It is not running the software, but I am stil intermittenlty patching code whose copyringht statement at the head (written by me) says "Copyright 1984. We still have users of that software, they still find bugs with new hardware, we still fix them. Admittedly, that 1984 software is not much in use, but 1994 software is still definitely mainstream support (the article regards Win98 as incredibly old).
Sadly, I think you are right about the cost element putting it out. Though int the cost ot tanks costing tens of millions and aircraft and ships costing billions, the cost of a bit of domestic polumbing really does not look very much. The fuel cost of one F15 sortie would probably pay to install private showers for half the maintainance crew. And the military is supposed to be getting smaller and smarter - fewer troops with better kit. So there should be a few free rooms to convert.
Another way to look at it that they guys and girls are supposed to be fighting for liberty, civil rights and good treatment of individuals all over the world - and yet you can't give them the privacy that just about any employee elsewhere in the country would demand. If it is the right thing to do, on Civil Liberties grounds, shouldn't you give it to the defenders of those liberties first, not last.
em>So here is a little thinking question...if it is simply 100% OK to admit gays into the military (and leave them in when they are discovered), is it OK to house men and women soldiers together in the same rooms? Make them use the same showers? Make them roomies?
In very few other circumstances are people forced into communal living for long lengths of time. Which begs the question - why aren't soldiers given the amount of privacy we give college students? Shouldn't we give the vast majority of service personnel, when in barracks, the same privacy most of us would expect if posted by our employers on some business task? Even more so for the Air Force, which tend to operate from prepared bases.
Obviously, in battle or simulated battle, you cannot afford those luxuries. But any soldier who cannot control his/her lust in such a situation should not, in my opinion, be running around with lethal weapons at the taxpayers expense.
I can see there could be a problem in submarines, where ot really is necessary to fit a large number of people into a small volume. But the vast majority of the armed forcs should be able to arrange things better, just on the basis of treating people properly.
Which also gives you a way to get out of being tested. Get az bit of metal embedded in you skull. Maybe some cosmetic jewellery, but put in so it cannot be take out without minor surgery. Then they cannot safely test you without surgery - which they would find very hard to enforce. Of course, if you need the scan - say you have a suspected brain tumour - minor surgery here we come.
Which is to assume that the only form of Communism that is possible is that practised by the Soviet Union. That wasn't Communism, that was state dictatorship wrapping itself in the flag of Communism. It would be perfectly possible to imagine something which followed the beliefs of the Paris Commune of 1870 or the writings of Karl Marx which was entirely compatible with democracy and personal freedom. I think that even Lenin believed that the iron control he instituted was a necessary evil in order to root out the evils of the past, which would wither away once the perfect Communist state was established. (Stalin was another matter).
So I think you reply epitomises the problems of this sort of blanket condemnation. There were some atrocious evils carried out by the Soviet Union and others in the name of communism. Those atrocities should be condemned, and anybody proposing them (forcel labour camps, wholsesal uprooting of poulations, deliberately induced famines) shoudl be shut up firmly, whatever their politivcal allegience. On the other side, there is the political idea that "from each according to his ability, to each according to his need" is the right way to run a country. It doesn't work, in my opinion, but it is a perfectly discussable point, and those proposing it, and legislation to implement it, should be allowed their voice - aven if that voice is raised in the name of Communism.
Reliability. You may have quite a small, relatively low loaded database for your small business. But if your business depends on quick response, you want 24/7 uptime. If somebody asks at a shop if you have something in stock or checks a reservation at a hotel, you want to be able to say yes/no quickly. How amny times to we go elsewhere if someone says "Sorry, the computer is down"? I got that at my doctor's the other day - due to building works, the one computer with the appointments on it had been powered off. OK, one PC will host the database without breaking into a sweat. But how reliable is that PC? Well, you can get UPS, raid drives etc. etc, which make failures less likely, but they cost money and are not infallible. Or how about you use this to cluster (for free) across all the PCs you have sitting around, mostly running screen savers? Orders of magnitude reliability increase for zero bucks.
The best way to knock over oracle is to start up a company that supports open source for a fee (which is cheaper than running oracle for a year).
Which is exactly what MySQL AB does for MySQL. Their support is not particularly cheap, (though I be that it is a lot less than Oracle's), but I recommend it highly. The original designers are still leading the development/support team (is that true for many of the alternatives?) and make a living *only* because of their superior product, not because some salesman conned the management.
It is understandable i the world we live in that the Bean Counters have no concept of the public good. If there is just the tiniest remote possibility of a financial return to the University - which did, after all, pay for it to be written - they want to keep that possibility alive.
So you need to put something in the other pan of the scales - something which will advantage the University from an OS release.
Presumably this software is (a) useful for your research, or you wouldn't have written it, (b) useful to others, because somebody thinks there is an outside chance of selling it, and (c) not bug free, because no software is ever bug free.
In my opinion, good, well documented bug reports are of value in their own right. My eexperience is that if someone provides a clean, repeatable bug report, the bug can usually be fixed in less than a day. Getting the bug to "repeatable" status can take many painful days of work. Therefore the presence of a body of users who provide quality reports, even if they never actually look at the code (and some will not only do so, they will submit patches) is a real, positive, benefit to the software - which, we have already said, is of benefit to your research, whish is what you are really paid for.
So you have three scenarios:
They don't let you release the softwarfe, but you never commercialise it. The lose-lose scenario.
You release the source commercially. You will have to put up a lot of money for packaging, advertising etc. if you want to make a success of it (don't believe the Better Moustrap Fallacy). The possibility of this happening, and succeeding, is pretty small, so weight down the return you would get if it did happen by quite a lot. If you do sell it, and a user reports a bug, he will probably not have the source and even if he does, won't expect to do the job he has paid you to do - so he will demand that you drop your research and fix his problem now.
You GPL the source. At worst, nothing happens (in which case the commercial path would have failed, too). But users probably report, and sometimes fix, bugs. If they report a bug and it doesn't worry you, you forget about it - if they want it fixed, they can fix it themselves. If it does worry you - great - someone else is fixing (or at least problem-spotting) your software for free; you can spend your time on research instead of debugging. And if it is a real success, you set yourselves up as a support organisation and sell support.
In short - play the accountants at their own game. Tell them why it benefits your research to release the source. The fact that it also benefits the world is a non-cash benefit to you - you can feel all warm and comfortable about the good you have done.
Very, very few of these "we might be able to sell it" options ever come off. So they would be trading a 50% chance of help with your research (with no downside) against a 1% chance of making significant bucks from the software (with the possibility of losing significant bucks instead).
Not only do I click on the ads, in Google and Google alone, I ask for them. If I want cheap flights, I ask Google for cheap flights and click on the resulting ads. Since Google knows I am in.uk, it filters the ads accordingly. It works. It may be contributing to a monopoly, but hey, I'm lazy.
No effect yet traced, but cautious researchers saying not finding something is not the same as proving it doesn't exist, which the worried then take as an assertion it does exist.
Mobiles are limited to ?1 watt?. A torch bulb is several watts, at higher (and conventionally more damaging) . I just don't see any mechanism for damage; and nobody (AFAIK) has followed up any suggestions with valid research.
Sunlight is about 500w/m^2. The top of my head is about about 20 cm round of this, so a sunny day gives 10-20W onto my skull. A mobile at a total of 1W, not all of which is radiated towards me? I am not worrying.
And, to keep on topic, I think WiFi is even less (?1/4 watt?) and you don't hold it close to you.
I would worry far more about exhaust fumes, myself. But those seem less dangerous to ordinary people, because you can "see" them, whereas you can't see this nasty electromagnetic radiation (big bad word there).
I would take issue with that. Certainly a cool aircraft - but the coolest ever? Its only special quality was being safe enough for passengers. When it was built, there were already military aircraft bigger and faster, and there have been many aircraft since better in many ways.
It was, fundamentaly, a mistake to build the thing. Once it was built and the development money spent, it was not necessarily a mistake to keep flying it. But it was a mistake in the first place. And not only one that could be foreseen, but one that was foreseen, by many people. But it was forced through at the height of dirigiste socialism in the UK and (more so) France.
I cannot call something that was a fantastic waste of money that could have been (a) spent on something worthwhile, or (b) not taxed in the first place (choose according to political taste) "cool".
The fact is that building Concorde destroyed the Eurpoean commercial aircraft manufacturers. Before Concorde, there was competition between Europe and the US, after it was between Boeing and McDonnel Douglas (and Lockheed, a bit). It took 30 years (and even more public money) to the European industry to get back off the floor with Airbus.
So some regrets at its passing, but not deep grief, from me at least.
All three articles are talking about highly specialised, basically single function, machines. As other posters have correctly pointed out , programming such machines is very, very difficult. When you manage to do so, they can be very powerful indeed. But they do only one job, even though they do it very, very well. Saying that they are likely to replace general putpose CPUS is like sayign that F1 cars of Indy racers about to replace pickups or family cars. They may do a job worth doing in their specialist area, and they may make money, bu they are never going to replace the VN machine in 90% of the places it is used.
One of them is a specialised web server. Fine, there are a lot of web pages out there that need serving. I can well believe that you can build an FPGA-based static-page web server which will beat the pants of a Sun/Intel server doing the same thing. But what about dynamic content? is their DBMS as good as the latest Oracle or MySQL? Willit, say, handle the internationalisation issues that those systems will? Bet it won't. Will it runs PHP or Python natively? I doubt it - I bet it hands that over to a traditional back-end processor.
As has also been said elsewhere, thus kind of hype is a repeated event. A specialist machine outperforms a generalist machine at its specialist task, and journalists claim that the world has turned upside down. Connection Machine, Deep Blue, GAPP, transputer... Just a few I can call to mind.
I don't know about the US, but in the UK the power company are required to make the total number of cycles in a day come out pretty close to right (weak memory says withing one second) but allowed larger margins of error within the day So yes, they do positively intervene to correct the number of cycles per day.
This is old news - it has been "true" for years. It is actually a corrolary of Clarke's law ("Sufficiently advanced technology is indistinguishable from magic"). If we understand how a system works normally, then any misbehaviour it shows is a fault. If we don't, then we can classify the misbehaviour as a "neurosis". Unskilled users often believe their computer sare sufferring from a neurosis. This usually means that at some time in the past they have installed some app or extension which is trying to do something they don't understand. A more skilled user can come along and "cure" that neurosis, because they understand the system at a deeper level.
A car I once had displayed what appeard to be a "neurosis" - it seemed to be frightened of going more than 30mph. It would run fine up to that speed, but if you went any faster it "paniced" and stalled. Dirt in the fuel line: at low flow rates, it lay flat and let fuel pass. At higher flow rates, it flipped up and blocked the flow completely, causing the engine to stall before it had time to flip down again. The point is, the first analysis of "neurosis" was corrected to "fault" once the problem was understood.
So the diagnosis of "neurosis" is relative - it means "I don't understand this failure mode". It can, of course, become absolute if nobody understands it.
So, are we building systems so large that nobody understands them? Definitely. Networks are already bordering on incomprehensible. Particularly, of course, the Internet. It would not surprise me at all if the Internet started showing "neurotic" behaviour. Indeed, it already does - if you ragard humans and their input as part of the net istelf. DOS attacks and the/. effect are both "twitches" in the body of the Internet. (And spam is a cancer which requires operating now) Thus far, these nervous ticks have expanded into full-scale neurosis - but they could.
I have to agree. I have got the BSc too, and I ducked out of science because it was too much like hard work. I went to play with computers because people would pay me to have fun while I found out what my career was really going to be. 25 years on, I still haven't found out. But some of my friends are Real Scientists. From what they tell me, it is a lot like creative software design. Of course there are dreary bits - so there are in programming, too. But the bit that distinguishes the Eagle from the Turkey is not those dreary bits, it is the insight that comes when you let the problem flood you mind and wait for the solution to rise to the surface.
Contrary to popular belief, programmers more frequently resemble artists than scientists. If you want to maximize the creative potential on your development team, you've got to start thinking about the psychology of the programmer and be willing to back it up with management policy.
Which shows that this guy doesn't know scientists. Scientists - true scientists, not technicians - are very like this guy (correctly) describes programmers. Both programming and scientific research are creative skills which, as the man says, require you to be "in the groove". He is not wrong about programmers - he is wrong about scientists. Techicians, to some extent, have less need to be "in the groove" - though much of what he says applies to any human being, with only the time constants varying.
OTOH. 3. Accommodate Reasonable Special Requests. When I get really stuck on a design problem, I go for a walk in some very beautiful woods about three miles from my office. An hours walk in the woods has about an 80% chance of delivering a solution to the problem. Even, curiously, if I don't spend much time conscioulsy thinking about the problem. In fact, I sometimes feel that by subconscious is telling my conscious to let go that problem and leave it to me. Dropping a problem for an hour or day and then coming back to it can be remarkably constuctive.
In fact, I sometimes feel embarasssed that the conscious "me" claims credit for the bundle of mad scientist, lechers, random thought generators, and idiots who inhabit my subconscious and do all the work.
As you say, OTEC is not that new in concept, there are just relatively few places that you can build shoreline OTEC plants, for reasons you give.
So how about off-shore OTEC plants? Then you have the entire surface of the Pacific to populate - plenty of room.
First question - how do you get teh energy back to land (I am not going for fresh water at this point). Easy - electrolyse seawater to Hydrogen for this new Hydrogen Economy we're all due to enjoy.
So you create a concrete (cheap) tube 1000 metres long. You insert buoyancy chanbers so it is roughly neutral density (and make the concrete with pumice, which is very light and widely available near volcanoes), but make the bottom denser than the the top so that it is naturally stable in an upright position. Add piping to pump heat from top to bottom: the tube becomes a giant chimney, and you generate electricity with turbines inside it. The thing free floats in the Pacific, mostly following the big eddies that swirl round the Pacific and occasionally using a little of its power to push itself into the right current. Tankers picking up Hydrogen have to chase it round the ocean.
Of course, the big problem with any installation at sea is storm damage. You have to make things incredibly strong so as not to be destroyed by "worst case" storms, which puts up costs enormously(and they still get damaged, if not destroyed, and are costly to repair). Not in this case. I said the tube is nearly neutral density. It is *very* nearly neutral density. When a storm approaches, it floods a few tanks and ducks down below the surface to lurk, perhaps a hundred feet down. It watches the suurface with sonar, and when the waves have subsided, bobs back up again and resumes work.
The bottom is denser than water, so the tube floats vertically, which means the rest is lighter than water for neutral buoyancy. The working bits are not at the surface, but far enough down to be out of harms way, but high enough up that they can be raised to the surface (weather permitting) for maintainance. The top few tens of feet are again denser than water. These are sacrificial: if, by mischance, they get knocked off, the average density rises, so that the whole assembly floats higher instead of sinking deeper, and can be repaired - thr erepairs, of course, beanng to the easily accessible top section. In fact, there are small explosive charges for use in an emergency, which can blow bits of this off to lighten the structure.
Assembly should be easy enough. It can be fabricated in sections, sections floated far enough offshore to get the right depth and to get into the eddy, assembled horizontally and then carefully rotated into position (arefully - this thing has the general geometry of a thin drinking straw, so you don't want it to flex too much). If the worst comes to the worst (e.g. it is about to run aground), ensure the expensive bits (turbing, pumps, heat exchangers) have positive buoyancy and blow the tube to bits. Collect up the floating bits and re-assmeble with a new tube (cheap concrete).
Ecologically interesting. Much of the open ocean of the Pacific is pretty barren, because of lack of nutrients at the top and lack of oxygen at the bottom. This is going to stir the whole thing up, which should increase the fertility of the ocean a lot. And it produces Oxygen as a waste product: I wonder what happens if you inject that at the bottom end instead of venting to the atmosphere...
Not really aimed at the geek, more aimed at maximising profit. It is rather similar to the way two identical seats on an aircraft can be sold for wildly differing pricesbecause of arcane booking rules.
While it is true that some chips will run faster than others, once the process settles down, the vast majority of chips off the line will fall into the same ballpark. But "faster" processors get a premium price. So they simly test a lot of the chips at low speeds and sell them at "ordinary" prices, then re-test some of them at high speeds and, provided they pass, sell them at premium pricecs. The ordinary price chips migt be just as good as the premium ones - or they might not. It is not that they are inherently slow, it is that they haven't been tested fast.
But if people keep overclocking, Intel will no longer be able to charge a premium price for an extra, say, 200MHz of speed.
It is scaleable - provided you have the right sort of coast, with very deep war relatively close to the shore. It will therefore be very suitable for most Pacific islands, many of which are the tops of seamounts. I don't think the coast of Calif is quite so precipitous - and the power consumption per mule of coast is probably thousands of times that of (low population, low power consumptoion, lots of beach) Pacific islands. So don't get too excited.
Do we know, mathematically, how much a kilogram is? What is the mathematical definition of a kilogram that does not depend upon fundamental constants whose value, in turn, depends upon the kilogram?
The current definition of the kilogram is "the mass of that lump of metal near Paris". We the define other constants, such as Avogadros's number, in term of that kilogram
There is no electrical contact *to* short out. It is using magnetism, not electicity. Pad generates alternating magnetic field above its surface. Put coil tuned to same frequency in field. Conductor in changing magnetic field generates current, so AC generated. Add rectifier, DC generated. I can see that the add-on to the mobile device will be cheap - a coil and 4 diodes. This could work.
One non-green point: How much power is dissipated with no devices on it, or with fully charged devices? But then, how much power is dissipated by plugged in "wall warts".
In order to air cool, you need quite a considerable volume of airspace above the CPU and probably a head sink as well, and a relativewly unobstructed airflow from front to back of the PCB. (Some opeople do side to side, but usually not more than once). This is fine for one or two CPUs, but it makes the CPU and its cooling space effectively about two inches tall. Also, you cannot have one CPU pehind another in the airflow direction, because the second will have its air preheated by the first. For supercomputing use, the processor only required ram and network IO. This means that each kernel can probably be squeezed into an area about 3 inches square - which means that, cooling aside, you coult pack them three or four deep between front and back of the back. The graphic looks like a blade architecture. With air cooling, I would have thought you would get 1 CPU per blade on a 2 in spacing, or about 8CPUs per 2U rack unit. With warer cooling, you might get CPUs per blade on a 3/4 in spacing, or about 60 CPUs per 2U rack unit (hmmm - I think that is a bit optimistic...).
Then, if you use air cooling, where does your heat come out? Into the server room. And if you have 256 CPUs at 50W each, that is over 10kW of head being dumped into your server room. You need some *serious* aircon to get that out of the room. And what is the aircon doing? Extracting the heat from the air and putting it back into some kind of fluid, which is piped to heat exchangers. So it makes sense to get the heat into fluid earlier rather than later.
The way high-performance processing is going nowadays, I can see fluid cooling making a modest return to favour at the high end of the scale. It is already returning at the low end, with high performace laptops using fluid cooling instead of fan cooling. So I don't think this is a distracion at all.
They the Fingerworks Touchstream keyboard - if you can afford it. The same area acts as keyboard (single finger touches), mouse, and gesture area (multiple finger touches). I haven't tries the keyboard version; I have the mouse/gesture only iGesture, and it does the job well. Not as well as a mouse, but better than al the other pointing devices I have triedl. The keyboard ought to be better, because it keeps your hands in the active area while moisung - and it is, of course, zero-force (and zero feel).
Slashdot Mirror
It is not running the software, but I am stil intermittenlty patching code whose copyringht statement at the head (written by me) says "Copyright 1984. We still have users of that software, they still find bugs with new hardware, we still fix them. Admittedly, that 1984 software is not much in use, but 1994 software is still definitely mainstream support (the article regards Win98 as incredibly old).
Sadly, I think you are right about the cost element putting it out. Though int the cost ot tanks costing tens of millions and aircraft and ships costing billions, the cost of a bit of domestic polumbing really does not look very much. The fuel cost of one F15 sortie would probably pay to install private showers for half the maintainance crew. And the military is supposed to be getting smaller and smarter - fewer troops with better kit. So there should be a few free rooms to convert.
Another way to look at it that they guys and girls are supposed to be fighting for liberty, civil rights and good treatment of individuals all over the world - and yet you can't give them the privacy that just about any employee elsewhere in the country would demand. If it is the right thing to do, on Civil Liberties grounds, shouldn't you give it to the defenders of those liberties first, not last.
em>So here is a little thinking question...if it is simply 100% OK to admit gays into the military (and leave them in when they are discovered), is it OK to house men and women soldiers together in the same rooms? Make them use the same showers? Make them roomies?
In very few other circumstances are people forced into communal living for long lengths of time. Which begs the question - why aren't soldiers given the amount of privacy we give college students? Shouldn't we give the vast majority of service personnel, when in barracks, the same privacy most of us would expect if posted by our employers on some business task? Even more so for the Air Force, which tend to operate from prepared bases.
Obviously, in battle or simulated battle, you cannot afford those luxuries. But any soldier who cannot control his/her lust in such a situation should not, in my opinion, be running around with lethal weapons at the taxpayers expense.
I can see there could be a problem in submarines, where ot really is necessary to fit a large number of people into a small volume. But the vast majority of the armed forcs should be able to arrange things better, just on the basis of treating people properly.
Which also gives you a way to get out of being tested. Get az bit of metal embedded in you skull. Maybe some cosmetic jewellery, but put in so it cannot be take out without minor surgery. Then they cannot safely test you without surgery - which they would find very hard to enforce. Of course, if you need the scan - say you have a suspected brain tumour - minor surgery here we come.
Which is to assume that the only form of Communism that is possible is that practised by the Soviet Union. That wasn't Communism, that was state dictatorship wrapping itself in the flag of Communism. It would be perfectly possible to imagine something which followed the beliefs of the Paris Commune of 1870 or the writings of Karl Marx which was entirely compatible with democracy and personal freedom. I think that even Lenin believed that the iron control he instituted was a necessary evil in order to root out the evils of the past, which would wither away once the perfect Communist state was established. (Stalin was another matter).
So I think you reply epitomises the problems of this sort of blanket condemnation. There were some atrocious evils carried out by the Soviet Union and others in the name of communism. Those atrocities should be condemned, and anybody proposing them (forcel labour camps, wholsesal uprooting of poulations, deliberately induced famines) shoudl be shut up firmly, whatever their politivcal allegience. On the other side, there is the political idea that "from each according to his ability, to each according to his need" is the right way to run a country. It doesn't work, in my opinion, but it is a perfectly discussable point, and those proposing it, and legislation to implement it, should be allowed their voice - aven if that voice is raised in the name of Communism.
Reliability. You may have quite a small, relatively low loaded database for your small business. But if your business depends on quick response, you want 24/7 uptime. If somebody asks at a shop if you have something in stock or checks a reservation at a hotel, you want to be able to say yes/no quickly. How amny times to we go elsewhere if someone says "Sorry, the computer is down"? I got that at my doctor's the other day - due to building works, the one computer with the appointments on it had been powered off. OK, one PC will host the database without breaking into a sweat. But how reliable is that PC? Well, you can get UPS, raid drives etc. etc, which make failures less likely, but they cost money and are not infallible. Or how about you use this to cluster (for free) across all the PCs you have sitting around, mostly running screen savers? Orders of magnitude reliability increase for zero bucks.
The best way to knock over oracle is to start up a company that supports open source for a fee (which is cheaper than running oracle for a year).
Which is exactly what MySQL AB does for MySQL. Their support is not particularly cheap, (though I be that it is a lot less than Oracle's), but I recommend it highly. The original designers are still leading the development/support team (is that true for many of the alternatives?) and make a living *only* because of their superior product, not because some salesman conned the management.
(You may gather that I am a fan of MySQL).
So you need to put something in the other pan of the scales - something which will advantage the University from an OS release.
Presumably this software is (a) useful for your research, or you wouldn't have written it, (b) useful to others, because somebody thinks there is an outside chance of selling it, and (c) not bug free, because no software is ever bug free.
In my opinion, good, well documented bug reports are of value in their own right. My eexperience is that if someone provides a clean, repeatable bug report, the bug can usually be fixed in less than a day. Getting the bug to "repeatable" status can take many painful days of work. Therefore the presence of a body of users who provide quality reports, even if they never actually look at the code (and some will not only do so, they will submit patches) is a real, positive, benefit to the software - which, we have already said, is of benefit to your research, whish is what you are really paid for.
So you have three scenarios:
In short - play the accountants at their own game. Tell them why it benefits your research to release the source. The fact that it also benefits the world is a non-cash benefit to you - you can feel all warm and comfortable about the good you have done.
Very, very few of these "we might be able to sell it" options ever come off. So they would be trading a 50% chance of help with your research (with no downside) against a 1% chance of making significant bucks from the software (with the possibility of losing significant bucks instead).
With all these fans of the old IBM keyboards, it looks like there is a market opportunity for someone.
Not only do I click on the ads, in Google and Google alone, I ask for them. If I want cheap flights, I ask Google for cheap flights and click on the resulting ads. Since Google knows I am in .uk, it filters the ads accordingly. It works. It may be contributing to a monopoly, but hey, I'm lazy.
No effect yet traced, but cautious researchers saying not finding something is not the same as proving it doesn't exist, which the worried then take as an assertion it does exist.
Mobiles are limited to ?1 watt?. A torch bulb is several watts, at higher (and conventionally more damaging) . I just don't see any mechanism for damage; and nobody (AFAIK) has followed up any suggestions with valid research.
Sunlight is about 500w/m^2. The top of my head is about about 20 cm round of this, so a sunny day gives 10-20W onto my skull. A mobile at a total of 1W, not all of which is radiated towards me? I am not worrying.
And, to keep on topic, I think WiFi is even less (?1/4 watt?) and you don't hold it close to you.
I would worry far more about exhaust fumes, myself. But those seem less dangerous to ordinary people, because you can "see" them, whereas you can't see this nasty electromagnetic radiation (big bad word there).
perhaps the coolest commercial plane ever to fly
I would take issue with that. Certainly a cool aircraft - but the coolest ever? Its only special quality was being safe enough for passengers. When it was built, there were already military aircraft bigger and faster, and there have been many aircraft since better in many ways.
It was, fundamentaly, a mistake to build the thing. Once it was built and the development money spent, it was not necessarily a mistake to keep flying it. But it was a mistake in the first place. And not only one that could be foreseen, but one that was foreseen, by many people. But it was forced through at the height of dirigiste socialism in the UK and (more so) France.
I cannot call something that was a fantastic waste of money that could have been (a) spent on something worthwhile, or (b) not taxed in the first place (choose according to political taste) "cool".
The fact is that building Concorde destroyed the Eurpoean commercial aircraft manufacturers. Before Concorde, there was competition between Europe and the US, after it was between Boeing and McDonnel Douglas (and Lockheed, a bit). It took 30 years (and even more public money) to the European industry to get back off the floor with Airbus.
So some regrets at its passing, but not deep grief, from me at least.
All three articles are talking about highly specialised, basically single function, machines. As other posters have correctly pointed out , programming such machines is very, very difficult. When you manage to do so, they can be very powerful indeed. But they do only one job, even though they do it very, very well. Saying that they are likely to replace general putpose CPUS is like sayign that F1 cars of Indy racers about to replace pickups or family cars. They may do a job worth doing in their specialist area, and they may make money, bu they are never going to replace the VN machine in 90% of the places it is used.
One of them is a specialised web server. Fine, there are a lot of web pages out there that need serving. I can well believe that you can build an FPGA-based static-page web server which will beat the pants of a Sun/Intel server doing the same thing. But what about dynamic content? is their DBMS as good as the latest Oracle or MySQL? Willit, say, handle the internationalisation issues that those systems will? Bet it won't. Will it runs PHP or Python natively? I doubt it - I bet it hands that over to a traditional back-end processor.
As has also been said elsewhere, thus kind of hype is a repeated event. A specialist machine outperforms a generalist machine at its specialist task, and journalists claim that the world has turned upside down. Connection Machine, Deep Blue, GAPP, transputer... Just a few I can call to mind.
I don't know about the US, but in the UK the power company are required to make the total number of cycles in a day come out pretty close to right (weak memory says withing one second) but allowed larger margins of error within the day So yes, they do positively intervene to correct the number of cycles per day.
This is old news - it has been "true" for years. It is actually a corrolary of Clarke's law ("Sufficiently advanced technology is indistinguishable from magic"). If we understand how a system works normally, then any misbehaviour it shows is a fault. If we don't, then we can classify the misbehaviour as a "neurosis". Unskilled users often believe their computer sare sufferring from a neurosis. This usually means that at some time in the past they have installed some app or extension which is trying to do something they don't understand. A more skilled user can come along and "cure" that neurosis, because they understand the system at a deeper level.
/. effect are both "twitches" in the body of the Internet. (And spam is a cancer which requires operating now) Thus far, these nervous ticks have expanded into full-scale neurosis - but they could.
A car I once had displayed what appeard to be a "neurosis" - it seemed to be frightened of going more than 30mph. It would run fine up to that speed, but if you went any faster it "paniced" and stalled. Dirt in the fuel line: at low flow rates, it lay flat and let fuel pass. At higher flow rates, it flipped up and blocked the flow completely, causing the engine to stall before it had time to flip down again. The point is, the first analysis of "neurosis" was corrected to "fault" once the problem was understood.
So the diagnosis of "neurosis" is relative - it means "I don't understand this failure mode". It can, of course, become absolute if nobody understands it.
So, are we building systems so large that nobody understands them? Definitely. Networks are already bordering on incomprehensible. Particularly, of course, the Internet. It would not surprise me at all if the Internet started showing "neurotic" behaviour. Indeed, it already does - if you ragard humans and their input as part of the net istelf. DOS attacks and the
I have to agree. I have got the BSc too, and I ducked out of science because it was too much like hard work. I went to play with computers because people would pay me to have fun while I found out what my career was really going to be. 25 years on, I still haven't found out. But some of my friends are Real Scientists. From what they tell me, it is a lot like creative software design. Of course there are dreary bits - so there are in programming, too. But the bit that distinguishes the Eagle from the Turkey is not those dreary bits, it is the insight that comes when you let the problem flood you mind and wait for the solution to rise to the surface.
Which shows that this guy doesn't know scientists. Scientists - true scientists, not technicians - are very like this guy (correctly) describes programmers. Both programming and scientific research are creative skills which, as the man says, require you to be "in the groove". He is not wrong about programmers - he is wrong about scientists. Techicians, to some extent, have less need to be "in the groove" - though much of what he says applies to any human being, with only the time constants varying.
OTOH. 3. Accommodate Reasonable Special Requests. When I get really stuck on a design problem, I go for a walk in some very beautiful woods about three miles from my office. An hours walk in the woods has about an 80% chance of delivering a solution to the problem. Even, curiously, if I don't spend much time conscioulsy thinking about the problem. In fact, I sometimes feel that by subconscious is telling my conscious to let go that problem and leave it to me. Dropping a problem for an hour or day and then coming back to it can be remarkably constuctive.
In fact, I sometimes feel embarasssed that the conscious "me" claims credit for the bundle of mad scientist, lechers, random thought generators, and idiots who inhabit my subconscious and do all the work.
As you say, OTEC is not that new in concept, there are just relatively few places that you can build shoreline OTEC plants, for reasons you give.
So how about off-shore OTEC plants? Then you have the entire surface of the Pacific to populate - plenty of room.
First question - how do you get teh energy back to land (I am not going for fresh water at this point). Easy - electrolyse seawater to Hydrogen for this new Hydrogen Economy we're all due to enjoy.
So you create a concrete (cheap) tube 1000 metres long. You insert buoyancy chanbers so it is roughly neutral density (and make the concrete with pumice, which is very light and widely available near volcanoes), but make the bottom denser than the the top so that it is naturally stable in an upright position. Add piping to pump heat from top to bottom: the tube becomes a giant chimney, and you generate electricity with turbines inside it. The thing free floats in the Pacific, mostly following the big eddies that swirl round the Pacific and occasionally using a little of its power to push itself into the right current. Tankers picking up Hydrogen have to chase it round the ocean.
Of course, the big problem with any installation at sea is storm damage. You have to make things incredibly strong so as not to be destroyed by "worst case" storms, which puts up costs enormously(and they still get damaged, if not destroyed, and are costly to repair). Not in this case. I said the tube is nearly neutral density. It is *very* nearly neutral density. When a storm approaches, it floods a few tanks and ducks down below the surface to lurk, perhaps a hundred feet down. It watches the suurface with sonar, and when the waves have subsided, bobs back up again and resumes work.
The bottom is denser than water, so the tube floats vertically, which means the rest is lighter than water for neutral buoyancy. The working bits are not at the surface, but far enough down to be out of harms way, but high enough up that they can be raised to the surface (weather permitting) for maintainance. The top few tens of feet are again denser than water. These are sacrificial: if, by mischance, they get knocked off, the average density rises, so that the whole assembly floats higher instead of sinking deeper, and can be repaired - thr erepairs, of course, beanng to the easily accessible top section. In fact, there are small explosive charges for use in an emergency, which can blow bits of this off to lighten the structure.
Assembly should be easy enough. It can be fabricated in sections, sections floated far enough offshore to get the right depth and to get into the eddy, assembled horizontally and then carefully rotated into position (arefully - this thing has the general geometry of a thin drinking straw, so you don't want it to flex too much). If the worst comes to the worst (e.g. it is about to run aground), ensure the expensive bits (turbing, pumps, heat exchangers) have positive buoyancy and blow the tube to bits. Collect up the floating bits and re-assmeble with a new tube (cheap concrete).
Ecologically interesting. Much of the open ocean of the Pacific is pretty barren, because of lack of nutrients at the top and lack of oxygen at the bottom. This is going to stir the whole thing up, which should increase the fertility of the ocean a lot. And it produces Oxygen as a waste product: I wonder what happens if you inject that at the bottom end instead of venting to the atmosphere...
Not really aimed at the geek, more aimed at maximising profit. It is rather similar to the way two identical seats on an aircraft can be sold for wildly differing pricesbecause of arcane booking rules.
While it is true that some chips will run faster than others, once the process settles down, the vast majority of chips off the line will fall into the same ballpark. But "faster" processors get a premium price. So they simly test a lot of the chips at low speeds and sell them at "ordinary" prices, then re-test some of them at high speeds and, provided they pass, sell them at premium pricecs. The ordinary price chips migt be just as good as the premium ones - or they might not. It is not that they are inherently slow, it is that they haven't been tested fast.
But if people keep overclocking, Intel will no longer be able to charge a premium price for an extra, say, 200MHz of speed.
It is scaleable - provided you have the right sort of coast, with very deep war relatively close to the shore. It will therefore be very suitable for most Pacific islands, many of which are the tops of seamounts. I don't think the coast of Calif is quite so precipitous - and the power consumption per mule of coast is probably thousands of times that of (low population, low power consumptoion, lots of beach) Pacific islands. So don't get too excited.
Do we know, mathematically, how much a kilogram is? What is the mathematical definition of a kilogram that does not depend upon fundamental constants whose value, in turn, depends upon the kilogram?
The current definition of the kilogram is "the mass of that lump of metal near Paris". We the define other constants, such as Avogadros's number, in term of that kilogram
There is no electrical contact *to* short out. It is using magnetism, not electicity. Pad generates alternating magnetic field above its surface. Put coil tuned to same frequency in field. Conductor in changing magnetic field generates current, so AC generated. Add rectifier, DC generated. I can see that the add-on to the mobile device will be cheap - a coil and 4 diodes. This could work.
One non-green point: How much power is dissipated with no devices on it, or with fully charged devices? But then, how much power is dissipated by plugged in "wall warts".
In order to air cool, you need quite a considerable volume of airspace above the CPU and probably a head sink as well, and a relativewly unobstructed airflow from front to back of the PCB. (Some opeople do side to side, but usually not more than once). This is fine for one or two CPUs, but it makes the CPU and its cooling space effectively about two inches tall. Also, you cannot have one CPU pehind another in the airflow direction, because the second will have its air preheated by the first. For supercomputing use, the processor only required ram and network IO. This means that each kernel can probably be squeezed into an area about 3 inches square - which means that, cooling aside, you coult pack them three or four deep between front and back of the back. The graphic looks like a blade architecture. With air cooling, I would have thought you would get 1 CPU per blade on a 2 in spacing, or about 8CPUs per 2U rack unit. With warer cooling, you might get CPUs per blade on a 3/4 in spacing, or about 60 CPUs per 2U rack unit (hmmm - I think that is a bit optimistic...).
Then, if you use air cooling, where does your heat come out? Into the server room. And if you have 256 CPUs at 50W each, that is over 10kW of head being dumped into your server room. You need some *serious* aircon to get that out of the room. And what is the aircon doing? Extracting the heat from the air and putting it back into some kind of fluid, which is piped to heat exchangers. So it makes sense to get the heat into fluid earlier rather than later.
The way high-performance processing is going nowadays, I can see fluid cooling making a modest return to favour at the high end of the scale. It is already returning at the low end, with high performace laptops using fluid cooling instead of fan cooling. So I don't think this is a distracion at all.
A bit of a dupe, but try this. Meets all your criteria except feel, and they claim to be good for RSI.
They the Fingerworks Touchstream keyboard - if you can afford it. The same area acts as keyboard (single finger touches), mouse, and gesture area (multiple finger touches). I haven't tries the keyboard version; I have the mouse/gesture only iGesture, and it does the job well. Not as well as a mouse, but better than al the other pointing devices I have triedl. The keyboard ought to be better, because it keeps your hands in the active area while moisung - and it is, of course, zero-force (and zero feel).