The basic idea behind this article seems to assume that as sequencing and synthesis technology and skills become widely available, there will be a parallel increase in the danger from the misuse of this technology. I beg to differ. Sequencing DNA does not give you that much insight into how things really work. Nor does tweaking out protein structure. That's the easy step. But the dynamic equilibrium of a cell is maintained by the DNA, the RNA and the proteins, all simultaneous interacting in an essentially stabilized chaotic system. Sure we can "knock-out" a gene here and there, replace one protein with another, but doing so is no more a display of knowledge then is pruning a tree. We're still a long (long long long) way from designing trees from scratch or people developing the new "super-bug" in the garage or even university lab.
That said, there is a real danger from people using the techniques described above to create hybrid strains (SARS+influenza etc.) to create new virulent strains based on existing virii and bacteria. Of course, even that is much harder than said, primarily because the only way to test which strains work, is to infect people. Any failure and your subject will develop resistance and be useless for future testing. So, you'd need a large number of subjects, or you'd need to develop on a disease which infects both humans and rats (or something) and then hope that the virulence will be analogous for humans. Fortunately, this is rarely the case, what kills rats like, well rats, often doesn't even faze humans and vice versa.
Hmm, I wonder if I should worry about men in dark suits showing up at my door now...
Did you read the article. (I know you didn't, just wanted to ask in a nice way).
unix based systems run many more daemons that are inherently more vunerable than microsoft products.
Proof? Most daemons nowadays are running as non-priviledged users or are explicitly chroot to prevent standard abuse. The only easy exploits are buffer overflow and those will only work on similar architectures and kernel versions. I'm not sure it's even technically feasible to write a virus that even comes close to spreading as quickly as SoBig. Oh, and do you have any idea how many daemons are running on a standard XP install?
it isn't the OS's fault, it is outlook...
Well, the OS (Windows) and the client (Outlook) are essentially running the same code. Whose fault is that?
if linux blows up, then "outlook for linux" would be just as vunerable ON TOP OF all the other client server bugs
Uhmm, again wrong, only a complete moron would run an mail readers as root, which is essentially what you are doing with Win+Outlook.
I just realized, damn it, I've been trolled again.
Read my journal entry where I talk about the exact same thing. I think we only differ on the size of the screen, you want one bigger than I do. Hell, I'd pay twice what I paid for my Handera for one that fits my requirements.
It's not a matter of perspective. This isn't a "Feel the elephant and guess what it is" problem. A perspective that steel doesn't ever melt is not a "perpective", it is a "incorrect view of the world". This is an engineering problem and the only valid perspective is one supported by analysis based on known facts and uncertainties. A chunk of foam fell off during the SS launch. What are the risks imposed by this? How can we improve our analysis of the risks involved. Do a hard-nosed analysis. What is the 10% best case? What is the 50% average case? What is the 10% worst case? Where is the most uncertainty in our model? How can we narrow our uncertainty bands?
Just because most people treat risk analysis like some grade school math problem doesn't mean that there hasn't been lots of research on how to do proper risk analyses for complex systems. It isn't simple, but you can do a rigorous risk analysis based on uncertain information. Such an analysis would show which missing information is contributing the largest amount of uncertainty to the end result. In this case, the largest uncertainty was "WHERE DID THE FOAM HIT". Given that this most basic uncertainty was never resolved until much later, there was no way that a proper analysis could have said with any certainty as to the safety of the Columbia given the foam strike.
"We think the foam was this big, we think it didn't hit a critical tile and we think our computer program is too pessimistic so the shuttle is safe" is utter and complete BULLSHIT. It doesn't matter how many numbers you wrap around those words. A bullshit perspective is still bullshit. And no real engineering manager would have let the Lockheed engineers get away with presenting the crappy analysis report. I have another post from a Feburary shuttle story about this.
And harddisk size, and backbone bandwidth, and the number of internet nodes, and the cost/performance of tech, etc. Any evolutionary process progresses at an exponential rate if you just observe history.
Exponential: You keep using that word. I do not think it means what you think it means. Go back and look at history, there may be rapid, pseudo exponential growth during the initial stages of technology and life (Cambrian explosion comes to mind) but inevitably, the laws of physics and the limits of resources cause a slowdown. I know of no examples of an old technology (or "evolutionary process") which is still undergoing exponential growth/improvement.
Actually NASA told the U.S. that they could fly n missions with x dollars. It turned out that they could barely fly n/50 missions with x*3 dollars. Why? Because the people doing the initial calculations were so intent on looking good that they ignored engineering realities. Caught in the lies they themselves had created in order to justify funding for the Space Shuttle to begin with, NASA started pushing safety limits issuing waivers to keep the launch schedule going.
I do blame the managers and I do blame congress. I blame NASA for failing to be truthful in it's own cost and safety reports. I blame Congress for not providing sufficient oversight and for forcing sub-par designs on NASA in order to appease pork barrel political hand-outs.
Also, I fail to see how you can blame "Billy boy" when he was busy fighting off impeachment and harrassment by a Republican congress when GBush I didn't do diddly for NASA either.
Part of a manager's job includes risk management and resource allocation. This means properly assessing the likelihood and impact of a risk. In this case, I would suggest that management considered the 'cost' of pursuing further investigation to be higher than the 'likelihood * impact' factor of doing nothing. They have probably made the same decision many times before, successfully, which would encourage them to make the same decision again. Only this time, they were wrong -- the statistics caught up with them.
Well I ain't gonna let this one just go by. Managers of massively complex billion dollar engineering systems with no-backup should not be making decisions based on half-assed, bullshit "average case analysis". What you are calling risk analysis is called "bullshit guessing" in the real world. Here's a hint, the value of the average case is not the average value, you fucktard. If on average every foam hit doesn't hit a critical part, that doesn't mean that a critical part will never get hit. That's like saying since the average case of a bullet shot in the air is not to hit anyone, we should be able to shoot bullets up into the air and never kill anyone.
Managers of systems like the space shuttle should be paranoid assholes who are constantly asking for proof and hard-numbered analyses from multiple parties. Managers of such systems should be asking for proof and thorough analyses, not reassurances. The fact that few (if any) such managers exist at NASA is an indication of how much the organization has changed from a hard-edged, truth-telling organization promoting engineering rigor and integrity to an organization filled with ass-kissing, platitude-spewing, mealy mouthed morons, bum sucking their respective ways to the top.
Advances in the evolution of all kinds of technology will continue to progress at an exponential pace;
You can measure "all technology" with a single variable (or each technology depending on how you parse your sentence)? Gimme a break. So, it's just a matter of time before we have more artificial memory capacity than there are atoms in the known universe? Kurzweil may be smart, but that doesn't mean everything he writes is correct or even reasonable. In some areas, he's a certifiable nut.
What do you mean "if it can be made to work?" Nature already does it, and "The principles of physics, as far as I can see, do not speak against the possibility of maneuvering things atom by atom." artificially.
But there are other principles than Physics involved. Organization and coordination being a big one (the big one in my opinion). How will nano-bots coordinate their activities? Radio waves? Too small. Chemicals? Too expensive and potentially complex. Tiny interconnects? Too fragile. How much memory/state can a theoretical nano-bot have? What messages will they need to coordinate? The organization and coordination of cells in nature (ie your body) is an insanely complex dance of chemical and biological triggers developed over billions of years of evolution. Duplicating something even remotely similar is a decidedly non-trivial task and one that has been neglected amongst all of the hype about the "comming age of nano-technology". Bah Humbug.
So, how about the primary point... that Dell supplies us with a real service manual with pictures and instructions on how to replace every part in the laptop
Big deal. What part do you need to replace? Memory? Hard-drive? Pretty straight forward to do and well documented for the PB. If you're replacing the screen/motherboard etc., then I would venture that a person with the required skills and parts could do either the i8500 or the PB with equal difficulty given the resources available on the web. The manual (or lack thereof) doesn't matter a twentieth as much to me as the fact that the PB comes with all of the software mentioned above out of the box and more and the Dell comes with WindowsXP.
How much time/effort have you saved because Dell supplies the manual and Apple doesn't? How much time/effort could you have saved if you bought a computer with all of the software you installed on your Dell already installed? What do you do with your laptop? YMMV, but if you aren't playing video games, then the PB beats the Dell by a mile for the vast majority of my computing needs.
So does your i8500 have a DVD-R? Does it have an ATI Mobility Radeon 9600 w/ 64MB DDR? Lighted keyboard? 80GB HD? Gigabit Ethernet? Built-in WiFi and Bluetooth? When I tried to price a similarly equipped i8500, the closest I could get was $2636. And no FireWire 800. Granted, the uber-high resolution screen is cool and I wish Apple would bump up their screen options. BTW, the PowerBook 15, with everything mentioned above, goes for $2600 and comes with iTunes, iDVD, Apache, Python, Perl, and bash installed.
BTW, I did price another i8500 bundle, and the best I could do was 2085 with the 2.6GHz and WUXGA option (I didn't see any WUXGA+ option either). Of course that was with 256MB of RAM, 30GB hard drive, DVD drive (no CD-RW, no DVD-R), no bluetooth, and a 32MB ATI Radeon 9000.
So have you really done the comparisons? Damn it, I think I've been trolled.
Ah, another language populist apologist. Language evolving (example "google" as a verb) is quite separate from moronic misspellings being propogated. The words "lose" and "loose" are two completely different words and to mix the two is not evolution, but a simple reflection of the spread of illiteracy and idiocy. I suppose you think that "there" and "their" should be spelled either way since so many people do that also. Perhaps you think that it will soon be acceptable to write "Their is know way that there horse could loose" instead of "There is no way that their horse could lose."
A pox on you and your AC guts. Just because a language "feature" is widespread doesn't mean it isn't a bug.
What the heck is wrong with people? Once you get on the internet, does everybody forget how to spell lose (rhymes with booze), meaning "to miss from one's possession", "to undergo defeat"? Why are people spelling it loose (rhymes with goose), which means "not rigidly fastened or securely attached", "slack" and "to release" (as in "loose your arrows)? I've seen it in manuals, websites, even brochures for supposedly profitable companies. Is the entire English speaking world in a conspiracy to drive me mad? Why is this spelling spreading? What the hell is wrong with people? Will we start spelling hear (to listen) as here? Don't answer that!
TO THOSE WHO CONTINUE TO MISSPELL 'LOSE': A POX ON YE, YE FINGERS, YE KEYBOARDS, AND YE MICE! LEARN HOW TO SPELL, YOU UGLY WARTS ON SALAMANDERS' TONGUES!
Sorry, I meant, purchase option, not boot option. Sigh, posting late at night and such. If you read the grandparent post of my post, it should be clear that I am well aware of how Microsoft "protected their innovation" through illegal OEM contracts. Thanks for clarifying that.
Putting BeOS on a system would increase support costs -- Everytime someone calls because they chose the wrong item on the boot menu, there goes pretty much the entire profit on the machine.
Nothing stopped the OEM's from offering BeOS as an unsupported option. And profit from the machine should have much higher for not having to pay for the OS license. Unfortunately, as we all can see, Microsoft managed to strangle that baby before it even had a chance to crawl.
If they were trying to prevent OEMs from shipping another OS altogether, then maybe I'd take up a pitchfork too. Be had a chance to be competitive and chose the whine and sue road to success.
Uhmmm, if you don't know shit about what happened, shut the fuck up. Sorry for the strong language, but that is precisely what Microsoft did. Not only were vendors prohibited from shipping dual boot machines, but if they tried to ship machines with only BeOS on them, Microsoft would have
Still charged them for the Windows license
Changed their "partner status" so they wouldn't get anymore kickbacks discounts etc., raising the costs of ALL of their copies of Windows.
Be offered their BeOS for FREE to any OEM who would install it on their machines. The end result was not a single top-20 pc manufacturer shipped any machines loaded with BeOS or dual booted with BeOS. A few manufacturers shipped with BeOS on hidden partitions requiring an arcane complex series to steps to activate.
$23.5 million? Chump change. They should have gotten at least $235 million if not $2.35 billion. There is no justice.
But if Apple cared to, they could actually just make X11 a fully functional component of the OS X desktop, indistinguishable from Carbon and Cocoa. That would involve some automatic clipboard and drag-and-drop translation (between X11 and OS X), making X11 executables double-clickable, and fixing Apple X11 window management.
Can you say Mac OS X 10.3 Panther? According to the page above: "Panther will include a final X11 window server for Unix-based apps, improved NFS/UFS, FreeBSD 5 innovations as well as support for popular Linux APIs, IPv6 and other important acronyms.
Let's face it. Most people are now used to visual as apposed to aural stimulation. Think of the mid 20th Century. When people relaxed at home, they were talking, reading or listening to the radio or a record. During the last third of the 20th Century, people were watching TV instead. People prefer to watch moving pictures instead of listening. And now the DVD is providing the same video sales revolution that the CD did for audio sales. Add video games to the mix and even less "entertainment" money is going to be spent on audio CD's. And that trend isn't going to stop. Ever.
When I went to a Large-Media-Store yesterday, I did a quick comparison of the CD section and the DVD section. The DVD section was mostly around $20 with a bunch of older titles at $10. DVD sets were around $40. So let's say that I went to this "L-M-S" with $40 and wished to spend it on something to entertain me. Among the bundles I could buy were:
4 old but popular movies (6-7 hours of video)
1 complete season N of TV show X (4-10 hours of video + extras)
2 recent movies (3 hours of video + commentary + extras)
1 current release video game (3-30 hours of game play)
2 older release video games (6-60 hours of game play)
2 current CD's with a few songs worth listening to. (at most 140 minutes of music)
3-4 older CD's with many songs worth listening to. (about 4 hours of music)
I don't know about you, but those last two look pretty fucking anaemic compared to the first five. That is why CD sales are down. And why they aren't going to ever recover to the levels they were during the 80's and 90's.
"Smaller generators are less efficent than larger ones"
Your basic premise if incorrect. Smaller power plants are generally less efficient, but home generation is much more efficient because the waste heat can be used effectively. From my recollection; a very efficient natural gas powerplant will have theoretical efficiencies approaching 60% while a home natural gas generator utilizing waste heat will be 15-20% more efficient.
Yeah, when I was there, NAVLAB 1 (or was it 2) was a huge oversized van with two cameras (or was it three) a laser scanner, 5 or 7 sun workstations and two extra generators to keep stuff powered. And it went around 10 feet a minute. We would occasionally see it creeping around the park.
I saw their latest thing on TV a few years back, the scaled down system consisted of a laptop and a tiny video camera clipped to the rear-view mirror. And the car was going 60 mph. Moore's Law + a decade and a half of software research + a goood chunk of DARPA research money can achieve pretty amazing results.
ThinkGeek has them also. I have one and I found the diaphram made of plastic too loud. It makes kind of a loud "PHWAPP" when you do a full power blast. To fix this (in true geek fashion) I modified the 'zooka. I disassembled the rear assembly, and replaced the plastic diaphram with a piece cut from a black T-shirt I never wore. Now, the blast is a little less powerful, but it makes a quiet "fwop". Don't have any measuring devices, but it's at least 3x quiter. Much more effective and much more fun.
Horse-pucky. As has been mentioned elsewhere, real-time mission critical software should be written in such a way that errors are handled gracefully, not propogated to other modules. Even with errors it should allow the system to continue functioning. Those "few bugs" indicate a fundamental flaw in the design of the software in addition to flaws in how it was implemented. Needing to "reboot" is frankly bullshit for the quality of software that should be in the Raptor or any combat military system in the integration stage and beyond.
Just because something is "insanely" complicated, doesn't mean that the complications are impossible to manage through proper design and development. Just because most software sucks doesn't mean that all software needs to suck. Of course, given the quality of most commercial software, that's what we have come to expect.
3 hours of use. Forget that
on
Garmin iQue 3600
·
· Score: 3, Insightful
To quote the msnbc review Garmin claims "approximately two weeks standby time or approximately 10 days if used an average of 30 minutes per day with backlight off. Battery life will vary depending upon temperature and individual use patterns."... In real life, expect two to three hours max if you're using the GPS features. Screw that. Is it just me or has battery life been getting worse and worse with each new batch of Palm devices? My Palm Personal did 5-6 weeks, and my four AAA Handera 330 has no problem going up to two months of light use, 1 month of moderate use. 10 days?
Yeah, the e-ink stuff looks really cool, low power and high contrast. Two things bother me about it: illumination and touch screen compatibility. From the looks of it, it looks like it'll need front lighting. Also, given it's reliance on electric field, I'd be worried about interference with the compatibility. The fact that application of the "ink" is straight-forward, however, enhances the useability factor alot. Perhaps a 'no-bezel' or all surface display PDA, where even the buttons are mini displays.
OCR+Translation is a cool idea. I think what you're looking for is probably closer to the realm of what PDA+PDA camera would be able to do. Digital cameras are too busy trying white-balance, filter, image enhance, etc. to think about OCI. Check out my journal entry for the first half of your idea. I'm adding your idea to it also, with references natch.
Darn my glasses. so, 1,200,000 km^2. Of course, if you improve efficiency to 50% you only need 24,000 km^2. And if you choose an area with better than average sunlight levels, say double, we're down to 12,000 km or 110 km * 110 km.
Slashdot Mirror
That said, there is a real danger from people using the techniques described above to create hybrid strains (SARS+influenza etc.) to create new virulent strains based on existing virii and bacteria. Of course, even that is much harder than said, primarily because the only way to test which strains work, is to infect people. Any failure and your subject will develop resistance and be useless for future testing. So, you'd need a large number of subjects, or you'd need to develop on a disease which infects both humans and rats (or something) and then hope that the virulence will be analogous for humans. Fortunately, this is rarely the case, what kills rats like, well rats, often doesn't even faze humans and vice versa.
Hmm, I wonder if I should worry about men in dark suits showing up at my door now...
I just realized, damn it, I've been trolled again.
Read my journal entry where I talk about the exact same thing. I think we only differ on the size of the screen, you want one bigger than I do. Hell, I'd pay twice what I paid for my Handera for one that fits my requirements.
Just because most people treat risk analysis like some grade school math problem doesn't mean that there hasn't been lots of research on how to do proper risk analyses for complex systems. It isn't simple, but you can do a rigorous risk analysis based on uncertain information. Such an analysis would show which missing information is contributing the largest amount of uncertainty to the end result. In this case, the largest uncertainty was "WHERE DID THE FOAM HIT". Given that this most basic uncertainty was never resolved until much later, there was no way that a proper analysis could have said with any certainty as to the safety of the Columbia given the foam strike.
"We think the foam was this big, we think it didn't hit a critical tile and we think our computer program is too pessimistic so the shuttle is safe" is utter and complete BULLSHIT. It doesn't matter how many numbers you wrap around those words. A bullshit perspective is still bullshit. And no real engineering manager would have let the Lockheed engineers get away with presenting the crappy analysis report. I have another post from a Feburary shuttle story about this.
Exponential: You keep using that word. I do not think it means what you think it means. Go back and look at history, there may be rapid, pseudo exponential growth during the initial stages of technology and life (Cambrian explosion comes to mind) but inevitably, the laws of physics and the limits of resources cause a slowdown. I know of no examples of an old technology (or "evolutionary process") which is still undergoing exponential growth/improvement.
I do blame the managers and I do blame congress. I blame NASA for failing to be truthful in it's own cost and safety reports. I blame Congress for not providing sufficient oversight and for forcing sub-par designs on NASA in order to appease pork barrel political hand-outs.
Also, I fail to see how you can blame "Billy boy" when he was busy fighting off impeachment and harrassment by a Republican congress when GBush I didn't do diddly for NASA either.
Well I ain't gonna let this one just go by. Managers of massively complex billion dollar engineering systems with no-backup should not be making decisions based on half-assed, bullshit "average case analysis". What you are calling risk analysis is called "bullshit guessing" in the real world. Here's a hint, the value of the average case is not the average value, you fucktard. If on average every foam hit doesn't hit a critical part, that doesn't mean that a critical part will never get hit. That's like saying since the average case of a bullet shot in the air is not to hit anyone, we should be able to shoot bullets up into the air and never kill anyone.
Managers of systems like the space shuttle should be paranoid assholes who are constantly asking for proof and hard-numbered analyses from multiple parties. Managers of such systems should be asking for proof and thorough analyses, not reassurances. The fact that few (if any) such managers exist at NASA is an indication of how much the organization has changed from a hard-edged, truth-telling organization promoting engineering rigor and integrity to an organization filled with ass-kissing, platitude-spewing, mealy mouthed morons, bum sucking their respective ways to the top.
You can measure "all technology" with a single variable (or each technology depending on how you parse your sentence)? Gimme a break. So, it's just a matter of time before we have more artificial memory capacity than there are atoms in the known universe? Kurzweil may be smart, but that doesn't mean everything he writes is correct or even reasonable. In some areas, he's a certifiable nut.
What do you mean "if it can be made to work?" Nature already does it, and "The principles of physics, as far as I can see, do not speak against the possibility of maneuvering things atom by atom." artificially.
But there are other principles than Physics involved. Organization and coordination being a big one (the big one in my opinion). How will nano-bots coordinate their activities? Radio waves? Too small. Chemicals? Too expensive and potentially complex. Tiny interconnects? Too fragile. How much memory/state can a theoretical nano-bot have? What messages will they need to coordinate? The organization and coordination of cells in nature (ie your body) is an insanely complex dance of chemical and biological triggers developed over billions of years of evolution. Duplicating something even remotely similar is a decidedly non-trivial task and one that has been neglected amongst all of the hype about the "comming age of nano-technology". Bah Humbug.
Big deal. What part do you need to replace? Memory? Hard-drive? Pretty straight forward to do and well documented for the PB. If you're replacing the screen/motherboard etc., then I would venture that a person with the required skills and parts could do either the i8500 or the PB with equal difficulty given the resources available on the web. The manual (or lack thereof) doesn't matter a twentieth as much to me as the fact that the PB comes with all of the software mentioned above out of the box and more and the Dell comes with WindowsXP.
How much time/effort have you saved because Dell supplies the manual and Apple doesn't? How much time/effort could you have saved if you bought a computer with all of the software you installed on your Dell already installed? What do you do with your laptop? YMMV, but if you aren't playing video games, then the PB beats the Dell by a mile for the vast majority of my computing needs.
BTW, I did price another i8500 bundle, and the best I could do was 2085 with the 2.6GHz and WUXGA option (I didn't see any WUXGA+ option either). Of course that was with 256MB of RAM, 30GB hard drive, DVD drive (no CD-RW, no DVD-R), no bluetooth, and a 32MB ATI Radeon 9000.
So have you really done the comparisons? Damn it, I think I've been trolled.
A pox on you and your AC guts. Just because a language "feature" is widespread doesn't mean it isn't a bug.
TO THOSE WHO CONTINUE TO MISSPELL 'LOSE': A POX ON YE, YE FINGERS, YE KEYBOARDS, AND YE MICE! LEARN HOW TO SPELL, YOU UGLY WARTS ON SALAMANDERS' TONGUES!
Sorry, I meant, purchase option, not boot option. Sigh, posting late at night and such. If you read the grandparent post of my post, it should be clear that I am well aware of how Microsoft "protected their innovation" through illegal OEM contracts. Thanks for clarifying that.
Nothing stopped the OEM's from offering BeOS as an unsupported option. And profit from the machine should have much higher for not having to pay for the OS license. Unfortunately, as we all can see, Microsoft managed to strangle that baby before it even had a chance to crawl.
Uhmmm, if you don't know shit about what happened, shut the fuck up. Sorry for the strong language, but that is precisely what Microsoft did. Not only were vendors prohibited from shipping dual boot machines, but if they tried to ship machines with only BeOS on them, Microsoft would have
Be offered their BeOS for FREE to any OEM who would install it on their machines. The end result was not a single top-20 pc manufacturer shipped any machines loaded with BeOS or dual booted with BeOS. A few manufacturers shipped with BeOS on hidden partitions requiring an arcane complex series to steps to activate.
$23.5 million? Chump change. They should have gotten at least $235 million if not $2.35 billion. There is no justice.
Can you say Mac OS X 10.3 Panther? According to the page above: "Panther will include a final X11 window server for Unix-based apps, improved NFS/UFS, FreeBSD 5 innovations as well as support for popular Linux APIs, IPv6 and other important acronyms.
Let's face it. Most people are now used to visual as apposed to aural stimulation. Think of the mid 20th Century. When people relaxed at home, they were talking, reading or listening to the radio or a record. During the last third of the 20th Century, people were watching TV instead. People prefer to watch moving pictures instead of listening. And now the DVD is providing the same video sales revolution that the CD did for audio sales. Add video games to the mix and even less "entertainment" money is going to be spent on audio CD's. And that trend isn't going to stop. Ever.
When I went to a Large-Media-Store yesterday, I did a quick comparison of the CD section and the DVD section. The DVD section was mostly around $20 with a bunch of older titles at $10. DVD sets were around $40. So let's say that I went to this "L-M-S" with $40 and wished to spend it on something to entertain me. Among the bundles I could buy were:
I don't know about you, but those last two look pretty fucking anaemic compared to the first five. That is why CD sales are down. And why they aren't going to ever recover to the levels they were during the 80's and 90's.
Your basic premise if incorrect. Smaller power plants are generally less efficient, but home generation is much more efficient because the waste heat can be used effectively. From my recollection; a very efficient natural gas powerplant will have theoretical efficiencies approaching 60% while a home natural gas generator utilizing waste heat will be 15-20% more efficient.
I saw their latest thing on TV a few years back, the scaled down system consisted of a laptop and a tiny video camera clipped to the rear-view mirror. And the car was going 60 mph. Moore's Law + a decade and a half of software research + a goood chunk of DARPA research money can achieve pretty amazing results.
ThinkGeek has them also. I have one and I found the diaphram made of plastic too loud. It makes kind of a loud "PHWAPP" when you do a full power blast. To fix this (in true geek fashion) I modified the 'zooka. I disassembled the rear assembly, and replaced the plastic diaphram with a piece cut from a black T-shirt I never wore. Now, the blast is a little less powerful, but it makes a quiet "fwop". Don't have any measuring devices, but it's at least 3x quiter. Much more effective and much more fun.
Just because something is "insanely" complicated, doesn't mean that the complications are impossible to manage through proper design and development. Just because most software sucks doesn't mean that all software needs to suck. Of course, given the quality of most commercial software, that's what we have come to expect.
I don't want more multimedia features, I want a better screen, LONGER battery life, and small and thin (Palm V). The latest trends are definitely away from the last two.
Yeah, the e-ink stuff looks really cool, low power and high contrast. Two things bother me about it: illumination and touch screen compatibility. From the looks of it, it looks like it'll need front lighting. Also, given it's reliance on electric field, I'd be worried about interference with the compatibility. The fact that application of the "ink" is straight-forward, however, enhances the useability factor alot. Perhaps a 'no-bezel' or all surface display PDA, where even the buttons are mini displays.
OCR+Translation is a cool idea. I think what you're looking for is probably closer to the realm of what PDA+PDA camera would be able to do. Digital cameras are too busy trying white-balance, filter, image enhance, etc. to think about OCI. Check out my journal entry for the first half of your idea. I'm adding your idea to it also, with references natch.
Darn my glasses. so, 1,200,000 km^2. Of course, if you improve efficiency to 50% you only need 24,000 km^2. And if you choose an area with better than average sunlight levels, say double, we're down to 12,000 km or 110 km * 110 km.