Especially sensitive devices such as medical and safety relevant devices should not be a black box where it is illegal to look into the inner workings. While third-party liability is nice this is still just based on trust and not on tests. My trust into these system would increase quite a bit if a hacker plays around with a utility meter and finds no obvious vulnerability.
I want all my devices unlocked, the liability can be linked to a tamperproof soft/hardware seal as it is already done today. This is fine with me, I do not expect the manufacturer to be liable if I took it apart, hacked it and reassembled it but I do not see any advantage in making hacking illegal.
The idea to reproduce important results is good and is part of the scientific method. In practice this is much harder to accomplish due to several constraints. I can only speak for my field but I think this applies to other fields as well that the reproduction is hard by itself.
This leads us to a bunch of problems. If it takes a graduate student a year to collect a data set on a custom made machine that is expensive and time consuming who has the resources to reproduce the results? In most branches we are limited by the available personnel. It is hard to imagine giving someone the task of 'just' reproducing someone else's result, as this does not generate high-impact publications nor can be used for grant applications.
The thought behind this would benefit the scientific progress, especially to weed out questionable results that can lead you far off track but someone needs to do it. And it better not be me, as I need time for my own research to publish new results. Any reviewer always asks him/herself whether this is really an achievement that it is worth publishing, which reviewer would accept a paper stating "We reproduced X and did not find any deviations from the already published results" ?
This is exactly what someone would come up with as some kind of "challenge" where the outcome is so obviously biased.
In none of these videos the speed of any smartphone or operating system is on the test. It is the speed of the individual people using their phones. I tried some of these challenges, like a local search for a restaurant. After a bit of practice I could get an answer on my 3 year old and slow smartphone after 6 seconds. From those 6 seconds I spend about 1 second waiting for my phone, the rest is the network connection and my typing speed. So any improvement on the hardware or software side can only influence the 1 second. Other than that the phone has to guess what I think to prevent the slow typing or speech recognition part. My first try took 20 seconds as I missed a few shortcuts and this is how you can beat almost everybody by claiming to have a faster phone when you just know exactly what to touch/type.
And it's *always* cheaper to in-source [...] You can either do it yourself, or you can pay someone their cost, which could be your cost, plus 20% or more overhead and profit.
I agree that in-sourcing can be cheaper as you do not have to pay for overhead and profit. Your argument relies on one premise, which is completely broken.
Your cost to do something is almost never their cost. So if they can get to the same result with 50% or sometimes far less due to economies of scale adding another 20% is still much cheaper than paying for in-source. Even if you in-source you have to pay for the whole infrastructure. Your email server needs a backup, UPS and staff, whereas the cost of a UPS in a large datacenter is split over millions of users.
Unfortunately the situation is not as simple. Just one example: With physical items it takes time and effort to share them. You own a book, the neighbor down the street might own the same book. The likelihood that you are reading it at the same time is pretty low, so why not share it? In a digital world this is no problem, just send some bits, use dropbox or something similar.
Or think of network licences in a company. For a lot of special programs we only have around 10 licences for 250 employees. It's never a problem usually, as very few people are working with the same software at the same time. So we only pay for 10 instead of 250 because sharing the licences is seamless. If the software company would charge the same amount for the 10 network licences than for 250 regular licences they would damage their business model quite a bit. In comparison we use a lot of reference books. I would guess we have around 100 copies of the most used ones, so that you don't waste your time looking for one. It would be painful to only have 10 and then search the whole place to get hold of one.
This is just one aspect but illustrates that there is a conflict. I personally also prefer a game that I can buy and sell as I want to like a physical item but I have tons of games that I only play for a few days per year. In principle I could give away all those games for most of the time on some kind of lending model but that would definitely influence sales of all those companies that produce games, software or other digital items.
While I do not want to defend the journals I think your comparison with Bernie Madoff does not work here.
While he might have been well respected, he had an incentive to cheat and abuse the trust by putting the money in his own pocket. Why should any journal profit from suppressing or pushing a certain kind of research? It is more the other way around: as an editor I would be looking for breakthroughs and unusual findings as they increase the influence of the journal.
The newspaper article is not giving any information that is not already included in the summary.
The paper is published in Optics Express, the abstract can be read here. The full article is behind a paywall unfortunately. The author claim that this concept could deliver random numbers at a rate of 100 GHz which is quite fast compared to other true random number generators out there that are based on thermal noise, radiation or other processes.
In short: THz penetrates your T-shirt (airport scanners) but not any thin drywall.
Roughly speaking for electromagnetic waves the higher the frequency the more light-like the radiation becomes. THz is close to infrared light, it will not penetrate much but can be used to transmit a lot of data because you can modulate it with a much higher frequency than standard 2.4 GHz wireless LAN. This comes at a price though, if a person walks through the line-of-sight between your notebook and the hypothetical THz wireless access point the signal will be cut off immediately. So it is a nice idea to replace HDMI cables or similar connections but might not work so well as a WLAN replacement.
There are also people working on modulating your LED lighting to transmit data without cables (OFDM Visible Light Wireless Communication Based on White LEDs) and this is a nice example that in the future we might use the whole available spectrum to transmit information and saturation will not be a big problem anymore.
So this would be the answer how it is possible to display so much information with so little memory. That means that the CPU basically has 320x240*24 bit = 225K of video memory to draw the picture in.
How is this possible with so little RAM? It has an image viewer app. A single 320x240 8 bit image is 75K and somehow the LCD needs that data for smooth scrolling at 25fps. So does the processor basically generates a few lines of pixels, sends those to the LCD and then continues rendering/reading from SD card the next few lines and manages all that at 25fps?
1. Yes, they can see each other. Why should they not? You can hear a fighter jet flying faster than the speed of sound easily, similar here. But don't get confused: you will only ever be able to see the past "image" of the other object, this image is traveling towards you with c.
2. More than 1 year. Your idea of a rod is not quite right. Think of it like a big rubber band, then make it stiffer and stiffer. If you pull too hard you would rip off one end but in any case they would probably not notice it for way more than a year (speed of sound).
While this sounds good in theory there is one big problem with this:
Until today it is not possible to predict a complex molecule because of the number of interactions between the atoms and electrons.
There are at least a hundred different specialized algorithms that may predict certain properties but can fail completely on others. Additionally the physics of a single molecule might be quite different from a cube of the same material. These challenges keep a lot of condensed matter theoreticians busy since over sixty years and are most likely not solved with any amount of computing power. On the other hand there are ideas to use quantum computers, because they are basically a model of the real thing and maybe that will result in a breakthrough.
So while your proposal is a respectable idea it won't work so easily in practice. If you look at big research institutions a lot of chemist already sit in front of their computers and try to get something useful out of computational chemistry software and sometimes they even succeed but this is far from a certain process that would lead to results with a genetic algorithm.
The interesting bit here is something different though. I have never seen that someone wants to use a phase change material for buildings, but why not? For coffee cups this already works nicely. The walls of the mug contain a material that is undergoing some phase transition (liquid to solid, different crystalline structure, magnetic, etc.) at a temperature that is slightly below really hot coffee but still a nice drinking temperature.
What happens is the following: the thermal energy of the coffee gets absorbed quickly by the material, therefore cooling it down fast from really hot to a lower temperature. The material can store a large amount of thermal energy and releases it slowly so that the coffee stays at a constant temperature for much longer (gizmag article).
For a whole building this makes a lot of sense as well. It more or less acts as a large thermal reservoir, so that your wall temperature does not increase during the day and falls too much during the night. You could achieve a somewhat similar effect by using 20 inch stone walls but this might be a bit easier to incorporate into modern buildings.
This sounds like a pretty nice and simple idea to me.
The extra amount of traffic does not matter, just a few bytes for the passwords and the delay does not really matter. Additionally that helps you if someone stole your phone as you could easily add some information about the current location.
One loophole is that you have to disable access/decryption instantly after your phone is missing, otherwise interception of the traffic would give the attacker the unencrypted password.
This is clearly a study that is not worth much without the raw data open and accessible for everyone.
Maybe a single household near a mountaintop would cost several million dollar to connect but quite a few others could be done for a thousand. So before someone can make any political conclusions it is definitely worth to look at the actual data behind this.
The specs of the N810 look really nice. I could not find any of them used but I will look further. How can you control your camera and motors with this? I guess the USB port does not work as a host. Can you access any GPIO pins or something similar to connect any hardware to it?
We can only hope that this device or a different version of it gets a small GPIO connector for the connection of external sensors, devices, etc. With a $10 device with display and a 400 MHz cpu is really incredible in my opinion. After looking into Arduino, Beagleboard and similar inexpensive and relatively easy to program and use boards I am still looking for something that already comes in a case, is more powerful than an AVR and has a builtin display.
Almost all test and measurement devices that I currently use in our research laboratory have much less computing power which limits their capabilities and increase their price quite a bit. A couple of years ago developing embedded applications looked like black magic to me, fiddling around to save a few bytes, using a lot of tricks to get it done somehow but today you can easily throw a much more powerful processor at the problem and instead of tuning you can just program in whatever language you like and it will probably be fast enough.
If they can keep up to their announced sales price I will order a couple as nice presents another couple to take apart.
It seems that WebGL is basically an experiment how to implement a subset of OpenGL that can be part of a webpage. This experiment more or less ended in a standard which allows the website to use the graphics card to it's full extend.
As shaders are turing-complete and we do not have a secure IOMMU in every computer but the real possibility of access a lot of memory which the website should not be allowed to. Yesterday a new exploit was published which underlines this point (Exploit from contextis.com) So this boils down to a nice idea for some internal stuff, kind of downloading an executable and showing the results right in your browser. If we do not want to repeat ActiveX for the GPU instead of the CPU and all it's problems with blacklists, etc. there are very few viable alternatives. Either something like Java or Google's native client which provide a more or less secure sandbox or a good security architecture in the graphics driver which prevents these exploits.
Until one of these security measures are in place it is hard not to agree that WebGL is a big security risk and should not be used for websites out of your control.
It is not only about refresh rate and shutter speed. One key problem is brightness and ghosting. If you do basically a time multiplex and want to achieve the same brightness level you have to boost the amount of light coming out of the screen quite a bit. Additionally all shutter glasses are not 100% dark, so even now you often see ghost images.
While the reasoning of Merkel's government seems to be based on fear and emotions of the general public the background behind this is the nuclear waste.
Fukushima is just an example that a complex technology like nuclear power can fail, even with a lot of safeguards in place and in a high-tech country like Japan. It is now obvious that Tepco did not do their homework correctly and that it is just a bad idea in general to build a power plant where a tsunami can hit the shore but this is only the catalyst for the debate in Germany. The main problem is and will be in the future the massive amounts of nuclear waste, with high and medium radiation levels. The situation in Germany for waste disposal is abysmal. In the 1960s due to political issues only two underground mines were seriously examined if they can keep the waste safe for eternity until the radiation levels are low enough to be harmless. These two mines are Asse and Gorleben.
It is now very clear that during the last decades a lot of negative security reports for both mines were downplayed or never published. Asse is currently more or less flooded from groundwater penetrating the salt and while Gorleben seems safe today serious cracks have been discovered. So there is no place in Germany were we could safely store nuclear waste at all. The consensus was for a while to search for better places and it was obvious that any politician will fight tooth and nail against a mine in his district.
At the same time Germany tries to increase the amount of renewable energy and is quite successful. Merkel's current move is certainly not completely ruled by reason but it fits into the bigger picture and the last thing she wants is large demonstrations and her being seen as a cold technocrat which almost brought her a defeat in the last election.
While I personally like nuclear power much more than polluting the air with coal power plants, were the emissions also contain a lot of radioactivity and of course CO2 it feels irresponsible to use a technology as long as the waste problem is completely unsolved, at least in Germany.
The HDMI spec requires a 55mA supply at 5V. This seems to be enough to power this little computer. It might not work with a lot of usb devices without a hub that has external power but a keyboard should be possible.
Slashdot Mirror
Unfortunately the press release is a little short on details. Here is the link to the actual article (paywalled):
"1.25 Gbit/s Visible Light WDM Link based on DMT Modulation of a Single RGB LED Luminary", opticsinfobase.org
Especially sensitive devices such as medical and safety relevant devices should not be a black box where it is illegal to look into the inner workings. While third-party liability is nice this is still just based on trust and not on tests. My trust into these system would increase quite a bit if a hacker plays around with a utility meter and finds no obvious vulnerability.
I want all my devices unlocked, the liability can be linked to a tamperproof soft/hardware seal as it is already done today. This is fine with me, I do not expect the manufacturer to be liable if I took it apart, hacked it and reassembled it but I do not see any advantage in making hacking illegal.
The idea to reproduce important results is good and is part of the scientific method. In practice this is much harder to accomplish due to several constraints. I can only speak for my field but I think this applies to other fields as well that the reproduction is hard by itself.
This leads us to a bunch of problems. If it takes a graduate student a year to collect a data set on a custom made machine that is expensive and time consuming who has the resources to reproduce the results? In most branches we are limited by the available personnel. It is hard to imagine giving someone the task of 'just' reproducing someone else's result, as this does not generate high-impact publications nor can be used for grant applications.
The thought behind this would benefit the scientific progress, especially to weed out questionable results that can lead you far off track but someone needs to do it. And it better not be me, as I need time for my own research to publish new results. Any reviewer always asks him/herself whether this is really an achievement that it is worth publishing, which reviewer would accept a paper stating "We reproduced X and did not find any deviations from the already published results" ?
This is exactly what someone would come up with as some kind of "challenge" where the outcome is so obviously biased.
In none of these videos the speed of any smartphone or operating system is on the test. It is the speed of the individual people using their phones. I tried some of these challenges, like a local search for a restaurant. After a bit of practice I could get an answer on my 3 year old and slow smartphone after 6 seconds. From those 6 seconds I spend about 1 second waiting for my phone, the rest is the network connection and my typing speed.
So any improvement on the hardware or software side can only influence the 1 second. Other than that the phone has to guess what I think to prevent the slow typing or speech recognition part.
My first try took 20 seconds as I missed a few shortcuts and this is how you can beat almost everybody by claiming to have a faster phone when you just know exactly what to touch/type.
And it's *always* cheaper to in-source [...] You can either do it yourself, or you can pay someone their cost, which could be your cost, plus 20% or more overhead and profit.
I agree that in-sourcing can be cheaper as you do not have to pay for overhead and profit. Your argument relies on one premise, which is completely broken.
Your cost to do something is almost never their cost. So if they can get to the same result with 50% or sometimes far less due to economies of scale adding another 20% is still much cheaper than paying for in-source. Even if you in-source you have to pay for the whole infrastructure. Your email server needs a backup, UPS and staff, whereas the cost of a UPS in a large datacenter is split over millions of users.
Unfortunately the situation is not as simple. Just one example: With physical items it takes time and effort to share them. You own a book, the neighbor down the street might own the same book. The likelihood that you are reading it at the same time is pretty low, so why not share it? In a digital world this is no problem, just send some bits, use dropbox or something similar.
Or think of network licences in a company. For a lot of special programs we only have around 10 licences for 250 employees. It's never a problem usually, as very few people are working with the same software at the same time. So we only pay for 10 instead of 250 because sharing the licences is seamless. If the software company would charge the same amount for the 10 network licences than for 250 regular licences they would damage their business model quite a bit. In comparison we use a lot of reference books. I would guess we have around 100 copies of the most used ones, so that you don't waste your time looking for one. It would be painful to only have 10 and then search the whole place to get hold of one.
This is just one aspect but illustrates that there is a conflict. I personally also prefer a game that I can buy and sell as I want to like a physical item but I have tons of games that I only play for a few days per year. In principle I could give away all those games for most of the time on some kind of lending model but that would definitely influence sales of all those companies that produce games, software or other digital items.
recent publications. Good luck trying to find and email an author from a 1980ies paper.
While I do not want to defend the journals I think your comparison with Bernie Madoff does not work here.
While he might have been well respected, he had an incentive to cheat and abuse the trust by putting the money in his own pocket. Why should any journal profit from suppressing or pushing a certain kind of research? It is more the other way around: as an editor I would be looking for breakthroughs and unusual findings as they increase the influence of the journal.
The newspaper article is not giving any information that is not already included in the summary.
The paper is published in Optics Express, the abstract can be read here. The full article is behind a paywall unfortunately. The author claim that this concept could deliver random numbers at a rate of 100 GHz which is quite fast compared to other true random number generators out there that are based on thermal noise, radiation or other processes.
In short: THz penetrates your T-shirt (airport scanners) but not any thin drywall.
Roughly speaking for electromagnetic waves the higher the frequency the more light-like the radiation becomes. THz is close to infrared light, it will not penetrate much but can be used to transmit a lot of data because you can modulate it with a much higher frequency than standard 2.4 GHz wireless LAN. This comes at a price though, if a person walks through the line-of-sight between your notebook and the hypothetical THz wireless access point the signal will be cut off immediately. So it is a nice idea to replace HDMI cables or similar connections but might not work so well as a WLAN replacement.
There are also people working on modulating your LED lighting to transmit data without cables (OFDM Visible Light Wireless Communication Based on White LEDs) and this is a nice example that in the future we might use the whole available spectrum to transmit information and saturation will not be a big problem anymore.
So this would be the answer how it is possible to display so much information with so little memory. That means that the CPU basically has 320x240*24 bit = 225K of video memory to draw the picture in.
How is this possible with so little RAM? It has an image viewer app. A single 320x240 8 bit image is 75K and somehow the LCD needs that data for smooth scrolling at 25fps. So does the processor basically generates a few lines of pixels, sends those to the LCD and then continues rendering/reading from SD card the next few lines and manages all that at 25fps?
1. Yes, they can see each other. Why should they not? You can hear a fighter jet flying faster than the speed of sound easily, similar here. But don't get confused: you will only ever be able to see the past "image" of the other object, this image is traveling towards you with c.
2. More than 1 year. Your idea of a rod is not quite right. Think of it like a big rubber band, then make it stiffer and stiffer. If you pull too hard you would rip off one end but in any case they would probably not notice it for way more than a year (speed of sound).
While this sounds good in theory there is one big problem with this:
Until today it is not possible to predict a complex molecule because of the number of interactions between the atoms and electrons.
There are at least a hundred different specialized algorithms that may predict certain properties but can fail completely on others. Additionally the physics of a single molecule might be quite different from a cube of the same material. These challenges keep a lot of condensed matter theoreticians busy since over sixty years and are most likely not solved with any amount of computing power. On the other hand there are ideas to use quantum computers, because they are basically a model of the real thing and maybe that will result in a breakthrough.
So while your proposal is a respectable idea it won't work so easily in practice. If you look at big research institutions a lot of chemist already sit in front of their computers and try to get something useful out of computational chemistry software and sometimes they even succeed but this is far from a certain process that would lead to results with a genetic algorithm.
Yes, every material absorbs and releases heat.
The interesting bit here is something different though. I have never seen that someone wants to use a phase change material for buildings, but why not? For coffee cups this already works nicely. The walls of the mug contain a material that is undergoing some phase transition (liquid to solid, different crystalline structure, magnetic, etc.) at a temperature that is slightly below really hot coffee but still a nice drinking temperature.
What happens is the following: the thermal energy of the coffee gets absorbed quickly by the material, therefore cooling it down fast from really hot to a lower temperature. The material can store a large amount of thermal energy and releases it slowly so that the coffee stays at a constant temperature for much longer (gizmag article).
For a whole building this makes a lot of sense as well. It more or less acts as a large thermal reservoir, so that your wall temperature does not increase during the day and falls too much during the night. You could achieve a somewhat similar effect by using 20 inch stone walls but this might be a bit easier to incorporate into modern buildings.
This sounds like a pretty nice and simple idea to me.
The extra amount of traffic does not matter, just a few bytes for the passwords and the delay does not really matter. Additionally that helps you if someone stole your phone as you could easily add some information about the current location.
One loophole is that you have to disable access/decryption instantly after your phone is missing, otherwise interception of the traffic would give the attacker the unencrypted password.
This is clearly a study that is not worth much without the raw data open and accessible for everyone.
Maybe a single household near a mountaintop would cost several million dollar to connect but quite a few others could be done for a thousand. So before someone can make any political conclusions it is definitely worth to look at the actual data behind this.
Of course it does not increase trust, that the website (Social science research network) is currently down.
Additionally, "well regulated" means "well trained and organized".
Interesting conclusion. Usually regulated means in this context according to the Merriam-Webster's dictionary:
to bring under the control of law or constituted authority
That implies something completely different from what you have in mind. More similar to the National Guard instead of some guys in the backyard.
On the Rybka website you can buy the current version of either Rybka 4 for 36 Euros or Deep Rypka 4 for 65 Euros.
If these downloads contain any of the stolen code in their binaries this definitely counts as distribution.
The specs of the N810 look really nice. I could not find any of them used but I will look further. How can you control your camera and motors with this? I guess the USB port does not work as a host. Can you access any GPIO pins or something similar to connect any hardware to it?
We can only hope that this device or a different version of it gets a small GPIO connector for the connection of external sensors, devices, etc. With a $10 device with display and a 400 MHz cpu is really incredible in my opinion. After looking into Arduino, Beagleboard and similar inexpensive and relatively easy to program and use boards I am still looking for something that already comes in a case, is more powerful than an AVR and has a builtin display.
Almost all test and measurement devices that I currently use in our research laboratory have much less computing power which limits their capabilities and increase their price quite a bit. A couple of years ago developing embedded applications looked like black magic to me, fiddling around to save a few bytes, using a lot of tricks to get it done somehow but today you can easily throw a much more powerful processor at the problem and instead of tuning you can just program in whatever language you like and it will probably be fast enough.
If they can keep up to their announced sales price I will order a couple as nice presents another couple to take apart.
It seems that WebGL is basically an experiment how to implement a subset of OpenGL that can be part of a webpage. This experiment more or less ended in a standard which allows the website to use the graphics card to it's full extend.
As shaders are turing-complete and we do not have a secure IOMMU in every computer but the real possibility of access a lot of memory which the website should not be allowed to. Yesterday a new exploit was published which underlines this point (Exploit from contextis.com) So this boils down to a nice idea for some internal stuff, kind of downloading an executable and showing the results right in your browser. If we do not want to repeat ActiveX for the GPU instead of the CPU and all it's problems with blacklists, etc. there are very few viable alternatives. Either something like Java or Google's native client which provide a more or less secure sandbox or a good security architecture in the graphics driver which prevents these exploits.
Until one of these security measures are in place it is hard not to agree that WebGL is a big security risk and should not be used for websites out of your control.
It is not only about refresh rate and shutter speed. One key problem is brightness and ghosting. If you do basically a time multiplex and want to achieve the same brightness level you have to boost the amount of light coming out of the screen quite a bit. Additionally all shutter glasses are not 100% dark, so even now you often see ghost images.
While the reasoning of Merkel's government seems to be based on fear and emotions of the general public the background behind this is the nuclear waste.
Fukushima is just an example that a complex technology like nuclear power can fail, even with a lot of safeguards in place and in a high-tech country like Japan. It is now obvious that Tepco did not do their homework correctly and that it is just a bad idea in general to build a power plant where a tsunami can hit the shore but this is only the catalyst for the debate in Germany. The main problem is and will be in the future the massive amounts of nuclear waste, with high and medium radiation levels. The situation in Germany for waste disposal is abysmal. In the 1960s due to political issues only two underground mines were seriously examined if they can keep the waste safe for eternity until the radiation levels are low enough to be harmless. These two mines are Asse and Gorleben.
It is now very clear that during the last decades a lot of negative security reports for both mines were downplayed or never published. Asse is currently more or less flooded from groundwater penetrating the salt and while Gorleben seems safe today serious cracks have been discovered. So there is no place in Germany were we could safely store nuclear waste at all. The consensus was for a while to search for better places and it was obvious that any politician will fight tooth and nail against a mine in his district.
At the same time Germany tries to increase the amount of renewable energy and is quite successful. Merkel's current move is certainly not completely ruled by reason but it fits into the bigger picture and the last thing she wants is large demonstrations and her being seen as a cold technocrat which almost brought her a defeat in the last election.
While I personally like nuclear power much more than polluting the air with coal power plants, were the emissions also contain a lot of radioactivity and of course CO2 it feels irresponsible to use a technology as long as the waste problem is completely unsolved, at least in Germany.
The HDMI spec requires a 55mA supply at 5V. This seems to be enough to power this little computer.
It might not work with a lot of usb devices without a hub that has external power but a keyboard should be possible.