The upside to that, is now we know what failed during the exercise, so policies and training can focus on those failed areas. I didn't expect things to go flawlessly, and I think that anyone who did is not a realist. The fact of the matter is that they were not prepared for such an event. That's fine (It really is!). What's imperative is that now that we know that breakdowns occurred, and more importantly where they occurred, we can start to fix those breakdowns... It's the natural progress of trying to prepare for the unknown. We'll never truly know if we are prepared until we come under fire (or perhaps realize that we are under fire). But blatant failures such as this one are invaluable in figuring out what should and what shouldn't work in an actual event.
I agree. I do disagree with one line in the summary tho,
Ideally, legal and regulatory framework must allow companies participating on commons-based R&D to generate intangible assets for their contribution to successful projects
Why should it be limited to successful projects? Since this is open source, even a failed project can be hugely beneficial to society in terms of code, ideas or even just experience. Plus, who would declare success? Would a "successful project" be one that gets 1000 downloads a month? Or would it be a project that has a certain amount of community involvement? These questions (and others) are way to vague to justify that clause. Simply allow companies to deduct a portion of taxes for time donated to an open source project as a charitable donation. Sure, there will be abuse, but you can't stop abuse, you can only try to limit it at a huge expense... Plus, I think that the benefits will outweigh the negatives to such a system...
Well, when you submit a content to one of these sites, you explicitly say that you have copyright over it and grant the site permission to use it. How are they supposed to verify every single work to see if you really did have the right to grant permission in the first place? It's not only inefficient, but for a sizable amount of content it's all but impossible (If you recorded a video of kids playing in a park, how can they know for sure if A. You actually took the video, B. You had permission to take the video, and C. You have the permissions of each and every person and object videoed to be recorded?)... You bring up news services as an example because clear all the rights to a video. That's different for two reasons. First, they are typically seeking the content (as opposed to openly receiving the content) and as such the onus comes on them to verify the source... Secondly, they are claiming the source as credible (and as such, they need to do the research to make sure it's not fake, pirated, etc)...
I should be able to go to sleep at night knowing that my content is safe wherever it is being distributed, because that distributor has taken the necessary precautions to clear all rights and compensate performers, artists, and creators appropriately
Let's say that I copy a work that you made, and then say that I made it. Unless the people I release it to recognize that it's yours, how in the world are they supposed to know that it's not my content? And how in the world should they be liable for me claiming that it's my own? Not only is that a logistical nightmare (maybe even impossible), but it's also a childish thing to ask for. What it boils down to is that you (or your guardian in the case where you are not mentally disabled) are responsible for the protection of your own rights. If I constantly harass you (phone calls, emails, whatever so long as it's non-violent), it's up to you to get the police to do something about it. If I threaten to kill you, it's up to you to take measures to protect yourself (calling the cops, hiring a body guard, etc). Freedom is not free, and neither are your rights. You need to fight for them, and to think that someone else should be doing that for you is ignorance at best...
This is honestly the first time in my life that I am ashamed of my heritage. This is absolutely ridiculous and reeks of corruption and incompetence. I seriously hope that --for all of our sakes-- this gets overturned in the appeal. But seriously, when did "having evidence of a crime" make you guilty of a crime (So long as you didn't try to tamper or hide it)? Did I miss the memo?
Fair point about the PCM. I confused terms... And as far as power steering controlled by the BCM, that's what I was told by the mechanics that worked on it. Could it have been pure hogwash, sure. But they did flash the software on it and it did fix the problem, so it had to at least have some level of control (even if just switching power on and off)...
When you're depending on the computer to operate the vehicle, such a thing is essentially impossible, aside from having triple redundancy and allowing the redundant units to vote on the outcome... and if the problem was a software bug, they'd probably all vote wrong together anyway, unless competing teams developed the code.
Let's take the recent Toyota gas pedal debacle as an example. How hard would it be to add a second circuit that would "enable" the electronic throttle (a separate system, outside of the main computer) whenever the gas pedal was pressed? For cruise control, you'd have to add a secondary interlock, but it would still add a fail-safe path without requiring redundancy. The same goes for the power steering issue (You could add a relay that would only let the power steering motor turn on if the ECM was in a certain phase -- such as boot -- so that if the engine is already running, and the motor had stopped for some reason, it can't be restarted unless you restart the vehicle). I'm not saying they are perfect or flawless, or that it's even possible for all systems, just that an added layer of checks can't be bad. What's the worst thing that could happen? So your gas pedal stops working. Which is the greater failure, losing the ability to press on the gas pedal, or having the computer put it to full throttle without you being able to intervene (short of shifting into neutral or shutting off the engine)? There is absolutely no excuse for a critical system to ever be able to fail in a method that compromises safety from the failure of a single component/system. Brakes have dual circuits, so that even if you do rupture a brake line, you do still have one set of brakes left (either front or back). Steering is still a mechanical linkage, so if the power assist fails, you can still have a level of control over direction. Heck, even (most) automatic transmissions will fail in such a way as to cut power output to the wheels (they won't fail "in gear", thanks to the torque converter). Sure, you can always have multiple failures that will cause entire systems to go out, but the key point is that it should require multiple failures, not just a single one, to cause it...
Easy, press the brakes in a turn on a slippery (wet|snow) road. There is little to no weight in the back of the vehicle (except for the spare tire and battery), so all you need is the little weight shift that comes from a moderate application of the brakes. To tell you the truth, I prefer rear wheel drive, because although it does tend to fishtail more than a FWD vehicle, it usually happens slower and more predictably. Every FWD vehicle I've spun (when I get a new car, I usually take it to a parking lot in a fresh snow to learn how it reacts to abnormal conditions), once the rear loses traction, it snaps out. Meaning that if you aren't expecting a fishtail, you almost have no time to react, and will likely spin it. Compare that to a typical RWD (again, those that I have driven, which is a fair number of them) where when the rear end does come around, it's usually quite slow and very easy to predict, hence why I find them "safer". Sure, if you're not paying attention or don't know how to react, a FWD vehicle will typically be "safer"... I just don't like the thought of when my power wheels lose traction (because of snow, etc), my directional control becomes greatly diminished (contrast that to RWD, where you always have directional control so long as you are not spinning). JMHO...
Well, as a prior commenter pointed out, steering systems designed for power assist are typically a lot different than those designed for non-power steering (Yes, I have driven vehicles without it). Variables such as the length of idler arms, and the design of the steering box. Don't forget, most cars without power steering (at least in modern times) have rack and pinion steering. Most without power steering have a gearbox... However, there is another issue. Part of the "adaptive" steering is power feedback. As you rounded a corner, depending on the speed, it may apply slight pressure against the turn (to give you something to "turn against", and it works REALLY well if you're expecting it. The first time I merged onto a highway with the car I almost put it into a guardrail because I was used to a Chevy Blazer, which would get harder in turns, not easier). If that all of a sudden kicks on, it's easy to lose control.
Go to a car dealer. Look. Every car sold since 1996 (At least in the US, and I assume the rest of the world) today has at least an ECM (Engine Control Module) which is just a fancy name for a computer controlling the engine. That's what the government mandated OBD-2 program was (OBD == On Board Diagnostics). The number of cars that are completely computer controlled (drive by wire) is far lower, but higher than you'd think.
I had an '05 Chevy Cobalt that had "computer assisted" electromechanical power steering. Basically, what I found out from the dealer after the computer controlling it failed (and I lost all power steering) is that the computer (BCM, Body Control Module) takes inputs from the ABS system, Traction control (if equipped), speedometer, accelerometers and about a dozen other sensors and computes the way it thinks you want to be steering. Then it provides an "intelligent" boost in that direction. I must say, it worked really well in the snow and when fishtailing (it made if VERY difficult to over-correct and put it into a spin). But when it failed, I'd be in the middle of a curve on the highway when all power steering went out... Luckily they were smart enough to put a kill switch in to prevent it from coming back on while the car was moving (I could just imagine struggling through a corner when all of a sudden it came back)... It turns out that it was a software issue in the first place (they updated the software, and it never happened again). I got rid of the car a few years later for other, more significant reasons...
The benefits of computer control are good, but there needs to be intelligent fail-safes put in place to prevent disaster when something does go wrong (not if, when)...
As reader donberryman points out, you can even put Android onto some Windows Mobile phones, now.
I had android (1.0) running on my UNROOTED winmo phone (AT&T Tilt, HTC Tyan II) way back when the G1 JUST came out (early in 2008). You booted into WinMo, then ran the kickstart app, and it unloaded WinMo, and booted Android. It worked quite nice until I got my G1. Why is the fact that you can load a ROM onto a rooted device even news (especially considering the OS has been running on those devices for nearly 2 years)?
I didn't suggest that a panel of this size would even be possible. What I did show is the power density that it should be capable of (based on their calculations). Since their claimed power density is far higher than what's possible for sunlight, getting 1W out of a 1" strand is impossible with sunlight (but quite possible with other forms of "generated" light...
Sunlight striking the earth's surface has about 1.353 kW per square meter. Therefore,
1.35300 (kW / (m^2)) = 0.00087290148 kW / (in^2)
So their 1W estimation would be off by nearly 4 orders of magnitude from what's possible from sunlight (without lenses focusing the incoming light onto the strand). And that's assuming 100% efficiency. So this estimation of 1W of power (again, they never said sunlight) would likely be off by between 4 and 5 orders of magnitude when powered by raw sunlight...
Dammit. I didn't realize/. stripped out the micro symbol until after I posted. That should have been: 1 hair width is 100 micrometer... That will teach me to trust valid html to be sanitized correctly...
That kind of density is amazing. If they could produce it cheap enough, it could have applications on the macro scale as well. 1 hair width is 100 m, so you could fit about 2540 of these wires side by side in 1 inch. So your power generation would be around 2.5kW per square inch. That means a 4 foot by 8 foot panel would be able to generate around 11.7 mW (yes, that's megawatts). So, that means one of four things.
This is a gross exaggeration of its capabilities
This is exactly what they are capable of
They cannot be placed right next to each other (or at least efficiency suffers if you do)
This argument is used all the time, but it really doesn't apply. Leaving your door unlocked is not consent, implied or otherwise, for anyone to waltz on in.
True, but this was more akin to walking in to a library, and finding confidential documents in the general section right next to the Sunday newspaper (AKA, not behind any doors at all). All it took was knowing (or figuring out) where to look. There was no door here (if there was, it would have been in the form of a password or a DNS block (only allowing internal IP addresses), etc)...
Well, based on my understanding of it (I am by no means an expert), you can get true randomness from conventional hardware. Each transistor that's produced in a chip will be ever so slightly different from every other one (different number of atoms, different alignment of crystals, etc). So there is a source of entropy there that's quite large. The problem with this is two fold. The first problem is getting the entropy out of the manufactured chip (which is what these people are claiming to do). The other problem, is the type of entropy. The randomness is introduced at fabrication. So (again, if I understand correctly) each chip would essentially be a pre-seeded random number generator. Sure, you couldn't define a general "pattern" across all chips, but you very well may be able to for a single chip.
That got me thinking. If indeed you could pattern a chip, could that be a "foolproof" and "tamperproof" method of fingerprinting a device? So a USB thumb drive (or anything with a chip) could become a secure authentication device (since you'd be able to fingerprint a chip on the device, generating random numbers with that fingerprint would "prove" you have that exact chip, not just one with an identical state)?
A source of income coupled with backup power? Right now, if you supply power to the grid, you get paid for it. So if these things generate say 50KW per unit, and your needs are 100KW, then buy 3. You get N+1 redundancy, as well as generating income (so long as the power company would pay more than the Nat-Gas costs in the first place). The added income will help offset the cost of the units, and vastly reduce the break even time.
If you were using grid power, you'd want N+1 backup generators as well. Reliable generators aren't cheap. Plus, they cost $$$ to have even when they are NOT running (Maintenance, etc). If you could tackle 2 birds with one stone, why not do it? Now, it's likely not to be economical to rip out an existing backup system now (unless it needs replacement), but if you're building a new DC, it's possible that this could make financial sense...
At the end of the metastable state, the contents of the memory are purely random. The researchers' experiments with an array of flip-flop units show that for small arrays the extra layer makes the random number almost twenty times more 'random' than conventional methods
If it's "purely" random (as they put it), then how can you measure the difference between it and a "conventional" method? Wouldn't comparing a pseudo random source to "true" randomness be like comparing a finite number to infinity? In that you "know" it's more random, but it's impossible to quantify with a finite value (the twenty times qualifier)?
I have to second this recommendation. I used Spry (VPSLink's parent company. VPSLink's offerings are unmanaged, Spry's are managed) for a few years before migrating to a colocated server at another DC. The big advantage that they have (I'm not sure if other VPS providers offer it) is console access. You can login to their admin client, and get full console access to the server. Plus, their customer service is EXCELLENT. Tickets were responded to in minutes, and resolved in usually 10 to 15 minutes (the longest I had was 1 hour). Their backup system rocks. I had to do a full server restore once. I called them up, and 10 minutes later the restored system was up and running.
Plus, the other thing that blew me away, is that twice my server was moved to another physical VM box (for planned maintenance on the original hardware). A full WEEK before hand, I got an email saying the server was going to be moved, and that I shouldn't experience any downtime (it was a live transfer), but that I should be aware that at 01:00 PST (they are based in Seattle, Washington) they would be doing "something" with my server. Then, I got another email right before they started, and another right when they finished. I would use them again in a heartbeat if I didn't own my own servers now...
Because most of them tell you that they do back up. I would never go with a host that didn't do their own backups. With that said, I would never trust them to backup my data (Meaning, I would do it myself anyway). Having it both ways is far better than one (If their system isn't stupid, it should be FAR faster to restore from their backup than yours for a non-trivial amount of data, but you still have a copy for yourself in case something fails (or they screw up))...
Read the fine print when you sign up for an unlimited account. I've seen "unlimited bandwidth" accounts that were capped by the 1.5 mbps uplink to the internet. Sure, it's "as much as you can fit out over our pipe", but the question to ask is how big that pipe is. Same goes for disk space. Same goes for CPU time. Actually, CPU time is the bigger resource, since more often then not, you'll get an angry call saying "pay more, or we'll shut off your site for excessive cpu usage"... Unlimited is just an advertising term. There are so many other limitations on it, that the advantage is on the side of the hoster.
That being said, I much prefer a VPS (Managed or unmanaged). Sure, it'll cost a few $$$ more, but you get everything spelled out in advanced. You get an x% share of a y CPU box guaranteed. You get a z mbps uplink, and i gb of san based disk storage. Plus, with a VPS, if your site does grow, you can just increase the vm's specs up to 1 full server without having to migrate (assuming a decent host)...
First off, that's Pascal's Wager: http://en.wikipedia.org/wiki/Pascal's_Wager>
Secondly, it's a political argument, not a scientific one. The important distinction is that it's an invalid argument. You could use it to justify just about any falsehood. As an example, you could justify the extermination of an entire country. If the people of that country are "evil", then you would have saved the world. If they were not, then there's no harm done to you. So by the argument you proposed, we should exterminate the country. Of course, by moral and economic theories, we should most definitely not do it. (Again, I only bring this up to show how the wager is not a valid argument)...
Finally, i never said that we shouldn't do anything. What I said was the impetuous shouldn't be on just the USA to do it (And that we shouldn't implement radical changes until we know more about what's going on). It should be on everyone... Show me a moral and valid argument that one player should be responsible for the actions of many (the entire world in this case), and I'll concede, but as I can see it, such an argument does not exist...
Quite true. However, what about the latency? Even if you neglect pure bandwidth, the time it takes for a packet to make it from point A to B would be around 1/10 the time on a 1000baseTX link as it would on a 100baseTX link. That's not important for internet browsing (where the total latency of each router it needs to go through will dominate the experience), but it is important if working off of a file server (with lots of small files). At my home, I have a file server that holds my music collection. I then mount the directory (well, automatically mount it) from my desktop to play music. NFS over 100mbit was ok for streaming, but it took a few seconds to get a directory listing (each directory could have hundreds of subdirectories or files). When I upgraded to 1000mbit, the whole experience was so much smoother and faster. Sure, it's not the average use case, but it definitely is an important consideration...
If you ask your child to pick something up, and he/she responds that his/her brother/sister put it there, do you accept that answer? No. You asked him/her to pick it up. Non-causation doesn't imply non-responsibility.
That's a flawed analogy. The better one would be (based on what's happening): in a playground with 300 kids, picking one to always have to pick up every single thing that's on the ground, regardless of who put it there. Not to be responsible for making sure everything is picked, up, but making them do it for everyone. Which clearly is not fair...
However, let me pose a different angle to you: If our impact didn't cause it, can we fix it? If so, I propose that we should do so, on the grounds that continuing to live is a desirable thing. Even if this extends beyond our own lifespans, how do you want your children and grandchildren and great-grandchildren to live?
Reread my OP. I didn't say we shouldn't do anything. I said that the onus shouldn't be on the USA to do something. It should be on EVERYONE...
I disagree with the radical approach that many are dictating, because there's no proof that it is either A, Causing the problem, or B, that it will make a difference. I'm 100% for supporting research on both the climate front, and the technology front. To not do so would be irresponsible. To rush in and mandate "resolutions" that are expensive and potentially damaging to economies would be short sighted. Add that to the fact that we don't know (conclusively) if any of these mandates would actually make a difference, and the resolutions become down right irresponsible. How could reducing CO2 emissions be bad you say? Well, there's always a give and take. You never get something for nothing. What's the tradeoff? Is it less profitability? If so, we can live with that. If it's increased emissions of other byproducts, then who's to say that the CO2 is the "greater" enemy? Sure, it's the social pariah now, but that doesn't mean in the long term it's the true enemy...
Everyone freaks out because of HOW we're investing the money. To spend $100 billion on research is great, so long as it's not "loaded research". One reason people flipped out about the ethanol research proposal, was that it pre-selected corn based ethanol as the topic of research. This unnaturally focused efforts on one alternative rather than alternatives in general. We need to invest in ALL research, not just those that look promising to politicians and the media. Is some of the money going to be "wasted"? Possibly. But the potential is there (and is pretty significant) for something that's currently discounted now as "not possible" or "not economical" to mature to the point of being a viable alternative.
True, but the average consumer is also pretty happy at DSL level speeds (Or at least won't justify spending more for the 100mbps service).
What I see happening if this really does pass, is service levels for different pricing. Everyone gets a 100mbps connection, but the base plans only guarantee 1mbps (in theory 100 people or so per 100mbps uplink, in practice possibly thousands of customers). The more expensive plans would allocate more and more of the core bandwidth to each customer. That way, they could still afford a $20 per month plan, and still rape customers with $200 per month plans...
Slashdot Mirror
The upside to that, is now we know what failed during the exercise, so policies and training can focus on those failed areas. I didn't expect things to go flawlessly, and I think that anyone who did is not a realist. The fact of the matter is that they were not prepared for such an event. That's fine (It really is!). What's imperative is that now that we know that breakdowns occurred, and more importantly where they occurred, we can start to fix those breakdowns... It's the natural progress of trying to prepare for the unknown. We'll never truly know if we are prepared until we come under fire (or perhaps realize that we are under fire). But blatant failures such as this one are invaluable in figuring out what should and what shouldn't work in an actual event.
Why should it be limited to successful projects? Since this is open source, even a failed project can be hugely beneficial to society in terms of code, ideas or even just experience. Plus, who would declare success? Would a "successful project" be one that gets 1000 downloads a month? Or would it be a project that has a certain amount of community involvement? These questions (and others) are way to vague to justify that clause. Simply allow companies to deduct a portion of taxes for time donated to an open source project as a charitable donation. Sure, there will be abuse, but you can't stop abuse, you can only try to limit it at a huge expense... Plus, I think that the benefits will outweigh the negatives to such a system...
Let's say that I copy a work that you made, and then say that I made it. Unless the people I release it to recognize that it's yours, how in the world are they supposed to know that it's not my content? And how in the world should they be liable for me claiming that it's my own? Not only is that a logistical nightmare (maybe even impossible), but it's also a childish thing to ask for. What it boils down to is that you (or your guardian in the case where you are not mentally disabled) are responsible for the protection of your own rights. If I constantly harass you (phone calls, emails, whatever so long as it's non-violent), it's up to you to get the police to do something about it. If I threaten to kill you, it's up to you to take measures to protect yourself (calling the cops, hiring a body guard, etc). Freedom is not free, and neither are your rights. You need to fight for them, and to think that someone else should be doing that for you is ignorance at best...
This is honestly the first time in my life that I am ashamed of my heritage. This is absolutely ridiculous and reeks of corruption and incompetence. I seriously hope that --for all of our sakes-- this gets overturned in the appeal. But seriously, when did "having evidence of a crime" make you guilty of a crime (So long as you didn't try to tamper or hide it)? Did I miss the memo?
Let's take the recent Toyota gas pedal debacle as an example. How hard would it be to add a second circuit that would "enable" the electronic throttle (a separate system, outside of the main computer) whenever the gas pedal was pressed? For cruise control, you'd have to add a secondary interlock, but it would still add a fail-safe path without requiring redundancy. The same goes for the power steering issue (You could add a relay that would only let the power steering motor turn on if the ECM was in a certain phase -- such as boot -- so that if the engine is already running, and the motor had stopped for some reason, it can't be restarted unless you restart the vehicle). I'm not saying they are perfect or flawless, or that it's even possible for all systems, just that an added layer of checks can't be bad. What's the worst thing that could happen? So your gas pedal stops working. Which is the greater failure, losing the ability to press on the gas pedal, or having the computer put it to full throttle without you being able to intervene (short of shifting into neutral or shutting off the engine)? There is absolutely no excuse for a critical system to ever be able to fail in a method that compromises safety from the failure of a single component/system. Brakes have dual circuits, so that even if you do rupture a brake line, you do still have one set of brakes left (either front or back). Steering is still a mechanical linkage, so if the power assist fails, you can still have a level of control over direction. Heck, even (most) automatic transmissions will fail in such a way as to cut power output to the wheels (they won't fail "in gear", thanks to the torque converter). Sure, you can always have multiple failures that will cause entire systems to go out, but the key point is that it should require multiple failures, not just a single one, to cause it...
Easy, press the brakes in a turn on a slippery (wet|snow) road. There is little to no weight in the back of the vehicle (except for the spare tire and battery), so all you need is the little weight shift that comes from a moderate application of the brakes. To tell you the truth, I prefer rear wheel drive, because although it does tend to fishtail more than a FWD vehicle, it usually happens slower and more predictably. Every FWD vehicle I've spun (when I get a new car, I usually take it to a parking lot in a fresh snow to learn how it reacts to abnormal conditions), once the rear loses traction, it snaps out. Meaning that if you aren't expecting a fishtail, you almost have no time to react, and will likely spin it. Compare that to a typical RWD (again, those that I have driven, which is a fair number of them) where when the rear end does come around, it's usually quite slow and very easy to predict, hence why I find them "safer". Sure, if you're not paying attention or don't know how to react, a FWD vehicle will typically be "safer"... I just don't like the thought of when my power wheels lose traction (because of snow, etc), my directional control becomes greatly diminished (contrast that to RWD, where you always have directional control so long as you are not spinning). JMHO...
Well, as a prior commenter pointed out, steering systems designed for power assist are typically a lot different than those designed for non-power steering (Yes, I have driven vehicles without it). Variables such as the length of idler arms, and the design of the steering box. Don't forget, most cars without power steering (at least in modern times) have rack and pinion steering. Most without power steering have a gearbox... However, there is another issue. Part of the "adaptive" steering is power feedback. As you rounded a corner, depending on the speed, it may apply slight pressure against the turn (to give you something to "turn against", and it works REALLY well if you're expecting it. The first time I merged onto a highway with the car I almost put it into a guardrail because I was used to a Chevy Blazer, which would get harder in turns, not easier). If that all of a sudden kicks on, it's easy to lose control.
Go to a car dealer. Look. Every car sold since 1996 (At least in the US, and I assume the rest of the world) today has at least an ECM (Engine Control Module) which is just a fancy name for a computer controlling the engine. That's what the government mandated OBD-2 program was (OBD == On Board Diagnostics). The number of cars that are completely computer controlled (drive by wire) is far lower, but higher than you'd think.
I had an '05 Chevy Cobalt that had "computer assisted" electromechanical power steering. Basically, what I found out from the dealer after the computer controlling it failed (and I lost all power steering) is that the computer (BCM, Body Control Module) takes inputs from the ABS system, Traction control (if equipped), speedometer, accelerometers and about a dozen other sensors and computes the way it thinks you want to be steering. Then it provides an "intelligent" boost in that direction. I must say, it worked really well in the snow and when fishtailing (it made if VERY difficult to over-correct and put it into a spin). But when it failed, I'd be in the middle of a curve on the highway when all power steering went out... Luckily they were smart enough to put a kill switch in to prevent it from coming back on while the car was moving (I could just imagine struggling through a corner when all of a sudden it came back)... It turns out that it was a software issue in the first place (they updated the software, and it never happened again). I got rid of the car a few years later for other, more significant reasons...
The benefits of computer control are good, but there needs to be intelligent fail-safes put in place to prevent disaster when something does go wrong (not if, when)...
I had android (1.0) running on my UNROOTED winmo phone (AT&T Tilt, HTC Tyan II) way back when the G1 JUST came out (early in 2008). You booted into WinMo, then ran the kickstart app, and it unloaded WinMo, and booted Android. It worked quite nice until I got my G1. Why is the fact that you can load a ROM onto a rooted device even news (especially considering the OS has been running on those devices for nearly 2 years)?
Sunlight striking the earth's surface has about 1.353 kW per square meter. Therefore,
So their 1W estimation would be off by nearly 4 orders of magnitude from what's possible from sunlight (without lenses focusing the incoming light onto the strand). And that's assuming 100% efficiency. So this estimation of 1W of power (again, they never said sunlight) would likely be off by between 4 and 5 orders of magnitude when powered by raw sunlight...
Dammit. I didn't realize /. stripped out the micro symbol until after I posted. That should have been: 1 hair width is 100 micrometer... That will teach me to trust valid html to be sanitized correctly...
Still, it's a cool concept...
True, but this was more akin to walking in to a library, and finding confidential documents in the general section right next to the Sunday newspaper (AKA, not behind any doors at all). All it took was knowing (or figuring out) where to look. There was no door here (if there was, it would have been in the form of a password or a DNS block (only allowing internal IP addresses), etc)...
Well, based on my understanding of it (I am by no means an expert), you can get true randomness from conventional hardware. Each transistor that's produced in a chip will be ever so slightly different from every other one (different number of atoms, different alignment of crystals, etc). So there is a source of entropy there that's quite large. The problem with this is two fold. The first problem is getting the entropy out of the manufactured chip (which is what these people are claiming to do). The other problem, is the type of entropy. The randomness is introduced at fabrication. So (again, if I understand correctly) each chip would essentially be a pre-seeded random number generator. Sure, you couldn't define a general "pattern" across all chips, but you very well may be able to for a single chip.
That got me thinking. If indeed you could pattern a chip, could that be a "foolproof" and "tamperproof" method of fingerprinting a device? So a USB thumb drive (or anything with a chip) could become a secure authentication device (since you'd be able to fingerprint a chip on the device, generating random numbers with that fingerprint would "prove" you have that exact chip, not just one with an identical state)?
A source of income coupled with backup power? Right now, if you supply power to the grid, you get paid for it. So if these things generate say 50KW per unit, and your needs are 100KW, then buy 3. You get N+1 redundancy, as well as generating income (so long as the power company would pay more than the Nat-Gas costs in the first place). The added income will help offset the cost of the units, and vastly reduce the break even time.
If you were using grid power, you'd want N+1 backup generators as well. Reliable generators aren't cheap. Plus, they cost $$$ to have even when they are NOT running (Maintenance, etc). If you could tackle 2 birds with one stone, why not do it? Now, it's likely not to be economical to rip out an existing backup system now (unless it needs replacement), but if you're building a new DC, it's possible that this could make financial sense...
If it's "purely" random (as they put it), then how can you measure the difference between it and a "conventional" method? Wouldn't comparing a pseudo random source to "true" randomness be like comparing a finite number to infinity? In that you "know" it's more random, but it's impossible to quantify with a finite value (the twenty times qualifier)?
Proof Read Much?
I have to second this recommendation. I used Spry (VPSLink's parent company. VPSLink's offerings are unmanaged, Spry's are managed) for a few years before migrating to a colocated server at another DC. The big advantage that they have (I'm not sure if other VPS providers offer it) is console access. You can login to their admin client, and get full console access to the server. Plus, their customer service is EXCELLENT. Tickets were responded to in minutes, and resolved in usually 10 to 15 minutes (the longest I had was 1 hour). Their backup system rocks. I had to do a full server restore once. I called them up, and 10 minutes later the restored system was up and running.
Plus, the other thing that blew me away, is that twice my server was moved to another physical VM box (for planned maintenance on the original hardware). A full WEEK before hand, I got an email saying the server was going to be moved, and that I shouldn't experience any downtime (it was a live transfer), but that I should be aware that at 01:00 PST (they are based in Seattle, Washington) they would be doing "something" with my server. Then, I got another email right before they started, and another right when they finished. I would use them again in a heartbeat if I didn't own my own servers now...
Because most of them tell you that they do back up. I would never go with a host that didn't do their own backups. With that said, I would never trust them to backup my data (Meaning, I would do it myself anyway). Having it both ways is far better than one (If their system isn't stupid, it should be FAR faster to restore from their backup than yours for a non-trivial amount of data, but you still have a copy for yourself in case something fails (or they screw up))...
Read the fine print when you sign up for an unlimited account. I've seen "unlimited bandwidth" accounts that were capped by the 1.5 mbps uplink to the internet. Sure, it's "as much as you can fit out over our pipe", but the question to ask is how big that pipe is. Same goes for disk space. Same goes for CPU time. Actually, CPU time is the bigger resource, since more often then not, you'll get an angry call saying "pay more, or we'll shut off your site for excessive cpu usage"... Unlimited is just an advertising term. There are so many other limitations on it, that the advantage is on the side of the hoster.
That being said, I much prefer a VPS (Managed or unmanaged). Sure, it'll cost a few $$$ more, but you get everything spelled out in advanced. You get an x% share of a y CPU box guaranteed. You get a z mbps uplink, and i gb of san based disk storage. Plus, with a VPS, if your site does grow, you can just increase the vm's specs up to 1 full server without having to migrate (assuming a decent host)...
First off, that's Pascal's Wager: http://en.wikipedia.org/wiki/Pascal's_Wager>
Secondly, it's a political argument, not a scientific one. The important distinction is that it's an invalid argument. You could use it to justify just about any falsehood. As an example, you could justify the extermination of an entire country. If the people of that country are "evil", then you would have saved the world. If they were not, then there's no harm done to you. So by the argument you proposed, we should exterminate the country. Of course, by moral and economic theories, we should most definitely not do it. (Again, I only bring this up to show how the wager is not a valid argument)...
Finally, i never said that we shouldn't do anything. What I said was the impetuous shouldn't be on just the USA to do it (And that we shouldn't implement radical changes until we know more about what's going on). It should be on everyone... Show me a moral and valid argument that one player should be responsible for the actions of many (the entire world in this case), and I'll concede, but as I can see it, such an argument does not exist...
Quite true. However, what about the latency? Even if you neglect pure bandwidth, the time it takes for a packet to make it from point A to B would be around 1/10 the time on a 1000baseTX link as it would on a 100baseTX link. That's not important for internet browsing (where the total latency of each router it needs to go through will dominate the experience), but it is important if working off of a file server (with lots of small files). At my home, I have a file server that holds my music collection. I then mount the directory (well, automatically mount it) from my desktop to play music. NFS over 100mbit was ok for streaming, but it took a few seconds to get a directory listing (each directory could have hundreds of subdirectories or files). When I upgraded to 1000mbit, the whole experience was so much smoother and faster. Sure, it's not the average use case, but it definitely is an important consideration...
That's a flawed analogy. The better one would be (based on what's happening): in a playground with 300 kids, picking one to always have to pick up every single thing that's on the ground, regardless of who put it there. Not to be responsible for making sure everything is picked, up, but making them do it for everyone. Which clearly is not fair...
Reread my OP. I didn't say we shouldn't do anything. I said that the onus shouldn't be on the USA to do something. It should be on EVERYONE...
I disagree with the radical approach that many are dictating, because there's no proof that it is either A, Causing the problem, or B, that it will make a difference. I'm 100% for supporting research on both the climate front, and the technology front. To not do so would be irresponsible. To rush in and mandate "resolutions" that are expensive and potentially damaging to economies would be short sighted. Add that to the fact that we don't know (conclusively) if any of these mandates would actually make a difference, and the resolutions become down right irresponsible. How could reducing CO2 emissions be bad you say? Well, there's always a give and take. You never get something for nothing. What's the tradeoff? Is it less profitability? If so, we can live with that. If it's increased emissions of other byproducts, then who's to say that the CO2 is the "greater" enemy? Sure, it's the social pariah now, but that doesn't mean in the long term it's the true enemy...
Everyone freaks out because of HOW we're investing the money. To spend $100 billion on research is great, so long as it's not "loaded research". One reason people flipped out about the ethanol research proposal, was that it pre-selected corn based ethanol as the topic of research. This unnaturally focused efforts on one alternative rather than alternatives in general. We need to invest in ALL research, not just those that look promising to politicians and the media. Is some of the money going to be "wasted"? Possibly. But the potential is there (and is pretty significant) for something that's currently discounted now as "not possible" or "not economical" to mature to the point of being a viable alternative.
True, but the average consumer is also pretty happy at DSL level speeds (Or at least won't justify spending more for the 100mbps service).
What I see happening if this really does pass, is service levels for different pricing. Everyone gets a 100mbps connection, but the base plans only guarantee 1mbps (in theory 100 people or so per 100mbps uplink, in practice possibly thousands of customers). The more expensive plans would allocate more and more of the core bandwidth to each customer. That way, they could still afford a $20 per month plan, and still rape customers with $200 per month plans...