This is a misconception. Your $5 Casio will be off by a tiny fraction; e.g. 1/2 a second per day. But, it will *always* be off by the same amount, so that the error will accumulate - it will be ~3 minutes off after a year.
...
So, a quality mechanical watch may vary forward and backward by more in a single day than the cheap Casio - but the errors will very often cancel themselves, so that after a year, the Omega may well keep much better time.
I see no reason why that would be correct, a good mechanical watch may keep better time than a bad quartz watch, but not for the reason you describe.
The basis of both systems is an imperfect oscillator which runs continuously at a certain resonant frequency and is divided or multiplied by the appropriate amount so that seconds/minutes/hours are the outputs. In a digital watch it is a quartz crystal oscillator, in a mechanical watch it is a balance wheel. Both will suffer from short term variations due to finite quality factor (Q factor) of the oscillator, and the digital watch will win due to the inherently enormous Q of a quartz oscillator versus a balance wheel.
The other big drivers of variation are temperature, mechanical perturbation (i.e. vibration and shock) and aging effects. Everything has a temperature dependence, though with enough effort compensation circuits can be built for quartz oscillators, and alloys with exceptionally low temperature coefficients of size and spring constant can be employed for mechanical oscillators. If the effort has been put in to design a very low temperature coefficient wristwatch, it might outperform a low-end digital watch despite the lower Q factor.
Note that pendulum clocks can have Q factors approaching that of a quartz oscillator, and that metrology grade pendulum clocks have been designed by NBS (now NIST) which are solidly in the realm of a quartz oscillator, though they are blown out of the water by a high quality crystal oscillator for much less cost.
If people search for Armagedon do you think they want the directly matching misspelled song title (or whatever) or Armageddon the best selling movie?
There's endless cases like this, direct matching leads to worse user experience than trying to infer intent.
And yet the only case you listed was a ridiculous corner case. Chances are the GP's books' titles are not just single letter misspellings of highly well-known and popular books, and he probably should be the top result. I have definitely seen this with the Amazon results where I type in an exact title in the search box and the book I'm looking for is several results down, below both related and unrelated titles.
You are missing my point, which is that internship is the same as apprenticeship and it should be viewed as an alternative to formal education.
Internships and apprenticeships are not the same, with the former referring more to white collar job and the latter referring more to blue collar and typically having a more formal training aspect to it. Though I admit the two are similar. Nevertheless your example is poorly chosen as apprentices (at least in the US) are typically paid, because again they are working and providing value to an employer.
So if I am getting an intern, I am getting a known quantity: somebody without prior experience, and so I am absolutely not going to pay them anything or I may pay them a token amount, something definitely lower than minimum wage.
News flash: you never get a known quantity when you hire somebody. As an employer you have chosen to take on certain risks in return for the rewards of being in management or ownership of a company. If you aren't willing to take on the risk of paying somebody without credentials then that's fine, but you're wasting their time if you ask them to work for you for free.
At my very first internship where I was paid quite nicely my first two weeks or so consisted largely of handing me some textbooks and papers on the subject I would be working on, telling me to teach myself to program in Matlab, and asking me to ask questions if I got stuck. I was paid for my time to train myself of course, because hiring somebody who already knew how to do what I ended up doing for them would have cost my employer even more.
Now, what is actually funny, is that in USA an intern can work for a company for free, but should he be hired for actual money, then he must be making minimum wage.
I don't find it funny at all that a company can legally get interns to work for free, but certain federal and state exemptions do exist which allow this to happen under specific conditions, such as receiving college course credit. Do you remember your history classes where companies would hire people and pay them such utter crap that they were practically starving in the street while working full time or more? Actually that's current events in some countries... If legal exemptions were created which allowed interns to be paid less than minimum wage then all of a sudden lots of jobs would stop being jobs and would become internships methinks, and that's not good for society.
As to the other complains, that intern positions must be paid with minimum wage - that's so fucking stupid. If you are forced to pay minimum wage for an intern, then you won't be hiring an intern, will you?
Yes they will. If the intern provides value to the company then they are worth compensating. Period. If they provide no value then the intern shouldn't be there because they are costing the company money to have a staff member oversee them. They are still much cheaper than staff at the same hourly rate because there are typically no benefits provided. Furthermore interns are great for temporary projects because internships are typically understood to be temporary, there is no fuss about not rehiring them the next summer if the company doesn't have the money for it.
You'll be hiring the new University graduates, after all, they do have the degree, so then if you are forced to pay them, why take somebody who does not have a degree, when there are so many new idiots who do?
What field are you in that the university graduates are making the minimum wage which many of us think should be paid to currently unpaid interns? It should be obvious that the responsibilities for a staff and intern are quite different. Staff will be around much longer and they are better compensated so they are giving longer term projects and greater accountability for the results thereof.
I agree with your points about many people going to university for useless degrees. But I don't think encouraging unpaid internships is the way to do it. Interns get on-the-job training in a particular area, emphasis on *on-the-job*. They are doing a job so they should be paid, though at a lower rate than an experienced staff member. I think that expansion of the acceptance of professional degrees and training programs for specific work areas would be a better solution to giving people practical knowledge for a job rather than getting a four-year degree to become an office drone.
I'm an electrical engineering student, and as an undergrad in my city (admittedly one with a high cost of living) most undergraduate level engineering internships I was aware of paid something like $15/hr for new hires. At my internship after my sophomore year I was making more than that upon hire, and the company gave me annual performance reviews with bonuses and raises so I was making a pretty decent amount by the time I graduated and they hired me as staff.
So I guess what I'm trying to say is that engineering is awesome:)
But seriously I think all interns should be paid, because if the intern provides no value to the company then the company wouldn't waste their paid employees' time by bringing them on in the first place. And since they must be providing value to the company, they should be compensate with minimum wage, or more based on their value. But these companies are probably happy getting their slave labor, they don't want it to change any time soon.
No. Headphone cables have no effect on sound (as long as they are not torn or shorted).
To be fair, expensive cables usually have easier to use more durable connectors, and more flexible cabling with a more durable jacket. Features like that are certainly worth a few extra bucks.
But it's true that there is negligible effect on sound quality.
>Thanks, but no thanks, we don't want your GMO anymore, we saw what it does. Feed billions of people?
If I could pay a 50% premium on my rice, wheat flour, oats, etc. in order to ensure that none of them were GMO, I probably would. GMO may or may not be harmful, but given the unethical conduct of large agribusiness I don't really trust them to be the ones saying if it's safe or not. The problem is that the Monsanto and farm lobbies have made sure that there are no labeling requirements for GMO foods, so consumers cannot easily make an informed choice about the source of their food.
The milk I buy (no hormones, all organic, etc) is about 20 cents cheaper per gallon than the milk at the chain store down the street and literally lasts 2 - 3x as long.
Most organic milk is ultra-pasteurized which, which involves heating to a higher temperature for a shorter time than a normal pasteurization which is done for non-organic milk. This is what gives organic milk the extremely long shelf-life, though some people argue that it decreases nutritional value. The plus side of organic milk is still that it requires that cows not get BGH and not get continuous antibiotics, plus they get organic feed though it's usually still grain rather than grass.
My guess is that within a year or two, there will be better open-source alternatives to Jacket, just like there are better open source alternatives to MATLAB alrady. I'll just wait, thank you very much.
I don't dispute that there are alternatives to Matlab, but "better" is still premature in my opinion. Over the past year I had an interest in removing my work's dependence on proprietary software, so I have researched the Matlab alternatives, and I have even been using Python for some of my work (instrument automation and simple data processing and plotting), but I would say that open source has a way to go before reaching Matlab's level. For a professional user the license costs are not terribly significant, for academic, personal, or consultants, the open alternatives might make more sense.
One of the usual problems with open source which is true here is fragmentation. Matlab has a wide range of nice add-ons available for purchase which integrate with Matlab and Simulink. Some of the features of these may be replicated in other products, but rarely are they replicated as well, and they certainly are not integrated into a single architecture. No, you have certain functions available in Python libraries, one of the scientific Python distributions, or Octave, or Scilab, and you must spend much more time finding each one, finding the documentation, and integrating it into your workflow.
The second usual problem of open source software is polish. Matlab is more polished and in my opinion has a much better UI than the open alternatives I've used. Yes it has some problems, but on the whole I think their interface is better, and the documentation is better. Open source programs like many Python libraries may be comprehensively documented, but the documentation is rarely as well organized. Of course the fragmentation comes up here again: some libraries may have excellent documentation, others very poor.
This does raise one of the benefits of open source: it's open. There was a Python library (mwavepy) which I wasn't understanding too well from the documentation alone, so I was able to dive into the source code. I even thought about contributing to the project, though I haven't gotten around to it.
So in summary, I like Matlab, and I like open source. But I think Matlab is simpler and easier to get the job done, and that is what professional users care about more than a couple $k in license costs.
With modern chip fabrication techniques they could potentially integrate a ferromagnetic material deposition into a stand chip process, and then possibly they could achieve densities and speeds of storage that are relevant in today's world.
But at the end of the day the target to beat is always silicon. Even if each bit of silicon non-volatile memory is less reliable, they can integrate vast numbers of bits in a given die area, and thus they have enough storage space that they can throw tons of error correction data into the storage. Industry's decision making process for a technology like this is basically "if it can be done in silicon, do it in silicon". So there needs to be some incredibly compelling benefit to magnetic core memory to make it more valuable than silicon in a given application.
1. By the time graphene is ready to be modulated at 500GHz we will almost certainly have the technology to modulate the graphene at 500GHz. It requires a transistor with unity gain at roughly 1000GHz to do that. State of the art university and military researchers are building both analog and (very simple) digital circuits in the 300+GHz region using transistors with unity gain frequencies at 1THz and above. They are using group III-V heterostructure devices such as InGaAs/InP heterojunction bipolar transistors for this, and people continue to push the envelope on scaling silicon transistors, though the group IV materials like silicon and silicon/germanium alloy are still well behind what III-V's can do and will probably never be as fast.
2. You demodulate the signal the same way current ultra high speed optical signals are demodulated: A really fast and expensive chip(s) de-multiplexes the wideband 500GHz signal to N signals each having 500GHz/N bandwidth, so that cheaper chips can handle the lower bandwidth signals and eventually send it to customers. If we can modulate it we can demodulate it.
It will definitely cause some increased multipath fading, but that's why new wireless routers often come with the 3 antennas. With the multi-antenna configuration at least one of the antennas will have a strong signal at the laptop (and reciprocally the laptop at the router) because multipath fading is position dependent.
Well the way I understand it, in DC circuits the resistance of the wire increases as the distance of it increases. In AC, this is not true.
That's completely incorrect. For both AC and DC the resistance is rho*L/A, where rho is the resistivity of the material in ohm-meters, L is length in meters, and A is cross-sectional area in meters^2. AC actually has it slightly worse off because there is the skin effect which causes current to bunch up on the outer edges of the conductor, but for most practical situations that would be encountered in building wiring the effect is negligible.
If you RTFA you will see that the problem occurred at elevated power levels, above what consumer level equipment normally outputs. Unlike the consumer electronics that you are in love with aviation equipment needs to be far more reliable because it is "mission critical", i.e. people's lives depend on the equipment performing correctly under a variety of circumstances. This is why they test it more stringently with elevated levels of interference which is higher than what is expected during operation.
Furthermore, as is evident from the summary alone, this has occurred during EMC testing of the aircraft, i.e. the part of the testing process which identifies problems exactly like this one so that they can be fixed before delivering the plane to customers. If you got your hands on alpha or beta versions of consumer electronics you would probably find it had bugs like that as well.
The Slashdot summary barely hypes this up, the post title is clearly exaggerated but the actual summary is pretty fair. And yet you jump up, swallow the hype whole, and comment without even reading the summary.
Light will travel 300 microns (about a quarter of a millimeter) in one picosecond. If it travels through a non-air medium such as fiber optic cable or ethernet cable then it will be slowed down even more. Even in a nanosecond light in air only travels about a foot in distance.
It is not physically possible for information to move an appreciable distance at a picosecond time scale, because the fastest speed that information can move is at the speed of light.
Station keeping is cheap. Just the fuel that a shuttle carries for getting back to Earth should be enough to keep it (and the ISS!) on-station for a few years. And sure, a shuttle isn't an ideal orbital transfer vehicle. (It's got the basics - fuel, engines, habitable environment - but it's got a lot of useless stuff too.) But when it's practically free to put it in position (just leave it up there!)... why not do it?
Since when is it practically free? Last time I checked a Shuttle launch costs around a half billion dollars, that's far away from free.
Your estimate of a "few years" of station keeping fuel is highly optimistic, considering that the ISS needs to be re-boosted every time a shuttle visits. Also consider that the Shuttle gets completely overhauled on the ground after every flight, do you expect it to run for a few years with only much trickier on-orbit maintenance? The liquid hydrogen and oxygen fuel will also not last that long even if none of it is burned.
There's a similar sounding arcade I've been to called Dave and Buster's which is basically a club/bar with tons of arcade machines, they're scattered in various places all over the US. Lots of fun, but I wish one was actually close to me:(
You're right that a significant issue with noise cancellation is the dynamic range / resolution of the reciever. A lock-in amplifier does just the job: http://en.wikipedia.org/wiki/Lock-in_amplifier [wikipedia.org]
Lock-in amps are very nice for very specific tasks, like trying to measure very faint signals which would ordinarily be swamped by noise, but they don't fit in to communication systems. The principal of a lock-in amplifier relies on correlating an unknown signal with a known modulating signal over many measurements, meaning that the transmitter has to transmit the same thing over and over again, which completely defeats the purpose of trying to increase wireless data throughput.
You're correct in identifying dynamic range as one of the major issues here. An A2D with an extremely high number of bits would help, as you could recognize both the faint and large signals simultaneously, and with prior knowledge of the large transmitted signal it could be removed from the data.
But there is another issue, which is the dynamic range of the RF hardware itself. Their experiments were conducted at 0dBm (1mW) transmit power, but that is not at all realistic if you want to get decent range, a more realistic power for longer range would be 10-100mW. With 20dB (factor of 100) cancellation there is then up to 1mW heading in to the receiver, which is way way too much for typical RF CMOS, and even for typical SiGe receiver chips, the output will be completely garbled by distortion at that high power. In fact the chip they're using to do the analog cancellation (QHX220) can only take 4 microwatts in at 2.4GHz before it is completely distorting. They probably want an IIP3 (a measure of dynamic range) of minimum 10mW to limit the distortion, which is fairly large for typical receiver hardware. It's not terribly hard to do with GaAs chips (but more expensive) and it will take some work to do it in CMOS/SiGe which is cheaper.
In other words they tested under very specific conditions, and currently there needs to be additional work done on the RF hardware end to make this viable in an actual product. Not that I'm criticizing them of course, their interest is exploring more the communication aspect. If a company wants to make a product of this all it takes is money to fix the RF dynamic range issue, they're not pushing the limits of possibility yet.
Reflections/multipath will reduce the actual isolation achieved by their antenna nulling technique, but they will still achieve some improvement in isolation as no reflections are 100% and there is two-way path loss involved. As long as they get enough isolation with the antenna nulling to keep the receiver front-end from saturating then they might be able to compensate for the multipath in the digital domain, depending on the processing power available. It would be pretty simple for it to do a self-calibration, though moving objects which reflect significantly would require more frequent cals.
GPS and CDMA use something completely different. Spread spectrum techniques like GPS and CDMA take a signal with (for example) 1MHz bandwidth and spread that data over a 100MHz bandwidth. Now up to 100 people employing this technique can transmit over that 100MHz bandwidth simultaneously, but there is no gain in throughput because it's the same in the end as those 100 users transmitting in a 1MHz bandwidth with user 1 at 1.000GHz, user 2 at 1.001GHz, and so on. The benefit of spread spectrum is that it's hard to segregate each radio into such a small bandwidth without interfering with adjacent users. It could not be used for full duplex single frequency radio because the transmitted signal would still swamp out the received signal, unless it were combined with isolation/nulling techniques like these Stanford guys are using.
The research page for the work in this article is here: http://sing.stanford.edu/fullduplex/ They are using multiple techniques to selectively null out the transmit signal at the receiver. Their main novelty is spatial nulling of the antenna. Two antennas transmitting the same signal will have points in space where the signals destructively interfere and cancel. If they are spaced by an odd number of half wavelengths then this includes the entire line between the two antennas, so this is where the receive antenna is placed. Then they use existing analog and digital techniques to further cancel out the component of the transmitter which appear at the receiver.
Although the techniques for this are well known the trick is getting it to actually work effectively, because you need to achieve very high isolation from your own transmitter to receiver in order to avoid the transmitter effectively jamming the receiver. Their antenna nulling is apparently what gave them that extra isolation they needed.
While I have to applaud Google for trying to keep their users' accounts safe, I have to say that this idea is really untenable. Not everyone has a cellphone, not everyone with a phone carries it all of the time, and you might not always have reception. Just this last summer, I had a month-long internship in Nebraska. The town I stayed at had zero reception on Sprint's network and the nearest cell tower was over an hour away. And last February, I was in Switzerland, where again, I had no cell service.
Clearly then you are not well-suited to this optional extra feature, or at the very least you should not enable it while travelling abroad or in poorly developed areas. I for one think it's great that I now have the option to make my Gmail account far more secure.
Furthermore, if my bank can authenticate me without requiring an SMS, then certainly my email provider can do the same.
Does your bank even implement two-factor authentication? Mine doesn't. Of course it can easily and securely be done with RSA key fobs, but those are are fairly expensive and would require much more effort for Google to implement since they would need to snail mail you the key. It hardly makes sense for a free email account. Otherwise a phone call or text is one of the best ways to cheaply implement two-factor authentication.
A nice idea, but in addition to other criticisms presented in above comments this would not prevent a criminal from using pre-arranged codewords to communicate with outside affiliates. If I was a mob boss or drug kingpin I would probably take the time to work out such a system with my subordinates just in case, not sure if all criminals would go through the trouble of course.
Slashdot Mirror
I see no reason why that would be correct, a good mechanical watch may keep better time than a bad quartz watch, but not for the reason you describe.
The basis of both systems is an imperfect oscillator which runs continuously at a certain resonant frequency and is divided or multiplied by the appropriate amount so that seconds/minutes/hours are the outputs. In a digital watch it is a quartz crystal oscillator, in a mechanical watch it is a balance wheel. Both will suffer from short term variations due to finite quality factor (Q factor) of the oscillator, and the digital watch will win due to the inherently enormous Q of a quartz oscillator versus a balance wheel.
The other big drivers of variation are temperature, mechanical perturbation (i.e. vibration and shock) and aging effects. Everything has a temperature dependence, though with enough effort compensation circuits can be built for quartz oscillators, and alloys with exceptionally low temperature coefficients of size and spring constant can be employed for mechanical oscillators. If the effort has been put in to design a very low temperature coefficient wristwatch, it might outperform a low-end digital watch despite the lower Q factor.
Note that pendulum clocks can have Q factors approaching that of a quartz oscillator, and that metrology grade pendulum clocks have been designed by NBS (now NIST) which are solidly in the realm of a quartz oscillator, though they are blown out of the water by a high quality crystal oscillator for much less cost.
And yet the only case you listed was a ridiculous corner case. Chances are the GP's books' titles are not just single letter misspellings of highly well-known and popular books, and he probably should be the top result. I have definitely seen this with the Amazon results where I type in an exact title in the search box and the book I'm looking for is several results down, below both related and unrelated titles.
Internships and apprenticeships are not the same, with the former referring more to white collar job and the latter referring more to blue collar and typically having a more formal training aspect to it. Though I admit the two are similar. Nevertheless your example is poorly chosen as apprentices (at least in the US) are typically paid, because again they are working and providing value to an employer.
News flash: you never get a known quantity when you hire somebody. As an employer you have chosen to take on certain risks in return for the rewards of being in management or ownership of a company. If you aren't willing to take on the risk of paying somebody without credentials then that's fine, but you're wasting their time if you ask them to work for you for free.
At my very first internship where I was paid quite nicely my first two weeks or so consisted largely of handing me some textbooks and papers on the subject I would be working on, telling me to teach myself to program in Matlab, and asking me to ask questions if I got stuck. I was paid for my time to train myself of course, because hiring somebody who already knew how to do what I ended up doing for them would have cost my employer even more.
I don't find it funny at all that a company can legally get interns to work for free, but certain federal and state exemptions do exist which allow this to happen under specific conditions, such as receiving college course credit. Do you remember your history classes where companies would hire people and pay them such utter crap that they were practically starving in the street while working full time or more? Actually that's current events in some countries... If legal exemptions were created which allowed interns to be paid less than minimum wage then all of a sudden lots of jobs would stop being jobs and would become internships methinks, and that's not good for society.
Yes they will. If the intern provides value to the company then they are worth compensating. Period. If they provide no value then the intern shouldn't be there because they are costing the company money to have a staff member oversee them. They are still much cheaper than staff at the same hourly rate because there are typically no benefits provided. Furthermore interns are great for temporary projects because internships are typically understood to be temporary, there is no fuss about not rehiring them the next summer if the company doesn't have the money for it.
What field are you in that the university graduates are making the minimum wage which many of us think should be paid to currently unpaid interns? It should be obvious that the responsibilities for a staff and intern are quite different. Staff will be around much longer and they are better compensated so they are giving longer term projects and greater accountability for the results thereof.
I agree with your points about many people going to university for useless degrees. But I don't think encouraging unpaid internships is the way to do it. Interns get on-the-job training in a particular area, emphasis on *on-the-job*. They are doing a job so they should be paid, though at a lower rate than an experienced staff member. I think that expansion of the acceptance of professional degrees and training programs for specific work areas would be a better solution to giving people practical knowledge for a job rather than getting a four-year degree to become an office drone.
I'm an electrical engineering student, and as an undergrad in my city (admittedly one with a high cost of living) most undergraduate level engineering internships I was aware of paid something like $15/hr for new hires. At my internship after my sophomore year I was making more than that upon hire, and the company gave me annual performance reviews with bonuses and raises so I was making a pretty decent amount by the time I graduated and they hired me as staff.
So I guess what I'm trying to say is that engineering is awesome :)
But seriously I think all interns should be paid, because if the intern provides no value to the company then the company wouldn't waste their paid employees' time by bringing them on in the first place. And since they must be providing value to the company, they should be compensate with minimum wage, or more based on their value. But these companies are probably happy getting their slave labor, they don't want it to change any time soon.
To be fair, expensive cables usually have easier to use more durable connectors, and more flexible cabling with a more durable jacket. Features like that are certainly worth a few extra bucks.
But it's true that there is negligible effect on sound quality.
If I could pay a 50% premium on my rice, wheat flour, oats, etc. in order to ensure that none of them were GMO, I probably would. GMO may or may not be harmful, but given the unethical conduct of large agribusiness I don't really trust them to be the ones saying if it's safe or not. The problem is that the Monsanto and farm lobbies have made sure that there are no labeling requirements for GMO foods, so consumers cannot easily make an informed choice about the source of their food.
Most organic milk is ultra-pasteurized which, which involves heating to a higher temperature for a shorter time than a normal pasteurization which is done for non-organic milk. This is what gives organic milk the extremely long shelf-life, though some people argue that it decreases nutritional value. The plus side of organic milk is still that it requires that cows not get BGH and not get continuous antibiotics, plus they get organic feed though it's usually still grain rather than grass.
I don't dispute that there are alternatives to Matlab, but "better" is still premature in my opinion. Over the past year I had an interest in removing my work's dependence on proprietary software, so I have researched the Matlab alternatives, and I have even been using Python for some of my work (instrument automation and simple data processing and plotting), but I would say that open source has a way to go before reaching Matlab's level. For a professional user the license costs are not terribly significant, for academic, personal, or consultants, the open alternatives might make more sense.
One of the usual problems with open source which is true here is fragmentation. Matlab has a wide range of nice add-ons available for purchase which integrate with Matlab and Simulink. Some of the features of these may be replicated in other products, but rarely are they replicated as well, and they certainly are not integrated into a single architecture. No, you have certain functions available in Python libraries, one of the scientific Python distributions, or Octave, or Scilab, and you must spend much more time finding each one, finding the documentation, and integrating it into your workflow.
The second usual problem of open source software is polish. Matlab is more polished and in my opinion has a much better UI than the open alternatives I've used. Yes it has some problems, but on the whole I think their interface is better, and the documentation is better. Open source programs like many Python libraries may be comprehensively documented, but the documentation is rarely as well organized. Of course the fragmentation comes up here again: some libraries may have excellent documentation, others very poor.
This does raise one of the benefits of open source: it's open. There was a Python library (mwavepy) which I wasn't understanding too well from the documentation alone, so I was able to dive into the source code. I even thought about contributing to the project, though I haven't gotten around to it.
So in summary, I like Matlab, and I like open source. But I think Matlab is simpler and easier to get the job done, and that is what professional users care about more than a couple $k in license costs.
With modern chip fabrication techniques they could potentially integrate a ferromagnetic material deposition into a stand chip process, and then possibly they could achieve densities and speeds of storage that are relevant in today's world.
But at the end of the day the target to beat is always silicon. Even if each bit of silicon non-volatile memory is less reliable, they can integrate vast numbers of bits in a given die area, and thus they have enough storage space that they can throw tons of error correction data into the storage. Industry's decision making process for a technology like this is basically "if it can be done in silicon, do it in silicon". So there needs to be some incredibly compelling benefit to magnetic core memory to make it more valuable than silicon in a given application.
1. By the time graphene is ready to be modulated at 500GHz we will almost certainly have the technology to modulate the graphene at 500GHz. It requires a transistor with unity gain at roughly 1000GHz to do that. State of the art university and military researchers are building both analog and (very simple) digital circuits in the 300+GHz region using transistors with unity gain frequencies at 1THz and above. They are using group III-V heterostructure devices such as InGaAs/InP heterojunction bipolar transistors for this, and people continue to push the envelope on scaling silicon transistors, though the group IV materials like silicon and silicon/germanium alloy are still well behind what III-V's can do and will probably never be as fast.
2. You demodulate the signal the same way current ultra high speed optical signals are demodulated: A really fast and expensive chip(s) de-multiplexes the wideband 500GHz signal to N signals each having 500GHz/N bandwidth, so that cheaper chips can handle the lower bandwidth signals and eventually send it to customers. If we can modulate it we can demodulate it.
It will definitely cause some increased multipath fading, but that's why new wireless routers often come with the 3 antennas. With the multi-antenna configuration at least one of the antennas will have a strong signal at the laptop (and reciprocally the laptop at the router) because multipath fading is position dependent.
That's completely incorrect. For both AC and DC the resistance is rho*L/A, where rho is the resistivity of the material in ohm-meters, L is length in meters, and A is cross-sectional area in meters^2. AC actually has it slightly worse off because there is the skin effect which causes current to bunch up on the outer edges of the conductor, but for most practical situations that would be encountered in building wiring the effect is negligible.
If you RTFA you will see that the problem occurred at elevated power levels, above what consumer level equipment normally outputs. Unlike the consumer electronics that you are in love with aviation equipment needs to be far more reliable because it is "mission critical", i.e. people's lives depend on the equipment performing correctly under a variety of circumstances. This is why they test it more stringently with elevated levels of interference which is higher than what is expected during operation.
Furthermore, as is evident from the summary alone, this has occurred during EMC testing of the aircraft, i.e. the part of the testing process which identifies problems exactly like this one so that they can be fixed before delivering the plane to customers. If you got your hands on alpha or beta versions of consumer electronics you would probably find it had bugs like that as well.
The Slashdot summary barely hypes this up, the post title is clearly exaggerated but the actual summary is pretty fair. And yet you jump up, swallow the hype whole, and comment without even reading the summary.
Light will travel 300 microns (about a quarter of a millimeter) in one picosecond. If it travels through a non-air medium such as fiber optic cable or ethernet cable then it will be slowed down even more. Even in a nanosecond light in air only travels about a foot in distance.
It is not physically possible for information to move an appreciable distance at a picosecond time scale, because the fastest speed that information can move is at the speed of light.
Since when is it practically free? Last time I checked a Shuttle launch costs around a half billion dollars, that's far away from free.
Your estimate of a "few years" of station keeping fuel is highly optimistic, considering that the ISS needs to be re-boosted every time a shuttle visits. Also consider that the Shuttle gets completely overhauled on the ground after every flight, do you expect it to run for a few years with only much trickier on-orbit maintenance? The liquid hydrogen and oxygen fuel will also not last that long even if none of it is burned.
There's a similar sounding arcade I've been to called Dave and Buster's which is basically a club/bar with tons of arcade machines, they're scattered in various places all over the US. Lots of fun, but I wish one was actually close to me :(
Lock-in amps are very nice for very specific tasks, like trying to measure very faint signals which would ordinarily be swamped by noise, but they don't fit in to communication systems. The principal of a lock-in amplifier relies on correlating an unknown signal with a known modulating signal over many measurements, meaning that the transmitter has to transmit the same thing over and over again, which completely defeats the purpose of trying to increase wireless data throughput.
You're correct in identifying dynamic range as one of the major issues here. An A2D with an extremely high number of bits would help, as you could recognize both the faint and large signals simultaneously, and with prior knowledge of the large transmitted signal it could be removed from the data.
But there is another issue, which is the dynamic range of the RF hardware itself. Their experiments were conducted at 0dBm (1mW) transmit power, but that is not at all realistic if you want to get decent range, a more realistic power for longer range would be 10-100mW. With 20dB (factor of 100) cancellation there is then up to 1mW heading in to the receiver, which is way way too much for typical RF CMOS, and even for typical SiGe receiver chips, the output will be completely garbled by distortion at that high power. In fact the chip they're using to do the analog cancellation (QHX220) can only take 4 microwatts in at 2.4GHz before it is completely distorting. They probably want an IIP3 (a measure of dynamic range) of minimum 10mW to limit the distortion, which is fairly large for typical receiver hardware. It's not terribly hard to do with GaAs chips (but more expensive) and it will take some work to do it in CMOS/SiGe which is cheaper.
In other words they tested under very specific conditions, and currently there needs to be additional work done on the RF hardware end to make this viable in an actual product. Not that I'm criticizing them of course, their interest is exploring more the communication aspect. If a company wants to make a product of this all it takes is money to fix the RF dynamic range issue, they're not pushing the limits of possibility yet.
Reflections/multipath will reduce the actual isolation achieved by their antenna nulling technique, but they will still achieve some improvement in isolation as no reflections are 100% and there is two-way path loss involved. As long as they get enough isolation with the antenna nulling to keep the receiver front-end from saturating then they might be able to compensate for the multipath in the digital domain, depending on the processing power available. It would be pretty simple for it to do a self-calibration, though moving objects which reflect significantly would require more frequent cals.
GPS and CDMA use something completely different. Spread spectrum techniques like GPS and CDMA take a signal with (for example) 1MHz bandwidth and spread that data over a 100MHz bandwidth. Now up to 100 people employing this technique can transmit over that 100MHz bandwidth simultaneously, but there is no gain in throughput because it's the same in the end as those 100 users transmitting in a 1MHz bandwidth with user 1 at 1.000GHz, user 2 at 1.001GHz, and so on. The benefit of spread spectrum is that it's hard to segregate each radio into such a small bandwidth without interfering with adjacent users. It could not be used for full duplex single frequency radio because the transmitted signal would still swamp out the received signal, unless it were combined with isolation/nulling techniques like these Stanford guys are using.
The research page for the work in this article is here: http://sing.stanford.edu/fullduplex/
They are using multiple techniques to selectively null out the transmit signal at the receiver. Their main novelty is spatial nulling of the antenna. Two antennas transmitting the same signal will have points in space where the signals destructively interfere and cancel. If they are spaced by an odd number of half wavelengths then this includes the entire line between the two antennas, so this is where the receive antenna is placed. Then they use existing analog and digital techniques to further cancel out the component of the transmitter which appear at the receiver.
Although the techniques for this are well known the trick is getting it to actually work effectively, because you need to achieve very high isolation from your own transmitter to receiver in order to avoid the transmitter effectively jamming the receiver. Their antenna nulling is apparently what gave them that extra isolation they needed.
This PHD Comics issue is particularly appropriate here:
http://www.phdcomics.com/comics.php?f=878
Clearly then you are not well-suited to this optional extra feature, or at the very least you should not enable it while travelling abroad or in poorly developed areas. I for one think it's great that I now have the option to make my Gmail account far more secure.
Does your bank even implement two-factor authentication? Mine doesn't. Of course it can easily and securely be done with RSA key fobs, but those are are fairly expensive and would require much more effort for Google to implement since they would need to snail mail you the key. It hardly makes sense for a free email account. Otherwise a phone call or text is one of the best ways to cheaply implement two-factor authentication.
A nice idea, but in addition to other criticisms presented in above comments this would not prevent a criminal from using pre-arranged codewords to communicate with outside affiliates. If I was a mob boss or drug kingpin I would probably take the time to work out such a system with my subordinates just in case, not sure if all criminals would go through the trouble of course.
Or you need to be William Tell.