The attorney-client privilege is a different beast than the government. The attorney is my agent, he's a servant, for a fee. He does not serve the government. Even in crime. Moreover I can act as my own attorney. I can't act as my own public notary, or clergy, and they don't answer to me in a way an attorney does, but, if they privately know me, we have entangled lives vested in each other's, in the same community, then they answer to me at least as much as they answer to the government, even to the point of willing to hide stuff for me. They are a less strict class for holding confidential information compared to attorneys, but they are an order of magnitude better than the government, because they can be local, and adhere to distributedness and noncentralization principles. Moreover, even in the strictest case, if I don't empower anyone with the power of attorney, nobody would have the right to store my DNA information for extended periods of time. At least legally. Just like possession of drugs is illegal. It still happens, but does that mean we should make it legal? Also, of course, who can you trust these days? Having anyone else other than you holding any of your secrets is, well, naive? Attorney won't tell? The priest won't tell after a confession? Come on now, who are you kidding man? Still, trust happens. There are different degrees of trust, different kinds of trust, and the issue is whether to place trust in an overwhelming power entity providing it with even more power, or distributing and dispersing the trust, and therefore eliminating the concentration of power, well that's something I'm all for. Power leads to abuse, and abuse to even more power. It's the nature of power to do so, and you have the power to check the power of powers trying to overpower you and make your life into hell by running it for you, or anyone else's.
The lab should not "help out", people should be respected and allowed to chose for themselves. You cite the worst kind of medical trespassing there possibly is on people's rights. We only make life changing decisions for others when they are deemed too imbecile to make it for themselves. Similar to validity of wills. It's about respect, about human dignity.
From the wikipedia's human genome page, a person's DNA is about 3 billion base pairs long, which is about 750 MB of data, about a CD. Quite a bit of data to be pulling across a network, quite a bit to locally search, but a distributed query, like distributed.net's math solving trials, could be workable. Hence another reason for a distributed DNA database, instead of a centralized one. This argument won't be valid in a few years if computing storage and speed keeps advancing, but the privacy arguments may still stand. Though if only diff's are stored per person, and there is a categorization based on diff's, which is probably already done, centralized storage might be very efficient.
Sometimes when executing a will, relatives cannot be easily found. They might be hiding, or be in witness protection. A distributed query looking for relatives based on DNA, could allow the local databases to compare their own data, and let the person know, hey, someone is looking for you in trying to execute a will. Now this can be used as a great trap to bait and lure anybody out of privacy, but it's still better, because first you know someone is looking for you, second, you don't have to answer, third you can send an agent like your attorney who doesn't have to release your identity. All this cannot be handled through the phone book, because sometimes you just can't find somebody through the phone book. Because they prefer to have an unlisted number. So how can you find people who prefer being private? Send them a question and let them respond if they want to. Blood is thicker than water they say, at least relatives should be able to get connected, with privacy and security present. A national open database would allow all kinds of relatives to harass you based on relativity. It should be your choice whether you want to reply to a search and initiate contact, it should take two to tango.
Or imagine a tool like this in the hands of Hitler. That's what's most wrong with this whole thing, the power it gives to someone over other people. Anonymity and privacy, being shielded and safe from some paternalistic overseeing power entity should be a right. Power should be given to government only as much as necessary. Such databases should be in the private sphere, held by someone like the clergy with the "seal of the confessional", or by attorneys in a fashion similar to attorney-client privilege. We need a system of internetized public notaries/attorneys holding confidential private information, regarding issues of identity, privacy, will/testament etc. Some kind of distributed database with confidentiality barriers. Queries run against it, and people having to give consent before answers are released. Government access to it should be absolutely limited, with very strict rights and needing a warrant. In fact no central databases should exist, but some kind of public key/private key system published by many attorneys, or public notaries, from which matches can be found, such as relatives, or criminals, without revealing identity. To a posted public key search the other local small databases should react, and if they find themselves to be a match to a request, they should ask the owner of that DNA whether he would like to reveal his identity to the query. Of course you would find no criminals this way, because who would confess, yes, I'm the one. But that's exactly what the 5th amendment is about, it's not about making law enforcement easy, to the contrary, protecting individual liberty at the price of "security", or "ease of law enforcement."
Compiling databases about everybody in the name of security - well, you know what the founding fathers said: those who sacrifice liberty for security will get neither. A social security number databases tagging everyone for tax collection purposes should suffice. Fingerprint databases feel already too private, but all they reveal is your physical presence at a location, if you didn't wear gloves. And they are harder to plant than dna samples of hair, blood, etc. Fingerprints in the name of law enforcement, I can agree to that, because they don't contain much else about you. DNA, that's a whole other beast than a fingerprint. Occasional DNA tests by police, comparing suspects to locally found evidence could be OK, with the data returned to the owners, or owner's assigned attorney/public notary after the completion of the trial. It should not be allowed to be archived, even if it means a whole lot of wasted work, and having to redo everything over and over. Or who do you trust?
You should not feel more secure because of the databases compiled in the name of security, if anything, fear some coup, some power takeover at the top by some mad men. Then imagine what power they will have over you to deride you and ride you to hell and back, simply because they feel like it. And you're at fault, who previously sacrificed your privacy and anonymity in the name of security. What security? If you have many small localized/secret databases exchanging information only as needed, in case of a power coup at the top, well, my neighborhood notary public might be willing to hide my DNA from the new government, just like some people were willing to hide jews during the Nazi regime. People being able to disobey laws is a prerequisite of liberty. Where is the guarantee that we will never have another criminal regime like that in power, coming up with laws that are criminal. The time to defend is now. The price of liberty is eternal vigilance. Law enforcement is important, but so is the 5th amendment, which is more important than law enforcement itself. So how can you hide your DNA, how can you stay anonymous, retain your privacy in this world? Is that even possible? No. But at least we could have a makebelieve, pretend to respect each others rights to privacy world. Anytime you give blood, or a hair sample, for a simple thing as a drug test, others have your DNA, if that sample is tied to you in
As long as the product performs somewhere near as well as the old stuff, and it's patentable, then there is money to be made. We just have to find sufficient fault with the old stuff, and bad mouth it enough to start making money. Wikipedia says that as with the other lanthanides, samarium compounds are of low to moderate toxicity, although their toxicity has not been investigated in detail. An MSDS sheet where you can put toxicity N/A, no data available sounds better than one where you know it's toxic, because at least with an unknown there is a chance that it's not toxic. There is money to be made with the patent, and money saved by not having regulations to deal with. Regulations regulate know toxic materials, not unknowns.
If we can go to outer space, and figure out self sustaining local ecosystem bubbles, dependent only on solar radiation to sustain its inner lives, just like our Earth does, but on a smaller scale, so if we create many such little bubbles, then at least in one of them, we might make it. Unless something smarter than us evolves in one of the other bubbles, and goes on a hunt for us. Then we're in trouble. But as long as we know of nothing smarter than us, we're in pretty good shape, it's only a matter of will, not a matter of can't. Space bubbles, space stations, that, if needed, can be brought down to the surface too, bubbles with technology to be completely self dependent, and completely enclosed and sealed, so even in the case of some massive die off of any and all species on the planet because of some genetic engineered virus by some malicious hackers, the virus could not get inside because it would be hermetically sealed, even on the Earth's surface. Biosphere 2 was a failure of an experiment. Mostly because of lack of will, because we felt we didn't have to. When you're in outer space, you have to. The issue with many mini bubbles is a divergent evolution, different species evolving in different bubbles over millions of years - but I don't mind having kangaroos in Australia, and lamas in South America, I like diversity.
If we go into outer space, far enough from a nova issue, we might even survive our Sun going nova. It's all a matter of technicality. Such as incubating some sperm and eggs and some robotic nannies, and sending them off to Alpha Centauri, where on arrival, the test tube babies are born and robotic nannies that have tits that feel like human tits nurture them, and smell human, and teach them all the traditions and knowledge they need to know in order to sustain a large local space station living off of that stars radiant energy. Taking along other lifeforms too, of course, how else can you grow food. Then instead of all bags in one basket, dependent on the Sun, you'd have at least two baskets, Alpha Centauri, and the Sun. And when they see this Sun go nova for whatever reason, they can say "bummer!" So on and so forth. It's all just a matter of technicality, from an engineers point of view.
Worker bees will sacrifice like suicide bombers. You and I are collections of cells, multicellular organisms. Our cells sometimes do apoptosis, programmed cell death, so that the organism, the group can survive. By meme you mean the group, the organism. Suicide is a social, group phenomenon, evolutionarily speaking. Otherwise it wouldn't have made it thus far. Of course there are situations when things just go haywire, such as a bad mutation in a cell causes it to go nuts and do "nonprogrammed" cell death, more exactly a nonproperly programmed one. Or someone get mentally messed up by some accident and make the wrong decisions, and commit some errant suicide. But even "sane" worker bees, and "sane" cells know how to do suicide, in the name of survival, not all suicide is things gone haywire, some suicide is "proper". The real question is which category suicide bombers fall into - group protecting category, or nongroupprotecting insanity. The other question is, us, as cells in a larger group, organism, or meme, do we protect the meme that has evolved as a layer on top of us, generally called cultures, or religions - do we trust the meme? Do we people die for a culture, a tradition or a religion, or do we let the cultures die, so we the people can survive? What if our cells in our bodies decided that they disagree on the current "culture" and suddenly transformed into some other organism they find it betters their prospects of survival. Picture a monkey suddenly turning into a bird because its cells voted so, and the bird form got 51% of the vote, 48% wanted to stay monkey shaped, and 1% was undecided? We humans, can do such sudden feats with our groups, and they sometimes end up being the "wrong?" decision, maybe.
As long as there is a doctor and the patient has been informed about what's going on, and then the go ahead is given to the robot by the doctor who obtained consent, yeah. I'm very weary of any kind of automatic surgery device. What if it malfuctions? Imagine an automated lobotomy robot: come here, Mr. Patient, I have diagnosed you with a rare brain disorder!
Solar power might be built up from silicon on the moon, but the cost of local mineral processing and purification is enormous, especially considering lack of water. That's what still would happen in the long run, from locally obtained silicon, but to get things started, instead of carrying up silicon panels from the Earth, it's a lot cheaper to carry up a compact and relatively easy to manage nuclear power device and get a lot of juice. It's a lot easier to manage a few ten cubic metres of a submarine-like power plant, than a few hundred acres of solar panels. The shipping cost into space per megawatt of power is probably 100x less for the nuclear plant, except it runs out of fuel in a few years, and by that time you hope you're able to switch over to locally created solar panels or even locally mined nuclear fuel. Or carry up more fuel. A satellite usually functions on solar power, because it doesn't need that much power. But on the moon, if you want to move anything, to get anything as simple as a bulldozer going, you need a whole lot of juice. Solar is not enough, at least not on small scale. You're talking a square mile of panels vs. 2 submarine engines. That's a whole lot of panels, and it's not so easily managed.
The only issue on the moon compared to Earth is cooling. Nuclear needs a whole lot of cooling power because it's so energy dense. Submarines and pretty much all nuclear power plants rely on water cooling, something that's not available on the moon. Also, considering the vacuum, even air cooling with fins, like in computers, is not an option either. Geothermal cooling is one option, but probably has limited capacity, unless you drill very long holes. The safest bet is ultrahigh temperature reactors, and radiative cooling, through black body radiation into outer space. That way you can always relocate the plant too, if it rolls on wheels, carry it to where you need power, follow the bulldozer's power cord around as opposed to being tied to a geothermal hole.
The spare plane could be provided by the FAA itself, and shared by the competitors, as needed. One scenario is where it's available for a fee, the fee being higher than a normal plane's cost, so only in need would it be used. However if the competitors don't care about canceled flights and losing customers and only care about saving money, the FAA plane would just sit there with no use, nobody touching it. Another scenario is where it would be freely available to one of the competitors under certain rules, such as given enough system load, and some fairness and non-cheating principles, if you can devise some. Now the airplane companies would hire people to figure out how to get the free plane, and this would create very good understanding of the underlying upsets in the system. The real problem is how to come up with rules that keep getting the free plane fair. In a sense, you could have mathematicians on one side update rules, as the free plane hungry people figure out the loopholes. Formula 1 has such measures, where after an invention shows up, such as turbochargers, aerodynamic fins, superhigh rpms, every few years the design problem is made more difficult and more sane by putting limits. For instance the maximum rpm allowed is 19000 in an engine, because otherwise competitors would go to 40000 rpm or some crazy value like that, and research advances in that direction have no practical value in regular 2000 rpm automobiles, but some recent bikes do function at 9000 rpm, such as the Kawasaki Ninja 250R. Similar limits have been put on aerodynamic shape sizes that create downforce and excessive "safe speeds" (such as front funnel shapes of the early 90's that worked with venturi effects under the body.) You could set up an opposite interest group between regulating mathematicians, and the competitors competing for the free plane, minds pitted against each other, just like in the legal system, you have prosecution and defense, and from trying to outdo the opposition comes something valuable.
There is probably some equilibrium spareness that operators could rely on to ease some of the avalanches. If there are gate holds in Detroit, that cause an avalanche in the schedules - customers seeing expected arrival times to jump back and forth half an hour as the computer updates the possibilities - having one spare plane in say St. Louis, or Atlanta, or some really busy place, even with no spare gates, and letting the system know about it, could cut one order of magnitude of the problems, maybe even justify the cost of the spare plane. Just how much lost value can you assign to pissing off the customer? If you lose no profit over that, and pissing them off to the max is the most profitable way to do business, then it's going to be hard to justify the spares, on competitive business arguments. Maybe human decency arguments? FAA regulation might demand some minimum level of spareness? A lot of laws are like that, introduce human decency into a picture. This could be a great subject for some mathematicians to get funding for, to come up with FAA rules for optimum balance. You basically have too many equations and too much math that database people aren't smart enough to deal with, but linear algebra researchers could get off on. It's a brain tickle with very serious practical value.
If you're overwriting many times, overwriting with random data, instead of 0's, would make even the most sensitive electronic detection moot, unless they can predict what random values were used. Now that's something that's really impossible. Deciphering hieroglyphs was difficult enough, let alone a 10000x overwrite with a random pattern. Not only would they have to measure minute differences in signal levels, but they'd have to correctly predict how many 0's and 1's were used for each spot, and in what sequence, in that random pattern. Still, there is nothing like heating your harddisk to a temperature over the Curie temperature, where random thermal motion takes care of the demagnetization for you.
Why is this even called "The Great Zero Challenge". How about "Last chance at $40 for reading folders from a dd'd harddrive." Is it possible? The answer is maybe, so you should overwrite the same areas a few times. Reading residual magnetization left over after a rewrite probably takes special equipment with sensitivity and positioning beyond the regular head included in the disk. You're basically talking differences between zeros - one of the zeros is not the same as another zero: this zero used to be a zero, that zero used to be a one. You'd need access to the signal levels coming from the head before the analog to digital conversion takes place in the harddrive. Does this conversion take place in one shot, from volts or amps to 0 and 1, or is there an intermediate digitization, say getting a voltage level value between 0 to 255, and then later a calibration in the factory sets the levels of 0 to say 0 to 40, and 1 being 200 to 255, 40 to 200 indeterminate or read error, these values being freely tunable for each different disk? Then if you could get access to that pre-zero-or-one 0 to 255 digitized signal, then you could say that zeros that are reading 10 used to be zeros, and zeros that are reading 30, used to be 1, and subsequent overwrites would get the surface closer to 10 and 19, then 10 and 15, and 10 and 11, until the two different values would be so close, 10 and 10, that you'd need a separate equipment to sense and digitize differences between 10.0000 and 10.0002, and even then the signal to noise ratio might be so small that it'd be impossible no matter how good your equipment is. Is the data still available to somebody with 40 bux? Nope. Is the data still theoretically available after a single dd with 0? For a US or Chinese spy agency willing to spend a few million bucks on the question, the answer is: most likely. But 10000 dd's would most likely make it unavailable to them too. What's the actual number of overwrites needed? That needs real data, at least from someone with a semi-cleanroom hobby shop, with a superprecise oscilloscope reading squarewaves off a surface. Otherwise we're just guessing. One dd might be enough, or 10 might be, or even 10000 may not be enough. So how secure does your data erasing have to be, as long as we're guessing?
Upward wind tunnels, or really tall chimneys are expensive contraptions meant to generate wind from solar heat. Using naturally occurring wind generated by solar heat that blows sideways in most places and is available freely is generally thought to be cheaper.
I think their reason is not environmental, but to quell the hype, and the waste of money and resources. You really have to be watching your money like a hawk. Solar is like the only viable renewable energy for the long run - wind is better, but only certain places have sufficient wind/can tolerate huge height towers -, but you can make some really bad mistakes by spending too much. Like the military did by buying a 100 million dollar solar plant, that should have cost 10 million if profitability is anywhere in sight. Just because you can get power from solar, there is still a price you should spend, and not overpay 10-100x.
Office is Simonyi's masterpiece, even though the clunky or not so clunky(opinions vary) Hungarian notation is all over win32.
When Simonyi left in 2002, you could tell it was over a disagreement in principle. Out of the original people at MS, he was pretty much the only technical brain, Gates was a Basic fan, and Ballmer a business bully, and who else? Simonyi was wrong about nonpreemptive multitasking, because the performance penalty isn't such a huge price to pay, compared to the benefit in stability you get. Simonyi was an all out performance freak. He went overboard, and his point of view was completely avoided, because he was wrong before. Behold Windows Vista as a result of not caring about performance. If anything, nonpreemptive multitasking might be the future in superperformance computing, where it's a compiler enforced rule that a program relinquishes control to the next one, instead of the operating system butting in, unloading the processor state, storing the stuff in a temporary space, getting the other program loaded, fiddling around with it, then repeat, unload, store. Nonpreemptive multitasking could run circles around such things, because it can properly decide when the impact of interruption is mininal, and performance stuff like this really mattered in the days of the 286, but by P3 days not so much, because of the difficulty to manage code. I still believe performance, not ease of programming, is the highest principle in computing. It's what separates the professionals from the amateurs and hobbyists. Getting more abstract to sound more professional and paying huge performance penalties for it, as in java, dotnet, even c++, doesn't seem very professional to me. As a professional computer programmer you have to account for how you spend the computer cycles your customer entrusted you with. If nothing else, because of the energy crisis in the world. Each wasted clock cycle costs oil, costs blood, costs lives. I believe there has to be a safe way to write high performance code. It's hard, but there has to be a way. There has to be a safe way to write C programs without buffer overflows. Dump all performance because of a few problems? Are buffer overflows really not fixable? And if anything, there has to be a safe way to write nonpreemptively multitasking and cooperating programs, that give very high performance. It's difficult, yeah. It's next to impossible to do, but not impossible, and preemptive multitasking is just soooooo much easier, whew, what a breath of fresh air, all those crap programs can't bring the system down. Still, you could have a pretend-nonpreemptive multitasking system where the programs consciously pass on the you're next bar, and the OS only butts in as a referee, when one of the programs doesn't behave. All it has to do is observe, and time, watch the rules, and give some programs a yellow card, a red card, like in soccer. Timeout on the bench. Realtime programming environments, with guaranteed response time could have an easy rule, easy way to tell if somebody doesn't relinquish in time. That's all it would care about, no complex rule of expected behavior like in most sports. History proves Simonyi was wrong about nonpreemptive multitasking, but I disagree with history and agree with him on that one. But I personally hate Hungarian notation and I'm Hungarian by birth. I prefer $, %, sigils, they are intuitive and don't need to be read, my brain switches subconsciously into that mode by seeing sigils, but hungarian notation prefixes need effort to read, and there is so many of them, I can't keep track, it takes brain wearing conscious effort to digest them, and they get me quickly fatigued. I guess if you stare at them for a lifetime, they become to you like sigils, subconscious. As a bilingual person I know that a language cannot be spoken unless its subconsciously processed, if you need to exert significant conscious effort, you merely kludge around and you only speak at a very slow rate. You have to think in the language you speak when you speak it, literal translations don'
And maybe one of the teams may decide to release Windows Shark - Cold Blooded, Lean and Mean(TM) and go back to the Windows 95 simplicity/speed/tight code, based on stripped version of Win2K code wise, but add glitter looks wise, with a small registry, and add small improvements like bigger hard drive support and bluetooth and stuff like that as it arrives to the market. Maybe they need just a small spin off company, to support the "legacy" with very few very smart people, who will babysit and support win2k/office97/vb 6/sql server 2000, and provide updates only as needed, as it makes sense. VB6 could use a hell of a lot of updates, but it's still the easiest thing to use to bang up a business program in a matter of minutes that actually works without making the computer crawl to its knees. Usability wise that setup from that point in time could last the test of time. Windows Raptor for desktop and gamerz, Windows Shark for business, Windows Predator for servers hosting sql server 2000, nothing newer. Remember when Tom's hardware guide use to post frame rates achieved by different games? People love speed, people love fast cars. That's what they need to put the spice back into computing again, and it shouldn't cost much to sustain. It's like when a plant needs a trimming job because it overgrew. Or a new shoot. There is nothing wrong with getting a database transaction that takes nanoseconds instead of microseconds. Or mouseclick reaction rates that go to microseconds from milliseconds. The end user doesn't need it, doesn't notice it? Yeah, like a Hayabusa's or Lamborghini's top speed is not needed by everyday commuters. Where's the fun, man? XP-experience even your grandma can appreciate. Granda doesn't appreciate the computer, she appreciates her knitting kit. Vista. Vista of what? Of where we're heading? How we're growing slower, older? Where do you want to go today? I want to go back to the days when computing used to rock my brains, when it was so much fun. I want real time computing, with guaranteed reaction time. I want something as easy to program as basic, easier to read than a good book, and it works faster than manually tuned C/assembler. Something almost as plain as english, something I can drop into my robot and make it frigging work without having to go through the steps of abstractualizing the conceptualizations of how the different frameworks interact from an objective perspective inside the whole thing. I just don't give a crap, I don't want to care about the latest hype. Gimme something I can get my hands on, work, form into something, something that's easy already, something that makes my life even easier. Something that ticks fast, and tells me how fast. Nanoseconds I love. Microseconds I can tolerate. Don't tell me milliseconds, that's so 1999. And where are we today, in 2009? I click an icon on my windows desktop, and not only do I not see stuff appear in less than an eyeblink, but I can actually count past 3 before Outlook 2003 is started up. My time is more precious than that. I only have some much time in a lifetime before it's up, and I don't have that much longer to live for a friggin mouse click to waste it for me.
They should have let Billy Clinton split MS into two way back in 2000, instead of unleashing Bush/Cheney/Lewinsky on his ass. Would have done everybody a lot of good. In a free market competition economy when you got nobody to compete against, you wither away. Divide and compete against yourself to keep yourself on your tiptoes. Should have done that long time ago, and even today it's not too late, because there is nobody in quite the same position. Split MS in half, like picking team members at a basketball court, I choose this programmer, your turn, you choose that one. Then everybody inherits the common sourcecode and intellectual property, and try to outdo the other from there. Works like a charm. Have Bill Gates job be the referee, to score how everybody is doing, and bitch about anything he doesn't like. Fix what he is complaining about. Fix what everybody is complaining about, take care of the customer, instead of this pie in the sky attitude of we do what we want because we can, and who the hell does the customer think he is, to imagine he can dictate to us what to do. We have the upper hand, and we can screw mister customer anytime and anyway we want, and he better adapt to that reality.
Not all species go extinct. Sharks, crocodiles, turtles have been around for quite a while, and are still doing quite well. So have some species of bacteria. It's a lot about luck. Most species that stay around do still mutate and evolve, so when somebody asks what was first the chicken or the egg, well, back in the day the chicken wasn't the same chicken that you have today, it was more like a snake or some egg laying reptile or small dinosaur that ended up continuously or not so continuously evolving into the chicken we know today, and the chicken of the next billennium isn't going to be the same chicken that you have today. Eggs were first, but not chicken eggs, and it depends on what your definition of a chicken is. Is a dinosaur close enough to a chicken? They are pretty closely related, most birds today are remnants of the dinosaur reptile family, except they are warm blooded at 42C, as opposed to mammals are 37C. So will humans a few million years from now be the same as today? Maybe, maybe not. Sharks, crocodiles and turtles are still pretty close to what they used to be.
I think a lot of the issues with data centers are about inefficient code. In my humble opinion, with no hard data, I feel that older MS software from the 2000-2003 days used to consume a heck of a lot less cpu cycles than the newer 2005-2008 dotnetizized/vistized versions that are just hogs when it comes to speed, and to get the same work done consume a lot more cycles and a lot more power. Maybe this will nudge the software companies to pay better attention to the fact that computing speed and efficiency also matters, not just security. The cpu companies seem to have done and are still doing their share of work when it comes to low power consumption, but it doesn't mean a thing if software has gotten so much worse that overall the power consumption went up for the same work getting done.
Yeah but that's too much leap of a logic for most people to make. Saying we do not track our power usage so we have no data stops the thought there, as opposed to sayin hey, someone else tracks our power usage, and who would that be? The power company sending us the electric bills! Duh!
Slashdot Mirror
The attorney-client privilege is a different beast than the government. The attorney is my agent, he's a servant, for a fee. He does not serve the government. Even in crime. Moreover I can act as my own attorney. I can't act as my own public notary, or clergy, and they don't answer to me in a way an attorney does, but, if they privately know me, we have entangled lives vested in each other's, in the same community, then they answer to me at least as much as they answer to the government, even to the point of willing to hide stuff for me. They are a less strict class for holding confidential information compared to attorneys, but they are an order of magnitude better than the government, because they can be local, and adhere to distributedness and noncentralization principles. Moreover, even in the strictest case, if I don't empower anyone with the power of attorney, nobody would have the right to store my DNA information for extended periods of time. At least legally. Just like possession of drugs is illegal. It still happens, but does that mean we should make it legal? Also, of course, who can you trust these days? Having anyone else other than you holding any of your secrets is, well, naive? Attorney won't tell? The priest won't tell after a confession? Come on now, who are you kidding man? Still, trust happens. There are different degrees of trust, different kinds of trust, and the issue is whether to place trust in an overwhelming power entity providing it with even more power, or distributing and dispersing the trust, and therefore eliminating the concentration of power, well that's something I'm all for. Power leads to abuse, and abuse to even more power. It's the nature of power to do so, and you have the power to check the power of powers trying to overpower you and make your life into hell by running it for you, or anyone else's.
The lab should not "help out", people should be respected and allowed to chose for themselves. You cite the worst kind of medical trespassing there possibly is on people's rights. We only make life changing decisions for others when they are deemed too imbecile to make it for themselves. Similar to validity of wills. It's about respect, about human dignity.
From the wikipedia's human genome page, a person's DNA is about 3 billion base pairs long, which is about 750 MB of data, about a CD. Quite a bit of data to be pulling across a network, quite a bit to locally search, but a distributed query, like distributed.net's math solving trials, could be workable. Hence another reason for a distributed DNA database, instead of a centralized one. This argument won't be valid in a few years if computing storage and speed keeps advancing, but the privacy arguments may still stand. Though if only diff's are stored per person, and there is a categorization based on diff's, which is probably already done, centralized storage might be very efficient.
Sometimes when executing a will, relatives cannot be easily found. They might be hiding, or be in witness protection. A distributed query looking for relatives based on DNA, could allow the local databases to compare their own data, and let the person know, hey, someone is looking for you in trying to execute a will. Now this can be used as a great trap to bait and lure anybody out of privacy, but it's still better, because first you know someone is looking for you, second, you don't have to answer, third you can send an agent like your attorney who doesn't have to release your identity. All this cannot be handled through the phone book, because sometimes you just can't find somebody through the phone book. Because they prefer to have an unlisted number. So how can you find people who prefer being private? Send them a question and let them respond if they want to. Blood is thicker than water they say, at least relatives should be able to get connected, with privacy and security present. A national open database would allow all kinds of relatives to harass you based on relativity. It should be your choice whether you want to reply to a search and initiate contact, it should take two to tango.
Or imagine a tool like this in the hands of Hitler. That's what's most wrong with this whole thing, the power it gives to someone over other people. Anonymity and privacy, being shielded and safe from some paternalistic overseeing power entity should be a right. Power should be given to government only as much as necessary. Such databases should be in the private sphere, held by someone like the clergy with the "seal of the confessional", or by attorneys in a fashion similar to attorney-client privilege. We need a system of internetized public notaries/attorneys holding confidential private information, regarding issues of identity, privacy, will/testament etc. Some kind of distributed database with confidentiality barriers. Queries run against it, and people having to give consent before answers are released. Government access to it should be absolutely limited, with very strict rights and needing a warrant. In fact no central databases should exist, but some kind of public key/private key system published by many attorneys, or public notaries, from which matches can be found, such as relatives, or criminals, without revealing identity. To a posted public key search the other local small databases should react, and if they find themselves to be a match to a request, they should ask the owner of that DNA whether he would like to reveal his identity to the query. Of course you would find no criminals this way, because who would confess, yes, I'm the one. But that's exactly what the 5th amendment is about, it's not about making law enforcement easy, to the contrary, protecting individual liberty at the price of "security", or "ease of law enforcement." Compiling databases about everybody in the name of security - well, you know what the founding fathers said: those who sacrifice liberty for security will get neither. A social security number databases tagging everyone for tax collection purposes should suffice. Fingerprint databases feel already too private, but all they reveal is your physical presence at a location, if you didn't wear gloves. And they are harder to plant than dna samples of hair, blood, etc. Fingerprints in the name of law enforcement, I can agree to that, because they don't contain much else about you. DNA, that's a whole other beast than a fingerprint. Occasional DNA tests by police, comparing suspects to locally found evidence could be OK, with the data returned to the owners, or owner's assigned attorney/public notary after the completion of the trial. It should not be allowed to be archived, even if it means a whole lot of wasted work, and having to redo everything over and over. Or who do you trust? You should not feel more secure because of the databases compiled in the name of security, if anything, fear some coup, some power takeover at the top by some mad men. Then imagine what power they will have over you to deride you and ride you to hell and back, simply because they feel like it. And you're at fault, who previously sacrificed your privacy and anonymity in the name of security. What security? If you have many small localized/secret databases exchanging information only as needed, in case of a power coup at the top, well, my neighborhood notary public might be willing to hide my DNA from the new government, just like some people were willing to hide jews during the Nazi regime. People being able to disobey laws is a prerequisite of liberty. Where is the guarantee that we will never have another criminal regime like that in power, coming up with laws that are criminal. The time to defend is now. The price of liberty is eternal vigilance. Law enforcement is important, but so is the 5th amendment, which is more important than law enforcement itself. So how can you hide your DNA, how can you stay anonymous, retain your privacy in this world? Is that even possible? No. But at least we could have a makebelieve, pretend to respect each others rights to privacy world. Anytime you give blood, or a hair sample, for a simple thing as a drug test, others have your DNA, if that sample is tied to you in
As long as the product performs somewhere near as well as the old stuff, and it's patentable, then there is money to be made. We just have to find sufficient fault with the old stuff, and bad mouth it enough to start making money. Wikipedia says that as with the other lanthanides, samarium compounds are of low to moderate toxicity, although their toxicity has not been investigated in detail. An MSDS sheet where you can put toxicity N/A, no data available sounds better than one where you know it's toxic, because at least with an unknown there is a chance that it's not toxic. There is money to be made with the patent, and money saved by not having regulations to deal with. Regulations regulate know toxic materials, not unknowns.
If we can go to outer space, and figure out self sustaining local ecosystem bubbles, dependent only on solar radiation to sustain its inner lives, just like our Earth does, but on a smaller scale, so if we create many such little bubbles, then at least in one of them, we might make it. Unless something smarter than us evolves in one of the other bubbles, and goes on a hunt for us. Then we're in trouble. But as long as we know of nothing smarter than us, we're in pretty good shape, it's only a matter of will, not a matter of can't. Space bubbles, space stations, that, if needed, can be brought down to the surface too, bubbles with technology to be completely self dependent, and completely enclosed and sealed, so even in the case of some massive die off of any and all species on the planet because of some genetic engineered virus by some malicious hackers, the virus could not get inside because it would be hermetically sealed, even on the Earth's surface. Biosphere 2 was a failure of an experiment. Mostly because of lack of will, because we felt we didn't have to. When you're in outer space, you have to. The issue with many mini bubbles is a divergent evolution, different species evolving in different bubbles over millions of years - but I don't mind having kangaroos in Australia, and lamas in South America, I like diversity.
If we go into outer space, far enough from a nova issue, we might even survive our Sun going nova. It's all a matter of technicality. Such as incubating some sperm and eggs and some robotic nannies, and sending them off to Alpha Centauri, where on arrival, the test tube babies are born and robotic nannies that have tits that feel like human tits nurture them, and smell human, and teach them all the traditions and knowledge they need to know in order to sustain a large local space station living off of that stars radiant energy. Taking along other lifeforms too, of course, how else can you grow food. Then instead of all bags in one basket, dependent on the Sun, you'd have at least two baskets, Alpha Centauri, and the Sun. And when they see this Sun go nova for whatever reason, they can say "bummer!" So on and so forth. It's all just a matter of technicality, from an engineers point of view.
Worker bees will sacrifice like suicide bombers. You and I are collections of cells, multicellular organisms. Our cells sometimes do apoptosis, programmed cell death, so that the organism, the group can survive. By meme you mean the group, the organism. Suicide is a social, group phenomenon, evolutionarily speaking. Otherwise it wouldn't have made it thus far. Of course there are situations when things just go haywire, such as a bad mutation in a cell causes it to go nuts and do "nonprogrammed" cell death, more exactly a nonproperly programmed one. Or someone get mentally messed up by some accident and make the wrong decisions, and commit some errant suicide. But even "sane" worker bees, and "sane" cells know how to do suicide, in the name of survival, not all suicide is things gone haywire, some suicide is "proper". The real question is which category suicide bombers fall into - group protecting category, or nongroupprotecting insanity. The other question is, us, as cells in a larger group, organism, or meme, do we protect the meme that has evolved as a layer on top of us, generally called cultures, or religions - do we trust the meme? Do we people die for a culture, a tradition or a religion, or do we let the cultures die, so we the people can survive? What if our cells in our bodies decided that they disagree on the current "culture" and suddenly transformed into some other organism they find it betters their prospects of survival. Picture a monkey suddenly turning into a bird because its cells voted so, and the bird form got 51% of the vote, 48% wanted to stay monkey shaped, and 1% was undecided? We humans, can do such sudden feats with our groups, and they sometimes end up being the "wrong?" decision, maybe.
I miss mammoths. I would so love to see them again.
As long as there is a doctor and the patient has been informed about what's going on, and then the go ahead is given to the robot by the doctor who obtained consent, yeah. I'm very weary of any kind of automatic surgery device. What if it malfuctions? Imagine an automated lobotomy robot: come here, Mr. Patient, I have diagnosed you with a rare brain disorder!
The only issue on the moon compared to Earth is cooling. Nuclear needs a whole lot of cooling power because it's so energy dense. Submarines and pretty much all nuclear power plants rely on water cooling, something that's not available on the moon. Also, considering the vacuum, even air cooling with fins, like in computers, is not an option either. Geothermal cooling is one option, but probably has limited capacity, unless you drill very long holes. The safest bet is ultrahigh temperature reactors, and radiative cooling, through black body radiation into outer space. That way you can always relocate the plant too, if it rolls on wheels, carry it to where you need power, follow the bulldozer's power cord around as opposed to being tied to a geothermal hole.
The spare plane could be provided by the FAA itself, and shared by the competitors, as needed. One scenario is where it's available for a fee, the fee being higher than a normal plane's cost, so only in need would it be used. However if the competitors don't care about canceled flights and losing customers and only care about saving money, the FAA plane would just sit there with no use, nobody touching it. Another scenario is where it would be freely available to one of the competitors under certain rules, such as given enough system load, and some fairness and non-cheating principles, if you can devise some. Now the airplane companies would hire people to figure out how to get the free plane, and this would create very good understanding of the underlying upsets in the system. The real problem is how to come up with rules that keep getting the free plane fair. In a sense, you could have mathematicians on one side update rules, as the free plane hungry people figure out the loopholes. Formula 1 has such measures, where after an invention shows up, such as turbochargers, aerodynamic fins, superhigh rpms, every few years the design problem is made more difficult and more sane by putting limits. For instance the maximum rpm allowed is 19000 in an engine, because otherwise competitors would go to 40000 rpm or some crazy value like that, and research advances in that direction have no practical value in regular 2000 rpm automobiles, but some recent bikes do function at 9000 rpm, such as the Kawasaki Ninja 250R. Similar limits have been put on aerodynamic shape sizes that create downforce and excessive "safe speeds" (such as front funnel shapes of the early 90's that worked with venturi effects under the body.) You could set up an opposite interest group between regulating mathematicians, and the competitors competing for the free plane, minds pitted against each other, just like in the legal system, you have prosecution and defense, and from trying to outdo the opposition comes something valuable.
There is probably some equilibrium spareness that operators could rely on to ease some of the avalanches. If there are gate holds in Detroit, that cause an avalanche in the schedules - customers seeing expected arrival times to jump back and forth half an hour as the computer updates the possibilities - having one spare plane in say St. Louis, or Atlanta, or some really busy place, even with no spare gates, and letting the system know about it, could cut one order of magnitude of the problems, maybe even justify the cost of the spare plane. Just how much lost value can you assign to pissing off the customer? If you lose no profit over that, and pissing them off to the max is the most profitable way to do business, then it's going to be hard to justify the spares, on competitive business arguments. Maybe human decency arguments? FAA regulation might demand some minimum level of spareness? A lot of laws are like that, introduce human decency into a picture. This could be a great subject for some mathematicians to get funding for, to come up with FAA rules for optimum balance. You basically have too many equations and too much math that database people aren't smart enough to deal with, but linear algebra researchers could get off on. It's a brain tickle with very serious practical value.
If you're overwriting many times, overwriting with random data, instead of 0's, would make even the most sensitive electronic detection moot, unless they can predict what random values were used. Now that's something that's really impossible. Deciphering hieroglyphs was difficult enough, let alone a 10000x overwrite with a random pattern. Not only would they have to measure minute differences in signal levels, but they'd have to correctly predict how many 0's and 1's were used for each spot, and in what sequence, in that random pattern. Still, there is nothing like heating your harddisk to a temperature over the Curie temperature, where random thermal motion takes care of the demagnetization for you.
Why is this even called "The Great Zero Challenge". How about "Last chance at $40 for reading folders from a dd'd harddrive." Is it possible? The answer is maybe, so you should overwrite the same areas a few times. Reading residual magnetization left over after a rewrite probably takes special equipment with sensitivity and positioning beyond the regular head included in the disk. You're basically talking differences between zeros - one of the zeros is not the same as another zero: this zero used to be a zero, that zero used to be a one. You'd need access to the signal levels coming from the head before the analog to digital conversion takes place in the harddrive. Does this conversion take place in one shot, from volts or amps to 0 and 1, or is there an intermediate digitization, say getting a voltage level value between 0 to 255, and then later a calibration in the factory sets the levels of 0 to say 0 to 40, and 1 being 200 to 255, 40 to 200 indeterminate or read error, these values being freely tunable for each different disk? Then if you could get access to that pre-zero-or-one 0 to 255 digitized signal, then you could say that zeros that are reading 10 used to be zeros, and zeros that are reading 30, used to be 1, and subsequent overwrites would get the surface closer to 10 and 19, then 10 and 15, and 10 and 11, until the two different values would be so close, 10 and 10, that you'd need a separate equipment to sense and digitize differences between 10.0000 and 10.0002, and even then the signal to noise ratio might be so small that it'd be impossible no matter how good your equipment is. Is the data still available to somebody with 40 bux? Nope. Is the data still theoretically available after a single dd with 0? For a US or Chinese spy agency willing to spend a few million bucks on the question, the answer is: most likely. But 10000 dd's would most likely make it unavailable to them too. What's the actual number of overwrites needed? That needs real data, at least from someone with a semi-cleanroom hobby shop, with a superprecise oscilloscope reading squarewaves off a surface. Otherwise we're just guessing. One dd might be enough, or 10 might be, or even 10000 may not be enough. So how secure does your data erasing have to be, as long as we're guessing?
Upward wind tunnels, or really tall chimneys are expensive contraptions meant to generate wind from solar heat. Using naturally occurring wind generated by solar heat that blows sideways in most places and is available freely is generally thought to be cheaper.
I think their reason is not environmental, but to quell the hype, and the waste of money and resources. You really have to be watching your money like a hawk. Solar is like the only viable renewable energy for the long run - wind is better, but only certain places have sufficient wind/can tolerate huge height towers -, but you can make some really bad mistakes by spending too much. Like the military did by buying a 100 million dollar solar plant, that should have cost 10 million if profitability is anywhere in sight. Just because you can get power from solar, there is still a price you should spend, and not overpay 10-100x.
Office is Simonyi's masterpiece, even though the clunky or not so clunky(opinions vary) Hungarian notation is all over win32. When Simonyi left in 2002, you could tell it was over a disagreement in principle. Out of the original people at MS, he was pretty much the only technical brain, Gates was a Basic fan, and Ballmer a business bully, and who else? Simonyi was wrong about nonpreemptive multitasking, because the performance penalty isn't such a huge price to pay, compared to the benefit in stability you get. Simonyi was an all out performance freak. He went overboard, and his point of view was completely avoided, because he was wrong before. Behold Windows Vista as a result of not caring about performance. If anything, nonpreemptive multitasking might be the future in superperformance computing, where it's a compiler enforced rule that a program relinquishes control to the next one, instead of the operating system butting in, unloading the processor state, storing the stuff in a temporary space, getting the other program loaded, fiddling around with it, then repeat, unload, store. Nonpreemptive multitasking could run circles around such things, because it can properly decide when the impact of interruption is mininal, and performance stuff like this really mattered in the days of the 286, but by P3 days not so much, because of the difficulty to manage code. I still believe performance, not ease of programming, is the highest principle in computing. It's what separates the professionals from the amateurs and hobbyists. Getting more abstract to sound more professional and paying huge performance penalties for it, as in java, dotnet, even c++, doesn't seem very professional to me. As a professional computer programmer you have to account for how you spend the computer cycles your customer entrusted you with. If nothing else, because of the energy crisis in the world. Each wasted clock cycle costs oil, costs blood, costs lives. I believe there has to be a safe way to write high performance code. It's hard, but there has to be a way. There has to be a safe way to write C programs without buffer overflows. Dump all performance because of a few problems? Are buffer overflows really not fixable? And if anything, there has to be a safe way to write nonpreemptively multitasking and cooperating programs, that give very high performance. It's difficult, yeah. It's next to impossible to do, but not impossible, and preemptive multitasking is just soooooo much easier, whew, what a breath of fresh air, all those crap programs can't bring the system down. Still, you could have a pretend-nonpreemptive multitasking system where the programs consciously pass on the you're next bar, and the OS only butts in as a referee, when one of the programs doesn't behave. All it has to do is observe, and time, watch the rules, and give some programs a yellow card, a red card, like in soccer. Timeout on the bench. Realtime programming environments, with guaranteed response time could have an easy rule, easy way to tell if somebody doesn't relinquish in time. That's all it would care about, no complex rule of expected behavior like in most sports. History proves Simonyi was wrong about nonpreemptive multitasking, but I disagree with history and agree with him on that one. But I personally hate Hungarian notation and I'm Hungarian by birth. I prefer $, %, sigils, they are intuitive and don't need to be read, my brain switches subconsciously into that mode by seeing sigils, but hungarian notation prefixes need effort to read, and there is so many of them, I can't keep track, it takes brain wearing conscious effort to digest them, and they get me quickly fatigued. I guess if you stare at them for a lifetime, they become to you like sigils, subconscious. As a bilingual person I know that a language cannot be spoken unless its subconsciously processed, if you need to exert significant conscious effort, you merely kludge around and you only speak at a very slow rate. You have to think in the language you speak when you speak it, literal translations don'
And maybe one of the teams may decide to release Windows Shark - Cold Blooded, Lean and Mean(TM) and go back to the Windows 95 simplicity/speed/tight code, based on stripped version of Win2K code wise, but add glitter looks wise, with a small registry, and add small improvements like bigger hard drive support and bluetooth and stuff like that as it arrives to the market. Maybe they need just a small spin off company, to support the "legacy" with very few very smart people, who will babysit and support win2k/office97/vb 6/sql server 2000, and provide updates only as needed, as it makes sense. VB6 could use a hell of a lot of updates, but it's still the easiest thing to use to bang up a business program in a matter of minutes that actually works without making the computer crawl to its knees. Usability wise that setup from that point in time could last the test of time. Windows Raptor for desktop and gamerz, Windows Shark for business, Windows Predator for servers hosting sql server 2000, nothing newer. Remember when Tom's hardware guide use to post frame rates achieved by different games? People love speed, people love fast cars. That's what they need to put the spice back into computing again, and it shouldn't cost much to sustain. It's like when a plant needs a trimming job because it overgrew. Or a new shoot. There is nothing wrong with getting a database transaction that takes nanoseconds instead of microseconds. Or mouseclick reaction rates that go to microseconds from milliseconds. The end user doesn't need it, doesn't notice it? Yeah, like a Hayabusa's or Lamborghini's top speed is not needed by everyday commuters. Where's the fun, man? XP-experience even your grandma can appreciate. Granda doesn't appreciate the computer, she appreciates her knitting kit. Vista. Vista of what? Of where we're heading? How we're growing slower, older? Where do you want to go today? I want to go back to the days when computing used to rock my brains, when it was so much fun. I want real time computing, with guaranteed reaction time. I want something as easy to program as basic, easier to read than a good book, and it works faster than manually tuned C/assembler. Something almost as plain as english, something I can drop into my robot and make it frigging work without having to go through the steps of abstractualizing the conceptualizations of how the different frameworks interact from an objective perspective inside the whole thing. I just don't give a crap, I don't want to care about the latest hype. Gimme something I can get my hands on, work, form into something, something that's easy already, something that makes my life even easier. Something that ticks fast, and tells me how fast. Nanoseconds I love. Microseconds I can tolerate. Don't tell me milliseconds, that's so 1999. And where are we today, in 2009? I click an icon on my windows desktop, and not only do I not see stuff appear in less than an eyeblink, but I can actually count past 3 before Outlook 2003 is started up. My time is more precious than that. I only have some much time in a lifetime before it's up, and I don't have that much longer to live for a friggin mouse click to waste it for me.
They should have let Billy Clinton split MS into two way back in 2000, instead of unleashing Bush/Cheney/Lewinsky on his ass. Would have done everybody a lot of good. In a free market competition economy when you got nobody to compete against, you wither away. Divide and compete against yourself to keep yourself on your tiptoes. Should have done that long time ago, and even today it's not too late, because there is nobody in quite the same position. Split MS in half, like picking team members at a basketball court, I choose this programmer, your turn, you choose that one. Then everybody inherits the common sourcecode and intellectual property, and try to outdo the other from there. Works like a charm. Have Bill Gates job be the referee, to score how everybody is doing, and bitch about anything he doesn't like. Fix what he is complaining about. Fix what everybody is complaining about, take care of the customer, instead of this pie in the sky attitude of we do what we want because we can, and who the hell does the customer think he is, to imagine he can dictate to us what to do. We have the upper hand, and we can screw mister customer anytime and anyway we want, and he better adapt to that reality.
Not all species go extinct. Sharks, crocodiles, turtles have been around for quite a while, and are still doing quite well. So have some species of bacteria. It's a lot about luck. Most species that stay around do still mutate and evolve, so when somebody asks what was first the chicken or the egg, well, back in the day the chicken wasn't the same chicken that you have today, it was more like a snake or some egg laying reptile or small dinosaur that ended up continuously or not so continuously evolving into the chicken we know today, and the chicken of the next billennium isn't going to be the same chicken that you have today. Eggs were first, but not chicken eggs, and it depends on what your definition of a chicken is. Is a dinosaur close enough to a chicken? They are pretty closely related, most birds today are remnants of the dinosaur reptile family, except they are warm blooded at 42C, as opposed to mammals are 37C. So will humans a few million years from now be the same as today? Maybe, maybe not. Sharks, crocodiles and turtles are still pretty close to what they used to be.
Aha! Now I finally figured out how I got so stupid these last few years!
I think a lot of the issues with data centers are about inefficient code. In my humble opinion, with no hard data, I feel that older MS software from the 2000-2003 days used to consume a heck of a lot less cpu cycles than the newer 2005-2008 dotnetizized/vistized versions that are just hogs when it comes to speed, and to get the same work done consume a lot more cycles and a lot more power. Maybe this will nudge the software companies to pay better attention to the fact that computing speed and efficiency also matters, not just security. The cpu companies seem to have done and are still doing their share of work when it comes to low power consumption, but it doesn't mean a thing if software has gotten so much worse that overall the power consumption went up for the same work getting done.
Yeah but that's too much leap of a logic for most people to make. Saying we do not track our power usage so we have no data stops the thought there, as opposed to sayin hey, someone else tracks our power usage, and who would that be? The power company sending us the electric bills! Duh!