A virus needs DNA (or RNA) and a protective shell that helps protect and deliver it's payload. Here there's no shell.
Also, any commercial hardware and all research hardware i know of would encrypt the data before writing to prevent this from happening intentionally. No implemented encryption scheme is perfect but it would be far easier to just order your "virus" DNA with a credit card then it would be to hack a DNA hard drive that only makes DNA a 100-200 bases long.
Actually I think that millions of times faster statement in the summary came from a misquote or misunderstanding. For archival storage(write once read VERY seldom). It should have been millions (or trillions really) more bytes. 20 hours to write an archive and 1 hour to read it isn't bad. I believe amazon glacier has a similar read latency.
This is an excellent question! In principle any (hetero)polymer would work.
For context, in the short term we're targeting archival storage and not high speed storage (it'll probably never be low latency in the same way an SRAM is). The amazing thing about DNA is that it's not only long lived under reasonable storage conditions but also eternally relevant. Try reading a 30 year old 8" floppy today. The data may still be okay but you'll have trouble finding the hardware. Since DNA is so important to humans in other contexts (medicine), we can be fairly certain that DNA reading (sequencing) will be easy and available in 30 years, or 300, or 3000.
DNA is also much easier to manipulate. Nature provides us with all the tools (enzymes) we need to copy and select specific sequences. Academia provides us with a better understanding of DNA than other polymers. And industry provides us with wonderful machines like high throughput DNA sequencers. Watson-Crick base paring also allows DNA to do computation, which in some cases can be done directly on the data encoding DNA.
As for safety, you are right in that many microbes are quite good at finding and incorporating DNA but such an even would still be quite rare and there are many things to consider: 1) DNA data payloads are stored as just DNA, typically frozen or dried. And DNA alone is not sufficient to be pathogenic. 2) Generally when scientists store data in DNA right now it's encrypted first. This make it hard to pick out payload data that will encode for something specific. 3) It's far easier for a malicious person to just get parts of genes synthesized from different vendors and stitch them together at home. That ship has already sailed. 4) Currently the best way to synthesize the data payloads is in short sequences of DNA that aren't long enough to encode for anything but short peptides. 5) Randomly coming across a working pathogenic DNA sequence is technically possible but practically impossible (much like it's possible that you'll spontaneously quantum tunnel to the center of the earth but that'll never actually happen). An average gene in e. coli is about 8000 bases long. Of that size alone there are 4^(8000) (a beyond astronomically large number) of DNA sequences and only a handful of those do anything at all. Of those most would be toxic to the microbe it self or be a useless metabolic load (proteins are expensive to make and useless genes get cut out or turned off VERY fast in a population---days in continuous culture). 6) If desired, you can also trade off a little density to insert stop codons occasionally, limiting the size of anything translated to only a few amino acids.
Saying tape has a longer life is silly. I'd have no idea where to get an 8-track player today even though it's an analog format.
Same with a record player, but I could make one pretty easily. (there's a reason why we shot a record into space instead of a tape)
Really, though a documented and uncompressed digital file, properly kept track of, could last forever similar to a record even if we lost our codecs it would be easy to write a new one.
Other respondents to this comment pointed out that these are fluorescent (ie need an excitation light source) not "glowing".
But another problem with this idea is that, in a population neutral alleles maintain their frequency (though can drift randomly) and deleterious alleles will decline in frequency. In other words, you'd have to release a LOT of rabbits before the glowing allele would be common enough to have an effect on average fitness and that allele would be unstable in the population. Unless you don't believe in evolution, but then you're better off praying for the rabbits to go away.
It is seriously about time! The D-pad on the current controllers is absolutely awful (for those who don't know, often times it registers the wrong direction due to poor mechanical design). It has made some games (especially live-arcade games like megaman 9) incredibly frustrating.
Its naive to think that buses are only good for storage. 1080p video is 1080*1920pixels/frame*32bits/pixel*60frames/second is roughly 4Gb (uncompressed). If history is any indication displays will get larger and denser (also remember bandwidth needed will scale by the square of the number of lines). Or what if you want an external GPU? 16x pci express is about 32Gbits/sec. Also more speed reduces latency, which may be helpful too depending on what you use it for.
This is not some boring super scaler! Nor is it some vector processor!
in fact this is a complete departure from a von Neuman architecture. The architecture is called a Dataflow architecture. In one sentence a dataflow architecture is one where instruction execution is based on the availability of the instructions inputs not a program counter.
The article does a very bad job at conveying the fact that this is a relatively new idea. Like most reporting they report something thats been in research for some time as a huge breakthrough without describing it at all. Instead its really just an incremental step in dataflow computing research.
I work in a lab at the University of Washington on another dataflow architecture. Its a really interesting idea but it will take some time to develop and you're not going to get one on your desk for some years to come.
It really is different. Its not simply a super scaler. It's a data-flow machine. What this means is that instructions are arranged in a graph based on dependency and execute as soon as all inputs are ready.
I work in a lab at the University of Washington where we are working on _implementing_ a different data flow machine that shares some of fundamentals with the UT machine.
I was deleting a programs registry and I got distracted...
Next thing I knew I clicked yes to "are you sure you want to delete HKLM"... needless to say windows died that day.
Re:Took mine apart
on
Old Toy Modding?
·
· Score: 5, Interesting
"If a battery contains liquids, as the battery provides charge, the concentrations of these reactants change over time. The change of these concentrations cause the voltage of the battery to drop over time. This is the case with your average consumer battery -- items are in solution, so concentrations change over time. And then the voltage drops. Hence, your motor receives a different voltage, and may provide less movement for the same amount of time being activated. "
Ah, Now I understand. This is an over simplification. A cell voltage will decrease as it becomes discharged. However, if this were the reason motors did not get the same ammout of power at time A as they did at time B then it could be easily fixed with electronics.
A battery can be moddled as voltage source with an internal resistance in series. As a battery discharges this resistance increses.
While this is not 100% accurate (As cell voltage does change too.) It models the important fact that A cells capacity to provide current decreases with time and while a cells voltage may go from 1.2v (full) to 1.0v (dead) (for NiCad) the internal resistance will change much more dramatically over the life span of the battery.
"However, I was a little disappointed you didn't provide another suggestion. Are accelerometers a good way of actually measuring movement? Would you recommend, perhaps placing a rigid shaft on the ground, rotating about that, and having coded marking on that shaft to indicate movement? Myself, all I know is that a shaft encoder greatly improves performance relative to no correction mechanism. "
The reason I did not say what can be done is that it is very hard to do. Typically what is done is there is some other system on board that tells the robot where to go. For example, you may have a couple of photocells "looking" for a light source or a vision system. These systems provide feed back that does not say "turn 90 deg" but rather "keep turning your not there yet".
MEMS accelerometers are not very good for measuring change in distance. The output is at best +/- 1mg which does not sound so bad but when you have to take the double integral to got position you'll soon realize that error accumulates very fast. Your latter suggestion would probably work well but would be a pain to implement. I don't think I've seen many people do what you suggest.
One thing I have seen work relatively well is hacking two optical mice to watch the floor. Two are needed to detect rotation as well as displacement.
"This is a running problem with most batteries when operating a motor. Unless the battery is entirely solid state and doesn't decay over time, the voltage it supplies drops as it gets used up. Problem continues to exist today with Lego Mindstorm robots as well. To correct for this, one would want to put a sensor which senses the rotating shaft on the motor -- keep rotating motor until it's gone through the proper amount of degrees instead of just supplying driving voltage for about the right amount of time "
Though I'm not sure what your trying to say about "solid state" batteries but adding shaft encoders is not going to make steering work 100%.
My area of research is in robotics and I can tell you that it is incredibly difficult to get any kind of robot to turn 90 degrees. No matter how you turn the wheels are always going to slip some. There is no way of knowing exactly how much they will slip. With just shaft encoders it not possible to get a robot to turn at any predictable angle.
It seems like every time an article like this is on slash dot a million people say "wow I can't wait for a computer using that technology".
What people _don't_ understand is this is not the same technology as is used in a microprocessor. CPUs used Field Effect Transistors. The advantage of FETs is that there is no gate-drain current when the transistor isn't switching so they take very little power. With a bi-polar transistor, you are using a current switch, which would take massive amounts of current if you put many of these into an IC.
A more realistic application would be in communications systems where your carrier frequency is at 500Ghz.
Sorry to burst your bubble but you won't see 500Ghz computers next year. Maybe not ever using CMOS.
I actually read the "article" and the author said that standard Digital Rights features would be included as well as epoxy encapsulation and case intrusion.
Case intrusion can be a useful feature by which you may be able to disable the device if opened (with or with out power).
It is my understanding epoxy encapsulation is when you bond the die directly to the PCB and embed the IC under a bubble of epoxy. This is standard practice.
IMHO this is all fake so I guess it doesn't really matter in the end.;)
"A lot of information can be gleaned from the timing of the keystrokes and some (relatively simple) packet decoding"
This can be avoided by altering your key timing. How do you do this? Don't use a Qwerty keyboard:P Sense I've switched to a Dvorak key map my finger movements have been diminished by >60%. This change in finger movements has got to also translate into different keystroke timings.
Tantalum caps are only one of the seemingly thousands or varieties of capacitors around. If they all went away today we could easily replace them with other varieties of capacitors. Sure tantalum caps are one of my favorite varieties because of their long life and their low leakage current but there not that essential and if you look at a lot of newer electronics you won't find any tantalums any way because there so dam expensive.
Slashdot Mirror
A virus needs DNA (or RNA) and a protective shell that helps protect and deliver it's payload. Here there's no shell.
Also, any commercial hardware and all research hardware i know of would encrypt the data before writing to prevent this from happening intentionally. No implemented encryption scheme is perfect but it would be far easier to just order your "virus" DNA with a credit card then it would be to hack a DNA hard drive that only makes DNA a 100-200 bases long.
Actually I think that millions of times faster statement in the summary came from a misquote or misunderstanding. For archival storage(write once read VERY seldom). It should have been millions (or trillions really) more bytes. 20 hours to write an archive and 1 hour to read it isn't bad. I believe amazon glacier has a similar read latency.
This is an excellent question! In principle any (hetero)polymer would work.
For context, in the short term we're targeting archival storage and not high speed storage (it'll probably never be low latency in the same way an SRAM is). The amazing thing about DNA is that it's not only long lived under reasonable storage conditions but also eternally relevant. Try reading a 30 year old 8" floppy today. The data may still be okay but you'll have trouble finding the hardware. Since DNA is so important to humans in other contexts (medicine), we can be fairly certain that DNA reading (sequencing) will be easy and available in 30 years, or 300, or 3000.
DNA is also much easier to manipulate. Nature provides us with all the tools (enzymes) we need to copy and select specific sequences. Academia provides us with a better understanding of DNA than other polymers. And industry provides us with wonderful machines like high throughput DNA sequencers. Watson-Crick base paring also allows DNA to do computation, which in some cases can be done directly on the data encoding DNA.
As for safety, you are right in that many microbes are quite good at finding and incorporating DNA but such an even would still be quite rare and there are many things to consider:
1) DNA data payloads are stored as just DNA, typically frozen or dried. And DNA alone is not sufficient to be pathogenic.
2) Generally when scientists store data in DNA right now it's encrypted first. This make it hard to pick out payload data that will encode for something specific.
3) It's far easier for a malicious person to just get parts of genes synthesized from different vendors and stitch them together at home. That ship has already sailed.
4) Currently the best way to synthesize the data payloads is in short sequences of DNA that aren't long enough to encode for anything but short peptides.
5) Randomly coming across a working pathogenic DNA sequence is technically possible but practically impossible (much like it's possible that you'll spontaneously quantum tunnel to the center of the earth but that'll never actually happen). An average gene in e. coli is about 8000 bases long. Of that size alone there are 4^(8000) (a beyond astronomically large number) of DNA sequences and only a handful of those do anything at all. Of those most would be toxic to the microbe it self or be a useless metabolic load (proteins are expensive to make and useless genes get cut out or turned off VERY fast in a population---days in continuous culture).
6) If desired, you can also trade off a little density to insert stop codons occasionally, limiting the size of anything translated to only a few amino acids.
jquery isn't really a "Google API"
but googleapis.com, which is what the GP was refferring to is google APIs, which allows google to track who's downloading APIs from googleapis
Specifically, how is this different from other projector based stereolithography printers such as the muve3d DLP (http://www.muve3d.net/press/)?
I think it's in quotes because it's not an established word for people outside your(physics?) discourse community.
In some places its even better. Mine is 40Mbit down/20Mbit up (attainable), which is near the top of what VDSL can do(AFAIK)
Saying tape has a longer life is silly. I'd have no idea where to get an 8-track player today even though it's an analog format.
Same with a record player, but I could make one pretty easily. (there's a reason why we shot a record into space instead of a tape)
Really, though a documented and uncompressed digital file, properly kept track of, could last forever similar to a record even if we lost our codecs it would be easy to write a new one.
Other respondents to this comment pointed out that these are fluorescent (ie need an excitation light source) not "glowing".
But another problem with this idea is that, in a population neutral alleles maintain their frequency (though can drift randomly) and deleterious alleles will decline in frequency. In other words, you'd have to release a LOT of rabbits before the glowing allele would be common enough to have an effect on average fitness and that allele would be unstable in the population. Unless you don't believe in evolution, but then you're better off praying for the rabbits to go away.
It is seriously about time! The D-pad on the current controllers is absolutely awful (for those who don't know, often times it registers the wrong direction due to poor mechanical design). It has made some games (especially live-arcade games like megaman 9) incredibly frustrating.
Its naive to think that buses are only good for storage.
1080p video is 1080*1920pixels/frame*32bits/pixel*60frames/second is roughly 4Gb (uncompressed). If history is any indication displays will get larger and denser (also remember bandwidth needed will scale by the square of the number of lines). Or what if you want an external GPU? 16x pci express is about 32Gbits/sec. Also more speed reduces latency, which may be helpful too depending on what you use it for.
This is not some boring super scaler! Nor is it some vector processor!
in fact this is a complete departure from a von Neuman architecture. The architecture is called a Dataflow architecture. In one sentence a dataflow architecture is one where instruction execution is based on the availability of the instructions inputs not a program counter.
The article does a very bad job at conveying the fact that this is a relatively new idea. Like most reporting they report something thats been in research for some time as a huge breakthrough without describing it at all. Instead its really just an incremental step in dataflow computing research.
I work in a lab at the University of Washington on another dataflow architecture. Its a really interesting idea but it will take some time to develop and you're not going to get one on your desk for some years to come.
It really is different. Its not simply a super scaler. It's a data-flow machine. What this means is that instructions are arranged in a graph based on dependency and execute as soon as all inputs are ready.
I work in a lab at the University of Washington where we are working on _implementing_ a different data flow machine that shares some of fundamentals with the UT machine.
I was deleting a programs registry and I got distracted...
Next thing I knew I clicked yes to "are you sure you want to delete HKLM"... needless to say windows died that day.
Ah, Now I understand. This is an over simplification. A cell voltage will decrease as it becomes discharged. However, if this were the reason motors did not get the same ammout of power at time A as they did at time B then it could be easily fixed with electronics.
A battery can be moddled as voltage source with an internal resistance in series. As a battery discharges this resistance increses.
While this is not 100% accurate (As cell voltage does change too.) It models the important fact that A cells capacity to provide current decreases with time and while a cells voltage may go from 1.2v (full) to 1.0v (dead) (for NiCad) the internal resistance will change much more dramatically over the life span of the battery.
The reason I did not say what can be done is that it is very hard to do. Typically what is done is there is some other system on board that tells the robot where to go. For example, you may have a couple of photocells "looking" for a light source or a vision system. These systems provide feed back that does not say "turn 90 deg" but rather "keep turning your not there yet".
MEMS accelerometers are not very good for measuring change in distance. The output is at best +/- 1mg which does not sound so bad but when you have to take the double integral to got position you'll soon realize that error accumulates very fast. Your latter suggestion would probably work well but would be a pain to implement. I don't think I've seen many people do what you suggest.
One thing I have seen work relatively well is hacking two optical mice to watch the floor. Two are needed to detect rotation as well as displacement.
Though I'm not sure what your trying to say about "solid state" batteries but adding shaft encoders is not going to make steering work 100%.
My area of research is in robotics and I can tell you that it is incredibly difficult to get any kind of robot to turn 90 degrees. No matter how you turn the wheels are always going to slip some. There is no way of knowing exactly how much they will slip. With just shaft encoders it not possible to get a robot to turn at any predictable angle.
These are good points. I hope people are a little more informed now. :)
It seems like every time an article like this is on slash dot a million people say "wow I can't wait for a computer using that technology".
What people _don't_ understand is this is not the same technology as is used in a microprocessor. CPUs used Field Effect Transistors. The advantage of FETs is that there is no gate-drain current when the transistor isn't switching so they take very little power. With a bi-polar transistor, you are using a current switch, which would take massive amounts of current if you put many of these into an IC.
A more realistic application would be in communications systems where your carrier frequency is at 500Ghz.
Sorry to burst your bubble but you won't see 500Ghz computers next year. Maybe not ever using CMOS.
I actually read the "article" and the author said that standard Digital Rights features would be included as well as epoxy encapsulation and case intrusion.
;)
Case intrusion can be a useful feature by which you may be able to disable the device if opened (with or with out power).
It is my understanding epoxy encapsulation is when you bond the die directly to the PCB and embed the IC under a bubble of epoxy. This is standard practice.
IMHO this is all fake so I guess it doesn't really matter in the end.
"A lot of information can be gleaned from the timing of the keystrokes and some (relatively simple) packet decoding"
:P Sense I've switched to a Dvorak key map my finger movements have been diminished by >60%. This change in finger movements has got to also translate into different keystroke timings.
This can be avoided by altering your key timing. How do you do this? Don't use a Qwerty keyboard
Takahashi
Tantalum caps are only one of the seemingly thousands or varieties of capacitors around. If they all went away today we could easily replace them with other varieties of capacitors. Sure tantalum caps are one of my favorite varieties because of their long life and their low leakage current but there not that essential and if you look at a lot of newer electronics you won't find any tantalums any way because there so dam expensive.