaters. it simply requires more processing power period.
Squeezing an Ogg decoder into the same space as an MP3 decoder is hard enough but getting it to not drain the batteries in Planck time adds a whole level of complexity on top of it. The more work a chip has to do the more power it consumes. I wouldn't buy an iPod that only played 3 hours worth of music before needing a nuclear generator to recharge it.
For the hardware vendors though it is a question of space. Can an Ogg codec fit into the same ROM space as an MP3 codec and only use the same resources as said MP3 codec? If not they will not use Ogg codecs. Nor will they use Ogg codecs if it halves the battery life of the device, if the Ogg needs so much processing muscle it uses twice the wattage as the MP3 encoder they can't really sell that to people. Who cares if the device holds twice as many songs if the battery life is only half of what it would be otherwise. If playing an Ogg made my iPod only last 5 hours there's no way in hell I'd ever use them better quality or not. I routinely run my iPod for 8-10 hour stretches any period of time less than that is unacceptable for me personally.
Work on Ogg is going to continue and some intepid soul or souls are going to make a super cool Ogg decoder that can run on a paper clip taped to a Dorito but until then MP3 and WMP are going to dominate because they fit on the existing hardware.
If it's so trivial why haven't you done it already? Integer only MP3 decoders are all over the places and MP3 decoding using only integer math is well understood. MP3 is also standardized such that anyone with the specs can write a decoding algorithm for them if they desire to. The Tremor codec has just recently been released which means there's still a bit of development time before you see it adapted to handhelds like the Rio and iPod. Even if you've got a strong processor, which most MP3 handhelds don't have, you need to get your decoder on a MIPS diet so your chip isn't running full bore and sucking power out of the batteries like an electricty vampire. Integer only MPEG decoding is a well understood practice while Ogg is still relatively new even though it shares many concepts. Decoding algorithms are one thing, decoding algorithms that don't require 100+ MIPS are another.
Despite evidense to the contrary, California and Arizona are not different countries. Neither are New York and Pennsylvania. Ergo a call from one state to the next state is merely going over an abundance of fiber optic lines between the states. The difference between France and Germany is they are indeed different countries. Any call between the two countries has to switch between different national telephone systems. It isn't the amount of area you roam in but the different networks you roam in. Try roaming around northern Louisiana or central Texas making calls with impunity and see how high your phone bill is. Off network roaming no matter what continent you're on is very very expensive.
Technology may have advanced but it still costs as much money to get something into orbit as it did in the Apollo days. Despite technological progress there's been no developments in getting big things high up cheaply. L1 is a long way from Floria.
Besides Chris and Ferdinan weren't sailing for the sake of sailing and discovering, they were sailing to be rich when they got home. They were heavily backed by various crowned rulers to find trade routes to China so the rest of Europe wouldn't have to deal exclusively with the Venicians and Turks when they wanted to buy silks and spices that had come off the Silk Road.
Thanks for the info, I wasn't sure if they were just keen X terminals or framebuffers like you describe.
I wonder if it'd be possible to have a small cluster of application hosts all offering their services via X. Then another cluster of boxes to just run X clients and store the framebuffers, whenever a particular program were used the X client systems would just forward the command to the app servers who'd then export their display instructions to the clients and the terminals could just run the VNC-ish client. The processing load of the hosted apps could then be separate from the X client load. A couple boxes would be enough to serve a decent sized network with framebuffers and X protocol forwarding. The app server cluster could dynamically change their hosted X app to meet demand, i.e. a bunch of people have Mozilla or Konq running but only a few have xv or asclock. I wonder if you could even get something like that working if it would be cost effective.
Are you positive? The one I used looked a lot like it had a CDE desktop on it.
Because Mozilla crashed before I submitted my original response to you this one will be shorter.
Sun already does this with their Sun Ray terminals. Your X session is attached to an ID card you stick in the terminal and when you move to another machine your session state moves with you. When you plug your card in your screen pops up where you left it and how you left it.
I don't have an Enterprise 45 and some Sun Rays lying around but I'll make a guess that in order to do this they just did a little transparent tweaking on top of X. The session could be attached not to a network address but instead to an alias. The terminals could send out an alias release packet to the alias manager on the server with the X client whe na card is removed and an alias set packet when the card is inserted. All the server would then need to do is reassign the alias wherever you moved to and your screen would show up on the next screen update.
X is already maintaining your state on the machine hosting the client. The tricky part is just attaching the session to something other than a specific network address or physical port. Resolution and colour depth can be translated by the server running on your terminal. I've never seen an open source implementation of whatever Sun does, maybe someone else can correct me if I was wrong about that or how it works. Just because I haven't seen it doesn't mean it doesn't exist.
These aren't their only rockets. The HOPE project was originally intended to be up and running now to be able to lend a hand with the ISS. Unfortunately Japan's economic climate makes it a bit difficult for the government to finish funding the project. It is indeed inefficient to send up manned crews to place a satellite into orbit which is exactly why most birds are launched using conventional rockets. In the cases of projects like the Hubble or ISS (which is a waste in itself) you need a reusable and manned craft.
You'd have more of a point if Mozilla wasn't already a huge framework of code. The parts of the Mozilla that make it a browser, mail client, or IRC client are very small compared to the rest of the Mozilla system. If you want just a browser load up Opera or Athena. Complaining about Mozilla being bloated is silly. It is an entire application framework, not merely a web browser app with a mail client.
Linux is not a very good way to design an operating system. Even as robust as Linux has become, Andrew said it best way back when to Linus when he said if Linux had turned in Linux as a project he would have failed the OS design and implementation class. That doesn't make Minix superior to Linux it merely makes it a better operating system to look at the code and learn the concepts from. It was written to teach people how to build a Unix system. A modern Linux system is not that simple anymore. Minix is and thus a much better choice to teach OS design neophytes.
Who is talking short term besides you? Planet hunters are well aware they're not likely to set foot on a planet they discover orbiting a distant star. They're not thinking they are going to find Earth 2 and fly there. They don't look because of that. They look because they just want to see if we're really as unique as we think we are. We've gone from thinking the Earth was flat and at the center of the universe to knowing it is shaped sort of like a pear and is housed inside of a normal galaxy which turned out to not even look like we originally thought it did. Now we're seeing that not only are we not at the center which the universe revolves but there are planets orbiting stars other than our own. The next step is to find out that we're not the only intelligent group of amino acids and sugar molecules putzing around our galaxy. Hell, we're probably not the only group of self organizing amino acids and sugars putzing around our solar system.
A moon colony has nothing to do with looking for extra solar planets. Compared to the cost of sending a bunch of stuff into space to crash into the moon in order to build stuff on it building stuff on Earth to look up at light coming from the sky is much more likely, possible, and useful in the near future.
Icelandic and Greek can both be represented in the ASCII character set and Korean was defined in the KSC 5601 specification so Hangul could be typed on ASCII based keyboards. I don't know anything about any languages spkoen or written in India but how easy would it be for any of them to be represented in ASCII, even with a shift methlod like RFC1557 for Korean? I know there's a character encoding for Hindi but not knowing the language it doesn't mean much to me. If your language isn't going to work on commodity computer components it is going to miss out on the enourmous production base of both hardware and software that already exists. Programming in many languages requires some knowlege or at least awareness of Latin based semantics if not outright knowlege of English. If you're asking people to program in C, Pascal, Java, or Ruby you pretty much need to teach them about the Roman character set. Write a Hindi front end for a compiler and more people who speak Hindi will program.
No it doesn't because a PV cell has a limited lifetime with diminishing returns as it ages. A 30 year old PV cell is practically useless for generating power for your household, by the time its thatold you might as well just rig it to your Palm Pilot and it MIGHT be able to charge its batteries. They're only being at most 15% or so efficient so it takes them a long long time to pay for themselves ad by the time they do their effiency has already dropped signifigantly. Calling it a zero sum process is a bit pessimistic on my part but the net gain is not as high as some would like you to believe. If you consider semiconductor manufacturers and industrial chemical suppliers dumping toxic waste where no one will find it (anytime soon) clean then thats your perogotive. I call it being duped by clean energy activists into thinking there is a magical solution to generating electricity cleanly.
Well a quick googling found this rather nifty site that might be of some use to you trying to shirk out of paying a measily $250.
Though I'm making an assumption that you just want something cheap (i.e. free as in beer) to respond to other posters I can perfectly understand why a Linux based plotter control system would be a good thing. Even running Windows the plotter/computer combination is a bit of a black box. If you can control that system easily and remotely it makes you workflow all that more efficient. I too have worked for a vinyl sign shop. We had a bunch of cheap Windows boxes built from crap found at computer fairs. I rigged up a thinnet network between them all so we could send finished files to the plotter box and print from there. It could be a pain sometimes working with Windows and its flakiness.
If I were setting up now and found find decent control programs for the plotters we used in Linux I'd strap a little Linux box to the plotter's stand and stick an Ethernet cable in it. From there you just need to mount a SMB or AFP mount from your design stations (running Windows or MacOS respectively) and then have a script on the Linux system chroned to scan the shared directory and if it finds any HPGL files to print them out. Most print shops have their proofing RIPs set up in this fashion. You just save to a network share and the RIP scans that folder looking for files to print out. It makes it pretty easy to use any design program and any RIP software because there is no need to low level communications, all they need to understand is PS. A design station/plotter server would be the same with HPGL instead of PS.
Yes, to made crystalline PV cells requires the same sort of chemicals and plant processes used in making semiconductor chips. It isn't necessarily dirty but very power intensive. Amorphous silicon PVs are also dirty to produce because of the amount of power needed and the chemicals used. Even if the chemicals are handled responsibly by the manufacturer there is no guarantee that the chemical's manufacturer handled the chemicals safely.
Calling PV power generation clean is an absurd falsehood by those promoting it, not to insult you but instead to point out the people who convinced you PV was the clean wave of the future. To generate power you need to spend power, on the whole it is a zero sum process, you don't get moreo ut of what was put in.
The reason oil is cheap easy and popular is because the energy it contains has been put there over the course of millions of years by microbes decomposing organic matter. The energy required to tap fossil fuels is much less than all of the energy contained in fossil fuels. The same goes for fisson power, the energy in the uranium was put there by a supernova billions of years ago. All we have to do is spend a little energy to tap that. Water, wind, and solar power sources are clean on the level they don't produce emissions themselves but the processes constructing them sure as hell do.
PV is clean in the same way electric cars are clean. Sure the eletric car doesn't produce emissions itself but it did take quite a bit of power to construct. There there is the fact that 55% of the nation's power comes from coal power plants, so for every kilowatt an eletric car uses you need to chalk up the fossil fuel emissions that generated that kilowatt. ULEV cars are cleaner overall than electric ones.
Hydroelectric and geothermic power generation is typically the cleanest IIRC all things considered. They are both just redirecting energy being emitted naturally and require a minimum amount of dirty processing to construct. They also last much longer than PV or wind generators and produce most power.
The only real way to clean up power usage is to make things more efficient and work with what you already have. PV cells require too much material alteration to be long term efficient. Lower power electronics, higher efficiency lighting, better industrial resource planning, solar heating, and efficient building design are all measures that can clean up power generation simply because less power is required. PVs can help lighten loads of the power grid by they are far from being a clean power source or an effective alternative to fossil fuels.
PV cells are clean? It sure would be nice if everyone could ignore the costs of manufacture. Just because it generates energy from sunlight you are already getting doesn't make it clean.
You obviously do not get the point if you're still discussing it. The original comment made the FSF out to be a bunch of guys who never did anything who want a say in the overall direction of a system they have contributed a great deal of time and effort working on. This is contrary to the fact the FSF has provided as important a contribution to Linux as the kernel developers. Without userland code a kernel is pretty much useless. You don't use a kernel you use userland code. You can't even compile the damn thing without the FSF's userland code. Without even GCC the Linux kernel is entirely irrelevant to the world at large. Linux then relies on everyone else's work to operate, saying those people don't have a say in the direction of the system as a whole is ludicrous and arrogant.